Rights / License: Research Collection In Copyright - Non ...492/eth-492-02.pdfAccepted on the recommendation of Prof. Dr. Christophe Lacroix, examiner Prof. Dr. Leo Meile, co-examiner

Research Collection

Doctoral Thesis

Novel Technologies for detection, production and screening ofstress tolerant bifidobacteria

Author(s): Reimann, Sebastian

Publication Date: 2009

Permanent Link: https://doi.org/10.3929/ethz-a-005939078

Rights / License: In Copyright - Non-Commercial Use Permitted

This page was generated automatically upon download from the ETH Zurich Research Collection. For moreinformation please consult the Terms of use.

ETH Library

https://doi.org/10.3929/ethz-a-005939078

http://rightsstatements.org/page/InC-NC/1.0/

https://www.research-collection.ethz.ch

https://www.research-collection.ethz.ch/terms-of-use

Diss. ETH Nr. 18627

NOVEL TECHNOLOGIES FOR DETECTION,

PRODUCTION AND SCREENING OF STRESS

TOLERANT BIFIDOBACTERIA

A dissertation submitted to

Swiss Federal Institute of Technology Zurich (ETH ZURICH)

for the degree of

Doctor of Sciences

presented by

SEBASTIAN REIMANN

Dipl. Natw. ETH

Born April 24, 1979

Citizen of Oberhof AG

Accepted on the recommendation of

Prof. Dr. Christophe Lacroix, examiner

Prof. Dr. Leo Meile, co-examiner

Prof. Dr. Paul Ross, co-examiner

2009

Table of contents

I

Table of contents

TABLE OF CONTENTS.................................. ................................................................. I

ABBREVIATIONS...................................... ..................................................................... V

SUMMARY..................................................................................................................... VI

ZUSAMMENFASSUNG .................................... ............................................................. IX

1. INTRODUCTION...................................................................................................... 1

1.1 Human gastrointestinal tract microbiota .................................................................................................... 1

1.2 Bifidobacterium ............................................................................................................................................. 3 1.2.1 General characteristics ................................................................................................................................ 3 1.2.2 Metabolism of Bifidobacteria ..................................................................................................................... 4 1.2.3 Role of Bifidobacteria in the GIT ............................................................................................................... 5 1.2.4 Transcriptomics of Bifidobacterium longum NCC2705 ............................................................................. 6

1.3 Probiotic concept ........................................................................................................................................... 7

1.4 Health promoting effects related to probiotics............................................................................................ 9

1.5 Potential risks associated with probiotics .................................................................................................... 2

1.6 Technological requirements of probiotic bacteria .................................................................................... 13

1.7 Technologies to improve cell viability and yields ..................................................................................... 15 1.7.1 Production technologies........................................................................................................................... 15

1.7.1.1 Optimization of growth medium ..................................................................................................... 15 1.7.1.2 Stress adaptation.............................................................................................................................. 16 1.7.1.3 Immobilized cell (IC) technology in continuous cultures................................................................ 20

1.7.1.3.1 Principle of cell immobilization ................................................................................................. 20 1.7.1.3.2 Techniques of cell immobilization ............................................................................................. 21 1.7.1.3.3 Physiological changes of immobilized cells ............................................................................... 22 1.7.1.3.4 Advantages of continuous IC cultures ........................................................................................ 23

1.7.2 Stabilization of probiotics ......................................................................................................................... 26 1.7.2.1 Microencapsulation ......................................................................................................................... 26 1.7.2.2 Spray drying .................................................................................................................................... 27

1.7.2.2.1 Principle of spray drying............................................................................................................ 28 1.7.2.2.2 Parameters affecting cell viability during spray drying .............................................................. 31 1.7.2.2.3 Parameters affecting storage of dried powders ........................................................................... 36 1.7.2.2.4 Rehydration ................................................................................................................................ 38

1.7.3 Enumeration of viable probiotic cells ....................................................................................................... 39

Table of contents

I

1.8 Hypothesis and background of this study.................................................................................................. 41 1.8.1 Background and general objective ............................................................................................................ 41 1.8.2 Specific objectives .................................................................................................................................... 42 1.8.3 Hypotheses................................................................................................................................................ 43

2 IMPROVED TOLERANCE TO BILE SALTS OF AGGREGATED B. LONGUM NCC2705 PRODUCED DURING CONTINUOUS CULTURE WITH IMMOBILIZED CELLS.............................................. ............................................................................. 44

2.1 Abstract ........................................................................................................................................................ 45

2.2 Introduction ................................................................................................................................................. 46

2.3 Material and Methods ................................................................................................................................. 48 2.3.1 Strain and medium .................................................................................................................................... 48 2.3.2 Cell immobilization .................................................................................................................................. 48 2.3.3 Continuous fermentations ......................................................................................................................... 49 2.3.4 Batch fermentation.................................................................................................................................... 51 2.3.5 Viable cells enumeration........................................................................................................................... 51 2.3.6 Dry biomass determination ....................................................................................................................... 52 2.3.7 Microscopic observation........................................................................................................................... 53 2.3.8 Metabolite analysis ................................................................................................................................... 53 2.3.9 Fatty acid analysis..................................................................................................................................... 53 2.3.10 Heat test................................................................................................................................................54 2.3.11 Bile salt test .......................................................................................................................................... 54 2.3.12 Tolerance to antibiotics ........................................................................................................................ 55 2.3.13 Statistical analyzes ............................................................................................................................... 56

2.4 Results........................................................................................................................................................... 56 2.4.1 Biomass production .................................................................................................................................. 56 2.4.2 Metabolic Activity .................................................................................................................................... 60 2.4.3 Cell sensitivity to heat and porcine bile salts ............................................................................................ 63 2.4.4 Tolerance to antibiotics............................................................................................................................. 64 2.4.5 Cell membrane fatty acid composition ..................................................................................................... 65

2.5 Discussion ..................................................................................................................................................... 67

3 DEVELOPMENT OF A REAL-TIME RT-PCR METHOD FOR ENUMER ATION OF VIABLE BIFIDOBACTERIUM LONGUM CELLS IN DIFFERENT MORPHOLOGIES . 72

3.1 Abstract ........................................................................................................................................................ 73

3.2 Introduction ................................................................................................................................................. 74

3.3 Materials and Methods................................................................................................................................ 76 3.3.1 Strain and growth conditions .................................................................................................................... 76 3.3.2 Preparation of bacterial samples for calibration curves ............................................................................ 76 3.3.3 Production of aggregated cells .................................................................................................................. 77 3.3.4 Heat stress ................................................................................................................................................. 77 3.3.5 Rifampicin test ..........................................................................................................................................78 3.3.6 RNA isolation and reverse transcription reaction ..................................................................................... 78 3.3.7 Determination of RNA recovery rate using internal reference ................................................................. 79 3.3.8 Quantitative PCR amplification ................................................................................................................ 80 3.3.9 Statistical analyzes .................................................................................................................................... 81

Table of contents

I

3.4 Results........................................................................................................................................................... 81 3.4.1 Morphology of B. longum NCC2705........................................................................................................ 81 3.4.2 Calibration curves ..................................................................................................................................... 82 3.4.3 Heat test .................................................................................................................................................... 84 3.4.4 Rifampicin test ..........................................................................................................................................85 3.4.5 Viable cells quantification of cell aggregates............................................................................................ 86

3.5 Discussion ..................................................................................................................................................... 87

4 DEVELOPMENT OF A RAPID SCREENING PROTOCOL FOR SELEC TION OF BIFIDOBACTERIUM LONGUM STRAINS RESISTANT TO SPRAY DRYING AND POWDER STORAGE.................................................................................................... 90

4.1 Abstract ........................................................................................................................................................ 91

4.2 Introduction ................................................................................................................................................. 92

4.3 Material and Methods ................................................................................................................................. 94 4.3.1 Bacterial strains and culture conditions .................................................................................................... 94 4.3.2 Preparation of bacterial suspensions used for survival tests ..................................................................... 95 4.3.3 Preparation of strain mix........................................................................................................................... 95 4.3.4 Spray drying.............................................................................................................................................. 96 4.3.5 Storage test................................................................................................................................................ 96 4.3.6 Determination of water activity and moisture content .............................................................................. 97 4.3.7 Viable cell counts in spray dried powders ................................................................................................ 97 4.3.8 Identification of best surviving strains using RAPD................................................................................. 98

4.4 Results........................................................................................................................................................... 99 4.4.1 Strain survival and selection during the first selection batches................................................................. 99 4.4.2 Strain survival and selection during the second selection round............................................................. 104 4.4.3 Validation of selection procedures using single strain preparations ....................................................... 106

4.5 Discussion ................................................................................................................................................... 107

5 GENERAL CONCLUSIONS AND OUTLOOK.................... ................................. 111

5.1 General conclusions ................................................................................................................................... 111

5.2 Outlook and perspectives .......................................................................................................................... 114

6 APPENDIX ........................................................................................................... 117

6.1 Supplementary material from Chapter 4 ................................................................................................ 117

6.2 Genome expression analysis of B. longum NCC2705 produced during continuous culture with immobilized cells and batch culture ....................................................................................................................... 118

6.2.1 Background............................................................................................................................................. 118 6.2.2 Material and methods.............................................................................................................................. 118 6.2.3 Results and discussion ............................................................................................................................ 121

6.2.3.1 Comparison of cells from ICC with batch cultures ....................................................................... 121 6.2.3.2 Comparison between ICC1 and ICC2 and within ICC2................................................................ 122

Table of contents

I

7 REFERENCES..................................................................................................... 128

ACKNOWLEDGEMENTS ................................... ........................................................ 154

Abbreviations

V

Abbreviations

bp Base pair

cDNA Complementary DNA (reverse transcribed RNA)

CFU Colony forming units

cysS Gene coding for cysteinyl-tRNA synthetase

Da Dalton

DNA Deoxyribonucleic acid

GIT Gastrointestinal tract

HPLC High pressure liquid chromatography

IC Immobilized cells

LAB Lactic acid bacteria

mRNA Messenger ribonucleic acid

MRS Man Rogosa and Sharpe medium

MRSC MRS supplemented with 0.5% L-cysteine

NCC Nestlé Culture Collection

OD Optical density

ORF Open reading frame

PCR Polymerase chain reaction

purB Gene coding for adenylosuccinate lyase

qPCR Quantitative PCR (real-time PCR)

qRT-PCR Quantitative reverse transcription real-time PCR

RAPD Randomly amplified polymorphic DNA

RNA Ribonucleic acid

rpm Revolutions per minute

RT Reverse transcription

SFA Saturated fatty acid

USFA Unsaturated fatty acid

w/v Weight per volume

Summary

VI

Summary

Probiotics are defined as “live organisms which when administered in adequate amounts confer a

health benefit on the host”. The delivery of live cultures t adequate levels at the target site in the

host represents a challenging task in the production of probiotic food products. Probiotic bacteria,

generally isolated from the gastrointestinal tract of humans, are fastidious microorganisms and

susceptible to environmental stresses that occur during production, storage and transit through the

gastro intestinal tract. Different strategies have been proposed to produce probiotics with

enhanced stress resistance such as efficient screening methods to identify probiotic strains with

intrinsic resistance to environmental stresses, genetic engineering, the pre-treatment of bacterial

cells with sub-lethal stresses or the use of immobilized cells (IC) and continuous culture.

The dissertation consisted of three hypotheses: The first hypothesis claimed that B. longum

NCC2705 cells produced as IC and continuous culture exhibits improved tolerance to

environmental stresses which increases with culture age. The second hypothesis states that qRT-

PCR can be used to quantify viable B. longum NCC2705 cells independently of morphological

state. The third hypothesis is that spray drying can be used for preparation of bifidobacteria, and

that random amplified polymorphic DNA (RAPD) technology can be used to identify the best

survivor strain.

In a first approach, the probiotic strain Bifidobacterium longum NCC2705 was produced with

(IC) technology in combination with continuous culture (Chapter 2). The production of

immobilized B. longum NCC2705 was carried out in a two-stage fermentation system consisting

of a first small reactor (R1) containing cells entrapped in gellan/xanthan gel beads for high cell

productivity and a second larger reactor (R2) inoculated with bacteria produced from the first

reactor. Reactor volumes of 360 ml (R1) and 1800 ml (R2) were chosen to yield a shorter

Summary

VII

residence time in R1 than in R2, in order to obtain high cell productivity in R1 (high immobilized

cell concentration at high dilution rates (2.25 h-1)) followed by a final growth in R2 with longer

residence time at a low dilution rate (0.2 h-1). Two continuous cultures were produced for 20 days

with and without glucose limitation. Both fermentations exhibited formation of macroscopic cell

aggregates. Cells in the effluent from the second reactors showed enhanced tolerance to porcine

bile salts and to aminoglycosidic antibiotics which were associated with modifications in the cell

membrane composition. However, there was no significant difference in terms of culture age and

heat resistance when compared to cells from free cell batch cultures.

Formation of large cell aggregates resulted in an underestimation of viable cell counts when

measured by the traditional plate counting method. This complicated the characterization of

physiological properties. Hence, we developed a new method to enumerate viable B. longum

NCC2705 which is not influenced by the morphological state of the cells (Chapter 3). The new

method was based on real time PCR technology and enabled the accurate quantification of cells

in the aggregated state. To achieve this, we measured the level of a constitutively expressed

housekeeping gene and calculated viable cell concentration using calibration curves from 7.0 ×

103 to 4.7 × 108 CFU/ml. This quantification of IC samples containing aggregates resulted in up

to 125-fold higher viable cell counts compared to traditional plate counts. The method was

further used for analyzes of tolerance to bile salts of IC samples (see above).

In the last part of the thesis, a different strategy was evaluated to produce robust bifidobacteria.

Instead of improving stress tolerance of one specific B. longum strain, we developed a high

throughput screening method to identify bifidobacterial strains with intrinsic resistance to stresses

occurring during spray drying and storage (Chapter 4). In this approach, twenty-two B. longum

strains were mixed in separate batches and subjected to spray drying and storage experiments

during two selection rounds. Unique identification of the twenty-two B. longum strains in the

Summary

VIII

mixed bacterial preparation was achieved using molecular-based methods. Within the mixed

strain preparation, a robust strain was identified and stability was confirmed by performing single

strain spray drying and storage tests. The selected strain exhibited a 250-fold better survival after

60 days of storage than the control strain. The method allowed faster screening of resistance of B.

longum strains to spray drying and storage compared to traditional screening performed on

individual strains.

In this thesis, IC and continuous cultures used to produce probiotic cells with improved stress

resistance resulted in the formation of large cell aggregates. Auto-aggregation is a known

phenotype in LAB and bifidobacteria and was associated with enhanced ability to adhere to host

epithelial cells. Therefore it would be of interest to test adhesion ability of the cells produced with

IC and continuous culture. Additionally, we designed and experimentally validated a high

throughput screening method to select for probiotic strains with intrinsic resistance to stresses

occurring during spray drying and storage. The same approach could be applied for selection of

robust strains in response to other stress responses.

Zusammenfassung

IX

Zusammenfassung

Die gesundheitsfördernde Wirkung probiotischer Bakterien u.a. Wiederherstellung und

Stabilisierung einer ausgewogenen Darmflora, Verdrängung von Krankheitserregern und

unerwünschten Bakterien im Darm, Beeinflussung des Immunsystems und Verbesserung der

natürlichen Abwehrkräfte des Körpers, Senkung der Konzentration gesundheitsschädigender

Stoffwechselprodukte und krebsfördernder Enzyme im Dickdarm und Verhinderung bzw.

Verkürzung von Durchfallerkrankungen, wird nach momentanem Wissenstand hauptsächlich mit

einer bestimmter Anzahl lebender Bakterien erzielt. Einzelne Stämme der Gattung

Bifidobacterium sind bekannt für ihre probiotische Wirkung und Produkte, welche diese Stämme

beinhalten, sind im Handel erhältlich. Da sich Bifidobakterien natürlicherweise im Darmtrakt von

Säugetieren (u.a.) befinden, sind sie anaerobe Mikroorganismen und sensibel gegenüber

Sauerstoff und anderen Umwelteinflüssen, welche bei der Produktion, Lagerung und Verzehr von

probiotischen Produkten auftreten können. In dieser Hinsicht erfreut sich die Entwicklung von

neuen Technologien, welche die Robustheit von probiotischen Bakterien gegenüber diesen

Umwelteinflüssen stärken, einer grossen Nachfrage von Produzenten probiotischer

Lebensmitteln.

Diese Dissertation befasst sich in erster Linie mit der Entwicklung neuer Strategien, welche die

Robustheit des probiotischen Modelbakteriums Bifidobacterium longum NCC2705 gegenüber

Umweltfaktoren stärken sollen. Vorangehende Studien haben gezeigt, dass kontinuierliche

Zellkulturen mit immobilisierten Bakterien in Polysaccharidgel-Kügelchen physiologische

Änderungen hervorrufen können, welche sich positiv auf die Toleranz gegenüber verschiedenen

Umweltfaktoren auswirken können. Diese Änderungen werden den speziellen Bedingungen,

welche innerhalb der Polysaccharid-Matrix und an der Oberfläche dieser Kügelchen

Zusammenfassung

X

vorherrschen, zugeschrieben. In dieser Arbeit, wurde der Stamm B. longum NCC2705 in

Polysaccharidgel-Kügelchen bestehend aus einer Mischung aus den Polysacchariden Gellan und

Xanthan immobilisiert und über eine Zeitdauer von 20 Tagen kontinuierlich kultiviert (Kapitel

2). Hierzu wurde ein bewährtes System verwendet, welches aus zwei Bioreaktoren verschiedener

Grössen besteht, die miteinander verbunden sind. Im ersten, kleineren Reaktor (360 ml) befinden

sich die immobilisierten Zellen, welche kontinuierlich mit einer hohen Durchflussrate (2.25 h-1)

frischer Nährlösung versorgt werden. Der zweite grössere Reaktor (1800 ml) ist in Serie mit dem

ersten Reaktor geschaltet und wird fortwährend mit Zellen angeimpft, welche im ersten Reaktor

produziert wurden. Das grössere Volumen des zweiten Reaktors bewirkt eine kleinere

Durchflussrate (0.2 h-1) und gewährt den Zellen eine längere Verweilzeit, so dass sie ihr

Wachstum im zweiten Reaktor abschliessen können. Proben wurden in zeitlichen Abständen dem

System entnommen, verschiedene Parameter getestet und mit einer Standard Referenz-Kultur

verglichen. Es konnte eine signifikant erhöhte Resistenz gegenüber Gallensalzen und gewissen

Antibiotika, ausgemacht werden, welche unabhängig vom Zeitpunkt der Probennahme waren und

welche dem veränderten Aufbau der Zellmembran bzw. deren Fettsäurestruktur zugeschrieben

wird. Diese Veränderung der Membranstruktur ist höchstwahrscheinlich auch der Grund, warum

sich nach einer gewissen Fermentationszeit makroskopische Zellklumpen von 0.5 – 1 mm

Durchmesser im Überstand beider Bioreaktoren gebildet haben. Verklumpung von Bakterien

kann durchaus interessante Eigenschaften für probiotische Bakterien bergen (physischer Schutz,

erhöhte Adhäsion an Epithelialzellen des Wirtes), führt jedoch zur Unterschätzung der Lebend-

Zell-Konzentrationen mit der traditionellen Methode des Verdünnens und Ausplattierens. Eine

physiologische Charakterisierung der verklumpten Bifidobakterien basierend auf dieser Methode

ist folglich erschwert oder gar unmöglich.

Zusammenfassung

XI

Um dennoch eine physiologische Analyse der immobilisierten Zellen durchzuführen, wurde im

zweiten Teil der Dissertation eine neue Methode entwickelt, um die Anzahle lebender B. longum

NCC2705 Zellen unabhängig von ihrer Morphologie bestimmen zu können (Kapitel 3). Die

Methode nutzt quantitative „echt Zeit“ Polymerasen Kettenreaktion (engl. qRT PCR) als ein

Mittel, um die Kopienzahl eines sogenannten Haushaltsgens zu messen, um anhand einer

Standardkurve, welche Zellkonzentrationen über eine Spannweite von 7.0 × 103 bis 4.7 × 108

Kolonien formende Einheiten pro ml abdeckt, auf die ursprüngliche Zellkonzentration in der

Zellprobe zu schliessen. Die Kopien eines Haushaltsgens wurden ausgewählt, weil zum einen die

Anzahl der Kopien pro Zelle konstant ist, unabhängig von der Umgebung und des

physiologischen Zustandes der Zelle, und zum andern weil die Kopien eines Gens (mRNA)

instabil sind und kurz nach der Bildung wieder abgebaut werden. Ein konstantes Niveau der

Kopienanzahl eines bestimmten Gens kann deshalb nur durch konstante Erneuerung und folglich

nur von metabolisch aktiven, also lebenden Zellen erreicht werden. Eine gewisse Anzahl

gemessener Kopien eines Haushaltgens kann deshalb mit einer bestimmten Anzahl lebender

Bakterien gleichgesetzt werden, was z.B. bei der stabileren DNA nicht der Fall ist, welche auch

Stunden oder gar nach Tage nach dem Zelltod nachgewiesen werden kann.

Die Methode wurde experimentell validiert und zur Konzentrationsbestimmung von Zellen,

welche in makroskopischen Zellklumpen vom der kontinuierlichen Kultur mit immobilisierten

Zellen stammen verwendet. Die erhaltenen Konzentrationen lebender Bakterien wurden dann mit

den Werten, welche mit der traditionellen Methode des Ausplattierens erhalten wurden,

verglichen. Das Resultat offenbarte, dass Lebend-Zell-Konzentrationen bestimmt mittels qRT-

PCR bis zu 125 Mal höher waren als jene, welche mittels Ausplattierens erhalten wurden. Mit

Hilfe dieser Methode war es nun möglich, die reale Konzentration von lebenden Zellen in

Zusammenfassung

XII

Zellklumpen, nachdem sie mit Gallensalzen oder Hitze behandelt wurden, zu berechnen und so

die Toleranz von diesen Zellen gegenüber Gallensalzen oder Hitze zu testen (siehe oben).

In der letzten Phase der Dissertation wurde ein neuer Ansatz gewählt. Anstatt robuste

Bifidobakterien zu produzieren, entwickelten wir ein effizientes Auswahlprüfverfahren, welches

die Identifizierung von B. longum Stämmen ermöglicht, die eine natürliche Toleranz gegenüber

Stressbedingungen aufweisen, die während der Sprühtrocknung und Lagerungskonditionen

auftreten können. Um zeitraubende Wiederholungen der Lagerungstests zu vermeiden, wurden

mehrere Stämme gleichzeitig getestet. Als notwendige Bedingung für ein solches simultanes

Verfahren gilt die Unterscheidbarkeit der einzelnen Stämme untereinander. Hierzu wurde ein

Verfahren verwendet, welches sich auf Polymerasen Kettenreaktion (engl. PCR) stützt und die

polymorphe DNA Sequenz von Stämmen derselben Unterklasse ausnutzt. Die Polymerasen

Kettenreaktion mit drei „Universalprimern“ vervielfältigt spezifische Sequenzen der bakteriellen

DNA (engl. RAPD). Diese spezifische Vervielfältigung variiert von Stamm zu Stamm und

ermöglicht die Fertigung von Stamm-spezifischen molekularen Vervielfältigungsmustern mittels

Gel-Elektrophorese, dem sogenannten „molekularen Fingerabdruck“. Die so identifizierten 22 B.

longum Stämme wurden in Ansätzen von 7 – 8 Stämmen sprühgetrocknet und das resultierende

Pulver mit den lebenden Bakterien einem Lagerungstest bei erhöhten Temperaturen unterzogen

und wöchentlich ausplattiert, um die Lebend-Zellzahl im Pulver zu bestimmen. Die DNA der

verbleibenden Kolonien (bestehend aus unzähligen, genetisch identischen Klonen) von den

Nährlösung-Agar-Platten (100 – 200 Kolonien) wurde extrahiert und mit RAPD Technologie

identifiziert. Mit diesem Verfahren konnte am Ende des Lagerungstests derjenige Stamm

(NCC572) bestimmt werden, von welchem die meisten Kolonien übrig geblieben sind, respektive

welcher die Bedingungen von Sprühtrocknung und beschleunigter Lagerungstest am besten

überlebt hat. Die Methode wurde experimentell verifiziert, indem Stamm NCC572 einzeln

Zusammenfassung

XIII

angezüchtet, sprühgetrocknet und gelagert und sein die Überlebensrate mit einem Kontrollstamm

(NCC2916) verglichen wurde. Tatsächlich stellte sich heraus, dass Stamm NCC572 nach 60

Tagen Lagerung in Pulverform rund 250 Mal besser überlebte als der Kontrollstamm NCC2916.

Die entwickelte Methode erlaubt die zeit- und kosteneffiziente Anwendung eines Auswahl-

prüfverfahrens, mit welcher probiotische Bakterienstämme mit inhärenter Stresstoleranz

identifiziert werden können.

Die Methoden, welche in der vorliegenden Dissertation entwickelt wurden, können mannigfaltig

angewendet werden. So kann z.B. das beschleunigte Auswahlprüfverfahren auf beliebig andere

Konditionen angewendet werden. Ferner wäre es interessant, die verklumpten Bifidobakterien

aus der kontinnuierlichen Kultur mit immobilisierten Zellen, auf ihre Adhäsionsfähigkeit an

menschlichen Epithelialzellen zu testen.

Chapter 1 - Introduction

- 1 -

1. Introduction

1.1 Human gastrointestinal tract microbiota

The human gastro intestinal tract (GIT) is a complex system composed of different

compartments: oral cavity, stomach, small intestine (duodenum, jejunum and ileum), and large

intestine (caecum, colon and rectum) exhibiting very different environmental conditions. The

intestinal epithelium has a combined surface area of 400 m2 and provides ecological niches for

the three domains of life: prokarya, archaea and eukarya. The GIT is one of the most dense

microbial ecosystems on earth (Whitman et al., 1998), and its population number of up to 1014

cells exceeds that of the human cells by a factor of 10 (Backhed et al., 2004). Despite this

tremendous number, the gut microbiota forms an exclusive community underlying a strict

selection pressure. Gut microbes must be able to react rapidly to environmental changes, survive

the harsh conditions during gastrointestinal transfer and during re-colonization of a new host,

evade bacteriophages, and be capable of growth under GIT conditions to avoid disappearance

(Ley et al., 2006). Gut inhabitants may also harbor a collection of enzymes for nutrient

degradation and contain a cell-surface composition which might enable attachment to host

epithelial cells (Ley et al., 2006).

The microbial colonization of the GIT varies with age and is characterized by temporary changes.

The sterile GIT of newborns is colonized mainly by enterobacteria, bifidobacteria, and

streptococci. The percental distribution depends on the environment, mode of birth, the diet and

possibly on the host genetic background (Edwards and Parret 2002; Vaughan et al., 2002; Fanaro

et al., 2003; Zoetendal et al., 2004). At older age, the microbial diversity increases to more than

1000 different subspecies of bifidobacteria, enterobacteria, streptococci, bacteroides, clostridiae

and cyanobacteria. However, a satisfying general representation of the ecosystem has not yet


- 2 -

been established because only a limited number of human gastrointestinal microflora has been

analyzed and each subject harbors a unique microbial community (Rajilic-Stojanovic et al.,

2007). The diversity of microflora develops with host maturation and gets more unstable with

increasing age (Harmsen et al., 2000, Egert et al., 2006). Diversity of microflora is generally host

specific and depends on factors such as digestion characteristics, gender, age and genotype, as

well as on environmental factors like nutritional diet, ingestion of microorganisms and drugs

(Zoetendal et al., 2006).

Host and gut microflora live in a symbiotic relationship; the host provides the microflora a

nutrition-rich and protective habitat, while the microorganisms ferment non-digestible dietary

substrates and indigenous mucus produced by epithelial cells, resulting in the production of short

chain fatty acids (SCFA) which are absorbed by the host (Shanahan, 2002). The generated plus in

dietary energy from fibre fermentation favors the adsorption of ions (Mg2+, Ca2+, Fe2+/3+) in the

caecum (Guarner et al., 2006). Additionally, the commensal flora provides a barrier against

colonization of exogenous and pathogenic bacteria by outcompeting them for nutrients and

binding sites (Guarner et al., 2006) and by producing antimicrobials such as bacteriocins

inhibiting (a.o.) growth of pathogens (Shanahan, 2002). Besides protection, the gut microbes are

involved in developmental and immunological processes of the host by modulating immune

responses and influencing proliferation and differentiation of epithelial cells (Hooper et al., 2001;

Falk et al., 1998, Guarner and Magdalena, 2003). Endogenous bacteria are also involved in the

production of vitamin B12, vitamin K, biotin, folic acid, in the synthesis of amino acids from

ammonium and urea, and in the inactivation of dietary carcinogens (Wollowski et al., 2001,

Hooper et al., 2002). However, the complex microflora with its huge diversity of different

species and subspecies is a major obstacle in the prediction of specific bacterial function within

the GIT.


- 3 -

The development of molecular techniques allowed the exploration of the human microflora and

revealed new insights during the last decade. In the future, advanced molecular tools will confer

on understanding of how single or groups of bacteria interact with the host, thereby providing

new leads for additional investigations. One of the key players in the transient microflora of

newborns and also for the development of probiotic products is the genus Bifidobacterium,

displaying an appearance ranging from 40 – 90 % of total bacteria counts in the intestine of

infants (Harmsen et al., 2000).

1.2 Bifidobacterium

1.2.1 General characteristics

Bifidobacteria were first isolated from faeces of breast-fed infants by Tissier in 1906. They are

among the first species to colonize the sterile GIT of newborns. In adults, Bifidobacterium is the

third most common genus, after bacteroides and eubacteria (Bezkorovainy, 2001). Bifidobacteria

are Gram-positive, heterofermentative, non-motile, non-sporulating bacteria. They vary in size

and exhibit rods of various shapes, often in ‘V’ or ‘Y’ patterns (Latin bifidus: cleft, divided).

Although considered strictly anaerobic, some strains exhibited poor growth in presence of

oxygen (Simpson et al., 2005). Most strains isolated from humans grow optimally at

temperatures of 36 – 38 °C. Bifidobacteria are often included in the group of LAB due to their

metabolic capacities and the sharing of ecological niches (Klein et al., 1998). However,

phylogenetically the genus Bifidobacterium belongs to the family of actinomycetaceae, a family

including Corynebacterium, Mycobacterium, and Streptomyces. Actinomycetaceae share the

common feature of having a high G+C content ranging from 42 to 67 %, which is sufficiently

higher to differentiate them from lactic acid bacteria (Biavati et al., 2000). The genus


- 4 -

Bifidobacterium is relatively small with ca. 32 identified species to date, mainly of human or

animal origin and has a low level of phylogenetic and genomic diversity (Ventura et al., 2006).

The availability of genome sequences of microorganisms from the human GIT and progress in

molecular tools will lead to a better understanding of the behavior and survival of these bacteria

in the GIT. Up to today, the two B. longum strains NCC2705 (Schell et al., 2002) and DJO10A

and two strains of B. longum subsp. infantis (B. longum subsp. infantis ATCC55813 and

ATCC15697) have been sequenced (Lee et al., 2008; Sela et al., 2008).

1.2.2 Metabolism of Bifidobacteria

Bifidobacteria can utilize a large variety of carbohydrate sources. Utilization of complex

carbohydrates from dietary compounds confers a competitive advantage for bifidobacteria in the

lower GIT (Hooper et al., 1999; Klijn et al., 2005). Bifidobacteria metabolize carbohydrates

exclusively and characteristically in the so-called “bifido-shunt” via the key enzyme fructose-6-P

phosphoketolase (F6PPK). This enzyme converts hexoses mainly to lactic and acetic acid

(Scardovi and Tovatelli, 1965). However, sugar availability and consumption rate affects

production of other metabolites, such as formic acid and ethanol, thereby altering the molecular

ratio of these acids (van der Meulen et al., 2006).

Progressive genome sequencing technologies have disclosed an insight in encoded enzymes

involved in metabolic pathways. Genome annotations of bifidobacteria revealed that 8 % of the

genes code for enzymes involved in the carbohydrate metabolism which exceeds the percentage

of other GIT inhabitants such as Enterococcus faecium and Echerichia coli or allochthonous

Lactococcus lactis (Ventura et al., 2007). Sequence in combination with physiological analyzes

confirmed that bifidobacteria utilize a broad range of indigestible polysaccharides like hog gastric


- 5 -

mucin, pectin, plant oligosaccharides and fructo-oligosaccharides (Schell et al., 2002;

MacConaill et al., 2003, Backhed et al., 2004, Matsuki et al., 2004; Parche et al., 2007). Taken

together, bifidobacteria have a unique potential to metabolize indigestible components, which

helps to prolong residence time in the competitive environment of the human gut.

1.2.3 Role of Bifidobacteria in the GIT

Each compartment of the GIT is a habitat for a specific bacterial community. Bifidobacteria can

be found all along the GIT and the main species present in humans are B. adolescentis, B.

bifidum, B. infantis, B. breve and B. longum (Crociani et al., 1996; Marteau et al., 2001; Nielsen

et al., 2003; Champagne et al., 2005). The preferred location of bifidobacteria inside the GIT

cannot be easily determined since numbers and species vary between individuals (Nielsen et al.,

2003). Furthermore, the majority of studies on human microbiota comes from analyzes of fecal

contents and are thus limited in reflecting the distribution of species in the different habitats of

the GIT (Tannock 1999; Marteau et al., 2001). Bifidobacteria can be autochthonous (found where

they grew) or allochthonous (found at places other than where they grew). The first colonizers of

the sterile GIT of newborns are mostly strains that reside autochthonously after colonization

(Vaughan et al., 2005).

In adults, species-diversity of microbiota increases and percentage of bifidobacteria decreases

with age depending on the health of the individuals (Hopkins and MacFarlane, 2002; Vaughan et

al., 2002; Hebtuerne et al., 2003; Woodmansey et al., 2004). This decrease was associated with

the fact that allochthonous bifidobacteria do not naturally occur in food bifidobacteria but

remains subject of discussions. However, bifidobacteria are added to food for their claimed health

benefits and their ingestion results in a temporary (elevated) appearance in the GIT with average


- 6 -

retention time believed to be less than a week (Kullen et al., 1996, Fukushima et al., 1998), while

other studies claim it to be more than one week (Alander et al., 2001, Ouwehand et al., 2004,

Bartosch et al., 2005, Su et al., 2005). Ouwehand and coworkers summarized in their review in

2004 that several factors to affect the residence time of allochtonous bacteria such as duration of

the consumption, amounts of the administered doses, and the methods of detection used.

1.2.4 Transcriptomics of Bifidobacterium longum NCC2705

The transcriptome is defined as the set of all messenger RNA (mRNA) molecules, also called

"transcripts," produced in a population of cells. Since it includes all mRNA transcripts in the cell,

the transcriptome reflects the genes that are being actively expressed. Therefore, this relatively

new molecular tool can give insights into global cellular mechanisms under different

environmental and/or growing conditions. However, transcriptome analyzes require the

availability of DNA based arrays, thereby complete genome sequence of the tested organism. The

strain B. longum NCC2705 was the first completely sequenced genome within the genus

Bifidobacterium (Schell et al., 2002) and a microarray covering its whole genome is available at

Nestlé Research Center. Transcriptomic analyzes of this strain have been used to characterize

metabolism and growth on various sugars (Parche et al., 2007) as well as cellular response to

environmental stresses. The exposure of B. longum NCC 2705 cells to oxygen and starvation

induced expression of genes that are known to participate to the common stress response in other

bacteria, whereas other up-regulated genes still remain unidentified (Klijn et al., 2005). In 2007,

Rezzonico and co-workers tested the gene expression patterns of B. longum NCC 2705 after

exposure to different heat periods and reported a 46 % of differentially expressed genes,

including common bacterial stress-responses such as the induction of chaperones and the down-


- 7 -

regulation of the translation, cell division and chromosome-partitioning machineries. In a recent

study, the effect of subinhibitory concentrations of bile on the expression levels of three different

bifidobacterial strains including B. longum NCC2705 were tested and lead to the identification of

a new multidrug resistance transporter in strain NCC2705 (BL0902) conferring increased

resistance to bile salts (Gueimonde et al., 2009).

These results emphasize the importance of transcriptome analyzes which enable new insights into

general cellular mechanisms and may help to design probiotic strains with enhanced functional

properties (Corcoran et al., 2008). Moreover, transcriptomics may also support the analysis of

complex physiological or morphological changes of bacteria as a result of changing

environments.

1.3 Probiotic concept

In 1907 the Russian scientist Elie Metchnikoff was the first to mention the beneficial effects of

bacteria in his book “The prolongation of life” and presented the concept of “replacement of

harmful microbes by useful microbes”. In 1906, Henry Tissier claimed that bifidobacteria are the

dominant species in the microflora of breast-fed infants and recommended the administration of

bifidobacteria to out-compete the putrefactive bacteria in infants suffering diarrhea. The term

“probiotic” (Greek: pro life) was introduced by Lilly and Stillwell (1965) and after several

redefinitions, the Joint Food and Agricultural Organization and World Health Organization

(FAO/WHO) working group defined the term “probiotic” in 2001 as “live microorganisms that

when administered in adequate amounts confer a health benefit on the host”. Probiotic bacteria

exert beneficial effects like inhibition of colonization of potential pathogens of the gut,

immunomodulation, alleviation of constipation, reduction of problems associated with


- 8 -

assimilation of lactose, and alleviation of antibiotic-associated diarrhea (reviewed by DeVrese

and Schrezenmeir, 2008; see “1.4 Health promoting properties related to probiotics”).

Growing scientific interest in probiotic bacteria is reflected in the increasing number of scientific

publications on the subject (Figure 1).

Figure 1 Number of publications obtained with the key word “probiotic” in the Pubmed

database as a function of time. Dashed bar is an extrapolated value based on number of

publication obtained after 6 months in 2009.

Nowadays, the prospering probiotic food market in Western Europe generates more than 1.4

billion Euros, one billion of which is exclusively from yogurt and desserts. Annual growth of

sales is forecasted to be 7 – 8 % over the next five years (Saxelin, 2008). Besides dairy products,

probiotics are also administered in fruit juices, berry soups, ice cream, soy- and cereal-based

fermented products as well as cheese. Probiotic foods are manufactured either by adding the

desired strains simultaneously with the starter culture in the fermentation tank or by running the

fermentation in a separate reaction and adding the probiotics afterwards in frozen liquid or

formulated state (Champagne et al., 2005). The increase of probiotic consumption raises a


- 9 -

demand by the food industry for more scientific research to develop new methods to improve

quality, safety, and functionality of probiotics.

1.4 Health promoting effects related to probiotics

Potential probiotic strains are screened for their health promoting qualities based on the following

established selection criteria (reviewed by DeVrese and Schrezenmeir, 2008)

• Origin from the intestinal tract of healthy persons, as such microorganisms are regarded

safe for humans and best adapted to the ecosystem of the gut

• Safe for humans, free of pathogenic and toxic effects

• Tolerance and resistance to gastrointestinal conditions and digestive enzymes enabling the

survival during the passage through stomach and upper intestinal tract eventually resulting

in health-promoting effects in the bowel

• Possess properties associated with a (positive) influence on the intestinal flora, like adhesion

to intestinal epithelial cells, survival and reproducing capacity in the human large intestine,

or production of antimicrobial substances

The most common strains used in probiotic food products are either lactobacilli or bifidobacteria.

Other strains exerting probiotic effects belong to the genus of Propionibacterium which is not

typical for human microbiota but strains of this genus were shown to produce vitamin B12, and

bacteriocins in addition to organic acetic and propionic acids from sugar metabolism.

Furthermore, yeasts from the genus of Saccharomyces have been show to exhibit antagonistic

effects against enteric pathogens (Klein et al., 1993) or are used as preventive and therapeutic

agents for the treatment of a variety of diarrheal disease.


- 10 -

The beneficial effects of strains of bifidobacteria have been attested in vitro in numerous studies.

However, probiotic properties have to be confirmed also in vivo before claiming them as

probiotic strains (Reid, 2008) and therefore a number of clinical studies on the therapeutic effects

of supplementation of bifidobacterial strains in mouse models and humans have been conducted

in the recent years as illustrated in Table 1.

Table 1 A selection of studies with beneficial effects on humans or mice of products containing bifidobacteria Strain(s) Benefit Hosta Reference B. longum B6 and ATCC 15708 Reduction of lactose intolerance Humans (15) Jiang et al., 1996

B. infantis Reduced risk of necrotizing enterocolitis (NEC) Rats, neonates Caplan et al., 1999

B. bifidum Protection against S. enteridis ssp. typhimurium Mouse, healthy and gnotobioic Silva et al.,1999

B. lactis HN019 Enhance natural immunity in healthy elderly subjects Human (25) Arunachalam et al., 2000

B. lactis Bb-12 Decrease of extent and severity of atopic eczema Human, infants (27) Isolauri et al.. 2000

Strain mix: 4 Lactobacillus; 3 Bifidobacterium and 1 Streptococcus

Prevent flare-ups of chronic pouchitis Human (40) Gionchetti et al., 2003

B. lactis Bb 12 Protective effect against acute diarrhea Humans, children (90) Choraqui et al., 2004

B. pseudocatenulatum DSM 20439, B. breve

Inhibit Shiga toxin-production of E. coli (STEC) 0157:H7

Mice Ashara et al., 2004

Strain mix: Lb. confusus; Lb. fermentum, Lb. plantarum; B. infantis PL9506

Suppression of Th2 cytokines during antigen sensitization

Mice Lee et al., 2004

Strain mix: Lb. gasseri PA 16/8; B. longum SP 07/3; B. bifidum MF 20/5

Reduced severity and duration of cold episodes Humans (479) de Vrese et al., 2005

Strain mix: B. infantis, S. thermophilus; B. bifidus

Reduction of incidences and severity of necrotizing enterocolitis (NEC)

Human, neonates (145) Bin-Nun et al., 2005

B infantis 35624 Alleviates symptoms of irritable bowel syndrome (IBS)

Humans with IBS (77) O'Mahony et al., 2005

B. lactis (Bb-12)

Fewer and shorter episodes of diarrhea Humans (200) Weizman et al., 2005

B. thermacidophilum RBL 71

Reduced severity of E. coli O157:H7 infection Mice Gagnon et al., 2006

B. lactis HN019

Enhanced aspects in cellular immunity in the elderly Humans, elderly (30) Ahmed et al., 2007

Strain mix: Lb.acidophilus NCFM, B. animalis ssp. lactis Bi-07

Reduced fever, rhinorrhea, and cough incidence and duration and antibiotic prescription incidence

Humans, children (326) Leyer et al., 2009

Strain mix: Lb. rhamnosus GG, B. lactis Bb-12

Reduced risk of early acute otitis media and recurrent respiratory infections during the first year of life

Humans, children (81) Rautava et al., 2009

a in brackets: number of persons tested


- 12 -

The particular molecular mechanisms underlying most of the probiotic effects of bifidobacteria

are still not completely understood (Leahy et al., 2005). Bifidobacterial strains exhibited

antimicrobial activities against other gut inhabitants which are attributed to production of

bacteriocins or organic acids that lower the colonic pH thereby inhibiting acid sensitive

pathogenic bacteria (Klijn et al., 2005; Cheikhyoussef et al., 2008). Bifidobacteria can modulate

the immune response by regulating the production of anti- and proinflammatory cytokines in

peripheral blood mononuclear cells of humans (Medina et al., 2007). Strains of bifidobacteria

were shown to produce linolenic acid in vitro (Coakley et al., 2009) or vitamins (Tamine et al.,

1995), both substances that can contribute a health benefit to the host. Genome analyzes revealed

the complete pathway for production of folic acid, thiamine, and nicotinate in B. longum

NCC2705 (Schell et al., 2002). Metabolic activity may moreover reduce serum cholesterol level

through bile salt hydrolase activity (De Smet et al., 1998).

1.5 Potential risks associated with probiotics

There are three theoretical concerns regarding the safety of probiotics (Snydman, 2008).

1. They may lead to disease, such as bacteremia or endocarditis.

2. Toxic or metabolic effects on the gastrointestinal tract

3. Transfer of antibiotic resistance in the gastrointestinal flora

In rare cases, Bifidobacterium species were shown to be involved in human dental caries,

pulmonary infections, bacteremia, abscesses, and blood stream infections which were not related

to consumption of probiotic products (Green, 1978; Gasser, 1994; Saarela et al., 2002; Modesto

et al., 2006). A tetracycline resistance gene (tet(W)) was reported on the chromosome of B. lactis

DSM 10140 (Kastner et al., 2006) which is problematic due to possible horizontal gene transfer


- 13 -

to pathogens in the gut. In patients with short small bowel syndrome, deconjugated bile acid

metabolites produced by bifidobacteria can accumulate leading to malabsorption of food nutrients

(Snydman, 2008). However, bifidobacteria and lactobacilli (with exception of L. rhamnosus)

used in probiotic food products are “generally accepted as safe” (GRAS) by the Food and Drug

Administration of the USA and classified as “class I” (absolutely safe) in Germany. Furthermore,

there is no evidence for a higher risk of bacteremia or fungemia due to the ingestion of probiotic

products in comparison with conventional fermented food products (Snydman, 2008; Salminen et

al., 2002).

1.6 Technological requirements of probiotic bacteri a

In order to provide a therapeutic effect to the consumer probiotic food products should contain a

critical concentration of viable bacteria per gram of product (Tamime et al., 1995). Studies with

non-viable cells or cell-components of probiotic strains also showed positive effects on health or

immunomodulatory effect (Salminen et al., 1999; Lammers et al., 2003). However, the average

recommended level of viable probiotic bacteria is suggested between 106 to 107 CFU per gram of

product at the time of consumption (Talkamar et al., 2004; Ishibashi and Shimamura, 1993).

Additionally, the probiotic food product should be regularly consumed in sufficient quantities to

deliver the relevant “dose” of live bacteria to the gut. Nevertheless, increasing knowledge of

strain survival during GIT transit and strain specific therapeutic effect will probably modify these

numbers in the near future (Champagne et al., 2005). The achievement of sufficient viable

probiotic cells at the time of consumption and at target site remains a challenge for the food

industry.


- 14 -

The preparation of a critical level of viable cells in probiotic food products and activity at the

target site depends on conditions during fermentation, downstream processing, storage and

during gastrointestinal transit (Figure 2). Exposure of microorganisms to different stresses

decreases bacterial viability and causes large fluctuations in viable cell counts of probiotic

products. In fact, numerous probiotic food products failed to meet the minimal level of viable

bacteria in probiotic food products (Masco et al., 2005, Huys et al., 2006). The selection criteria

of probiotic candidate strains are thus mainly focused on their technological properties, which

means that a lot of strains with promising health properties are probably missed (Lacroix and

Yildirim, 2007).

Figure 2 Main factors affecting the viability of probiotics from production to the

gastrointestinal tract (Lacroix and Yildirim, 2007).

Economic reasons raise the demand of food industries for new technologies that enable high

viable cell yield at large scale production. Major drawbacks in propagation and scale-up process

are sensitivity to oxygen and acidic conditions of numerous strains of intestinal origin.

Technologies which ensure probiotic activity and stability in food products are thus of highest


- 15 -

interest. Improved survival of strains during manufacturing may enable production of

technologically less favorable candidate strains but with more relevant probiotic properties,

consequently resulting in higher product efficacy. Therefore, the selection of adequate strains and

the improvement of technologies for production of probiotics is crucial in the development of

new probiotic food products.

1.7 Technologies to improve cell viability and yie lds

1.7.1 Production technologies

1.7.1.1 Optimization of growth medium

Probiotic strains are preferentially added to dairy products after fermentation because the fast

growth and acidification rate of typical starter cultures are disadvantageous for probiotic culture

development and maintenance (Champagne et al., 2005). Probiotics are delivered as frozen liquid

or as dry state cultures produced with freeze- or spray drying technology (see “Stabilization of

probiotics”). The cultivation conditions prior to freezing or drying have an impact on growth,

culture stability and activity as well as on drying and subsequent storage (Reilly and Gilliland

1999; Desmond et al., 2002; Carvalho et al., 2004). Modulation of the culture conditions can

therefore improve stability and activity of probiotics in food products.

Traditional large scale production of cells is performed in large batch fermentations. Approaches

to enhance yields in biomass and enhance cell stability have ranged from designer growth

medium to alternative fermentation technologies. In research laboratories, expensive media such

as MRS broth (de Man, Rogosa, and Sharpe, 1960) is commonly used for cultivation of LAB or

bifidobacteria. Biomass production from ultra-filtered skim milk with different protein

concentrations or supplementation of milk with nitrogenous substrates such as whey and casein


- 16 -

fractions from human or cow milk resulted in higher cell counts of bifidobacterial strains

compared to growth in skim milk only (Ventling and Mistry, 1993; Petschow and Talbott, 1990).

In contrast, bifidobacteria grown in soymilk or animal-product-free vegetable medium (based on

soy peptone, glucose and yeast extract (YE)) displayed lower yields or lower viability during

storage conditions due to the low buffer capacity of vegetable medium (Heenan et al., 2002,

Shimakava et al., 2003). Recently, a study showed a significantly increased yield of cell after 24

h growth of a B. animalis strains or a B. longum strain in modified vegetable medium compared

to standard MRS medium (Mättö et al., 2006). Growth medium for bifidobacteria is generally

supplemented with redox-reducing compounds such as cysteine to improve growth (Doleyres and

Lacroix, 2005). However, the growth promoting properties of such supplements lose the effect

when the disulfide bonds are reduced, for example by performic acid oxidation or reduction-

alkylation (Poch and Bezkorovainy, 1991; Ibrahim et al., 1994). Summarized, higher biomass

yields of bifidobacteria can be achieved by modulation of the medium components and opens

possibilities to develop low cost production methods.

1.7.1.2 Stress adaptation

The capacity of microorganisms to adapt to adverse environments is associated with the

expression of stress proteins and the development of cross resistances to numerous stresses (Ross

et al., 2005; Lindner et al., 2006). Besides medium composition, growth conditions like pH,

fermentation time or addition of protectants can induce cell responses that particularly affect

stability during downstream processing, storage, and transit through stomach and duodenum. The

enhanced tolerance to environmental stresses of cells from stationary growth phase compared to

exponentially growing cells is well established (Kolter, 1999) and reflects the bacterial ability for


- 17 -

rapid adaptation to changing conditions. For example, improved tolerances to acidic conditions in

L. acidophilus could be linked to an over-expression of a F1F0-ATPase, an enzyme which is

actively involved in proton extrusion out of the cytoplasm (Kullen and Klaenhammer, 1999).

Exposure of oxygen tolerant B. longum strains to oxygen caused an induction of an Osp-protein

expression as well as changes in the bacterial membrane compositions such as an increase of

short- and cyclopropane fatty acids in the cell membrane to protect the cell from damage (Ahn et

al., 2001).

Using a proteomic approach, several up-regulated stress associated enzymes have been identified

in B. longum cells upon exposure to bile salts (Sanchez et al., 2005; Savijoki et al., 2005) or high

temperatures (Savijoki et al., 2005). These included general stress-related chaperones, proteins

involved in transcription and translation and some proteins from different metabolic pathways. In

a recent study of global gene expression, stress response upon heat treatment of B. longum

NCC2705 showed that 46 % of genes exhibited altered gene-expression. These included the

classical heat shock stimulon including chaperones DnaK, DnaJ or GroEL/ES which assure

correct folding of proteins (Sugimoto et al., 2008) and enzymes involved in repression of cell

division process or proteases to digest heat-induced misfolded proteins (Rezzonico et al., 2007).

Proteomic analyzes of B. longum and B. adolescentis upon bile treatment showed higher levels of

chaperones and enzymes involved in fatty acid synthesis, resulting in changes in membrane fatty

acid composition (Sanchez et al., 2007). These changes may favor tolerance to bile salts by

protecting cells from passive diffusion of bile salt into cytoplasm (Begley et al., 2005, Ruiz et al.,

2007). A model of a number of adaptation mechanisms to bile salt stress of bifidobacteria was

recently developed by Sanchez et al. (2008) based on data acquired from proteomic analyzes

(Figure 3).


- 18 -

The ability of microbes to adapt to stress conditions resulting in improved survival has been

widely reported and exploited in yogurt starter culture Streptococcus thermophilus or probiotic

cultures (Schmidt and Zink 2000; Wouters et al., 1999; Saarela et al., 2004). Bifidobacteria

treated with various stress conditions such as low pH, heat or combined stresses resulted in

increased stress responses to homologous stresses or cross-protection (Saarela et al., 2004).

Increasing sublethal treatment with NaCl, bile salts or heat exposure enhanced survival of several

bifidobacterial strains to lethal heat stress or freeze-thawing (Schmidt and Zink, 2000). Starvation

of B. longum for 30 or 60 min or exposure of B. lactis to pH 5.2 under starvation conditions

increased survival to prolonged cold storage in growth medium and lethal acidic conditions,

respectively (Maus and Ingham, 2003). However, cellular responses to pre-conditioning and the

subsequent tolerance to lethal stress depend on the sublethal stress (duration, severity,

combination) and the particular strain used (Simpson et al., 2005). The molecular mechanisms

involved in stress resistance have still to be clarified in detail with the help of global stress

response analyzes such as proteomics and transcriptomics (Sanchez et al., 2008). Consequently,

optimization of sublethal treatment has to be assessed strain by strain which is time consuming

and tedious. Recently, a system has been developed in our laboratory using continuous culture in

combination with a two-stage fermentation system which allows the efficient screening of a

number of sublethal stresses under controlled conditions (Mozzetti, 2009). Nevertheless, bacterial

stress response is complex and the number of molecular mechanisms rendering stress resistance

remains to be explored. An alternative to sublethal stress adaptation of probiotics is the use of

immobilized cell technology. In previous studies it was shown that immobilized cells during

continuous culture can be used to produce probiotic strains with improved functional and

technological properties (Lacroix and Yildirim, 2007).


- 19 -

Figure 3 A model of main physiological mechanisms involved in bile salt tolerance of

bifidobacteria. Conjugated bile acids/salts diffuse to the bifidobacterial cytoplasm (1), and are

cleaved by the bile salt hydrolase (BSH; bile adaptation and acid adaptation protein) and

rendering one amino acid (glycine or taurine) and one deconjugated bile acid moiety (2).

Unconjugated bile salts can also enter the cytoplasm by passive diffusion (3), being deprotonated

at the slightly acid pH of the cytoplasm (4). Ionized bile acids are non-permeable, and must be

excreted by the action of certain transporters, such as the cholate transporter “Ctr” of B. longum

(5) (Price et al., 2006). In addition, the process of tolerance to bile salts is associated with an

increase in the synthesis of molecular chaperones (6) and a decrease in the synthesis of long-

chain-fatty-acid CoA ligase (bile response and adaptation protein) together with a shift in the

fatty acid composition (7). Bile acid deprotonation cause cytoplasm acidification, which is

counteracted with mechanisms such as ammonia production from glutamine deamination (8) or

proton pumping by the F1F0-ATPase (acid response and bile adaptation protein) (9). The amounts

of ATP needed for feeding these mechanisms are provided through the bifid shunt (10) (adapted

from Sanchez et al., 2008).

Abbreviations: AA: amino acid; BCAA: branched chain amino acid; Ctr: cholate transporter


- 20 -

1.7.1.3 Immobilized cell (IC) technology in continuous cultures

Fermentation technologies, such as continuous culture and immobilized cell systems, are not

widely established in food industries, yet they have potential for enhancing the performance of

fastidious probiotic organisms. These technologies might be employed to develop strains with

improved physiology and functionality when introduced into the gut and to enlarge the range of

commercially available probiotics, as well as expanding product applications (Lacroix and

Yildirim, 2007). In continuous cultures, bacteria grow in a bioreactor with continuous feed of

fresh medium and removal of fermented broth at a given dilution rate. Steady conditions in

continuous cultures allow analyzes of metabolism, growth rate and gene expression of bacteria

under constant conditions during long time period (Hoskisson and Hobbs, 2005). Continuous

culture of B. longum SH2 exhibited increased volumetric biomass productivity compared to

traditional batch cultures (Kim et al., 2003). Recently, a B. longum NCC2705 culture grown in

prolonged chemostat culture for 200 h showed metabolic and transcriptomic stability throughout

entire culture time enabling to apply these cells in a new method to screen for sublethal

treatments (Mozzetti, 2009). Nevertheless, a substantial drawback of continuous cultures is the

increased susceptibility to contamination. The permanent feeding of fresh medium must be

maintained at dilution rates lower than the maximum specific growth rate to prevent wash out of

active biomass. Low dilution rate on the other hand increases probability of a competitive

contaminant strain to establish in the bioreactor.

1.7.1.3.1 Principle of cell immobilization

The principle of immobilized cell fermentation is based on the retention of microorganisms in a

discrete location of the fermentation to yield high biomass and/or to protect bacteria from an


- 21 -

antagonistic environment. Several immobilization techniques have been applied in dairy

fermentations to produce starter or probiotic cultures and metabolites, including attachment or

adsorption to a preformed carrier, membrane entrapment, microencapsulation and physical

entrapment in polymeric beads (Lacroix et al., 2005).

1.7.1.3.2 Techniques of cell immobilization

A promising technology using immobilized cells to produce probiotic cultures is the entrapment

of microorganisms in a food-grade porous polymeric matrix (Lacroix et al., 2005; Lacroix and

Yildirim, 2007). Cell entrapment is assessed by inoculating liquid polymeric matrix with fresh

cell culture followed by droplet formation of the polymer-cell mixture using extrusion or

emulsification. Spherical polymer gel beads of diameters between 0.3 – 3 mm are formed either

by thermal gelation (κ-carrageenan, locust bean gum (LSB), gellan, agarose, gelatin) or

ionotropic hardening (alginate, chitosan). After gelation, polymeric beads with the entrapped

viable cells are incubated in growth medium to colonize the beads (Lacroix et al., 2005).

When immobilized cells are incubated in growth medium, diffusion limitations occurring in gel

beads for both substrates and inhibitory products, mainly lactic and acetic acid in the case of

LAB, confer a more favorable environment for cell growth close to the bead surface than at the

bead center (Arnaud et al., 1992; Masson et al., 1994; Lamboley et al., 1997; Cachon et al.,

1998; Doleyres et al., 2002a). The very high density of cells at the bead surface can be compared

to bacterial biofilm structures (Lamboley et al., 1997). Colonized polymeric gel-beads incubated

in stirred reactors release large numbers of cells from their peripheral layers to the growth

medium upon pressure due to cell expansion, collision and shearing forces (Sodini et al., 1997;

Lamboley et al., 1999).


- 22 -

1.7.1.3.3 Physiological changes of immobilized cells

Several studies have shown that immobilized and released bacteria exhibit changes in growth,

morphology and physiology compared with cells produced in conventional free-cell cultures for

LAB and probiotics (Lacroix and Yildirim, 2007). For instance, immobilized Lb. plantarum

showed changes in cell morphology from rod to coccoid shape along with a shift from homo- to

heterofermentative metabolism (Krishnan et al., 2001). Saccharomyces cerevisiae and

Acetobacter aceti exhibited enhanced tolerance to ethanol and acetic acid upon cell

immobilization (Krisch and Szajani, 1997). Similar effects were reported for immobilized

Escherichia coli to phenol and antibiotics (Heipieper, et al., 1991; Jouenne et al., 1994) and

Lactococcus and Leuconostoc strains to quaternary ammonium sanitizers (Trauth et al., 2001).

These alterations in cell physiology or metabolism may have different origins. The biofilm-like

growth on bead surface induces changes in cell response. Natural biofilms comprise of a mixture

of bacteria, proteins, nucleotides and exopolysaccharides which offers protection from

restrictional environment in the host (Stanley and Lazazzera, 2004). Biofilm-embedded cells

exhibit resistance to environmental stresses like high temperatures, low pH, osmolarity and

biocides including antibiotics due to metabolic slowdown and physical protection by the cell

community (Costerton et al., 1999; Fux et al., 2005). When biofilms reach a certain size,

complex structures are formed composed of pillars of water channels to allow nutrient influx and

waste efflux (Davey and O’Toole, 2000). Such a complex community requires some kind of

inter-bacterial communication known as quorum sensing. Bacteria accumulate specific

compounds (peptides, bacteriocins) to a certain threshold, thereby controlling expression of

various genes involved in regulation and adaptation to the exterior environment within the

biofilm (Stanley and Lazazzera, 2004). Although gel beads are artificially produced, the high cell

density on the bead surface are comparable to biofilm environment and could induce quorum-


- 23 -

sensing responses, leading to improvement in physiological and technological characteristics that

are important for industrial applications of probiotics. The adaptation process might provide a

competitive advantage for the population, more effective adherence properties and rapid

responses to changing environmental conditions when introduced to a host (Sturme et al., 2002).

Alternatively, changes in cell morphology and physiology can be triggered by non-specific stress

adaptation during maturation of cells caused by the steep gradients of pH, nutrients and inhibitors

(organic acids) dominating the interior of the colonized gel beads (Doleyres et al., 2002b).

1.7.1.3.4 Advantages of continuous IC cultures

Continuous IC systems are usually processed at high dilution rates because physical retention of

cells prevents wash-out of active biomass and provide efficient biomass production. High

dilution rate ensures high substrate supply for cell growth and metabolism and rapid removal of

metabolites to prevent product inhibition, and diminishes risk of contamination. Continuous

culture with IC might also favor selection of stress tolerant subpopulations during prolonged

fermentation time (Doleyres et al., 2004a). Other advantages over free cell fermentations are the

reuse of biocatalysts, less sensitivity to contamination and bacteriophage attack, enhanced

plasmid stability, and physical and chemical protection (Doleyres and Lacroix, 2005).

Immobilized starter or probiotic bacteria have been produced using repeated batch or continuous

culture and exhibited various benefits compared to traditional batch free cell cultures (Table 2).

The complex cell response upon growth in polymeric gel beads remains to be elucidated. Further

knowledge about the mechanism responsible for the stress adaptation on a physiological and

global cell response level (transcriptomics, proteomics, metabolomics) may lead to the

production of probiotic strains with enhanced technological and biological properties during cell

production, gut transition and colonization in the human GIT (Lacroix and Yildirim, 2007).

Table 2 Fermentation systems with immobilized LAB or bifidobacteria and the corresponding bacterial phenotype Fermentation system/ immobilization matrix/ growth medium

Species Benefit of IC Reference

1-stage continuous/ polyacrylamid lattice/ Milk

Lb. casei

Investigate effect of IC on cell metabolism - Stable production of metabolites - Bead stability

Diviès and Siess, 1976

2-stage continuous/ fed batch/ alginate/ milk or cream

Lb. bulgaricus St. thermophilus St. diacetylactis; St. lactis; St. cremoris Lc. Lactis; Lc. diacetylactis

Yogurt/Cheese production and cream fermentation - Constant yogurt production, small volume - Cheese with constant properties, small volume - Cream with improved storage properties

Prévost et al., 1985; Prévost and Diviès, 1987 and 1992

1-stage continuous/ κ-carrageenan, locust soy bean (LSB)/ reconstituted skim milk, yeast extract (YE)

B. infantis Testing applicability of IC in milk fermentation - high bead stability - high biomass/cell volumetric productivity

Ouellette et al., 1994

2-stage continuous/ κ-carrageenan, LSB/ whey permeate, YE

Lb. helveticus Kinetic study of lactic acid production with IC - high lactic acid production - high biomass/cell volumetric productivity

Norton et al., 1994

1-stage continuous/ κ-carrageenan, LSB/ milk

Lc. lactis, Lc. diacetylactis, Lc. lactis; Leuconostoc mesenteroides

Continuous fermentation for fresh cheese production - feasible system for constant milk pre-fermentation in fresh cheese manufacture

Sodini et al., 1997

1-stage continuous/ alginate/ MRS (no acetate, lactose)

Lc. diacetylactis Comparison of chemostat IC vs free cell batch - high cell productivity - change of redox state

Cachon et al., 1998


2 strains Lc. lactis and 1 Lc. diacetylactis

Production of mixed starter culture - long-term biological stability for mixed starter culture

Lamboley et al., 1999

Repeated batches/ chitosan treated polypropylene/ medium of glucose, YE minerals

Lb. plantarum Physiological analysis of IC - shift from homo- to heterofermentative metabolism - changes in cell morphology

Krishnan et al., 2001

Repeated batches/ κ-carrageenan, LSB/ MRSC + CaCl2

B. longum; Lc diacetylactis Study effect of IC on co-cultivation - successful co-cultivation of mixed culture containing a non-competitive B. longum strain

Doleyres et al., 2002a

Table continued

1-stage continuous/ gellan/ MRS and cysteine (MRSC), whey permeate (WP)

B. longum Comparison of chemostat IC vs free cell batch growth - higher volumetric productivity

Doleyres et al., 2002b

2-stage continuous κ-carrageenan, LSB/ MRSC, lactose, CaCl2

B. longum; Lc. diacetylactis Effect of co-culture in chemostat IC on cell physiology - increased stress tolerance to freeze drying, H2O2 and gastrointestinal conditions

Doleyres et al., 2004a

2-stage continuous/ silicone rubber solid carriers/ whey permeate, YE

Lb. rhamnosus Biomass production chemostat IC vs free cells - induction of cell aggregation

Bergmaier et al., 2005


Lb. helveticus Modelling lactic acid production with chemostat IC - high lactic acid production - physiological adaptation

Schepers et al., 2006

1-stage continuous/ κ-carrageenan, LSB/ milk

Lb. rhamnosus; Lc. diacetylactis; P. acidilactici; Lc. cremoris

Test effect of mixed cultures with antagonistic properties in chemostat IC - enhanced acidification and tolerance to nisin

Grattepanche et al., 2007

2-stage continuous/ gellan, xanthan/ MRS

Lb. delbrüeckii Study the effect of chemostat IC and culture time - changes in cell morphology/aggregation - enhanced survival to freeze drying

Koch, 2006

1-stage continuous/ gellan, xanthan/ MRS + H2O2

B. longum System to apply selective pressure to continuous IC - selection of H2O-resistant mutant strain(s)

Mozzetti, 2009


- 26 -

1.7.2 Stabilization of probiotics

1.7.2.1 Microencapsulation

Microencapsulation is a technology for packing solids, liquids or gaseous material in small sealed

capsules, which can release their contents at controlled rates under specific conditions (Shahidi

and Han, 1993). The principle of microencapsulation has initiated numerous efforts in the last

two decades to embed sensitive metabolically active probiotic strains in microcapsules or

microparticles in order to increase stability during shelf-life and consumption. Encapsulation was

shown to protect probiotic bacteria from oxygen, acidic environments in food products,

refrigerated storage, and simulated gastro-intestinal conditions (Doleyres and Lacroix, 2005;

Anal and Singh, 2007).

Different techniques and materials have been used for microencapsulation of probiotic cultures.

Cells encapsulated in gel beads are achieved by entrapping probiotic cells within a polymeric

matrix such as pectin, gellan gum, κ-carrageenan-locust bean gum and alginate. Gel capsules

containing probiotics can be produced by various methodologies such as emulsion or extrusion

techniques (Champagne and Fustier, 2007; Talwalkar and Kailasapathy, 2003). Several studies

reported protective effects of microencapsulation and high survival of sensitive bifidobacteria and

lactobacilli in plant polymer matrices (Krasaekoopt et al., 2003), particularly against oxygen

(Talwalkar and Kailasapathy, 2003), acidic environment (Sun and Griffiths, 2000), freezing

(Shah and Ravula, 2000), refrigerated storage of ice cream and milk (Hansen et al., 2002) and

yoghurt (Adhikari et al., 2003; Adhikari et al., 2000; Sultana et al., 2000) and during simulated

gastro-intestinal tests (Lee and Heo 2000; Ding and Shah, 2007; 2009). Although promising on a

laboratory scale, the technologies developed to produce gel beads present serious difficulties for

large-scale production such as low production capacity and large bead diameters (2-5 mm), that


- 27 -

can affect food texture (Hansen et al., 2002). Moreover, the addition of these polysaccharides is

not permitted in yogurts or fermented milks in some European countries (Picot and Lacroix,

2004). Beside the limitations in up-scaling and interference with textural and sensorial properties

of food products (Anal and Singh, 2007), the proper release of viable cells at the target site is also

an obstacle in the probiotic approach. In their recent review, de Vrese and Schrezenmeir (2008)

reported that alginate capsules containing probiotics did not release their content after

administration to human and pigs and considered this type of encapsulation although frequently

used, as unsuitable as carrier for probiotic bacteria.

Another technique to encapsulate probiotics is the spray-coating technology for large scale

production, which is widespread in industries for diverse commercial products (Anal and Singh,

2007; Champagne and Fustier, 2007). But since the technology is difficult to master and most

information on the technology is proprietary to industry, there is only limited access for academic

research (Champagne and Fustier, 2007). Spray coating consists of suspending, or fluidizing,

particles of a core material in an upward stream of air and applying an atomized coating material

(lipid-based, proteins or carbohydrates) to the fluidized particles (Augustin and Sanguansri,

2008). One particular patent claims improved stability of spray-coated cells to simulating gastric

conditions, high temperature and during storage in powdered milk compared to uncoated freeze

dried cells. In addition the coated cells had better resistance to compression (US patent

2003/0109025).

1.7.2.2 Spray drying

Dairy starter and probiotic cultures are mainly preserved and distributed in frozen or dried form.

Frozen cultures occupy large volume and require storage at sub-zero temperatures causing high


- 28 -

costs of storage, shipping and energy and hence preservation in dried form is sometimes preferred

(Johnson and Etzel, 1995). Transformation of bacteria to a dried form through the physical

removal of water is a critical step since bacteria require a water activity (aw) of about 0.98 in the

product matrix for survival and growth. Preservation of a viable state requires either high aw for

metabolic activity or low aw to survive in dormant state in powders (Paul et al., 1993). Drying of

bacterial suspension is mainly assessed by freeze- or spray drying.

Freeze drying is widely used for formulation of starter and probiotic cultures but it is an

expensive process with low yields, and as such spray drying offers an alternative inexpensive

approach yielding higher production rates (Zamora et al., 2006). Spray drying is energy-efficient

and an established tool in food industries for the production of milk powders and instant coffee.

Despite restrictive process conditions for microorganisms (inlet reaching ≥ 180 °C), the rapidity

of drying combined with the possibility to dry large amounts of bacterial cultures has caught the

attention of research and industry in recent years (Zamora et al., 2006; Chàvez and Ledeboer,

2007; Meng et al., 2008).

1.7.2.2.1 Principle of spray drying

During spray drying, feed solution is transformed from a fluid state into a dried form by spraying

the feed into a hot drying medium (Figure 4). The spray drying process can be divided into two

phases. First, the constant drying rate period just after the feed solution is dispersed by a fluid

nozzle or atomizer and encounters the hot air in the drying chamber. The interior of the droplets

is not challenged to heat exposure due to the cooling effect of evaporation. During the subsequent

falling drying period, the temperature of the particles increases but does not reach values of the

inlet temperature, because the remaining moisture content compensates exposure to heat until the


- 29 -

entire water is evaporated. Hence, the residence time of the droplets in the drying chamber drives

the moisture content of the particles, as well as the survival of the microorganisms within the

particles (Santivarangkna et al., 2007). It is generally accepted that bacteria in the dried particle

are exposed to no higher temperatures than the outlet temperature, which in turn constitutes the

crucial factor for bacterial survival to spray drying (Gardiner et al., 2000; Lian et al., 2002). The

construction of the spray cylinder is designed to allow adequate residence time and droplet

trajectory distance for achieving the heat and mass transfer. After the drying of the spray in the

chamber, the majority of the dried product falls to the bottom of the chamber. The dried product

can either be discharged continuously from the bottom of the spray cylinder or passed into a

solid-gas separator (cyclone) where the solids are recovered from the gas stream. The powder is

then collected at the bottom of the glass container. Very fine particles which do not precipitate in

the glass container are usually collected in an outlet filter connected in series to the cyclone. The

operating conditions and dryer design are selected according to the drying characteristics of the

product and powder specification. The main parameters for powders are the residual moisture

content and the particle size. Moisture content depends mainly on evaporation rate and the dryer

∆ T (inlet air temperature minus the outlet air temperature), which in turn dictates the amount of

drying air needed and ultimately the sizing and cost of almost all of the system components. The

particle size requirement affects the choice of atomization method and can also affect the size of

the dryer.

The system allows many heat sensitive products (enzymes, microorganisms, volatile aroma

compounds, etc.) to be spray dried and transformed to powders with low moisture content.


- 30 -

Figure 4 Pathways of air and feed during spray drying

(Left) Path of the drying air in the spray dryer Büchi B290. Cold air is aspirated through the air

inlet tunnel (1) and then electrically heated (2) prior to the atomization of the feed solution by the

two-fluid nozzle (3). Dispersed feed enters the spray cylinder (4), in which drying of the droplets

into solid particles takes place. Dried powder is separated from fine particles in the cyclone (5)

and collected in the glass container (6). The outlet filter (7) is placed in series with the cyclone to

remove fine particles and to prevent them from entering aspirator (8) which generates the air

flow. Temperatures are measured in the entrance of spray cylinder (9, inlet temperature) and in

the intermediate piece between spray cylinder and cyclone (10, outlet temperature).

(Right) Path of the feed solution and pressurized air in the spray dryer in co-current mode. Feed

solution (A) conveyed by peristaltic feed pump (B) and atomizing air (D, inlet) are passed

separately to the nozzle head (C), where the atomization of the feed solution into fine droplets

takes place. The co-current two-fluid nozzle is located at the centre of the upper part of the spray

cylinder. Atomization is created by compressed hot air at a pressure of 0.5 to 2 bar. Powders

produced with this adjustment have particle size ranged from 5 to 15 µm on average.


- 31 -

1.7.2.2.2 Parameters affecting cell viability during spray drying

The drying process causes cell damage and decreased viability due to mechanical, heat, osmotic

stress, exposure to oxygen, and the removal of bound water (Meng et al., 2008; To and Etzel,

1997). The decrease of water availability inside or in the vicinity of the dried cells affects the

physiology of bacteria. Removal of water eventually results in slow down of bacterial

metabolism or stops it entirely after which bacteria enter a dormant state. Possible damages

caused by the dehydration are:

• increased permeability of cell wall/ cytoplasmic membrane to extracellular or leakage of

cytoplasmic compounds due to lesions in cell envelope (Texeira et al., 1996).

• transformation of liquid crystalline structure of phospholipid bilayer to gel phase and

incomplete reverse rehydration may cause leaking and disturbance of molecule transport

systems (Oliver et al., 1997).

• peroxidation of lipid cause membrane leakage causing loss of plasmids or increased influx of

DNase into cells affecting DNA integrity (Meng et al., 2008).

• conformational changes of enzymes may block or modify active binding sites resulting in

inactivation of the enzyme (Prestrelski et al., 1993; Burin et al., 2004).

• changes in the lipid bilayer of cell membrane causing displacement, denaturation or loss of

membrane-bound proteins (Texeira et al., 1996).

Poor survival of LAB and bifidobacteria during drying are substantial draw-backs for this

preservation method and strategies to optimize cell viability are thus required (Meng et al., 2008;

Santivarangkna et al., 2007). Survival of heat sensitive LAB and bifidobacteria during spray

drying depends on several factors such as the drying carrier, bacterial growth phase, the particular


- 32 -

probiotic strain used and mainly on the outlet temperature of the drier. Adjustment and

optimization of these parameters affects bacterial survival to spray drying as shown in Table 3.

Stabilizing compounds for drying and shelf live are hydrophilic polyhydroxy substrates like

sugars which partly substitute missing water molecules in membranes and proteins (Leslie et al.,

1995, Oliver et al., 1997, Corcoran et al., 2004). Reconstituted skim milk (RSM) as a carrier

contains lactose as a possible water substituent and was shown to be suitable for drying of

probiotics (Corcoran et al., 2004; Ananta et al., 2005). Besides lactose, skim milk proteins in

parallel with calcium may prevent cellular injury by stabilizing phospholipid bilayers and

shielding the membrane-bound proteins by formation of a protective coating (King and Su, 1993;

Castro et al., 1995).

In recent years, numerous studies on the optimization of spray drying of probiotics have been

performed to either increase cell robustness or to modify the drying parameters to enhance the

yields of viable cells in spray dried powders (Table 3).

Table 3 Survival rates of lactobacilli and bifidobacteria during spray drying

Microorganism Survival (%) a T outlet (Tout)

T inlet (T in) Feed solution Reference Special remarks

Lb. acidophilus Tout 75, 80, 85 °C RSM (25%) 8.2, 2.4, 0.5 RSM (40%) 3.7, 0.5, 0.3

75, 80, 85 170 25 % solid milk 40 % solid milk

Espina and Packard, 1979

Moisture content of powder (M.C.) 3.5 - 6.1%

S. salivarius ssp. thermophilus, Lb. delbrueckii ssp. bulgaricus

2 - 0.1 60, 70, 80

140 – 180

yoghurt, 13,63% solid non fat

Kim and Bhownik, 1990

Lb. helveticus 15 82, 120 220 19% maltodextrin Johnson and Etzel, 1995

Tout adjustment using feed flow rate

Lb. bulgaricus n. i. 62-105 200 skim milk Texeira et al., 1995 Lb. bulgaricus n. i. 80 200 40% maltodextrin in

water or RSM 11% Texeira et al., 1995

Lb. bulgaricus

n. i. 70 200 RSM Texeira et al., 1996

Lc. lactis, Lb. casei, S. thermophilus

3.0 – 0.4 14.7 – 13.0 34 – 5.0

70 – 90 220 25% (w/w) solid maltodextrin

To and Etzel, 1997

Lb. lactis 14.5 – 0.6 68 160 – 200

20% RSM Mauriello et al., 1999

Flow rate feed: 10, 13, 17 ml/min

S. thermophilus, Lb. delbrueckii

54.7 – 51.6 15.8 – 13.7

60 – 80 180 yoghurt Bielecka and Majkowska, 2000

Water evaporation capacity 1645 kg/h

Lb. paracasei Lb. salivarius

65.0 1.0

60-120 170 20% RSM Gardiner et al., 2000

M.C. ca. 4%

Lb. paracasei Heat: Control 4 Adapted 23 Salt: Control 7 Adapted 33

85 – 105 170 20% RSM Desmond et al., 2001

M.C. 1.7 - 3.3 % Tout adjusted with feed flow rate

Bifidobacteria 16.6 45 100 10% (encapsulation) O’Riordan et al., 2001

B. infantis (2 strains), B. longum (3 strains)

82.6 – 0.1 50-60 100 Gelatin, gum arabic, skim milk, soluble starch, 10% RSM

Lian et al., 2002 M.C. 7-9% Carrier dependent survivalc: RSM> Gelatin> gum> starch

Table continued

Lb. salivarius, Lb. sakei

100 41.1

70 200 11% milk Silva et al., 2002

Lb. paracasei NFBC338 Tout 100, 105 °C 20% RSM: 1.7, 1.4 10% RSM+gum: 0.9,< 0.01

105 100

170°C 20% RSM or 10% RSM with 10% gum accacia

Desmond et al., 2002

M.C. 2.5 - 3.2% Tout adjusted with feed flow rate

Lb. acidophilus B. lactis Bb12

Tin 130 °C, Tout 75 °C Lb. acido. 1.9 B. lactis 72.2

75 – 125

130 – 190

Cellulose acetate phthalate, Glycerol, Maltodextrin, Raftilose, milk

Favaro-Trindale and Grosso, 2002

Avg. particle size 22 µm

Lb. rhamnosus GG, Lb. rhamnosus E800, Lb. salivarius

RSM, RSM + polydextrose 50, 30b 25, 41b 0.7, 80.1b

85-90 170 RSM 20% or 10% RSM with 10% polydextrose

Corcoran et al., 2004

M.C. <4%, Tout adjustment using feed flow rate

B. brevis B. longum

B. brevis R070 25.7 B. longum R023 1.4

80 160 10 % Whey protein Picot and Lacroix, 2004

M.C.: B. brevis 2.0% B. longum 2.1%

Lb. delbrueckii ssp. bulgaricus

n.i. 70 200 11% RSM Silva et al., 2004 Sucrose in growth medium: anti-oxidant for powder storage

St. thermophilus Lb. acidophilus B. infantis, B. longum

Tout 75 °C B. infantis mixed with - Lb.acido 2.9 - St. thermo. 3.4 B. longum mixed with - Lb.acido. 11.1 - St. thermo. 13.1

60 – 90 100 Fermented soy milk Wang et al., 2004 M.C. 4.2 – 10.6 %; Study about influence of mixed cultures on survival to spray drying

Lb. rhamnosus GG RSM 20% ca. 65 b RSM/ ca. 55 b Polydextrose RSM/ ca. 65 b Raftilose

70 – 100 varying RSM 20% modified with oligofructose and polydextrose

Ananta et al., 2005 M.C. 2.7 – 4.5 % Tout adjusted by changing Tin,

Table continued

Bifidobacterium subsp. 12 – 102 85 – 90 170 RSM 20% (w/v) or RSM (10%, w/v) and gum acacia (10%, w/v)

Simpson et al., 2005

M.C. 2.7 – 4.2 %

Lb. salivarius BetL+ 1.4 BetL- 0.3

80 – 85 170 RSM Sheehan et al., 2006

“BetL” plasmid containing listerial betaine uptake system BetLd

Lb. acidophilus Tin 120 °C, Tout 79 °C

1.5

58 – 79 100 – 120

RSM, whey permeate with guarana gum

Riveros et al., 2009 M.C. 6.3 %

Lb. rhamnosus GG; Lb. rhamnosus E-97800

MSG., RSM 69, 75 23, 55

60 – 75 varying Trehalose (20%, w/v) supplemented with 12.5 g/l monosodium glutamate (MSG)

Sunny-Roberts and Knorr, 2009

M.C. 3.6 – 4.4 %

a Different survival rates correspond to different spray drying parameters (e.g. strain, carrier, Tout etc.) b Pproximate values estimated from data in figures. c Series of descending survival for each carrier for the majority of tested strains d Genetically modified organism n.i.: not indicated


- 36 -

1.7.2.2.3 Parameters affecting storage of dried powders

During powder storage, microbial stability is affected by parameters like temperature, oxygen

content, light exposure and storage materials. Moreover, powder characteristics such as moisture

content, powder composition and relative humidity can influence the glassy state of the carrier

medium and eventually the viability of microorganisms during shelf life. Glass transition

temperature (Tg) is the critical temperature at which a glass-forming liquid transforms from a

rubbery to a glassy state. The specific Tg of sugar containing powders plays an important role in

the stability of powders containing microorganisms and is dependent on storage temperature and

moisture content of the powder (Figure 5). High viscosity of sugar glasses below or around their

glass transition temperature retards molecular mobility and reactivity resulting in a stabilizing

effect of biological systems (Buitink et al., 2000). Consequently, the storage of dried cultures at

temperatures below the glass transition temperatures of sugars in powder mixtures enhance shelf-

life stability (Vega and Roos, 2006; Santivarangkna et al., 2008). High storage temperature and

remaining water in dried powders have thus a detrimental effect on cell viability, as shown for

dried lactobacilli and bifidobacteria in skim milk (Gardiner et al., 2000; Silva et al., 2002; Zayed

and Roos, 2004; Simpson et al., 2005).

Cell damage during storage may be caused by oxidation of membrane lipids. The presence of

allyl-groups in unsaturated fatty acids causes instable membrane properties during storage since

double-bonds are targets to oxidation and lead to formation of cytotoxic hydroperoxides (Texeira

et al., 1996). Progressive changes of lipid composition were observed during storage of

lactobacilli (Castro et al., 1996) and cells with damaged bacterial membranes became

consequently more sensitive to NaCl (Texeira et al., 1996). The detrimental effects on cell

viability during storage require treatments to stabilize and improve stability of spray dried

probiotic powders to make spray drying a cost efficient alternative to traditional freeze drying.


- 37 -

Figure 5 A simplified state diagram shows the glass transition curve which relates the glass

transition temperature and moisture content. Molecular mobility and deleterious reaction rates in

the glassy state are extremely low, while they increase at the storage conditions above the curve

(rubbery state) (Santivarangkna et al., 2008).

Studies on optimization of storage stability of bacteria in dried powders focused on

supplementation or substitution of the carrier medium (usually RSM) by oligosaccharides or

oxygen scavenger compounds. Although sugars are believed to be the major components

affecting glass formation, it has also been shown that polypeptides can significantly alter the

glass properties of sugars (Buitink et al., 2000) and therefore play the important role in glass

formation. This could explain why skimmed milk powder containing a large number of milk

proteins is considered as an efficient desiccation protectant (Hubalek, 2003). Powder stability of

lactobacilli dried in RSM supplemented with prebiotic substances or polydextran did not improve


- 38 -

bacterial viability during storage compared to RSM alone (Corcoran et al., 2004; Ananta et al.,

2005). Addition of anti-oxidants such as ascorbic acid and monosodium glutamate to powders of

Lb. delbrueckii ssp. bulgaricus stored at different temperatures showed ambiguous results and

even destabilizing effect at 20 °C (Texeira et al., 1995) concluding that protective agents during

the dehydration process do not necessarily support viability of cells during storage (Crowe et al.,

1987). Stability of probiotic powders can be improved by proper packaging. The use of

aluminium-based packaging material that protects the powder from exposure to light, moisture

and oxygen supports storage stability of dried probiotics (Nagawa et al., 1988; Ishibashi and

Shimamura, 1993). Shelf-life stability of probiotic bacteria is strain dependent and not related to

the survival during drying (Gardiner et al., 2000; Simpson et al., 2005; Sunny-Roberts and

Knorr, 2009). Hence, a careful selection of robust strains should be performed with a view to

obtained strains which have stable shelf life properties.

Variation of stability during drying and storage among strains of the same species can impede the

ability to extrapolate results across entire subspecies or species. Predictions are thus of limited

significance. Consequently, studies on the effect on survival of changing parameters in drying

and storage applications have to be performed strain by strain. This is laborious and time

consuming, particularly during storage tests, and hence new technologies for efficient

simultaneous screening methods are required to increase output of such experiments.

1.7.2.2.4 Rehydration

The rehydration of probiotic powders is the final critical step during stabilization of probiotic

powders and revitalization conditions can have significant influence on the recovery of viable

cells since bacterial membranes become more susceptible to environmental stresses (Texeira et

al., 1996). Revitalization of cells and can be divided into four steps: wetting, submersion,


- 39 -

dispersion and dissolving (Freudig et al., 1999). Wetting of the particles is very often the

reconstitution controlling step and the rate of recovery depends on various factors including

osmolarity, pH, and nutritional energy as well as on temperature and volume (Carvalho et al.,

2004, Vega and Roos, 2006). Slow rehydration under controlled conditions was shown to

improve cell viability compared to immediate rehydration (Poirier et al., 1999). The medium to

rehydrate cells also affects recovery rate. For instance, complex media such as 10% (w/v) RSM

and PTM media consisting of peptone, tryptone and meat extract positively affected bacterial cell

recovery compared to minimal media such as phosphate buffer, sodium glutamate and water

(Costa et al., 2000). In other studies the use of the same cryopreservation and rehydration

solution resulted in increased viability (Ray et al., 1971; Abadias et al., 2001). A possible

explanation is that such a solution provides a high osmotic pressure environment which controls

the rate of hydration, and thus avoids osmotic shock (Morgan et al., 2006).

1.7.3 Enumeration of viable probiotic cells

Apart from advanced production and downstream processing technologies, the criterion of

minimal levels of viable cells within probiotic products requires also adequate quantification

methods to detect the viable probiotic strains in functional food products. The application of

rapid and reliable methods for enumeration of probiotics in mono- and mixed cultures is

important to validate quality of products containing probiotics. Cell quantification is generally

performed by plate counting on selective media to distinguish the microorganisms. However, this

method has several limitations including cell clumping, inhibition by neighboring cells, and it is

time consuming (Auty et al., 2001; Breeuwer and Abee, 2000; Bergmaier et al., 2005). Another

aspect concerns the lack of a standard medium to distinguish bifidobacteria from other LAB and


- 40 -

expressed thus the need for alternate, non-culture based technologies for adequate cell

enumeration (Talkalwar and Kailaspathy, 2003; Masco et al., 2005). The use of fluorescent stains

alone or in combination with enzymatic assays enabled the differentiation between viable,

metabolic active, damaged, dormant, viable but not cultivable, and dead bacterial cells and were

used to quantify viable bacteria (Breeuwer and Abee, 2000; Alakomi et al., 2005). Molecular

tools like fluorescence in situ hybridization (FISH) and flow cytometry were successfully applied

to estimate viable cells in probiotic products (Maukonen et al., 2006). Moreover, flow cytometry

can distinguish between subpopulations of stressed versus unstressed bifdobacteria (Ben Amor et

al., 2002). However, culture-based and fluorescent methods demand single cells suspension

because cell in clusters or chains lead to biased quantification of viable cell counts (Daley et al.,

1977; Nebe-von-Caron et al., 2000; Bibiloni et al., 2001).

There exists however alternative methods to detect or quantify bifidobacteria in fecal samples and

pharmaceutical or probiotic products. Technologies to detect bifidobacteria are based on DNA

fingerprinting approaches such as denaturing gradient gel electrophoresis (DGGE), pulsed-field

gel electrophoresis (PFGE). Quantitative approaches relate to the use of quantitative real time

PCR (qRT-PCR) (Vitali et al., 2003; Gueimonde et al., 2004; Masco et al., 2005; Matsuda et al.,

2007; Masco et al., 2007). But these methods are unable to provide information on the metabolic

activity or viability of the microorganisms since DNA or rRNA are too stable to be used as

viability markers (Josephson et al., 1993; Sheridan et al., 1998). mRNA on the other hand with

short half-life times can be considered as accurate viable cell marker (Belasco et al., 1986;

Hellyer et al., 1999). The quantification of mRNA using qRT-PCR has been recently applied in

medical applications to enumerate viable pathogenic microorganisms (Birmingham et al., 2008,

Coutard et al., 2005), but to date no such approach with mRNA has been applied to probiotic

bacteria. Growing numbers of accessible genome sequences of food related microorganisms and


- 41 -

progressing microarray technology will enlarge the knowledge about gene expression profiles of

bacteria in different environments which will raise the number adequate target mRNAs to

accurately quantify viable bacteria in various environments.

1.8 Hypothesis and background of this study

1.8.1 Background and general objective

Bifidobacteria are fastidious bacteria and sensitive to various environmental stresses during

production, downstream processing, incorporation in probiotic food products, storage and transit

through upper GIT. There is a general demand by food industries for advanced technologies to

enhance cell robustness and to improve selection procedure of candidate strains to meet minimal

viable cell numbers in products of prospering probiotic food market. Furthermore, the limitations

of accurate enumeration of viable cells require alternative methods to accurately quantify viable

cells independent of their morphological state.

The aim of the first part of the dissertation was to produce and characterize stress tolerant B.

longum NCC2705 strains using IC during continuous culture (Chapter 2). Continuous IC culture

induced formation of large cell aggregates which made physiological characterization with

traditional plate count method very difficult. Hence, we developed a new method to enumerate

viable B. longum NCC2705 irrespective of the morphological state (Chapter 3). The method

based on the quantification of the mRNA of a constitutively expressed housekeeping gene with

real-time PCR technology. With the help of this method, we were able to characterize survival of

aggregated cells produced with IC technology to lethal bile and heat stresses (Chapter 2).

In the last part of the doctoral thesis, we chose a different approach. Instead of producing stress

tolerant B. longum strains, we developed a high throughput screening method to identify B.


- 42 -

longum strains with intrinsic resistance to stresses occurring during spray drying and storage

(Chapter 4). Molecular biological tools allowed a differentiation of 22 B. longum strains of which

we were able identify the best survivor strain after spray drying and storage. The method was

then validated by comparing survival of the selected strain with a poor surviving strain.

1.8.2 Specific objectives

a. Characterization of B. longum NCC2705 produced with ICT in continuous culture:

� Perform a long-term fermentation with immobilized cells in a two-stage fermentation

system for 20 days

� Characterize biomass production, cell metabolism, and membrane composition

� Analyze the effect of cell immobilization and continuous culture on cell physiology:

tolerance to antibiotics, heat and porcine bile compared to free cell batch culture

b. Development of a real time PCR method to quantify viable B. longum NCC2705 cells in large

aggregates:

� Generate calibration curves with level of transcript of two housekeeping genes as a

function of cell concentration

� Validate target fragments of transcripts as viability marker

� Apply the method to quantify viable cells in large aggregates from continuous IC culture


- 43 -

c. Development of a fast screening method for selection of Bifidobacterium longum strains

resistant to spray drying and storage in dry powder:

� Establish strain specific molecular profiles using randomly amplified polymorphic DNA

(RAPD)

� Perform spray drying and storage experiments of mixtures of B. longum strains

� Validate the selection of resistant strain with single strain spray drying and storage

experiment

1.8.3 Hypotheses

� Immobilization and continuous culture of Bifidobacterium longum NCC2705 improves

volumetric productivity and tolerance to bile salt, heat and antibiotics.

� Quantification of transcript levels of a housekeeping gene with qRT-PCR allows accurate

enumeration of viable Bifidobacterium longum NCC2705 cells of different morphologies.

� The simultaneous spray drying and storage of mixed B. longum strains allows the rapid

identification of the most robust strains using RAPD PCR technology.

Chapter 2 - Immobilized cells

- 44 -

2 Improved tolerance to bile salts of aggregated

B. longum NCC2705 produced during continuous

culture with immobilized cells


- 45 -

2.1 Abstract

The effect of cell immobilization and continuous culture was studied on selected physiological

and technological characteristics of Bifidobacterium longum NCC2705. B. longum NCC2705 was

entrapped in gellan/xanthan gel beads and cultivated for 20 days with and without glucose

limitation in a two stage continuous fermentation system, with a first reactor (240ml) containing

immobilized cells, and a second reactor (1800ml) inoculated with free cells produced in the first

reactor. Continuous immobilized cell (IC) cultures with and without glucose limitation exhibited

formation of macroscopic cell aggregates after 12 and 9 days, respectively. Auto-aggregation

resulted in underestimation of viable cell counts by plate counts by more than 2 log units CFU/ml

compared to cell counts obtained with a cell-aggregate-independent method using real time PCR

technology. Cells in the effluent of continuous IC culture showed enhanced tolerance to porcine

bile salts and to aminoglycosidic antibiotics compared to stationary-phase cells produced during

free-cell batch fermentations. However, tolerance to lethal heat treatment was not different than

that from batch control culture. Changes in the cell membrane composition of continuously

produced cells such as the decrease of the ratio of unsaturated divided by saturated fatty acid

content of 0.57 ± 011 were observed compared to 1.74 ± 0.00 of membranes from batch control

culture.

Cells produced during continuous immobilized-cell cultures exhibited changes in cell membranes

which may induce formation of cell aggregates and tolerance to bile salts and aminoglycosidic

antibiotics. Cell immobilization in combination with continuous culture can be used to produce

probiotic bacteria with improved stress tolerance.


- 46 -

2.2 Introduction

Probiotics are “living micro-organisms which upon ingestion in certain numbers exert health

benefits beyond inherent basic nutrition” (WHO/FAO 2001). Bifidobacteria are widely used as

probiotics in functional food products (Leahy et al., 2005). Bifidobacteria are sensitive to various

factors during production and downstream processing and need to adapt to competitive and

changing environment in the human gastrointestinal tract to maintain viability (Lacroix and

Yildirim, 2007). Apart from safety issues, selection of potential probiotic candidate strains relies

on technological as well as functional properties. Among others, resistance to oxygen, heat and

gastro-intestinal conditions during production and consumption as well as the ability to adhere to

intestinal mucosa are thought to be important to provide beneficial health effects for the host

(Ouwehand et al., 2002; Salminen et al., 1999).

Different strategies such as pre-treatment of bacterial cells with sub-lethal stresses (Desmond et

al., 2002; Maus and Ingham, 2003) or physical protection using whole cell encapsulation

(Champagne et al., 2007) have been proposed to enhance stability of probiotics to stressful

conditions. Doleyres et al. (2004a) reported for continuous co-culture of immobilized B. longum

and L. lactis strains that cells produced in the effluent medium of the immobilized cell reactor

exhibited significantly increased tolerance to hydrogen peroxide, freeze-drying, and simulated

gastrointestinal conditions compared to free cells cultured in batch mode. Enhanced physiological

properties upon cell immobilization were also reported for lactobacilli where tolerance to nisin Z

and higher acidification activity were observed for continuously produced cells in mixed culture

(Grattepanche et al., 2007). The enhanced stress tolerance of immobilized cells were tentatively

attributed to changes in the microbial cell membranes of LAB and/or to non-specific stress

adaptations due to steep gradients of inhibitory products, pH and biomass within gel beads


- 47 -

(Masson et al., 1994; Trauth et al., 2001; Lacroix et al., 2005; Schepers et al., 2006). However,

the exact mechanism conferring increased stress tolerance remains unclear.

In this work, we studied the production of B. longum NCC2705 and time stability of continuous

cultures with immobilized cells (IC) in a two-stage fermentation system similar to that used by

Doleyres et al. (2004a) for mixed cultures with a first small reactor containing cells immobilized

on gel beads and a second large reactor connected in series and continuously inoculated from

cells produced in the first reactor. B. longum NCC2705 was chosen as model strain because its

genome has been sequenced (Schell et al., 2002). Furthermore, the probiotic characteristics,

technological properties, and stress tolerance and adaptation of this strain have been thoroughly

analyzed in the review of Klijn et al. (2005). The effect of cell immobilization and fermentation

time on cell production, metabolism, and selected physiological characteristics (morphology,

tolerance to bile salt, heat and antibiotics and membrane fatty acid composition) in the effluent

broth of continuous IC cultures were measured and data were compared to that of stationary-

phase cells produced from traditional free cell batch fermentations. Growth medium not

supplemented with glucose induced starving conditions in both reactors and thus a second

continuous fermentation was carried out in growth medium supplemented with extra-glucose.


- 48 -

2.3 Material and Methods

2.3.1 Strain and medium

Bifidobacterium longum NCC2705 was provided by the Nestlé Culture Collection (NCC; Nestlé

Research Centre, Lausanne, Switzerland). Cells were grown under anaerobic conditions using

atmosphere generation system packs (AnaeroGen, Oxoid, Pratteln, Switzerland) at 37 °C in de

Man, Rogosa and Sharp medium (MRS; de Man et al., 1960) broth (BioLife, Milan, Italy)

supplemented with 0.5 % (w/v) L-cysteine (Sigma-Aldrich, Buchs, Switzerland) (MRSC) and

inoculated at 1 % (v/v) with an overnight pre-culture before use.

2.3.2 Cell immobilization

Gel beads were produced in a two phase dispersion process using gellan and xanthan gum

according to Cinquin et al. (2004). Briefly, a polysaccharide mix was prepared containing gelrite

gellan (2.5 % (w/v), xanthan (0.25 % (w/v)) powders and sodium citrate (0.2% (w/v)) (Sigma-

Aldrich) resuspended in 500 ml pre-heated distilled water (90 °C) and mixed in a blender until

dissolution. The polymer solution was then autoclaved for 15 min at 121 °C just prior to use for

immobilization. The autoclaved polymer solution was cooled to 43 °C and inoculated aseptically

with B. longum NCC2705 pre-culture (2%; v/v). Gel beads were formed using a 2 phase

dispersion process by pouring the polymer solution into sterile hydrophobic phase (commercial

sunflower oil) at 43 °C. Beads were hardened by cooling the suspension to 25 °C and by

replacing the hydrophobic phase with 0.1 M solution of CaCl2 (Sigma-Aldrich). Beads with

diameters in the 1.0–2.0 mm range were selected by wet sieving and used for fermentation. The

entire process was completed in aseptic conditions within 1 h.


- 49 -

2.3.3 Continuous fermentations

Two IC continuous cultures were performed for 20 days. Before starting the continuous

fermentation, a bead colonization step was carried out in batch mode in a 500 ml bioreactor

(Sixfors, Infors-HT, Bottmingen, Switzerland) containing 370 ml MRSC and 80 ml of inoculated

gel beads. The colonization step was carried out at 37 °C for 16 h with CO2 flushed in the

headspace to ensure anaerobic conditions and agitation rate at 150 rpm by an inclined flat blade

impeller. pH was maintained at 6.0 by addition of 5 M NaOH. Colonized gel beads were then

aseptically wet sieved and transferred to a 500 ml bioreactor (R1) (Sixfors) containing 160 ml

medium for continuous fermentation for a total volume of 360 ml. Flow rate of fresh medium

was set at 360 ml/h using peristaltic pumps (Ismatec Reglo, Ismatec, Glattbrugg, Switzerland)

during the 20 days of culture, corresponding to a dilution rate of 2.25 h-1. Stirring was set at 200

rpm. A second reactor (R2) (Bioengineering, Wald, Switzerland) with a working volume of 1800

ml (dilution rate 0.2 h-1) was operated in series with R1 and continuously inoculated by

fermented broth from R1 (Figure 6). Stirring with two rushton type impellers in R2 was set at 250

rpm. The reactor volumes were chosen to yield a shorter residence time in R1 than in R2 (ca. 27

min and 5 h, respectively) for a high cell productivity in R1 containing a high immobilized cell

concentration and operated at high dilution rate and final growth in R2 with longer residence time

(Doleyres et al., 2004b). The pH in both reactors was maintained at 6.0 by addition of 5 M

NaOH. Temperature was set at 37 °C and CO2 was flushed in the headspace of both reactors to

provide anaerobic conditions.

Fermented broth samples were taken daily from the two reactors for microscopy, dry biomass,

HPLC analyzes, and viable cell enumeration by plate counts. Samples for stress tolerance assays

were taken every 2 days, centrifuged (6,000 × g for 15 min at 4 °C), and pellets were resuspended


- 50 -

pH 6.0Temp. 37 °CVol. 1800 mlDil. rate 0.2 h -1

MRSCor

MRSC +40 g/l glucose



MRSCor

MRSC +40 g/l glucose


in an equal volume of MRSC containing 10 % (w/v) glycerol (Sigma-Aldrich) prior to freezing in

liquid nitrogen and storage at -80 °C. This step was required to test samples in blocks and correct

for assay variability. HPLC samples were prepared by centrifuging 2 ml samples (10,000 × g for

1 min) and 1.5 ml of the supernatant was stored at -20 °C until analysis.

The first culture (ICC1) was run with unsupplemented MRSC leading to complete sugar

consumption in R1 and R2 whereas the second culture (ICC2) was operated with MRSC

supplemented with glucose (final concentration of 60 g/l) and anhydrous CaCl2 at 1.5 g/l (Sigma-

Aldrich). CaCl2 was added to ensure gel bead-stability throughout the entire fermentation time

because gel beads appeared soaked and consistency was weak at the end of ICC1. After

sterilization, initial glucose concentration measured in fresh medium for ICC1 and ICC2 was

13.6 and 52.8 g/l, respectively.

Figure 6 Diagram of the two-stage reactor setup for continuous fermentation with

immobilized cells including tubings for feed inlet, reactor connection, medium outlet, emergency

out, sampling ports, NaOH addition (pH: 6.0) and CO2 injection for anaerobic conditions.


- 51 -

2.3.4 Batch fermentation

Batch fermentation was carried out in a glass bioreactor (Bioengineering) with a working volume

of 2000 ml similar to R2 of the continuous fermentation. Unsupplemented MRSC was inoculated

at 2 % (v/v) with twice sub-cultured B. longum NCC2705 and incubated for 16 h at 37 °C under

anaerobic conditions by flushing CO2 in the headspace. Mixing was set at 250 rpm and pH

controlled at 6.0 by addition of 5 M NaOH. Growth was monitored by OD600 and growth phases

were set as follows: end exponential (9.5 h, OD600 4.3 ± 0.1), early stationary (12.5 h, OD600 of

5.7 ± 0.4) and mid stationary growth (16 h, OD600 5.6 ± 0.4). Samples were collected and

proceeded as described above. Batch fermentations were carried out in triplicate.

2.3.5 Viable cells enumeration

Viable cells in the fermented broth were enumerated by serially diluting fresh effluent of both

reactors in 12 mM phosphate buffer saline with 0.05 % cysteine (w/v) (PBSC). Appropriate

dilutions were drop plated in duplicate on MRS agar (1.5 %) (Becton Dickinson, Basel,

Switzerland) supplemented with 0.05 % (w/v) cysteine. Plates were incubated anaerobically

using atmosphere generation system packs (AnaeroGen, Oxoid, Pratteln, Switzerland) at 37 °C

for 48 h.

Viable cell enumeration using qRT-PCR method was only performed for samples from R2 of

ICC2 and bile- and heat tests of the same samples by quantifying levels of a 400 bp mRNA

fragment of a housekeeping gene, purB, (see Chapter 3.3). This method was used to estimate

viable cell counts in aggregated cell samples. Briefly, 0.5 ml of washed cell samples were put in

2-mL screw-cap tubes containing a mix of glass beads, phenol/chloroform 1:1 (v/v), sodium

dodecyl sulfate 10 % (w/v) and 3 M sodium acetate (pH 5.2) (Sigma-Aldrich) and immediately


- 52 -

frozen in liquid nitrogen. Cells were disrupted by bead-beating in a homogenizer (FastPrep FP

120, Q-Biogene, Carlsbad CA, USA), four times for 40 sec at speed 4.0 intercalated with cooling

periods on ice. Samples were then centrifuged to remove cell debris and glass beads. The aqueous

phase containing RNA was transferred to a new tube and supplemented with a known amount of

exogenous luciferase mRNA (Promega, Dübendorf, Switzerland) as a reference. Traces of phenol

were removed by extraction with pre-chilled chloroform. The resulting aqueous phase was mixed

with an equal volume of 70%-ethanol and applied to RNA purification column using RNeasy

Mini Kit (Qiagen, Basel, Switzerland). Further purification was performed according to the

manufacturer’s protocol including on-column DNase I digestion for 30 min followed by elution

in 100 µl H2O. Purified RNA (max 100 ng) was applied to reverse transcription using the high

capacity cDNA reverse transcription commercial kit (Applied Biosystems, Rotkreuz,

Switzerland). cDNA was then applied to quantitative PCR using an ABI PRISM 7700 equipment

with SYBR® Green PCR Master Mix (Applied Biosystems). The normalized levels of the 400 bp

amplicon were then measured to determine viable cell concentrations along with prepared

calibration curves.

2.3.6 Dry biomass determination

Dry biomass was determined according to Bergmaier et al. 2005 with slight modifications. 6 ml

of thawed sample were centrifuged at 10,000 × g for 15 min at 4 °C. The pellet was washed twice

with distilled water and transferred to a pre-weighed aluminum dish. The cells were dried for 20

h in an oven at 99 °C and weighed after cooling to room temperature (25 °C) in an exsiccator

(Duran Group GmbH, Mainz, Germany). Values are single determinations.


- 53 -

2.3.7 Microscopic observation

Pictures of fresh cells from the fermentation were obtained using an optical microscopy at 1000-

fold magnification (Leica DM, Leica Microsystems, Wetzlar, Germany) by an attached camera

(Leica DFC280) without any further treatment.

2.3.8 Metabolite analysis

Consumption of glucose and concentration of main metabolites, lactic and acetic acids, were

determined by HPLC (Merck Hitachi, Darmstadt, Germany). Samples were diluted in MilliQ

water and filtered through a 0.45 µm nylon membrane filter (Infochroma AG, Zug, Switzerland).

Separation was carried out in an Aminex HPX-87-H column (Bio-Rad Laboratories, Reinach,

Switzerland) with 10 mM H2SO4 as mobile phase at a flow rate of 0.6 ml/min and a temperature

of 40 °C. Compounds were detected using a refractive index detector (Merck Hitachi, Darmstadt,

Germany). Reported data are means of duplicate analysis.

2.3.9 Fatty acid analysis

The fatty acid composition of cell membrane was determined by gas chromatography at the

DSMZ external service (German Collection of Microorganisms and Cell Cultures, Braunschweig,

Germany) using the Microbial Identification System (MIDI, Newark, DE; http://www.midi-

inc.com). Frozen cell samples of end exponential growth phase cells produced with batch cultures

and from R2 of ICC2 at day 4, 12 and 20 were thawed at room temperature and washed (6,000 ×

g for 3 min) with PBSC and pellets were frozen using liquid nitrogen. Fatty acids were converted

to methyl esters and extracted in a five-step procedure according to the DSMZ protocols.

Analyzes were performed in duplicate.


- 54 -

2.3.10 Heat test

Heat survival tests were carried out according to Simpson et al. (2005) with some modifications.

Briefly, frozen cell samples of ICC2 were thawed at room temperature and centrifuged at 6,000 ×

g for 2 minutes. Supernatant was discarded and pellet resuspended in the same volume of fresh

MRSC and incubated for 4 h at 37 °C to reactivate cells. This treatment was selected based on

preliminary screening tests to obtain sufficient cell activity for RNA degradation. Cell

suspensions were split in 1 ml aliquots and mixed with either 9 ml pre-warmed MRSC at 56 °C

or room temperature (control) and incubated in a water bath at the corresponding temperature for

30 min. After the stress, cells were immediately cooled to room temperature in a water bath for

three minutes, washed in fresh MRSC and pellets suspended in 1 ml MRSC. To quantify viable

cells, 30 µl of cell solution was applied to a dilution row in PBSC and appropriate dilutions were

plated on MRSC agar.

To enumerate viable cells from macroscopic aggregates containing samples with qRT-PCR, 0.5

ml of the suspended cells solution was shock frozen in liquid nitrogen and stored at -80 °C for

RNA-extraction and qRT-PCR analysis as described above. For all tests, controls consisted of

cells produced during batch cultures. Experiments were performed in duplicate.

2.3.11 Bile salt test

The effects of bile salts were tested as previously described by Saarela et al. (2004) with some

modifications. Samples were prepared as for the heat test. After a reactivation step at 37 °C for 4

h, cell samples from ICC1 were split in aliquots of 1 ml and immediately added to 9 ml 1.5 %

(w/v) porcine bile salt (Sigma-Aldrich) dissolved in PBSC or in 9 ml PBSC without bile salts and

incubated for 17 min at 37 °C. After the stress, cells were cooled to room temperature in a water


- 55 -

bath for three minutes and washed twice in fresh MRSC to remove bile salts. To quantify viable

cells, 30 µl of cell solution was applied to a dilution row in PBSC and appropriate dilutions were

plated on MRSC agar.

Bile tolerance of samples from ICC2 was evaluated as described above, but lethal bile treatment

was carried out with 1.0 % (w/v) bile solution for 12 min at 37 °C. Less restrictive stress

conditions were chosen, because the detection limit of the qRT-PCR method (ca. 4.4 × 104

CFU/ml) was higher than for plate counts and could not detect viable after bile stress at 1.5 %

(w/v) for 20 min of control cells. Survival was determined with standard plate counts and qRT-

PCR as for the heat test. Experiments were performed in duplicate.

2.3.12 Tolerance to antibiotics

Antibiograms were performed using an antimicrobial disc susceptibility assay according to the

National Commitee for Clinical Laboratory Standards (1991) with some modifications. Cell

samples were diluted in MRSC to an OD600 of 0.0125 and streaked out with a cotton tip on

MRSC agar plates. Cells from IC cultures were only tested before formation of macroscopic

aggregates because OD values of samples containing cell clusters could not be accurately

measured and were not representative of cell counts. Discs of 6 mm diameter containing

antibiotics were then placed on agar plates. The used antibiotic-discs were: Penicillin G (P 10),

Norfloxacine (NOR 10), Nalidixic acid (NA 30), Ampicilline (AM 10), Cefotaxime (CTX 30),

Erythromicin (E 15), Gentamicin (GM 10), Tetracycline (TE 30), Vancomicin (VA 30),

Chloramphenicol (C 30) (BioMérieux Suisse SA, Geneva, Switzerland) and Neomicin (N 10),

Bacitracin (B 10), Teicoplanin (TEC 30) (Oxoid AG, Switzerland)). Plates were incubated


- 56 -

anaerobically at 37 °C for 48 h and the inhibition zone was measured in mm. Experiments were

performed in duplicate.

2.3.13 Statistical analyzes

Means of log10-transformed cell counts for heat- and bile tests as well as inhibition diameters of

antibiograms and molecular acetate /lactate ratios were compared with Student’s t-test using JMP

In Software version 6.0 (SAS Institute Inc., Cary, NC) with a level of significance at 0.05.

2.4 Results

Two continuous IC cultures were carried out in a two-stage fermentation system (Figure 6) for 20

days and effluents of both fermentations and reactors were tested for cell morphology, cell and

metabolites concentration, and cell survival to lethal heat and bile salt stresses to assess the

effects of immobilization in combination with continuous culture time and different glucose

concentrations.

2.4.1 Biomass production

Continuous culture of immobilized B. longum NCC2705 induced formation of large cell

aggregates in both reactors of ICC1 after 12 days of fermentation and to an even higher extent in

ICC2 where macroscopic aggregates of ca. 0.5 – 1 mm diameter were observed after only 9 days

of continuous culture (Figure 7).


- 57 -

Figure 7 Optical micrographs of fresh cells collected from fermented broth (1000 ×

magnification) of B. longum NCC2705 produced in batch culture (A) and in ICC2 from day 8

(B), day 12 (C) and day 20 (D).

Free cell population for ICC1 with limiting glucose tested in the fermented broth medium with

plate counts were very similar in both reactors and decreased progressively with culture time

from 1.4 ± 0.1 × 109 to 3.4 ± 0.7 × 108 CFU/ml between day 2 and 20 (Figure 8). Cell dry weight

(CDW) of ICC1 also decreased progressively but to a lower extent with time in R1 from 1.9 at

day 4 to 1.3 g/l at day 20, but remained constant in R2 at 1.7 ± 0.2 g/l throughout the entire 20

days (Figure 8). High maximal volumetric productivities were reached after 2 days of culture,

with 6.2 × 1011 for R1 and 2.5 × 1011 CFU/l h for the two stage culture (Table 4).

In ICC2 operated with glucose supplementation, populations of both reactors tested with plate

counts showed large fluctuations with fermentation time. From day 1 to 7 cell counts increased

steadily from 1.2 × 107 to 1.0 × 109 and from 1.7 × 107 to 2.3 × 109 CFU/ml for R1 and R2,

respectively (Figure 8). After detection of macroscopic cell aggregates at day 9, cell counts

decreased continuously until the end of fermentation to 2.4 × 108 and 6.3 × 108 CFU/ml for R1

and R2, respectively (Figure 8). In contrast, viable cell concentration determined for R2 with

aggregation independent qRT-PCR method increased steadily from 1.1 ± 0.4 to 8.6 ± 4.4 × 109

CFU/ml between day 4 and 12 and remained stable afterwards (Figure 8 B). Similarly, CDW

increased steadily from 0.3 to 2.7 g/l and from 0.3 to 4.6 g/l in R1 and R2, between day 2 and 20,

respectively (Figure 8). CDW correlated well with viable cell counts tested with qRT-PCR but


- 58 -

not with plate counts. Maximal volumetric productivities reached 5.9 × 1011 for R1 and 4.2 × 1011

for the two stage culture determined by plate counts or 1.5 × 1012 CFU/l h for the two stage

culture determined with qRT-PCR method (Table 4).

The maximal cell concentration for batch cultures was 1.6 ± 1.1 × 109 CFU/ml after 12.5 h

growth in MRSC at an OD600 of 5.7 ± 0.4 (early stationary growth phase) corresponding to a

volumetric productivity of 1.2 × 1011 CFU/l h (Table 4).

Table 4 Maximal cell counts and volumetric productivities of B. longum NCC2705 in

different fermentation systems. Values obtained by plate counts or by qRT-PCR method.

Fermentation systems Fermentation time

Dilution rate (h-1)

Cell counts (CFU/ml)

max. vol. productivity (CFU/l h)

Batch 13.5 h - 1.6 × 109 1.2 × 1011

ICC1 R1 2 days 2.25 1.4 × 109 6.2 × 1011

ICC1 R1 + R2 2 days 0.18 1.4 × 109 2.5 × 1011

ICC2 R1 6 days 2.25 1.3 × 109 5.9 × 1011

ICC2 R1 + R2* 7 days 12 days (qRT-PCR)

0.18 2.3 × 109

(7.9 × 109) 4.2 × 1011

(1.5 × 1012)

Continuous IC1 n.i. 0.5 4.9 × 109 2.5 × 1012

Free cell continuous2 105 h 0.1 1.1 × 109 1.1 × 1011

* values in brackets from qRT-PCR 1 for B. longum ATCC15707 in Doleyres et al. (2002) 2 Mozzetti (2009)


- 59 -

Figure 8 Viable cell concentrations and cell dry weight (CDW) in samples from R1 (A) and

R2 (B) during continuous immobilized cell cultures. Cell concentrations determined by plate

counts (full line, □, ■) or by qRT-PCR (semi dashed line, ●) and cell dry weight (dashed line, ∆,

▲). Data from ICC1 (open symbols) and ICC2 (solid symbols) are means of two determinations.


- 60 -

2.4.2 Metabolic Activity

Glucose consumption in R1 of ICC1 was stable throughout the entire fermentation at 12.5 ± 0.2

g/l and resulted in almost no sugar in the effluent of R1 at a mean residual concentration of 1.1 ±

0.2 g/l (Figure 9). Lactic and acetic acid production were stable during 20 days, with 5.4 ± 0.4

and 6.2 ± 0.6 for R1 and 0.3 ± 05 and 0.8 ± 0.9 g/l for R2, respectively. Therefore the molar ratio

of main metabolites acetic and lactic acids in R1 remained also constant during the entire

fermentation at 1.72 ± 0.07. The low sugar concentration in the effluent from R1 of ICC1 caused

starving conditions in R2 and hence formation of metabolites was not relevant.

In ICC2, glucose consumption varied with time in both reactors. In R1, glucose consumption

increased steadily from 16.0 ± 1.3 at day 6 to 39.7 ± 0.2 g/l at day 13 after an adaptation phase

during the first five days. Afterwards, consumption decreased to reach 29.9 ± 0.6 g/l at day 20. In

contrast, glucose consumption in R2 increased steadily between day 1 and 7 from 1.8 ± 0.0 to

23.0 ± 0.5 g/l, then decreased sharply to 4.1 ± 0.2 g/l at day 12. From day 13 until the end of

fermentation, consumption increased again to reach 19.5 ± 0.0 g/l. Residual glucose

concentrations in ICC2 were never lower than 13.1 ± 0.2 and 1.8 ± 0.1 in R1 and R2,

respectively, indicating no sugar limitation. The acetic and lactic acid production in R1 increased

from 7.0 ± 0.1 and 5.0 ± 0.1 at day 3 to 15.3 ± 0.2 and 14.3 ± 0.3, respectively, and then

decreased until the end of fermentation to 11.4 ± 0.4 and 11.0 ± 0.4 g/l, respectively. The

production of acetic and lactic acid in R2 varied like the sugar consumption and ranged between

a maximum of 9.4 ± 0.9 and 7.9 ± 0.8 at day 7 and a minimal value of 1.7 ± 1.1 and 0.9 ± 1.0 g/l

at day 12, respectively. The molar ratio of acetic to lactic acids showed a slight but significant

decrease in R1 after an adaptation phase in the first 3 days, from 1.75 at day 4 to 1.56 at day 20.


- 61 -

R2 showed large fluctuations over time in the range of 1.58 to 2.85, mirroring glucose

consumption (Figure 9 B).

Glucose in batch culture was entirely consumed when cells reached early stationary growth phase

after 12.5 h growth. The levels of acetic and lactic acid reached the maximum at mid stationary

growth phase (16 h) with 7.3 ± 0.0 and 5.6 ± 0.0 g/l, respectively, corresponding to a molar ratio

of acetic to lactic acids of 1.93.


- 62 -

Figure 9 Consumption of glucose and molar ratio of acetic to lactic acid in samples from

R1 (A) and R2 (B) during continuous immobilized cell cultures. Glucose consumption (full line;

□, ■) and molar ratio of acetate and lactate (dashed line; ∆, ▲). Data from ICC1 (open symbols)

and ICC2 (solid symbols) are means of two determinations.


- 63 -

2.4.3 Cell sensitivity to heat and porcine bile sal ts

Cell tolerance to porcine bile salt stress was first tested by exposing cells from ICC1 and control

batch for 20 min to 1.5 % (w/v) porcine bile salt solution and measured by plate count

enumeration. Samples from ICC1 at day 4, 12 and 20 showed significantly less loss of viable

cells of 3.5 ± 0.4, 2.8 ± 0.1 and 3.0 ± 0.1 for R1 and 2.6 ± 0.3, 2.6 ± 0.4 and 1.8 ± 0.2 log10

CFU/ml for R2, respectively, compared to batch stationary phase cells with 5.7 ± 1.2 log10 loss of

CFU/ml (Table 5). Indeed, the viable cell counts of batch control samples after exposure to 1.5 %

(w/v) bile salt solution were too low to be detected with the qRT-PCR method (detection limit ca.

4.5 × 104 CFU/ml) and hence the conditions for the bile assay were set at 1.0 % (w/v) bile salt

solution to test samples from ICC2 and batch control cells. Viability loss of ICC2 cells was tested

for cells from R2 at day 4, 12 and 20, using plate counts and qRT-PCR method. The bile assay

was carried out for 15 min at 1.0 % (w/v) bile salt solution, Viability loss of cell samples

evaluated with plate counts was 3.3 ± 0.0, 1.7 ± 0.1, 3.0 ± 0.2 log10 CFU/ml for cells from R2 day

4, 12, 20 and significantly smaller than 4.0 ± 0.2 log10 CFU/ml of batch control cells. When

tested with qRT-PCR, viability losses for cells from day 4 and 20 were 0.6 ± 0.2 and 0.5 ± 0.2

log10 CFU/ml, respectively, and also significantly smaller than 1.0 ± 0.1 log10 CFU/ml for batch

control cells, whereas no significant difference was detected for cells from day 12 (Table 5).

Heat tolerance was tested by heating cell samples from ICC2 R2 at 56 °C for 30 min. Viability

losses tested with plate counts of cells collected at day 4, 12 and 20 range between 4.1 – 3.4 log10

CFU/ml and were not significantly different to 3.4 ± 0.4 log10 CFU/ml of batch control cells.

Similarly, no significant differences between cells from continuous and batch were observed

using qRT-PCR method (Table 5).


- 64 -

Table 5 Cell viability loss after exposure to porcine bile salt and heat lethal stresses of

cells from batch, and continuous immobilized cell cultures at day 4, 12 and 20. Values are given

as loss of log10 (CFU/ml) determined by plate counts or qRT-PCR. Data are means of two

independent repetitions.

Treatment Detection method

Batch ICC1*

day 4 Day 12 Day 20

20 min in 1.5 % porcine bile

Plate counts

5.7 ± 1.2a

R1 3.5 ± 0.4b R2 2.6 ± 0.3b

2.8 ± 0.1b 2.6 ± 0.4b

3.0 ± 0.1b 1.8 ± 0.2C

Batch ICC2 (R2)

day 4 Day 12 Day 20

30 min at 56°C Plate counts

qRT-PCR

3.4 ± 0.4a

2.3 ± 0.3a

4.1 ± 0.1b

2.5 ± 0.2a

3.1 ± 0.3a

2.0 ± 0.3a

3.4 ± 0.0ab

2.5 ± 0.3a

15 min in 1 % porcine bile §

Plate counts

qRT-PCR

4.0 ± 0.2a

1.0 ± 0.1a

3.3 ± 0.0b

0.6 ± 0.2b

1.7 ± 0.1c

0.7 ± 0.1ab

3.0 ± 0.2b

0.5 ± 0.2b

Different letters on each line (p ≤ 0.05). * Values from R1 and R2

2.4.4 Tolerance to antibiotics

Tolerances to antibiotics were tested using disc-assay and only applied to samples from

continuous cultures before formation of aggregates. No inhibition zone was observed for cells

from ICC1 and 2 and batch cultures to norfloxacine and nalidixic acid (Table 6). Cells from ICC1

collected at day 2 and 10 showed significant increased tolerance to aminoglycosidic gentamicin

and neomycin (p ≤ 0.05) compared to batch control cells, and a slight but not significant

increased tolerance with culture time for both antibiotics (Table 6).

A significant increase in tolerance of cells from ICC2 was detected between day 2 and 8 for

penicillin, erythromycin, gentamicin, tetracycline, vancomycin, neomycin, bacitracin and


- 65 -

teicoplanin. Furthermore, cells collected at day 8 showed significantly smaller diameters (p ≤

0.05) for gentamicin and neomycin, compared to cells from control batch (Table 6).

Table 6 Inhibition diameters measured by the disc assay method of different antibiotics on

cells from control batch cultures and produced by continuous immobilized cell cultures ICC1 and

ICC 2 tested at different days of fermentation before formation of macroscopic cell aggregates.

Values are means of two repetitions.

Antibiotics Batch ICC1 ICC2

Day 2 Day 10 Day 2 Day 8

Penicillin G [10] 42.0 ± 1.3 a - - 52.5 ± 0.7b 47.5 ± 0.7c Norfloxocine [10] 6.0 ± 0.0* a - - 6.0 ± 0.0* a 6.0 ± 0.0* a Nalidixic acid [30] 6.0 ± 0.0* a - - 6.0 ±0.0* a 6.0 ± 0.0* a Amplicillin [10] 41.0 ± 2.3 a - - 48.0 ± 2.8b 42.5 ± 0.7ab Cefotaxime [30] 41.0 ± 2.9 a 40.0 ± 1.3 a 38.8 ± 1.0a 45.5 ± 3.5b 37.5 ± 4.9a Erythromycin [15] 45.0 ± 1.4 a - - 51.5 ± 2.1b 44.5 ± 0.7a Gentamicin [10] 16.0 ± 1.3 a 10.2 ± 2.4b 9.3 ± 2.3b 23.0 ± 1.4c 8.0 ± 2.8b Tetracycline [30] 41.0 ± 1.3 a - - 46.5 ± 3.5b 41.0 ± 0.0a Vancomycin [30] 31.0 ± 1.0 a - - 38.5 ± 0.7b 32.0 ± 0.0a Chloramph. [30] 44.0 ± 1.3 a - - 50.0 ± 4.2b 46.0 ± 0.0ab Neomycin[10] 14.0 ± 1.3a 9.4 ± 1.7b 8.6 ± 1.5b 16.5 ± 0.7a 6.0 ± 0.0b Bacitracin[10] 23.0 ± 2.2 a - - 31.0 ± 2.8b 24.0 ± 0.0a Teicoplanin [30] 30.0 ± 0.0 a - - 36.5 ± 2.1b 29.0 ± 1.4a *6 mm diameter refers to no inhibition (diameter of antibiotic disc). - : not tested Different letters on each line (p ≤ 0.05).

2.4.5 Cell membrane fatty acid composition

Fatty acid composition of membranes of cells produced in ICC2 exhibited no differences between

R1 and R2 (data not shown) but showed large fluctuations compared to membranes of batch

culture cells (Table 7). Major changes in fatty acid composition were observed in saturated 14 : 0

and 16 : 0, as well as in unsaturated 18 : 1w9c fatty acids which account for more than 60 % of

the total fatty acids of membranes from cells of all tested samples (Table 7). The ratio of


- 66 -

unsaturated to saturated fatty acids (USFA/SFA) of cells collected during continuous IC culture

was 0.58 ± 0.09 and significantly lower than 1.74 ± 0.00 for cells from free cells batch culture.

Table 7 Fatty acid membrane composition in % of total fatty acids of cells from ICC2

collected at day 4, 12, 20 and batch culture cells. Data are means of two determinations.

* Fatty acids also including DMA forms. n.d., not detected; USFA, unsaturated fatty acids; SFA, saturated fatty acids; BCFA, branched chain fatty acids; DMA, dimethyl acetal fatty acids

Fatty acid Batch ICC2 R2

Day 4 Day 12 Day 20

12:0 0.33 ± 0.01 0.77 ± 0.09 0.40 ± 0.06 0.66 ± 0.04

14:0* 2.91 ± 0.04 14.36 ± 0.82 10.38 ± 0.92 13.30 ± 1.89

16:0 17.57 ± 0.08 27.73 ± 2.81 29.74 ± 1.15 32.72 ± 0.29

18:0 10.16 ± 0.06 6.94 ± 0.51 9.14 ± 0.12 6.02 ± 0.70

19: 0 cyclo* 1.95 ± 0.00 10.34 ± 0.39 5.65 ± 0.14 3.02 ± 0.72

19 cycloprop 1.36 ± 0.04 6.01 ± 0.90 1.89 ± 0.25 4.20 ± 0.48

13:1 none 0.21 ± 0.01 3.25 ± 0.43 1.91 ± 0.09 1.46 ± 0.28

16:1 w9c 0.42 ± 0.01 1.60 ± 0.02 0.80 ± 0.02 1.37 ± 0.01

17:1w9c 5.59 ± 0.10 1.13 ± 0.04 2.98 ± 0.05 1.93 ± 0.27

18.1 w9c* 50.25 ± 0.13 18.89 ± 0.43 29.02 ± 0.15 29.30 ± 2.82

18:1w7c* 5.70 ± 0.03 1.71 ± 1.55 1.76 ± 0.14 2.07 ± 0.42

18:1w6c n.d. 3.48 ± 0.83 1.54 ± 0.11 1.32 ± 0.05

18:0 OH 5.10 ± 0.03 2.91 ± 1.20 0.95 ± 0.40 0.91 ± 0.27

Total BCFA (ISO) 1.46 ± 0.01 0.73 ± 0.23 0.44 ± 0.03 0.57 ± 0.09

Total DMA 32.63 ± 0.01 25.15 ± 0.89 28.61 ± 0.99 20.26 ± 3.92

USFA 62.17 ± 0.08 30.05 ± 1.33 37.99 ± 0.09 37.11 ± 2.43

SFA 35.72 ± 0.07 66.88 ± 0.73 57.41 ± 0.42 60.49 ± 1.22

USFA/SFA 1.74 ± 0.00 0.45 ± 0.02 0.66 ± 0.01 0.61 ± 0.05

Labeled 97.89 ± 0.15 96.93 ± 0.63 95.41 ± 0.35 97.60 ± 1.39

Not labeled 2.12 ± 0.15 3.07 ± 0.63 4.60 ± 0.35 2.40 ± 1.39


- 67 -

2.5 Discussion

Immobilized cell technology (ICT) in combination with continuous culture has previously been

suggested for the production of starter or probiotic cultures with enhanced functional and

technological properties and high volumetric productivities (Doleyres and Lacroix, 2005). In the

present study, two continuous IC cultures were carried out for 20 days with and without glucose

limiting conditions. The supplementation of additional glucose resulted in a high maximal cell

production of 7.9 × 109 CFU/ml and volumetric productivity of 1.5 × 1012 CFU/l h for the two

stage culture measured with qRT-PCR was reached in ICC2 at day 12. This value is similar to

that reported by Doleyres et al. 2002 during continuous culture with immobilized B. longum

ATCC15707 in MRS medium supplemented with whey permeate at a dilution rate of 0.5 h-1 and

approximately ten fold higher than that of continuous culture with free cells of B. longum

NCC2705 in MRSC operated at a dilution rate of 0.1 h-1 as reported in a by Mozzetti (2009) and

controlled batch cultures (Table 4).

Formation of large cell aggregates was observed in both continuous cultures leading to

considerable underestimation of cell concentration with plate counts. Indeed, accurate viable cell

enumeration of B. longum NCC2705 using the recently developed qRT-PCR method which is

independent of cell aggregates (see Chapter 3) detected higher viable cell counts compared to

traditional plate counts. Extensive formation of cell aggregates has already been reported during

continuous cultures with immobilized lactobacilli (Bergmaier et al., 2005; Koch, 2006), whereas

no cell aggregates were reported for the B. longum NCC2705 grown for more than 8 days in

continuous mode with free cells in MRS (Mozzetti, 2009). Aggregation of bacterial cells can be

induced by changing environmental factors for instance pH, level of divalent cations in the


- 68 -

medium and/or nutrients availability (Koop et al., 1989; Canzi et al., 2005, De Schryver et al.,

2008).

In this study, formation and variation of size of cell aggregates in both IC cultures may have

different origins. Although the pH in the bioreactors was constant at 6.0 during all fermentations

(ICC and batch), the effect of acidic conditions in gel beads could be involved in cell

aggregation. Indeed, steep pH gradients can develop within gel beads (Masson et al., 1994;

Schepers et al., 2006). Larger size of aggregates in ICC2 may be a result of the supplementation

of growth medium with CaCl2 because Ca2+ ions in the growth medium were previously shown to

induce auto-aggregation in bifidobacteria (Ibrahim et al., 2005). Divalent cations may act as a

bridging compound for negatively charged functional groups on the cell surface and may promote

adhesion between bacterial cells (Higgins and Novak, 1997). Another factor affecting size of

aggregates is the supplementation of growth medium with additional glucose. Higher sugar

concentration in fermentation broth yielded higher biomass concentration and consequently more

cell growth and larger sized cell aggregates. The sugar limiting conditions of ICC1 may also

explain a metabolic shift towards production of acetic acid resulting in elevated molecular ratio

of acetic and lactic acid. Enhanced acetic acid production was shown to improve the ATP-yield

in B. animalis (Ruas-Madiedo et al., 2005), thereby generating higher energy resources for

starving cells from equal amount of glucose. The molar ratio in batch cultures was higher than in

continuous cultures, but comparison of the ratio is limited due to metabolite accumulation in

batch fermentation systems.

Physicochemical characteristics of the cell surface such as hydrophobicity may also affect auto-

aggregation and adhesion of bacterial cells to different surfaces (Perez et al., 1998; Del Re et al.,

2000; Zavaglia et al., 2002). Changes in membrane characteristics upon growth in adverse micro-

environment created within gel structures were already reported for Lactococcus lactis ssp. lactis


- 69 -

SL03 and Lactococcus lactis ssp. cremoris SC09 (Trauth et al., 2001) or resulted in lower

USFA/SFA ratio in fatty acid composition of membranes of immobilized Saccharomyces

cerevisiae (Jirku et al., 1999). In the present study, immobilized B. longum exhibited notable

modifications in fatty acid profiles (e.g. lower USFA/SFA ratio) of cell membranes compared to

free cells grown in batch cultures. Alterations in membrane composition may occur as a result of

an adaptation to changing environment and can lead to changes in membrane hydrophobicity as

shown for starved or acid-adapted Listeria innocua cells (Moorman et al., 2008).

The observed changes in membrane composition of cells released from gel beads may also play

an important role in their improved resistance to bile salts and the aminoglycosidic antibiotics

gentamycin and neomycin. A similar decrease of the ratio of unsaturated to saturated fatty acids

of cell membrane from IC cultures was reported for a bile adapted mutant strain of B. animalis

compared to its mother strain (Ruiz et al., 2007). These changes might be involved in restricted

diffusion of bile salts into the cytoplasm (Begley et al., 2005). The higher cell loss tested with

aggregated cells and plate counts compared to qRT-PCR were not expected because in theory one

viable cell in aggregate can give a colony, leading to an overestimation of cell survival after the

bile test. The observed difference in viability-loss between plate count and qRT-PCR method can

be a result of a bile induced sublethal injury within the Bifidobacterium population, possibly

through a reversible and transient membrane permeabilization which resulted in a loss of

viability, as defined by plate counts, but these cells could regain growth after being sorted and

resuscitated as reported by Ben-Amor et al. (2002).

Modifications in the membrane properties might also prevent passive diffusion of

aminoglycosides into the cell (Bryan et al., 1984). The integrity of membranes of B. longum was

shown to be involved in the tolerance to aminoglycosides, where susceptibility to


- 70 -

aminoglycosides increased when cells were previously subjected to membrane corrupting oxgall

or H2O2 stress (Kheadr et al., 2007).

Doleyres et al. (2004a) reported a progressive increase of survival of B. longum ATCC 15707 to

simulated intestinal conditions and tolerance to nisin Z with fermentation time induced by cell

immobilization in continuous co-culture with Lc. lactis MD. The increase of bile tolerance with

time correlated with increased tolerance to nisin Z, which targets bacterial cell membranes. These

findings support the observations in this study of increased tolerance to bile salts due to

membrane modifications induced by IC and continuous culture. The effect of culture age on

stress tolerance can be strain dependent and/or may originate from an adaptation process with

fermentation time of B. longum to restrictive environments produced by the competitive Lc. lactis

strain (Doleyres et al., 2004a). In a recent study, a co-culture of B. longum and B. breve was

cultivated in a compartmentalized system where the cells grew separately but the supernatants

were mixed together (Ruiz et al., 2009). The authors reported an increased expression of stress

response proteins such as chaperones in B. breve which may protect bacterial cells from

environmental stresses.

Cells from ICC2 did not exhibit detectable changes in tolerance to heat compared to batch cells.

Although, changes in membrane composition can influence bacterial membrane fluidity and

hence cell tolerance to heat (Guillot et al., 2000; Denich et al., 2003), the increase in shorter fatty

acids (acyl-chains of less than 18 carbons) and the decrease of unsaturated fatty acids in

membranes of cells from ICC2 did not affect survival to lethal heat stress. It is noteworthy to

mention that preliminary results from gene expression analyzes of cells from ICC2 compared

with control batch cells showed that genes coding for stress related proteins were not

differentially expressed as reported for heat stressed B. longum NCC2705 cells (Rezzonico et al.,

2007).


- 71 -

The use of aggregated cells in probiotic food industries is still unfavorable in product applications

due to underestimation of viable cell counts with standard plate counts that is the standard

method to verify product quality (see Chapter 3). However, the ability to form aggregates could

be a beneficial functionality for probiotic strains. Indeed, aggregation ability of Bifidobacterium

as well as Lactobacillus cells has been associated with enhanced adhesion to epithelial cells from

human or pigs (Del Re et al., 2000; Kos et al., 2003), coaggregation with pathogenic bacteria

(Schachtsiek et al., 2004; Collado et al., 2007) and improved protective effect against colitis in

mice (Castagliuolo et al., 2005). Further experiments are required to test whether observed

morphological and physiological changes of cells from IC cultures affect survival during GIT

transit and adhesion-ability to epithelial cells.

Chapter 3 - qRT-PCR

- 72 -

3 Development of a real-time RT-PCR method for

enumeration of viable Bifidobacterium longum cells in

different morphologies

Data presented in this chapter were published in 2009 in Food Micriobiology with the following authors: Reimann S., Grattepanche F., Rezzonico E., and Lacroix C.

Chapter 3 - qRT-PCR

- 73 -

3.1 Abstract

Viability of probiotic bacteria is traditionally assessed by plate counting which has several

limitations, including underestimation of cells in aggregates or chains morphology. We describe a

quantitative PCR (qPCR)-based method for an accurate enumeration of viable cells of

Bifidobacterium longum NCC2705 exhibiting different morphologies by measuring the mRNA

levels of cysB and purB, two constitutively expressed housekeeping genes. Three primer-sets

targeting short fragments of 57-bp of cysS and purB and one 400-bp fragment of purB were used.

Cell quantification of serially diluted samples showed a good correlation coefficient of R2 0.984

± 0.003 between plate counts and qRT-PCR for all tested primer sets. Loss of viable cells

exposed to a lethal heat stress (56 °C, 10, 20 and 30 min) was estimated by qRT-PCR and plate

count. No significant difference was observed using qRT-PCR targeting the 400-bp fragment of

purB compared to plate count indicating that this fragment is a suitable marker of cell viability. In

contrast, the use of the 57-bp fragments led to a significant overestimation of viable cell counts

(18 ± 3 and 7 ± 2 fold for cysB and purB, respectively). Decay of the mRNA fragments was

studied by treatment of growing cells with rifampicin prior qRT-PCR. The 400-bp fragment of

purB was faster degraded than the 57-bp fragments of cysB and purB. The 400-bp fragment of

purB was further used to enumerate viable cells in aggregate state. Cell counts were more than 2

log10 higher using the qRT-PCR method compared to plate count.

Growing interest in probiotic characteristics of aggregating bacteria cells make this technique a

valuable tool to accurately quantify viable probiotic bacteria exhibiting heterogeneous

morphology.

Chapter 3 - qRT-PCR

- 74 -

3.2 Introduction

Probiotics are defined as “Live microorganisms which when administered in adequate amounts

confer a health benefit on the host” (FAO/WHO 2001). It is generally accepted that probiotic

food products should contain a minimal level of viable cells of 106 per gram or milliliter of

product (Ouwehand et al., 1998), although this value is relative since beneficial effect depends on

the strain and targeted health benefit (Lacroix and Yildirim, 2007). Viability of probiotic bacteria

is traditionally assessed by plate counting which has several limitations, including

underestimation of cells in aggregates or chains morphology. Auto-aggregation trait of

Bifidobacterium longum strains isolated from human gastric juice and intestine has been

associated with their ability to adhere to epithelial cells (Del Re et al., 2000) which is considered

as an important functionality of probiotic bacteria for some beneficial health effects such as to

prevent attachment of pathogens to gut mucosa (Resta-Lenert and Barrett, 2003; Cesena et al.,

2001; Castagliuolo et al., 2005).

However, bacterial cell aggregates lead to underestimation of cell counts with traditional, culture-

dependent methods (Davey et al., 1977; Auty et al., 2001). Enumeration techniques with

fluorescent stains targeting bacterial membrane potentials or enzymatic activities were proposed

as alternative methods to plate counts (Breeuwer and Abee, 2000; Hewitt and Nebe-Von-Caron,

2001; Alakomi et al., 2005). The accuracy of these methods using fluorescence probes, generally

combined with flow cytometry can be affected by clustering and chaining of cells (Wallner et al.,

1995; Nebe-von-Caron et al., 2000; Bibiloni et al., 2001).

The use of real-time PCR techniques to quantify bifidobacteria and lactic acid bacteria by

specific DNA or rRNA targeting showed good correlation with plate counts (Masco et al., 2007;

Friedrich and Lenke, 2006; Matsuda et al., 2007). However, their use as accurate viability

Chapter 3 - qRT-PCR

- 75 -

markers is limited since rRNA and DNA can be detected hours or even days after cell death

(TolkerNielsen et al., 1997; Sheridan et al., 1998; Hellyer et al., 1999; Aellen et al., 2006; Keer

and Birch, 2003; Lahtinen et al., 2008). Molecules of mRNA in contrast have short half-life

times ranging between 0.5 and 50 min (Belasco et al., 1986; Hellyer et al., 1999; Takayama and

Kjelleberg, 2000) and are therefore a more suitable viability marker for bacterial cells. Yeasts and

pathogenic bacteria were quantified using mRNA in combination with quantitative reverse

transcription PCR (qRT-PCR) (Bleve et al., 2003; Jacobsen and Holben, 2007; Coutard et al.,

2007; Gonzalez-Escalona et al., 2009). The use of instable mRNA as target molecule however

requires special attention to minimize losses during RNA extraction and purification (Bustin

2002, Johnson et al., 2005).

In this study, we describe the development of a qRT-PCR-based method for the quantification of

viable B. longum NCC2705 cells. Serial dilutions of bacterial culture aliquots were used to

generate qRT-PCR calibration curves targeting three mRNA fragments (two short amplicons of

57-bp of cysS or purB and one of 400-bp of purB) and plotted against plate counts. Spiked

internal reference mRNA was used to monitor mRNA loss during sample processing and thus to

provide accurate mRNA comparison across different experimental conditions (Johnson et al.,

2005). Stability of the three mRNA fragments was tested by comparing cell counts obtained by

qRT-PCR and plate counts of culture exposed to lethal heat stress. Instability of the different

mRNA fragments was also assessed by measuring mRNA decay after rifampicin treatment. The

mRNA fragment selected was then used to estimate viable cell numbers in culture samples

containing macroscopic cell aggregates.

Chapter 3 - qRT-PCR

- 76 -

3.3 Materials and Methods

3.3.1 Strain and growth conditions

Bifidobacterium longum NCC2705 (Nestlé Culture Collection, Nestlé Research Center,

Lausanne, Switzerland) was grown at 37 °C in rubber-lid flasks containing 100 ml MRS broth

(de Man et al., 1960) (BioLife, Milan, Italy) supplemented with 0.5 g/l L-cysteine (Sigma-

Aldrich, Buchs, Switzerland) (MRSC) and inoculated at 2 % (v/v) with an overnight MRSC pre-

culture. The headspace of flasks was flushed with CO2 to ensure anaerobic conditions. Colony

forming units (CFU) were determined by plating serial dilutions of culture aliquots in duplicate

on MRSC agar plates 1.5 % (w/v) (Becton Dickinson, Basel, Switzerland) and incubating at

37 °C for 48 h in anaerobic jars containing atmosphere generation system packs (AnaeroGen,

Oxoid, Pratteln, Switzerland).

3.3.2 Preparation of bacterial samples for calibrat ion curves

Cultures were grown at 37 °C in a 2000 ml reactor (Bioengineering, Wald, Switzerland)

containing 1800 ml MRSC inoculated at 0.5% (v/v) with pre-culture. pH was maintained at 6.0

by addition of NaOH (5 M), agitation rate was set at 200 rpm using a flat blade impeller and

anaerobic conditions were maintained by flushing CO2 in the head space of the reactor. Cells

were harvested in end-exponential growth phase and centrifuged at 6,000 × g for 2 min. Pellets

were suspended in an equal volume of a cryoprotectant solution composed of MRSC

supplemented with glycerol 10% (w/v) (Fluka, Buchs, Switzerland), shock frozen in liquid

nitrogen and stored at -80 °C before RNA extraction and plate counts. Standards with cell

concentrations ranging from about 3.2 × 104 to 4.7 × 108 CFU/ml determined by plate counts

were prepared by serial dilution in MRSC before RNA extraction.

Chapter 3 - qRT-PCR

- 77 -

3.3.3 Production of aggregated cells

B. longum NCC2705 cells aggregates were produced during continuous culture with immobilized

cells grown in a two-stage reactor system as described by Doleyres et al. (2004b), with some

modifications. Briefly, the first reactor (R1) was a 500 ml glass reactor (Sixfors, Infors-HT,

Bottmingen, Switzerland) with a working volume of 240 ml, containing 80 ml of gel beads pre-

colonized with B. longum NCC2705 as described by Cinquin et al. (2004) and stirred between

150 – 250 rpm. The second 2000 ml reactor (Bioengineering) was run in series at a volume of

1800 ml, agitated at 250 rpm and inoculated with cells produced in R1. CO2 was injected in the

headspace of the two reactors to maintain anaerobic conditions during culture. The continuous

system was fed with MRSC medium supplemented with 1.5 g/l CaCl2 (Sigma-Aldrich) and 40

g/l glucose at a flow rate of 360 ml/h. Continuous fermentation was run for 20 days at 37 °C and

pH was maintained at pH 6.0 by addition of NaOH (5 M) in both reactors. Culture samples from

the effluent of both reactors were periodically collected and processed as described above.

3.3.4 Heat stress

A 100-ml volume of MRSC was inoculated at 1% (v/v) with B. longum NCC2705 and incubated

in a rubber-lid flask at 37 °C for 16 h. Cells were collected by centrifugation at 6,000 × g for 2

min. Supernatant was discarded and pellet suspended in the same volume of fresh MRSC. Cell

suspensions were split in aliquots of 1 ml and immediately added to 9 ml MRSC either pre-

warmed at 56 °C for 10, 20 and 30 min or kept at room temperature. After heating, cells were

immediately cooled to room temperature in a waterbath for 3 min, collected by centrifugation,

washed and suspended in the initial volume of MRSC for plate counting. 0.5 ml of the cell

Chapter 3 - qRT-PCR

- 78 -

suspension was added to prepared screw cap tubes as described below and immediately frozen in

liquid nitrogen for RNA extraction. Experiments were performed in triplicate.

3.3.5 Rifampicin test

MRSC (100 ml) in a rubber-lid flask was inoculated at 5% (v/v) with preculture and incubated at

37 °C until the culture reached mid-exponential phase (OD600 0.5 – 0.8). Rifampicin (Sigma-

Aldrich) was then added to a final concentration of 400 µg/ml. Samples were taken 2.5, 7.5, 14,

20, 25 and 45 min after rifampicin-addition. Cells were collected by centrifugation, washed in

fresh MRSC and suspended in the initial volume of MRSC for RNA extraction. The half-life time

(T½) of an mRNA was determined from the degradation rate constant (k) corresponding to the

slope of a semi-logarithmic plot of mRNA amount as a function of time with the relation: T½ =

ln2/k. Experiments were performed in duplicate.

3.3.6 RNA isolation and reverse transcription react ion

RNA extraction was performed according to Stevens et al. (2008). Briefly, 0.5 ml of fresh or

stored cell samples was centrifuged and the pellet was washed once in MRSC, transferred to 2-ml

screw-cap tubes containing a mix of 500 mg glass beads of 100 µm diameter (Sartorius,

Goettingen, Germany), 500 µl of a 1:1 solution of phenol/chloroform (Sigma-Aldrich), 30 µl of

10% (w/v) sodium dodecyl sulfate (Sigma-Aldrich) and 30 µl of 3 M sodium acetate (pH 5.2) and

immediately frozen in liquid nitrogen. Frozen cells were then disrupted by bead-beating in a

homogenizer (FastPrep FP 120, Q-Biogene, Carlsbad CA, USA), four times for 40 sec at speed

4.0 intercalated with cooling periods on ice. Afterwards, the mixtures were centrifuged (21,000 ×

g, 3 min at 4 °C) to remove glass beads and cell debris. The aqueous phase was transferred to a

Chapter 3 - qRT-PCR

- 79 -

new tube and traces of phenol were removed by extraction with pre-chilled chloroform. The

resulting aqueous phase was mixed with an equal volume of 70%-ethanol and applied to RNA

purification column using RNeasy Mini Kit (Qiagen, Basel, Switzerland). Further purification

was performed according to the manufacturer’s protocol including on-column DNase I digestion

for 30 min followed by elution in 100 µl H2O. Total RNA quantity and purity was determined

using Nanodrop ND-1000 (Peqlab, Erlangen, Germany).

Reverse transcription (RT) was performed using the high capacity cDNA reverse transcription

commercial kit (Applied Biosystems, Rotkreuz, Switzerland) according to manufacturer’s

protocol. A maximum amount of 100 ng of purified total RNA was applied per RT reaction.

Reaction was carried out in a thermocycler (Biometra T-personal, Göttingen, Germany) using the

following cycling conditions: 10 min at 25 °C, 120 min at 37 °C, 5 sec at 85 °C followed by a

cooling step to 4 °C.

3.3.7 Determination of RNA recovery rate using inte rnal reference

Luciferase mRNA (Promega, Dübendorf, Switzerland) was used as internal reference to estimate

loss of RNA during extraction and purification procedures. 2.03 × 108 copies of luciferase mRNA

were added to the aqueous phase, after centrifugation of lysed cells. Copy numbers of luciferase

mRNA added per volume were calculated using transcript sizes of 1.8 kb and an average

molecular mass of 330 Da/nucleotide. The number of luciferase mRNA copies recovered from

spiked samples after RNA extraction was determined with qRT-PCR and standard curves

prepared from serial dilutions of luciferase mRNA. Recovery rate was evaluated by dividing the

measured level by the initial amount of spiked reference mRNA and used as normalization factor

to correct for loss of mRNA during the RNA isolation procedure.

Chapter 3 - qRT-PCR

- 80 -

3.3.8 Quantitative PCR amplification

Primers used for quantitative PCR amplification (Table 7) were designed with PrimerExpress

software Version 3.0 (Applied Biosystems). Genes cysS and purB were selected for their

constitutive expression under different culture conditions (Rezzonico et al., 2007). Quantitative

PCR was performed with an ABI PRISM 7700 equipment using SYBR® Green PCR Master Mix

(Applied Biosystems) according to the manufacturer’s protocol. Each 25 µl PCR reaction

contained 12.5 µl of qPCR master mix, 5 µl of template from the RT-reaction, and 200 nM

forward and reverse primers. PCR cycling conditions were as follows: 2 min at 50 °C, 10 min at

95 °C and 40 cycles of 15 sec at 95 °C and 1 min at 60 °C. Measurement of SYBR green

fluorescence was performed at the end of each amplification step and continuously during the

melt-curve analysis to control for amplicon specificity. The qPCR results and melting curve were

analyzed using the SDS 2.1 Software (Applied Biosystems). Control reactions from non-reverse

transcribed samples were included to check for residual genomic DNA in the RNA samples.

Difference between cycle threshold of reverse transcribed and non reverse transcribed samples

was typically higher than 3 cycles indicating that residual DNA accounted for less than 10% in

the final result.

Efficiency of PCR amplification was measured using serial dilutions of cDNA from luciferase

mRNA or total bacterial RNA after plotting the cycle threshold values versus the log10-

transformed number of copies of luciferase mRNA or cells per reaction. Efficiency of PCR

amplification was typically in the range of 94 to 99%.

Chapter 3 - qRT-PCR

- 81 -

3.3.9 Statistical analyzes

Means of log10-transformed cell counts determined by plate counts and qRT-PCR of heat- and

rifampicin tests were compared with Student’s t-test using JMP In Software version 6.0 (SAS

Institute Inc., Cary, NC, USA) with a level of significance at 0.05.

Table 8 Primers used in qRT-PCR to quantify mRNA fragments of cysS and purB

transcripts

Gene or target (locus)

Primer sequence (5’ to 3’) Amplicon length

Transcript length (location of fragment)

Reference

cysS (BL0301) 5’-CAACCGCCGCGATCTTC-3’ 5’-CCAGCTGTGAAAGCAACGTATT-3 57 bp 1641 (1271 – 1327) bp

Rezzonico et al., 2007

purB (BL1800) 5’-CATGGGCGGCCTTGAGT-3’ 5’-TCAAGCTCACGCTCGATGAC-3’ 57 bp 1446 (254 – 310) bp

Rezzonico et al., 2007

purB (BL1800) 5’-AGCAGCGGCATATCCTTGAA-3’ 5’-TTCTGGCCAACGGCTTTG-3’ 400 bp 1446 (893 – 1292) bp This study

Luciferasea 5’-TACAACACCCCAACATCTTCGA-3’ 5’-GGAAGTTCACCGGCGTCAT-3’ 67 bp 1751 (257 – 323) bp

Johnson et al., 2005

a reference mRNA

3.4 Results

3.4.1 Morphology of B. longum NCC2705

Light microscopy of B. longum NCC2705 cells from batch cultures exhibited rods in various

shapes with the typical V- or Y-pattern. Cells appeared in single cell state and few cell clusters of

small numbers were observed (Figure 10 A). Cells originating from continuous cultures with

immobilized cells displayed similar shapes but aggregation increased with fermentation time

from small aggregates at day 4 (Figure 10 B) to macroscopic aggregates of uncountable cells with

a diameter of 0.5 – 1 mm after 9 days of fermentation and remained stable for 20 days of

continuous culture (Figure 10 C). Calibration curves for qRT-PCR and stability tests were carried

out with free cells from batch cultures.

Chapter 3 - qRT-PCR

- 82 -

Figure 10 Optical micrographs of fresh cells collected from fermented broth (1000 ×

magnification) of B. longum NCC2705 produced by batch culture (A), and continuous

immobilized cell culture at day 4 (before macroscopic aggregation) (B) and day 16 (after

macroscopic aggregation) (C).

3.4.2 Calibration curves

Calibration curves of cysS 57-, purB 57- amd purB 400-bp mRNA fragments were generated by

serially diluting cell samples with concentrations ranging from 7.0 × 103 to 4.7 × 108 CFU/ml,

corresponding to 9 to 1.7 × 105 cells per qPCR reaction volume (25 µl). The internal reference of

luciferase mRNA was used to account for the variability induced by the RNA extraction

procedure with RNA recovery rates ranging between 6 and 70 %.

CT values for the three amplicons were plotted as a function of the log10-transformed CFU/ml

determined by plate counting multiplied by the recovery rates of spiked luciferase mRNA (Figure

11). Linear amplification of target mRNAs was measured over a range between 4 and 5.5 log10

values, with detection limits of 9 and 20 cells/reaction volumes for short amplicon length of purB

and cysS, respectively, and of 40 cells/reaction volumes for purB 400-bp fragments.

Chapter 3 - qRT-PCR

- 83 -

Figure 11 Calibration curves obtained by plotting CT values for cysS 57-bp (A), purB 57-bp

(B) and purB 400-bp (C) primer sets as a function of log10 transformed cell concentrations

determined by plate counts. Data points are from three independent repetitions.

Chapter 3 - qRT-PCR

- 84 -

3.4.3 Heat test

Cells were exposed to lethal heat stress and bacterial survival assessed with plate counts was

compared with survival obtained with qRT-PCR method. Plate counts showed a reduction of

viable cells of 2.7 ± 0.2, 3.2 ± 0.1 and 3.7 ± 0.2 log10 CFU/ml after 10, 20 and 30 min exposure at

56 °C, respectively (Figure 12). Cell numbers after heat stress determined by qRT-PCR targeting

the short amplicons of cysS and purB mRNAs were significantly higher by an average factor of

18.0 ± 3.3 and 7.0 ± 2.0, respectively, compared to plate counts (p ≤ 0.05). In contrast, cell

numbers measured with qRT-PCR and purB 400-bp fragments were the same as plate counts for

all three tested heat stress periods (p > 0.05).

Figure 12 Viable cell counts of B. longum NCC2705 cells measured with qRT-PCR using 3

primer sets and plate counts after heat treatment at 56 °C for 10 (dashed), 20 (grey), 30 min

(white) compared to untreated cells (black). Standard deviations of each column are represented

by error bars. Different letters indicate significantly different values for each treatment (p ≤ 0.05).

Chapter 3 - qRT-PCR

- 85 -

3.4.4 Rifampicin test

Cells were grown in batch mode until cultures reached OD600 0.5 - 0.8 at which point rifampicin

was added to stop transcription The decay of transcripts encoding fragments of cysS and purB of

B. longum growing at 37 °C was followed at 2.5, 7.5, 14, 20, 25 and 45 min after arrest of

transcription (Figure 13).

Figure 13 Decay of mRNA fragments of cysS 57-bp (▲), purB 57-bp (■) and purB 400-bp (●)

as a function of time after addition of rifampicin at 400 µg/ml to exponentially growing cells.

Values are given as corresponding viable cell concentrations derived from calibration curves.

Standard deviations of each point are represented by error bars.

The levels of cysS and purB fragments in treated samples were translated into viable cell

concentrations using calibration curves. Amplicon levels for cysS 57- and purB 57-bp decreased

rapidly within the first 2.5 min after addition of rifampicin (log10 loss 1.19 ± 0.14, T½ 38.4 ± 4.4

sec) followed by a slower decline until 15 min (1.61 ± 0.25). In comparison, the levels of purB

400-bp amplicons decreased significantly faster (p ≤ 0.05) (log10 loss 1.47 ± 0.03; T½ 30.7 ± 0.7

Chapter 3 - qRT-PCR

- 86 -

sec) and remained at a lower level until 15 min after addition of rifampicin (log10 loss 1.86 ±

0.25). After 15 min, the levels of all three fragments increased steadily. The long fragment of

purB reached similar level as its shorter counterpart 20 min after addition of rifampicin.

3.4.5 Viable cells quantification of cell aggregate s

The qRT-PCR method targeting purB 400-bp fragments was used to quantify viable B. longum

NCC2705 cells of samples containing macroscopic cell aggregate. Formation of macroscopic cell

aggregates (0.5 – 1 mm) was observed after ca. 9 day continuous culture and remained until the

end of fermentation (Figure 10 C). Viable cell numbers obtained by measuring purB 400-bp

fragment level were compared to plate counts method from samples before and after extensive

cell aggregation. Plate counts from samples taken before aggregation (day 4) were lower by 0.5

log10 (approximately 3.2 fold) compared to cell numbers determined with qRT-PCR (Table 8).

Samples taken after formation of large aggregates at day 12, 16, and 20 showed larger differences

ranging between 2.1, 1.3 and 1.2 log10 units, respectively.

Table 9 Cell concentrations of B. longum NCC2705 during continuous immobilized cell

culture determined by plate counts and qRT-PCR (purB 400-bp). Values from day 4 are before

formation of macroscopic cell aggregates. Reported values are means of two independent

repetitions ± standard deviation.

Fermentation day

Cell count (plate count)a [CFU/ml]

Cell count (qRT-PCR)a [Cells/ml]

Ratio qRT-PCR/plate counts

4a 3.3 ± 0.3 × 108 1.1 ± 0.4 × 109 3.2 ± 1.0

12 6.6 ± 1.9 × 107 8.6 ± 4.4 × 109 124.8 ± 26.1

16 1.8 ± 0.7 × 108 4.3 ± 2.2 × 109 22.5 ± 5.7

20 5.0 ± 2.3 × 108 7.4 ± 3.3 × 109 14.7± 0.2 a before formation of macroscopic aggregates

Chapter 3 - qRT-PCR

- 87 -

3.5 Discussion

Levels of mRNA fragments of housekeeping genes cysS and purB were used for quantification of

viable cells of B. longum NCC2705 in preparations displaying heterogeneous morphology. These

two genes are constitutively expressed in various conditions and used to normalize gene

expression analyzes (Rezzonico et al., 2007). The use of unstable mRNA as viability marker is

difficult and strongly dependent on the efficiency of RNA extraction because loss during sample

preparation significantly biases the output of the experiment (Bustin, 2000). Addition of an

exogenous reference mRNA circumvents this bias by monitoring the losses during sample

processing. The importance of normalizing mRNA levels with the recovery of exogenous

references is highlighted by the large variation in extraction yields ranging from 6 – 70 %. High

regression coefficients were obtained for all three mRNA fragments using normalized data with

recovery rates. A mean detection limit of 2.7 ± 1.7 × 104 cells per ml of culture was achieved for

all three mRNA used. This low level is comparable to that reported for bifidobacteria quantified

in fecal samples using real-time PCR targeting DNA (Gueimonde et al., 2004). A detection limit

of 5 × 104 cells of Salmonella per gram of seeded soil or chicken manure can be achieved using

real-time PCR method targeting mRNA extracted with magnetic capture hybridization (Jacobsen

and Holben, 2007).

The short mRNA fragments of both housekeeping genes were too stable to be used as accurate

viability markers and yielded an overestimation of viable cell counts after the three heat

treatments when compared to standard plate counts. In contrast, results obtained with long

fragments of purB coincided well with plate counts irrespective the duration of the stress

treatment. Differential stability of specific mRNA fragments was confirmed by treating B.

longum NCC2705 with rifampicin, an inhibitor of RNA-transcription (Campbell et al., 2001).

Chapter 3 - qRT-PCR

- 88 -

The 400-bp fragments of purB were degraded faster than short fragments of both genes. Similar

inverse relation between length of RNA or DNA molecules and their stability was recently

reported in other bacterial species (Aellen et al., 2006, Soejima et al., 2008). Quantification of

rRNA levels after antibiotic treatment in Streptoccus gordonii showed better correlation between

qRT-PCR values and plate counts when targeting a 427-bp compared to a 119-bp fragment which

remained stable after drug killing (Aellen et al., 2006). DNA was used as a viability marker of

Listeria monocytogenes in combination with a DNase-assay and showed that long DNA

fragments (894-bp) of heat killed cells were not detectable after treatment, whereas short

fragments (113-bp) were still evident causing false positive results (Soejima et al., 2008).

The rapid decrease of levels of the three fragments in the first minutes after addition of rifampicin

yield very small half-life times (33.6 ± 3.4 sec) compared to that reported for Escherichia coli,

Bacilllus subtilis and Lactococcus lactis of 3-8 min (Bernstein et al., 2002; Hambraeus et al.,

2003; Redon et al., 2005). However, mRNA half-lives in these studies were calculated from the

degradation rate constant obtained from microarray analyzes. The use of sensitive qRT-PCR

technique can affect the determination of RNA levels compared to microarray results. Decay

rates of transcripts determined with microarray technology were shown to be slower than values

obtained by Northern blots or real time PCR (Hambraeus et al., 2003; Hundt et al., 2007).

The increase of transcript levels shortly after the start of the treatment can be attributed to rapid

bacterial adaptation to rifampicin or selection for resistant subcultures within the same culture as

reported for bifidobacteria from human samples treated with this antibiotic (Mangin et al., 1994).

Proliferating cells regenerate new intact mRNA transcripts resulting in equal levels of de novo

synthesized transcripts of short and long purB fragments as observed 20 min after beginning of

treatment.

Chapter 3 - qRT-PCR

- 89 -

Although auto-aggregation might improve adhesion of probiotic bacteria to epithelial cells (Del

Re et al., 2000), application of aggregated cells in functional food products suffers from

underestimation of viable cell counts obtained with standard plate count since the method cannot

detect individual cells in aggregates. The use of cell aggregates is thus still unfavorable in

product applications with plate counts being the standard method to verify product quality. In

fact, the use of the qRT-PCR method with aggregated cells showed that plate counts largely

underestimated viable cell numbers with factors ranging from 15 to 125.

In conclusion, we developed an accurate real time PCR method determining viable cell counts of

B. longum NCC2705 with different morphologies. The careful selection of the target gene and

fragments length as viability markers is essential to avoid biases in viable cell counting.

Comparison of primer sequences of purB 400-bp with sequenced genomes showed direct hits

with genomes of B. longum DJO10A and B. longum subsp. infantis ATCC 55813, CCUG 52486

and ATCC 15697, an auto-aggregating strain (Rahman et al., 2008) and could therefore be used

as viability marker in these strains. qRT-PCR may be used for rapid quantification of viable cells

in mixed cultures provided that primer specificity and sample processing is not affected by mixed

culture conditions. Moreover, the method may be developed in combination with DNA

quantification to evaluate the proportions of dead cells within cell aggregates which may provide

a protection layer for viable cells inside the core of aggregates during transit through the GIT.

Growing interest in probiotic characteristics of aggregating bacterial cells make this technique a

valuable tool to accurately quantify viable probiotic bacteria exhibiting heterogeneous

morphology.

Chapter 4 - Spray drying

- 90 -

4 Development of a rapid screening protocol for

selection of Bifidobacterium longum strains resistant

to spray drying and powder storage

Data presented in this chapter were submitted and accepted with minor revision in 2009 in

Beneficial Microbes with the following authors: Reimann S., Grattepanche F., Baggenstos C.,

Rezzonico E., Berger B., and Lacroix C.


- 91 -

4.1 Abstract

An efficient screening method for selection of Bifidobacterium longum strains resistant to spray

drying and storage was developed based on Randomly Amplified Polymorphic DNA (RAPD) for

identification of the best survivors in mixed strains bacterial preparations. Three different primers

were used to generate RAPD profiles of 22 B. longum strains. All strains were distinguished

according to their RAPD profiles except for the strain NCC2705 and its H2O2 resistant derivative

variant. The 22 strains were grouped in 3 batches of 7, 7 and 8 strains and subjected to spray

drying and storage at 30 and 37 °C under anaerobic conditions. Batch survival rates after spray

drying reached 17.1 ± 4.4 %. Strains showing the highest prevalence and/or resistance to storage

at 37 °C were selected from individual batches for subsequent spray drying and storage testing.

After 67 days of storage, NCC572 was identified as dominant strain in powder. The stability of

strain NCC572 was confirmed by performing single spray drying and storage tests. Out of 22 B.

longum strains, a robust strain was identified by combining RAPD with a simultaneous screening

test for survival to spray drying and storage. The method allowed a fast screening of B. longum

strains in mixture for resistance to spray drying and storage compared to traditional screening

procedures carried out with individual strains, in the same conditions. This approach could be

applied to other stress conditions.


- 92 -

4.2 Introduction

Bifidobacteria are widely used as probiotics in functional food products. Industrial standards

currently used for products containing viable probiotic strains require presence of minimal

bacterial viable counts in order to confer health benefits to consumers. Probiotic bacteria in

functional food products are usually incorporated as frozen concentrates or freeze dried powders

(Meng et al., 2008; Saarela et al., 2006). Freezing and freeze drying are however expensive

stabilisation technologies due to the infrastructure and energy consumption during production,

transport and storage (Knorr, 1998). Spray drying has gained popularity over the last decade for

large scale production of starter cultures and probiotics (Corcoran et al., 2004; Desmond et al.,

2002; Gardiner et al., 2000; Lian et al., 2002; Meng et al., 2008; O’Riordan et al., 2001; Simpson

et al., 2005). Although spray drying is an economical process, between 5 – 10 times cheaper than

freeze drying (Santivarangkna et al., 2007), bacterial cells suffer lethal damages attributed to the

effect of exposure to heat and dehydration (Ross et al., 2005; Teixeira et al., 1995a). These have

deleterious effects on survival of sensitive bacteria such as probiotics. Survival rates of probiotic

bacteria during spray drying is strain-dependent and influenced by several factors, the most

important being the drying conditions applied, in particular outlet temperature (Gardiner et al.,

2000; Lian et al., 2002), the carrier media used (Corcoran et al., 2004; Johnson and Etzel, 1993;

Lian et al., 2002;), presence of protective compounds and the growth phase of cells (Meng et al.,

2008).

Loss of bacterial viability during storage of dry cultures is also a major limiting factor for many

cultures (Ross et al., 2005; Texeira et al., 1995). Parameters such as temperature, moisture

content of spray dried powders, water activity, carrier media, exposure to light and oxygen have

been shown to impact bacterial survival during storage in powder form (Ananta et al., 2005;


- 93 -

Chàvez and Ledeboer, 2007; Gardiner et al., 2000; Meng et al., 2008; Simpson et al., 2005;

Teixeira et al., 1996). Spray drying and storage survival rates do not concur and vary from strain

to strain (Corcoran et al., 2004; Gardiner et al., 2000; Simpson et al., 2005) and therefore careful

strain selection is important to ensure high survival during spray drying and subsequent storage

of dried cultures. The underlying molecular mechanisms responsible for viable cell losses during

drying and storage remain unclear and thus empirical approaches for efficient strain screening are

of utmost value to develop probiotic strains with good technological and functional properties

and to avoid labour-intensive, lengthy and costly strain-by-strain screening procedures.

Molecular typing techniques such as restriction fragment length polymorphism, pulse field gel

electrophoresis, repetitive element sequence-based PCR fingerprinting and randomly amplified

polymorphic DNA (RAPD) allow the differentiation of bacteria at strain level without prior

knowledge of genome sequence (Ben Amor et al., 2007; Collado et al., 2006; Gevers et al., 2001;

Masco et al., 2003; Rossetti and Giraffa, 2005; Vincent et al., 1998).

In this study, we applied an efficient method using RAPD-based strain identification to identify

the best survivor strains after spray drying and storage of mixed bacterial preparations during two

selection rounds followed by a final confirmation step (Figure 14). Twenty two B. longum were

divided into three batches of 7 or 8 strains and exposed to a first selection consisting of a spray

drying and storage experiment. Best survivors of each batch preparation were identified by

RAPD after 4 log10 viability loss during storage and applied to a second selection round to

identify the most resistant B. longum strain. The survival of the selected strain was tested

separately and compared to a low resistant strain for validation.


- 94 -

Figure 14 Flow chart of experimental procedure. The 22 B. longum strains were divided into

3 batches for spray drying and storage. After the first selection round, 2 - 3 best survivor strains

were selected for a subsequent spray drying and storage batch. After the second selection round,

the best survivor strain and a control strain were selected for a single strain spray drying and

storage confirmation batch.

4.3 Material and Methods

4.3.1 Bacterial strains and culture conditions

The 22 B. longum strains used in this study were provided by Nestlé Culture Collection (NRC,

Lausanne, Switzerland) (Table 10). Strains were routinely cultured in 20 ml-McCartney bottles

containing de Man, Rogosa, Sharpe (MRS) broth (BioLife, Milan, Italy) supplemented with 0.05

% (w/v) L-cysteine (Sigma-Aldrich, Buchs, Switzerland) (MRSC), and incubated for 16 h at 37

22 B. longum strains

First selection round

Second selection round

Single strainconfirmation batches

8 strains

1 strain 1 strain

22 B. longum strains

First selection round

Second selection round

Single strainconfirmation batches

8 strains

1 strain 1 strain


- 95 -

°C in anaerobic conditions using atmosphere generation system packs (AnaeroGen, Oxoid,

Pratteln, Switzerland). Cultures from frozen stock were sub-cultured twice before use as

inoculum at 1% (v/v). Cell counts of pure liquid cultures were determined by serial dilutions of

samples in 12 mM phosphate buffer saline with 0.05 % (w/v) L-cysteine (PBSC) followed by

plating in duplicate on MRSC agar 1.5% (w/v) (Becton Dickinson, Basel, Switzerland). Plates

were incubated anaerobically at 37 °C for 72 hours.

4.3.2 Preparation of bacterial suspensions used for survival tests

Individual strains were grown separately at 37 °C in rubber-lid flasks containing 250 ml MRSC

inoculated at 2 % (v/v). The headspace of flasks was flushed with CO2 to ensure anaerobic

conditions. Growth was monitored by measuring OD600 and cells were harvested 3 h after

reaching stationary growth phase. Volumes of 40 ml of cell broth were centrifuged (15,000 × g,

15 min at 4 °C), and cell pellets washed once with PBS-C and resuspended in 100 ml ice-cold 20

% (w/v) reconstituted skim milk (RSM) (Oxoid Ltd, Hampshire, United Kingdom). A sample of

1 ml was used for plate counts.

4.3.3 Preparation of strain mix

For the first and second selection batches, 100 ml of each bacterial suspension (7 or 8 strains per

batch) in RSM were mixed together in a 1 l-bottle containing a magnetic stirrer (Table 10). The

solution was kept at 4 °C in an ice bath and stirred for at least one hour before spray drying. For

single strain confirmation tests, a similar procedure was used with pure cultures to produce 400

ml ice-cold RSM cell suspension used for spray drying.


- 96 -

4.3.4 Spray drying

Spray drying was conducted using a laboratory scale spray dryer (Büchi model B-290, Büchi,

Flawil, Switzerland). The feed suspension was pneumatically atomized into a vertical, co-current

drying chamber using a two-fluid nozzle. The inlet air temperature was maintained at 170 °C and

the outlet air temperature was adjusted to 78.2 ± 1.9 °C by levelling the feed pump rate. A

maximum volume of 300 ml of bacterial RSM mix was used per batch corresponding to a

maximum of 3 h operation time. The spray dried powder was collected in a single cyclone

separator.

Spray drying was performed once for the first selection batches whereas second selection batch

and confirmation experiments using single strain preparations were carried out in duplicate.

4.3.5 Storage test

Spray dried powder was aliquoted in 2-ml polypropylene tubes (Eppendorf, Hamburg, Germany)

sealed with sterile cotton wool. Tubes were then placed in 2.7 l desiccators (DURAN Group

GmbH, Mainz, Germany) containing 150 ml of 5 M saturated potassium acetate as desiccant and

kept in the dark. Storage temperatures of 30 and 37 °C were tested for first selection batches

whereas powders from second selection and confirmation batches were stored at 37 °C only.

Desiccators were evacuated and then flushed with nitrogen to ensure anaerobic conditions.

Powders from selection batches were sampled 2-3 times per week or once a week towards the

end of storage or in confirmation batches for cell counts and water activity measurement. Storage

was stopped when a threshold loss in cell counts of ca. 4 log10 of number of colony forming units

(CFU) per gram of powder was reached. Storage experiments were performed once per spray


- 97 -

drying batch for the first selection batches and in duplicate per spray drying experiment for the

second selection batches.

4.3.6 Determination of water activity and moisture content

Water activity of powder was measured using a hygrometer (AquaLab, Gerber Instruments SA,

Effretikon, Switzerland) according to the manufacturer’s recommendations. Values are means of

two determinations per sample taken at a reference temperature of 25 °C.

Moisture content expressed as percentage of spray dried powder weight was determined by

measuring the difference of powder weight before and after heating 10 g of powder overnight at

102 °C. Dry matter of bacterial suspension in RSM was determined by weighing 10 ml of

suspension oven-dried for 36 hours at 80 °C. Determination of moisture content and dry matter

was performed in duplicate.

4.3.7 Viable cell counts in spray dried powders

Survival of B. longum strains to spray drying and storage was performed according to Gardiner et

al. (2000). Briefly, 0.1 g powder was resuspended in 9.9 ml PBS-C, homogenized using a

stomacher 80 (Müller and Krempel AG, Bülach, Switzerland) for 30s and stored at room

temperature for 1 hour to allow rehydration. Cell suspensions were serially diluted in PBSC and

appropriate dilutions were plated in duplicate on MRS-C agar. Percentage of survival after spray

drying was calculated as the CFU per gram powder after spray drying divided by the initial

number of CFU per gram dry matter in RSM bacterial suspensions before spray drying.


- 98 -

4.3.8 Identification of best surviving strains usin g RAPD

Genomic DNA was extracted from pure cultures grown in liquid media according to

Goldenberger et al. (1997). Briefly, 0.5 ml of culture were centrifuged and resuspended in 0.2 ml

digestion buffer containing 50 mM Tris-base (Sigma-Aldrich), 1 mM EDTA (pH 8.5) (Sigma-

Aldrich), 0.5 % SDS (Sigma-Aldrich), 200 µg/ml proteinase K (AppliChem GmbH, Darmstadt,

Germany). After incubation at 55 °C for 3 h under agitation (Thermomixer, Eppendorf,

Hamburg, Germany), the mixture was held at 95 °C for 10 minutes to inactivate proteinase K,

cooled to 4°C and centrifuged (12,000 × g, 10 min, at 4 °C). The supernatant was collected and

nucleic acid concentration was determined using Nanodrop ND-1000 (Peqlab, Erlangen,

Germany). Samples were diluted in distilled water to reach a final nucleic acid concentration of

25 ng/µl and stored at – 20 °C until PCR amplification.

RAPD analysis was carried out according to Vincent et al. (1998), i.e., one µl of DNA sample

(25 ng/µl) was mixed with 24 µl mastermix containing 0.2 mM dNTP (GE Healthcare Europe,

Glatttbrugg, Switzerland), 0.2 pmol of each primer, 10 x Taq buffer (Euroclone, Siziano, Italy),

1.5 mM MgCl2 (Sigma-Aldrich), 2 % (w/v) Tween 20 (Sigma-Aldrich) and 2.5 U Taq

Polymerase (Sysmex Digitana, Horgen, Switzerland). Three random single primers were used:

OPA-02 (5’ TGCCGAGCTG 3’), OPL-07 (5’ AGGCGGGAAC 3’) and OPL-16 (5’

GGGAACGTGT 3’) (Microsynth, Balgach, Switzerland). These three primers were selected

because they permit differentiation of all 22 strains except for NCC2705 and its H2O2-resistant

variant NCC2916 (data not shown) which could be differentiated according to physiological

H2O2 resistance tests (Figure S1). These two strains were distributed in two different batches

during the first selection batch experiments to avoid identification problems (Table 1). PCR

amplification was run in a Biometra®TGradient thermocycler (BioLabo, Châtel-St. Denis,


- 99 -

Switzerland) under the following cycling conditions: 94 °C for 3 min, 45 cycles of 94 °C for 1

min, 30 °C for 1 min and 72 °C for 2 min. PCR products were run on 2 % (w/v) agarose gel

stained with ethidium bromide. The RAPD profiles were recorded using AlphaImager™ system

(Alpha Innotech Corporation, California, USA).

Survival of individual strains after spray drying and storage of dry powder containing mixed

strains was measured after reaching 4 log10 loss in total bifidobacterial counts. A total of at least

hundred colonies from 2 - 4 plates harboring between 5 and 100 colonies were then randomly

picked and separately cultivated in MRSC broth at 37 °C for 24 h under anaerobic conditions and

stored at -20 °C until DNA extraction.

4.4 Results

4.4.1 Strain survival and selection during the firs t selection batches

22 B. longum strains were divided in three batches each consisting of 7 - 8 strains (Figure 14) for

spray drying and storage experiments (Table 10). Mean initial cell concentration of cell

suspensions in RSM for the three batches before spray drying was 1.5 ± 0.4 × 109 CFU/g dry

matter. Viable total cell counts of spray dried powders measured immediately after spray drying

were 2.2, 3.6 and 1.8 × 108 CFU per gram of powder, corresponding to survival rates of 12.6,

21.4 and 17.4 % for Batch 1, 2 and 3, respectively. Average moisture content of powders was 5.4

± 0.7 %. Aliquots of dried powder were then stored in two desiccators at 30 and 37 °C with a

water activity of 0.23 ± 0.01.


- 100 -

Table 10 Distribution of the 22 B. longum strains in the three batches of the first selection

round and origin of the strains.

Straina Batch number Source NCC284 1 Infant feces NCC293 1 Infant feces

NCC 305 1 Infant feces





NCC 417 (ATCC 15707) 2 Fermented milk (yogurt) NCC 435 (ATCC 15708) 2 Adult feces

NCC 444 (NCIMB 8809) 2 Infant feces

NCC 450 (DSM 20097) 2 Infant feces

NCC 461 2 Calf feces


NCC 4012 2 Commercial probiotic product

NCC 490 3 Infant feces NCC 510 3 Infant feces

NCC 521 3 Adult feces




NCC224 3 Infant feces

NCC 2916 3 H2O2 resistant mutant of NCC2705 a Alternative strain denotation indicated in brackets. ATCC: American Type Culture Collection; DSM: Deutsche Sammlung von Mikroorganismen; NCIMB: National Collection of Industrial, Marine and Food Bacteria; NCC: Nestlé culture collection A non-linear decrease of cell counts (log10) was observed in spray dried powders during storage

at both temperatures (Figure 2). Powder storage at 37 °C resulted in faster viability loss with a 4

log10 loss threshold in cell counts reached after 46, 40 and 49 days at 37 °C (Figure 2 B)

compared to 64, 65 and 53 days at 30 °C (Figure 2 A) for Batch 1, 2 and 3, respectively. After

this storage period, at least 100 individual colonies isolated from different plates of the highest

plating dilutions were collected and identified using RAPD.


- 101 -

Figure 15 Cell counts in spray dried powders of batches from first selection round during

storage at 30 °C (A) and 37 °C (B) for Batch 1 (full line, ●), 2 (semi-dashed line, ▲) and 3

(dashed line, ■).


- 102 -

Proportion of individual strains found within the three batches before and after spray drying and

storage at 37 °C are shown in Figure 16. Different fractions were already measured for individual

strains before spray drying as a result of different growth of cells in the culture used for

preparation of the bacterial strain mixtures in RSM. After spray drying and storage of Batch 1 at

37 °C, all seven strains present in the initial cell suspensions were detected at the end of the

storage period. NCC3001 and NCC293, which were predominant before spray drying showed

only a small decrease of their relative proportion after storage, from 32 and 45 % to 30 and 40 %,

respectively (Figure 16 A). NCC284 was the only strain showing increased relative proportion

after storage (from 4 to 10 %). Only three of the seven strains present in the initial cell

suspension were detected after storage at 30 °C. The relative percentage of strain NCC3001

increased from 32 to 56 % whereas NCC293 remained stable at about 44 %. NCC284 was not

detected at the end of storage at 30 °C (Figure 16 A).

Similarly to Batch 1, strain diversity in Batch 2 and 3 was also better conserved after storage at

37 °C compared to 30 °C. In Batch 2, strains NCC417 and NCC435 were already over-

represented before spray drying and displayed the highest stability since their relative proportion

increased after storage at both 30 and 37 °C, from 29 to 36 and 50 % for NCC417 and from 23 to

30 and 43 % for NCC435, respectively (Figure 16 B). Surprisingly, NCC461 doubled its relative

presence from 13 to 29 % in the powder stored at 30 °C, but its fraction decreased to only 1 % at

37 °C. In Batch 3, strain NCC572 initially accounting for 22 % of total bacterial counts, largely

dominated after storage, reaching 95 and 98 % at 30 and 37 °C, respectively (Figure 16 C).


- 103 -

Figure 16 Relative presence (%) and corresponding cell counts (CFU per gram) of B.

longum strains in RSM before spray drying and in spray dried powders after end of storage at 30

and 37 °C for the first selection Batches 1 (A), 2 (B) and 3 (C). Storage was stopped after 4 log10

loss of CFU counts in powder (in brackets: storage time). For RSM, distribution-profiles were

calculated using cell concentration of each strain in pure culture before mixing. Individual

populations in spray dried powders were estimated using their relative abundance (%) determined

by RAPD and total cell counts in powder.


- 104 -

In order to reduce the number of strains for the second selection round, only 2 or 3 strains from

each batch were selected based on their high survival in powders stored at 37 °C. Strains with

highest percentage (NCC293) or increased relative percentage (NCC 284) were selected among

the 7 strains detected in the powder for Batch 1 (Figure 16 A). For Batch 2, among the 5 strains

detected in the powder, strains NCC417 and NCC435 were selected due to their high fraction in

the RSM suspension (indicating good growth characteristics in MRSC) and their increased

relative proportion after storage. NCC450 was also selected because its fraction in the RSM cell

suspension and in the powder after storage remained stable (Figure 16 B). For Batch 3, among 3

strains detected in the powder after storage at 37 °C strains NCC572 and NCC585 were chosen

due to their dominant appearance (NCC572) and stable percentage (NCC585) (Figure 16 C).

Strain NCC2916 which showed low survival in Batch 3, was chosen as a “negative control” for

inclusion in the second selection round.

4.4.2 Strain survival and selection during the seco nd selection round

Selected strains were spray dried and stored at 37 °C. The moisture content and water activity in

spray dried powders from the second selection round with 8 strains were 5.4 ± 0.2 % and 0.21 ±

0.01, respectively. Mean initial cell concentration before spray drying in RSM was 1.2 ± 0.2 ×

109 CFU/g dry matters. Survival during spray drying was 16.2 ± 9.1 % and cell counts in the

powder before storage were 2.0 ± 1.3 × 108 CFU/g. The log10 transformed cell counts in spray

dried powder decreased non-linearly with storage time (Figure 17), but at a slower rate compared

to the first selection batches at 37 °C (Figure 2 B). Therefore, cell counts did not reach the 4 log10

loss threshold per gram of powder after 67 days of storage (Figure 17).


- 105 -

Figure 17 Cell counts in spray dried powders of second selection batch (8 strain mix) (full

line, ○) and confirmation batches NCC572 (semi-dashed line, ■) and NCC2916 (dashed line, ▲)

during storage at 37 °C. Data are means of two independent experiments. Standard deviations of

each point are represented by error bars.

Individual strains accounted for 6 to 20 % of total cell counts in the RSM mix (Figure 18 A).

RAPD identification of 239 colonies picked after 67 days storage at 37 °C indicated strain

NCC572 as best survivor after spray drying and storage, with a large proportional increase in

powder from 12 % in RSM before spray drying to 88 % in the powder (Figure 18 B). NCC585

also exhibited good survival in powder with constant proportion of total cells of about 9 ± 2 % in

the RSM mix and powder. All the other strains decreased in percentage or were not detected in

the powder (Figure 18 B).


- 106 -

Figure 18 Relative presence (%) and corresponding cell counts (CFU per gram) of the B.

longum strains of the second selection batch in RSM before spray drying (A) and in spray dried

powders after storage for 67 days at 37 °C (B). For RSM, distribution-profiles were calculated

using cell concentration of each strain in pure culture before mixing. After storage, individual

populations in spray dried powders were estimated using their relative abundance (%) determined

by RAPD and total cell counts. Data are means of two trials ± standard deviation. a No standard deviation since only one colony was picked.

4.4.3 Validation of selection procedures using sing le strain

preparations

To confirm NCC572 as good spray drying and storage survivor, NCC572 and NCC2916 (low

survivor control) were tested in pure culture spray drying and storage experiments. Average

moisture content and water activity of powders was 5.5 ± 0.6 % and 0.17 ± 0.01, respectively.

Survival rate of NCC572 and NCC2916 during spray drying was 18.8 ± 9.6 and 4.8 ± 3.9 % with

an initial viable cell count of 2.3 ± 2.7 × 109 and 9.7 ± 2.0 × 108 CFU/g dry matter, respectively.

Strain NCC2916 showed a 4 log10 loss threshold in cell counts after 32 days of storage whereas


- 107 -

strain NCC572 exhibited only 2 log10 viability loss after 34 days and 3.4 log10 after 69 days

(Figure 17). Viable cell counts in powders of NCC572 after 62 days of storage were 1.5 ± 0.5 ×

105 CFU/g powder and exceeded counts of NCC2916 after 60 days by a factor of 269.

4.5 Discussion

Spray drying and storage stability is an important bacterial characteristic for the use in food

application at industrial scale because spray drying is a low cost alternative to freeze drying.

Survival to drying and storage is strain dependent and selection of robust strains with good

functional properties is required for successful application. Screening of optimum process

conditions and strains with good resistance is laborious, expensive and limited by variability

between experiments. Simultaneous screening of multiple strains can be used to increase

screening throughput. RAPD is a fast, practical, easy to perform and inexpensive technique to

distinguish bacterial strains. Furthermore, its application does not require a high level of technical

skills (Rossetti and Giraffa, 2005). RAPD analysis was successfully used in this study to

discriminate 22 B. longum strains, except strain NCC2705 and its H2O2 resistant derivative

NCC2916, in mixtures of up to 8 strains. It is theoretically possible to apply all 22 strains in one

single spray drying and storage batch provided that uniform treatment for each strain is

guaranteed.

Spray drying parameters were chosen from previous studies with bifidobacteria and lactic acid

bacteria and kept constant for all batches. RSM as a carrier medium is commonly used for spray

drying of lactic acid bacteria due to its protective effect against heat (Ananta et al., 2005;

Desmond et al., 2002). The inlet and outlet temperatures were set at 170 °C and ca. 80 °C,

respectively, according to Gardiner et al., (2000). Survival rate of B. longum cells determined


- 108 -

after spray drying in mix or pure cultures ranged from 5 to 21 % under these conditions and

concurred well with survival rate of 20 ± 1 % reported for B. longum biotype longum NCIMB

8809 in 20 % (w/v) RSM solution (Simpson et al., 2005), a strain also used in this study

(NCC444). The different survival rates between batches are most likely caused by different strain

compositions. Strain dependent survival to spray drying was previously reported for

bifidobacteria ranging between 12 – 102 % with an outlet temperature of 85 – 90 °C (Simpson et

al., 2005). Conditions were set to achieve a moisture content of the powder close to the optimal

value of 4 % for stability during storage (Gardiner et al., 2000; Simpson et al., 2005). However, a

slightly higher moisture value of 5.4 ± 0.7 % of powders was obtained in our study, partly

explained by a lower outlet temperature of 78 °C. Increasing the outlet temperature to produce a

cell powder with 4 % moisture content would likely decrease survival rate after spray drying as

reported by Gardiner et al. (2000) and Lian et al. (2002) for spray drying of Lactobacillus

paracasei, Lactobacillus salivarius and bifidobacteria. Water activity of 0.23 ± 0.01 during

storage for all batches containing strain-mixes lies within the recommended range for stability of

dried lactic acid bacteria (Champagne et al., 1996; Teixeira et al., 1995b).

The log10 transformed cell counts in powder decreased non-linearly as shown for lactobacilli

stored in RSM powder at 30 °C (Gardiner et al., 2000). Storage temperatures of 30 and 37 °C

selected to accelerate the loss in cell viability caused a faster decrease in cell viability at 37 than

at 30 °C and reached a 4 log10 units loss within 5 - 10 weeks. This effect of temperature is in

agreement with other studies on survival of spray dried probiotic bacteria during storage

(Corcoran et al., 2004; Desmond et al., 2002; Gardiner et al., 2000; Simpson et al., 2005;

Teixeira et al., 1995b).

Survival of the single strains varied among powders stored at 30 and 37 °C and resulted in

differences of strain distribution in the two powders. Remarkably, a higher number of different


- 109 -

strains was detected after storage at 37 compared to 30 °C for Batch 1 and 2, and to lesser extent

for Batch 3. The time period of storage may have a noticeable impact on survival of each strain in

dried powders because time duration between collection of samples at 30 and 37 °C differed

between 18 and 25 days for Batch 1 and 2, respectively compared to Batch 3 where samples from

storage at 30 °C where taken after only 53 days of storage, 4 days after the sampling from storage

at 37 °C. These findings suggest that longer storage periods at lower temperatures may raise

selection pressure to strains in powder state leading to a lower variety of strains detectable at the

end of storage at 30 °C.

The second selection round aimed at identifying the best survivors among strains selected from

the first round. The assorted strains in the second selection batch exhibited as expected prolonged

survival in powder and did not reach the 4 log10 loss threshold after 67 days of storage compared

to first selection batches where thresholds were reached within less than 50 days. Survival to

spray drying and storage for 67 days at 37 °C varied largely among single strains and led to

identification of strain NCC572 as best survivor strain among 22 tested B. longum strains. Its

high survival was confirmed by single strain drying and storage tests.

Successful application of spray drying of probiotic strains as an economic alternative to freezing

or freeze drying depends on the careful selection of appropriate strains harbouring intrinsic

resistance to drying and powder storage in addition to aspects of safety and functionality. Our

study showed the potential of RAPD technology to screen strains for survival to drying and

storage in combination with a time-effective mixed strain screening approach. A similar method

combining strain mixture testing and RAPD detection could be used to screen probiotic strains

simultaneously for resistance to different lethal stresses related to technological and functional

properties such as gastric, bile, oxygen, low pH, heat, and osmotic stress. Moreover, this

approach allows direct comparison of stress response of single strains and avoids variation


- 110 -

between experiments and time consuming repetitions. Further analyses of the strains exhibiting

intrinsic tolerance to drying and storage using comparative genomics and transcriptomics tools

could be followed in order to try to elucidate mechanisms conferring microbial stability during

drying and storage.

Chapter 5 - Conclusion and Outlook

- 111 -

5 General conclusions and outlook

5.1 General conclusions

The PhD thesis “Novel technologies for detection, production and screening of stress tolerant

bifidobacteria” consisted of three hypotheses. The first hypothesis claimed that B. longum

NCC2705 produced with immobilized cell technology and continuous culture had an improved

stress resistance progressing with culture age. In previous studies, cell immobilization in

combination with continuous probiotic culture showed positive effects on environmental stress

tolerance improving with culture time. Our studies confirmed the hypothesis partly, since we

observed better tolerance of immobilized cells (IC) of B. longum NCC2705 to porcine bile salts

and some antibiotics compared to free cells from batch culture independent of the culture age,

whereas tolerance to heat stress did not change, neither with culture age nor compared to batch

control cells. However, a time dependent change of cell morphology was detected and resulted in

formation of macroscopic cell aggregates of B. longum NCC2705. This extensive morphological

changes lead to great difficulties in accurately determining viable cell numbers which are

prerequisite for testing cell stress tolerance. Continuous immobilized cell fermentation in a two-

stage process lead to high cell production, similar to bacterial batch cultures and increased cell

volumetric productivity by approximately ten fold.

The immobilization of B. longum NCC2705 did not concur with the observations of Doleyres et

al. (2004a) where immobilized cells of B. longum ATCC 15707 produced continuously in co-

culture with Lc. lactis showed increased tolerance to various environmental stresses and no

formation of cell aggregates. The differences in stress resistance as well as the induction of auto-

aggregation may be attributed to the strain specificity in addition to difficult viable cell

determination. Moreover, the enhanced robustness of B. longum ATCC 15707 may be a result of


- 112 -

growth in the immobilized matrix in combination with an adaptation process to co-culture

conditions rather than only from cell immobilization. In fact, a recent study showed that only the

supernatant of co-culture in a compartmentalized fermentation with a B. longum and B. breve

strain affected bacterial physiology (Ruiz et al., 2009). Improved tolerance to porcine bile salts

and auto-aggregation of B. longum NCC2705 produced in our study could confer advantages in

the probiotic approach. Large cell aggregates might provide physical protection from adverse

environmental parameters during transit through GIT and could improve adhesion abilities to

mucosal epithelium from the host. Although the acquired characteristics could be partly

explained by the observed changes in the fatty acid composition, other factors may also

contribute to altered cell physiology. Preliminary results from microarray analyzes showing

transcriptional regulation of membrane associated transporter proteins as well as ribosomal

proteins may reflect a deep impact of the culture conditions on transport mechanisms and

translation machinery (see Appendix 6.2).

The acquired physiological and morphological characteristics of cells from ICC should be tested

for reversibility by repeated batch cultures and the predicted advantage in adhesion capacity

could be testing using in vitro assays with intestinal cells such as Caco-2 by testing large cell

aggregates versus batch control cells. However, physiological analyzes of bacteria in large

aggregates are impossible to perform with traditional plate count method.

In the second part of the thesis, we developed thus a new protocol to measure viable B. longum

NCC2705 cells using real time PCR technology. The method was based on measurement of

transcript levels of a housekeeping gene using quantitative real time PCR (qRT-PCR). The

development of the protocol is based on the detection of mRNA of a housekeeping gene whose

constitutive expression was confirmed with microarray analyzes (data from Nestlé Research

Center Lausanne, Switzerland). The hypothesis was verified in that an adequate viability marker


- 113 -

was identified as an amplicon of 400 bp length of purB gene of B. longum NCC2705. Batch free

cells quantified with qRT-PCR showed good correlation with plate counts before and after a

lethal heat stress treatment. The quantification of cells in the aggregated state with qRT-PCR

confirmed that the traditional plate count method resulted in major underestimation of viable cells

counts.

Quantification of viable cells in macroscopic aggregates revealed the limitation of the traditional

plate count method and emphasized the need of new technologies to quantify viable bacteria. The

qRT-PCR method developed in this method is not limited to quantify viable B. longum NCC2705

cells in pure cultures only. The method could be applied to other bifidobacterial strains or used to

quantify Bifidobacterium subsp. in mixed cultures with bacteria belonging to different genera,

provided that enzyme activity and primer specificity remains unaffected. Drawbacks of the

method is the requirement of well-trained labor as well as the high cost of enzymes, fluorescent

stains and RNA-extraction which limit the extensive application of the method as a standard

protocol. Progressive use of such consumables for gene expression analyzes such as RT-PCR or

microarray analyzes might lower the cost in the near future and therefore paves the way for broad

application. Beyond the quantification of probiotic bacteria, the method could be a valuable tool

to examine the viability of pathogenic microorganism in biofilms. The growing number of

sequenced microorganisms will increase the knowledge of gene expression patterns and broaden

the applicability of rapid and accurate molecular quantification tools which will lead to

sophisticated quantification of viable bacteria independent of the morphological state.

In the last part of the thesis, we used a different approach to identify stress tolerant B. longum

strains. We aimed at developing an efficient strategy to screen a number of B. longum strains

with intrinsic resistance to conditions of spray drying and accelerated powder storage. We

hypothesized the possibility to apply a number of B. longum strains to spray drying and storage


- 114 -

experiment simultaneously and identify the best survivor strain(s) using random amplified

polymorphic DNA (RAPD) method at the end of the storage test. Twenty-two strains of

bifidobacteria belonging to B. longum species could be distinguished using this approach that

allowed identifying one strain with enhanced survival rate to drying and storage. This higher

resistance was confirmed by single strain spray drying and storage experiment and by comparing

the results with a poor survivor strain.

The rapid screening protocol developed for the spray drying and storage approach can be

projected to various other stress tests relevant for production and consumption of probiotic food

products. An additional empirical approach with the same 22 B. longum strains applied to freeze

drying and subsequent accelerated storage might deliver important insights about parallels and

differences of survival of each strain to spray- and freeze-drying followed by storage.

5.2 Outlook and perspectives

The technologies and protocols applied in the thesis can be also applied alone or combined in

new approaches to produce or select stress tolerant probiotic strains. Continuous culture with

immobilized cells for instance can be applied to other probiotic strains to test whether induction

of aggregation and bile tolerance is strain dependent or not. The high throughput screening

protocol could be used to monitor survival to simulated gastrointestinal conditions or even in

vivo, with humans or gnotobiotic mice being administered with a number of probiotic strains

which are distinguishable by their RAPD-fingerprint and samples of feces or biopsies from the

GIT of mice could be analyzed to identify the strain with the longest residence time within the

host. If administered to individuals with an intact microflora, a pre-selection step with selective

media in combination with morphological analyzes would be necessary to select B. longum


- 115 -

strains before applying picked colonies to RAPD analyzes. Another approach exploits the

potential of probiotic in co-cultures. Co-cultures with bifidobacteria were shown to induce stress

tolerance or production of chaperones in B. longum and thus the effect of different co-culture

combinations or ratios could be analyzed with immobilized cell technique and continuous

culture. Biomass of probiotic strains with different probiotic properties could be produced in

separate continuous reactors containing bacteria in gel beads and produced under controlled

conditions. Biomass produced in immobilized chemostat cultures would be used for controlled

inoculation of reactors run in batch mode in co-culture and simultaneously for monoculture in

batch mode as reference (Figure 19 A). The system would allow efficient analysis of effect of co-

cultures on cell physiology or on gene-/ protein expression. One important benefit of cell

immobilization is the ability to control and stability of different strains within a mixed

fermentation process (Doleyres et al., 2004b; Lacroix and Yildirim, 2007).

Another experiment would be the exposure of sub-lethal stress additional to cell immobilization

in a two-stage fermentation system and cell continuous culture (Figure 19 B). The first reactor is

used to efficiently produce active biomass with cells immobilized in gel beads. The released cells

from gel beads are pumped in a second reactor where they are exposed to sub-lethal stress

treatment similar as used with continuous free cell system recently described and validated by

Mozzetti (2009). The advantage of using two-stage continuous system with immobilized cells is

the stability and robustness of the system combined with an adjustable (high) dilution rate

without running the risk of washing out active biomass. The different stresses cells encounter

during growth in gel beads and sub-lethal treatment may lead to a synergistic stress response and

result in cross-protection.


- 116 -

Figure 19 Possible fermentation systems to produce probiotic bacteria with enhanced

technological and functional characteristics. (A) A number strains are produced using continuous

culture with IC in separate systems and effluent used to inoculate mixed batch co-cultures at

different proportions for fast screening of strain interaction in the second reactor stages. (B) One

strain is continuously produced in a first reactor containing IC and used to inoculate a second

reactor, where cells are exposed to sub-lethal stresses (Lacroix and Yildirim, 2007).

Chapter 6 - Appendix

- 117 -

6 Appendix

6.1 Supplementary material from Chapter 4

Figure S1 RAPD patterns of 22 B. longum strains. Patterns obtained with the primers OPA-

02; OPL-07 and OPL-16. ML: Low mol. weight DNA marker; MH: High mol. weight DNA

marker.


- 118 -

6.2 Genome expression analysis of B. longum NCC2705

produced during continuous culture with immobilized cells

and batch culture

6.2.1 Background

Global gene expression or comparison of protein profiles are valuable tools in the field of

comparative analysis between suspended and immobilized bacteria. Numerous studies compared

differences of gene expression profiles from cells within biofilms and in planktonic state and

reported changes only in small numbers of genes (Whiteley et al., 2001, Schembri et al., 2003).

However, only few studies describe the comparison of artificially immobilized microbial cells

and cells from free cell culture. Global proteomic analyzes of Gram positive and Gram negative

bacteria of sessile (immobilized) cells on various carriers compared to free (planktonic) cells

revealed considerable alterations with different expression of detected protein spots ranging from

3 to more than 50 % (Junter and Jouenne, 2004).

In our study, transcriptomic analyzes of samples from both ICC cultures and free cells batch

cultures were compared among each other to characterize the effect of 2-stage continuous IC

cultures with respect to sugar content of the feed, to culture age and to growth in immobilized

state. Gene expression patterns were compared with morphological and physiological

modifications of cells from IC continuous culture.

6.2.2 Material and methods

Samples were collected from both ICC as described in Chapter 2. Genome wide expression

analysis between continuously produced IC cells from R2 was carried only as single experiments


- 119 -

for the following hybridizations: “ICC1 day 8 - Batch (ES)”; “ICC1 R2 day 8 –ICC2 R2 day 8”

and “ICC2 R2 day 20 – ICC2 R2 day 8” (Figure S2).

Figure S2 Hybridization scheme. Each arrow represents one hybridization experiment.

Abbreviations: “ES” early stationary and “MS” mid stationary growth phase

Microarray design and RNA extraction DNA based arrays, produced by Agilent Technologies

(www.agilent.com), were obtained by in situ synthesis of 60 mer oligonucleotides on glass slides

(Wolber et al., 2006). For each gene, 3 to 6 different probes were randomly distributed on the

array. Total RNA was extracted with the Macaloid method and purified as previously described

(Parche et al., 2006).

For each hybridization, cDNA was synthesized starting from 4 µg of total RNA and

subsequently labeled using the Array 900MPX Genisphere kit (Genisphere Inc., Hatfield, PE,

USA), following the protocol provided by the supplier. Luciferase and kanamycin control

ICC2day 8

ICC2day 20

BatchES

ICC1day 8

BatchMS

ICC2day 8ICC2day 8

ICC2day 20ICC2day 20

BatchES

BatchES

ICC1day 8ICC1day 8

BatchMS

BatchMS


- 120 -

mRNA (Promega, Zurich, Switzerland) at 1 and 10 ng, respectively, were mixed with total RNA

before labeling to allow balancing of the two channels during scanning. After the hybridization

procedure, array slides were scanned at 10 µm using a Scanarray 4000 (Packard Biochip

Technologies, Billerica, MA, USA). Parameters (laser power and photomultiplier tube gain) were

set in order to prevent saturation of any spot, except the probes corresponding to rRNA.

Data extracted with Imagene 5.6 (Biodiscovery, El Segundo, CA, USA) were treated with

homemade scripts in Python language (www.python.org) and a local installation of the ArrayPipe

web server (Hokamp et al., 2004). Probes showing a signal smaller than twice the standard

deviation of the local background were considered without signal. Probes showing no signal or

saturated signals in both channels were discarded from the analysis. Assuming an intensity-

dependent variation in dye signal, (limma) loess global normalization was applied on signal

ratios. To calculate average gene expression values, data from different probes and hybridization

duplicates (with dye swap) were combined as follows. Within each hybridization data set, gene

fold changes were calculated from the median of the corresponding probes values. The

expression value of a gene was retained if a signal was detected in at least 50 % of its probes. The

average gene expression values were then calculated by combining data from two independent

hybridizations. Genes were considered to be differentially expressed if their log2-transformed

signal ratios were higher than 1.0 or smaller than -1.0. B. longum genes organized in operons

were identified according to Price et al. (2006) (http://www.microbesonline.org/operons/

OperonList.html).


- 121 -

6.2.3 Results and discussion

All hybridization experiments were only carried out once without dye swap confirmation and

hence data are of limited significance, but general trends can be elucidated.

6.2.3.1 Comparison of cells from ICC with batch cultures

Cells from the second reactor produced during continuous cultures with immobilized cells were

compared with free cells batch cultures from early- and mid stationary growth phase. These

comparisons were carried out to find a correlation between gene expression profiles and

physiological changes including fatty acid compositions of cell membrane. Gene expression of

cells from ICC1 collected at day 8 were compared to cells from early stationary phase and

showed a relative expression of 1587 (92 %) ORFs of which 229 (14 %) genes were differently

expressed. When compared with cells from mid stationary phase a relative expression of 1347

ORFs (78 %) was reported among which 149 (11 %) were differentially expressed (Table S1).

Relative changes in gene expression of cells from continuous IC culture were more pronounced

when compared with cells from early stationary batch cells than when compared with cells from

mid stationary growth phase (Tables S2 a, b).

Gene expression of cells from ICC1 exhibited less differently expressed ORFs when compared

with batch cells from mid- than with cells from early stationary growth phase. This was expected

both cultures, from R2 of ICC1 and from batch mid stationary growth phase, face comparable

time periods in starvation conditions. ICC cultures indeed showed an increased expression of

proteins involved in the ribosomal complex compared to batch free cell cultures and may be a

result of sensing environmental changes and lead to modulated synthesis of metabolic enzymes

affecting cell physiology (Chen et al., 2003, Wilson and Nierhaus, 2005). Bile tolerant ICC cells


- 122 -

did not show increased transcript levels of neither BL1102 (Na+ dependent nucleoside

transporter), a putative cholate transporter, nor BL0796 (choloylglycine hydrolase), a bile salt

hydrolase. Bile tolerance is therefore most likely attributed to changes in membrane composition.

These first gene expression profile comparisons showed that cells from ICC did neither show up-

regulation of ORFs related to stress response as reported for heat adapted B. longum NCC2705

(Rezzonico et al., 2007), nor different expression of (identified) enzymes involved in fatty acid

production (Schell et al., 2002). However, it is known that transcript level and proteins content

do not necessarily correlate since their relationship depends strongly on time, cellular location,

stability of the molecules (Ghigo, 2003) and thus an enzyme-mediated bile resistance in ICC cells

cannot be excluded.

Table S1 Summary of hybridization experiments. Absolute and relative number of totally

expressed, under- and over expressed genes.

a ES: early stationary growth phase; MS: mid stationary growth phase * Differentlly expressed, when over-/underexpression ≥ 2 fold.

6.2.3.2 Comparison between ICC1 and ICC2 and within ICC2

Gene expression profiles of cells from continuous cultures with immobilized cells with and

without glucose limitation collected at day 8 were compared and a relative expression value of

approximately 58 % (1007 genes) of the ORFs was found with 170 genes (17 %) differently

Hybridization a Control Treated Expr. Expr. (%)

Over-expr.

under-expr.

Diff. expr.*

ICC1 day 8 - Batch (ES) Batch (ES) ICC1 day8 1587 91.9 102 127 229

ICC1 day 8 - Batch (MS) Batch (MS) ICC1 day8 1347 78.0 85 64 149

ICC2 day 8 –ICC1 day 8 ICC1 day 8 ICC2 day 8 1007 58.3 69 101 170

ICC2 day 20 – ICC2 day 8 ICC2 day 8 ICC2 day 20 1541 89.2 52 47 99


- 123 -

expressed (Table S1). The most over- and underexpressed genes were both hypothetical protein

BL1055a (hypothetical protein) and BL0981 (hypothetical protein with helix turn helix motif),

respectively, (Table S3). Among the 10 most repressed genes of cells from ICC2, 4 genes are

involved in sugar transport or metabolism. The sugar limitation in ICC1 may induce a

considerable increase of sugar transporter proteins compared to cells which are not sugar limited

(ICC2) in order to open gates for possible energy resources in the environment.

DNA micro array analysis between cells from ICC2 collected at day 20 containing large cell

aggregates were compared to cells collected before aggregation at day 8 and enabled assessment

of relative expression value of approximately 89% (1541) of ORFs (Table S1). In total, 99 genes

(6 %) were differentially expressed. The most over- and underexpressed genes were BL1040

(hypothetical protein) and BL1694 (probable sugar binding protein of ABC transporter for

pentoses), respectively (Table S4). Comparison of aggregated cells from ICC2 day 20 showed a

down regulation of transporter proteins of pentoses. The membrane proximity of high cell density

environment in macroscopic cell aggregates may induce quorum sensing (Sturme et al., 2002)

and result in a down-regulation of sugar transport proteins.

Although gene-expression experiments are only preliminary results, the data support some

physiological observations and may confer the understanding of molecular mechanisms

underlying increased bile tolerance and formation of auto-aggregation. However, further

hybridizations are required (including dye-swap experiments) to confirm obtained global gene

expression patterns of cells from ICC compared to cells from batch control cultures.


- 124 -

Table S2a The 10 most over- and underexpressed genes represented by the factor of

induction/repression of cells from ICC1 Day 8 compared to Batch ES.

Fold change

Locus Product Name Operon

Over-expressed 9.3 BL1631 D-Glucose-proton symporter BL1630-1631

7.5 BL0105 beta-fructofuranosidase (sucrase/invertase); possible inulinase BL0105-0107

6.7 BL0106 sucrose transport protein BL0105-0107 4.9 BL1493 possible WhiB-like transcription factor BL1493-1494 4.9 BL1494 hypo. protein BL1494 BL1493-1494

4.6 BL1492 hypo. protein with similarity to the Par protein of B. breve plasmid pcibb1

-

4.3 BL1100 30S ribosomal protein S12 BL1097-1100 3.7 BL0624 hypo. protein BL0624 BL0624-0625 3.7 BL0573 hypo. protein BL0573 - 3.5 BL1586 50S ribosomal protein L16 BL1577-1604

Under-expressed -9.8 BL0036 probable ABC transporter permease protein for sugars BL0033-0036 -9.8 BL0188 solute binding protein of ABC transporter system BL0188-0190 -9.8 BL0035 probable ABC transport system permease protein for sugars BL0033-0036 -8.0 BL1329 alpha-mannosidase BL1330-1333 -7.5 BL1330 probable solute-binding protein of ABC transporter system BL1330-1333 -7.0 BL0034 ATP binding protein of ABC transporter BL0033-0036 -6.5 BL1331 probable sugar permease of ABC transporter system BL1330-1333 -6.5 BL0813 hypo. protein BL0813 -

-6.1 BL1335 probable endo-beta-N-acetylglucosaminidase; di-N-acetylchitobiosyl beta-N-acetylglucosaminidase BL1334-1335

-5.7 BL1165 probable solute binding protein of ABC transporter system for sugars -


- 125 -

Table S2b The 10 most over- and underexpressed genes represented by the factor of

induction/repression of cells from ICC1 Day 8 compared to Batch MS.

Fold change


Over-expressed 6.6 BL1631 D-Glucose-proton symporter BL1630-1631 6.1 BL0555 possible DO serine protease -

5.8 BL0105 beta-fructofuranosidase (sucrase/invertase); possible inulinase

BL0105-0107

5.4 BL0106 sucrose transport protein BL0105-0107 5.1 BL1594 30S ribosomal protein S8 BL1577-1604 4.9 BL1595 50S ribosomal protein L6 BL1577-1604 4.8 BL1598 50S ribosomal protein L30 BL1577-1604 4.6 BL1596 50S ribosomal protein L18 BL1577-1604 4.2 BL0358 ATP synthase gamma chain BL0356-0363 4.2 BL1586 50S ribosomal protein L16 BL1577-1604

Under-expressed -9.3 BL0036 probable ABC transporter permease protein for sugars BL0033-0036 -9.3 BL0034 ATP binding protein of ABC transporter BL0033-0036 -8.0 BL0035 probable ABC transport system permease protein for sugars BL0033-0036

-6.3 BL0033 probable solute binding protein of ABC transporter system possibly for sugars

BL0033-0036

-6.1 BL1340 probable transcriptional repressor in the Rok (NagC/XylR) family

BL1339-1340

-4.3 BL1772 sugar kinase in PfkB family BL1772-1773 -3.9 BL1771 C4-dicarboxylate transporter BL1770-1771

-3.6 BL1325 narrowly conserved hypo. protein with possible DNA interaction motif

-

-3.2 BL1770 widely conserved hypo. protein in the inosine-uridine preferring nucleoside hydrolase family

BL1770-1771

-3.2 BL1773 hypothetical protein BL1773 BL1772-1773


- 126 -

Table S3 The 10 most over- and underexpressed genes represented by the factor of

induction/repression of cells from ICC2 Day 8 compared to ICC1 Day 8.

Fold change


Over-expressed 6.2 BL1055a hypo. protein BL1055a BL1054-1055a 5.1 BL1161 solute bind. protein of ABC transp. system for peptides BL1157-1161 4.1 BL0945 ATP-dependent Clp protease proteolytic subunit 1 BL0943-0945 4.1 BL1183 FtsE-like ATP binding protein involved in cell division BL1182-1184 4.0 BL0813 hypo. protein BL0813 - 4.0 BL1182 FtsX-like protein involved in cell division BL1182-1184 3.5 BL0065 elongation factor P BL0065-0066 3.5 BL1062 acetylglutamate kinase BL1058-1068

3.0 BL1714 solute bind. protein of ABC transporter for branched-chain amino acids BL1714-1718

2.9 BL1071 narrowly conserved hypo. protein BL1069-1071

Under-expressed -39.4 BL0981 hypo. protein with helix turn helix motif - -22.3 BL0106 sucrose transport protein BL0105-0107

-17.4 BL0105 beta-fructofuranosidase (sucrase/invertase); possible inulinase BL0105-0107

-13.8 BL0980 hypo. protein BL0980 -

-11.6 BL1164 prob. solute binding protein of ABC transporter system for sugars BL1163-1164

-8.2 BL1445 narrowly conserved hypo. protein ? -8.0 BL0428 hypo. protein BL0428 BL0428-0431 -8.0 BL0595 hypo. protein BL0595 -

-7.8 BL1694 probable sugar binding protein of ABC transporter for pentoses

BL1694-1696

-7.2 BL1664 widely conserved protein in universal stress protein family -


- 127 -

Table S4 The 10 most over- and underexpressed genes represented by the factor of

induction/repression of cells from ICC2 Day 20 compared to ICC2 Day 8.

Fold change


Over-expressed 19.3 BL1040 hypo. protein BL1040 - 17.4 BL1041a possible GntR-family transcriptional regulator BL1041-1041a 13.6 BL1041 ATP binding protein of ABC transporter BL1041-1041a 7.4 BL0435 possible ammonium ion transporter BL0433-0435 6.9 BL1076 glutamine synthetase 1 - 5.4 BL1177 permease protein of ABC transporter system BL1176-1179 5.3 BL0011 cytosine deaminase - 4.0 BL1499 isocitrate dehydrogenase - 3.3 BL0002 chaperonin GroEL - 3.3 BL1404 hypothetical protein BL1404 -

Under-expressed

-5.8 BL1694 probable sugar binding protein of ABC transporter for pentoses

BL1694-1696

-5.3 BL1695 ATP binding protein of ABC transporter for pentoses BL1694-1696 -3.8 BL0923 hypo. protein BL0923 - -3.8 BL0932 possible ATP binding protein of ABC transporter BL0931-0932 -3.7 BL1055a hypo. protein BL1055a BL1054-1055a -3.6 BL0924 hypo. protein BL0924 - -3.4 BL0950 pyruvate formate-lyase 1 activating enzyme BL0947-0951 -3.2 BL1696 probable ABC transport system permease protein for sugars BL1694-1696 -3.2 BL0925 hypo. protein BL0925 - -3.1 BL0980 hypo. protein BL0980 -

Chapter 7 - References

- 128 -

7 References

Abadias, M., Teixido, N., Usall, J., Benabarre, A. and Vinas, I., 2001. Viability, efficacy, and storage stability of freeze-dried biocontrol agent Candida sake using different protective and rehydration media. Journal of Food Protection 64, 856-861.

Adhikari, K., Mustapha, A., Grun, I. U. and Fernando, L., 2000. Viability of microencapsulated

bifidobacteria in set yogurt during refrigerated storage. Journal of Dairy Science 83, 1946-1951.

Adhikari, K., Mustapha, A. and Grun, I. U., 2003. Survival and metabolic activity of

microencapsulated Bifidobacterium longum in stirred yogurt. Journal of Food Science 68, 275-280.

Aellen, S., Que, Y.A., Guignard, B., Haenni, M. and Moreillon, P., 2006. Detection of live and

antibiotic-killed bacteria by quantitative real-time PCR of specific fragments of rRNA. Antimicrobial Agents and Chemotherapy 50, 1913-1920.

Ahmed, M., Prasad, J., Gill, H., Stevenson, L. and Gopal, P., 2007. Impact of consumption of

different levels of Bifidobacterium lactis HN019 on the intestinal microflora of elderly human subjects. Journal of Nutrition, Health and Aging 11, 26-31.

Ahn, J. B., Hwang, H. J. and Park, J. H., 2001. Physiological responses of oxygen-tolerant

anaerobic Bifidobacterium longum under oxygen. Journal of Microbiology and Biotechnology 11, 443-451.

Alakomi, H. L., Mättö, J., Virkajärvi, I. and Saarela, M., 2005. Application of a microplate scale

fluorochrome staining assay for the assessment of viability of probiotic preparations. Journal of Microbiological Methods 62, 25-35.

Alander, M., Mättö, J., Kneifel, W., Johansson, M., Kogler, B., Crittenden, R., Mattila-

Sandholm, T. and Saarela, M., 2001. Effect of galacto-oligosaccharide supplementation on human faecal microflora and on survival and persistence of Bifidobacterium lactis Bb-12 in the gastrointestinal tract. International Dairy Journal 11, 817-825.

Anal, A. K. and Singh, H., 2007. Recent advances in microencapsulation of probiotics for

industrial applications and targeted delivery. Trends in Food Science & Technology 18, 240-251.

Ananta, E., Volkert, M. and Knorr, D., 2005. Cellular injuries and storage stability of spray-dried

Lactobacillus rhamnosus GG. International Dairy Journal 15, 399-409. Arnaud, J. P., Lacroix, C. and Castaigne, F., 1992. Counterdiffusion of lactose and lactic acid in

κ-carrageenan locust bean gum gel beads with or without entrapped lactic acid bacteria. Enzyme and Microbial Technology 14, 715-724.


- 129 -

Arunachalam, K., Gill, H. S. and Chandra, R. K., 2000. Enhancement of natural immune function by dietary consumption of Bifidobacterium lactis (HN019). European Journal of Clinical Nutrition 54, 263-267.

Asahara, T., Shimizu, K., Nomoto, K., Hamabata, T., Ozawa, A. and Takeda, Y., 2004. Probiotic

bifidobacteria protect mice from lethal infection with shiga toxin-producing Escherichia coli O157 : H7. Infection and Immunity 72, 2240-2247.

Augustin, M.A., and L. Sanguansri, 2008. Encapsulation of Bioactives. In Food Materials

Science, Principles and Practice, edited by J. M. Aguilera and P. J. Lillford. New-York: Springer, pp.577-601.

Auty, M. A. E., Gardiner, G. E., McBrearty, S. J., O'Sullivan, E. O., Mulvihill, D. M., Collins, J.

K., Fitzgerald, G. F., Stanton, C. and Ross, R. P., 2001. Direct in situ viability assessment of bacteria in probiotic dairy products using viability staining in conjunction with confocal scanning laser microscopy. Applied and Environmental Microbiology 67, 420-425.

Backhed, F., Ding, H., Wang, T., Hooper, L. V., Koh, G. Y., Nagy, A., Semenkovich, C. F. and

Gordon, J. I., 2004. The gut microbiota as an environmental factor that regulates fat storage. Proceedings of the National Academy of Sciences of the United States of America 101, 15718-15723.

Bartosch, S., Woodmansey, E. J., Paterson, J. C. M., McMurdo, M. E. T. and Macfarlane, G. T.,

2005. Microbiological effects of consuming a synbiotic containing Bifidobacterium bifidum, Bifidobacterium lactis, and oligofructose in elderly persons, determined by real-time polymerase chain reaction and counting of viable bacteria. Clinical Infectious Diseases 40, 28-37.

Begley, M., Gahan, C. G. M. and Hill, C., 2005. The interaction between bacteria and bile. FEMS

Microbiology Reviews 29, 625-651. Belasco, J. G., Nilsson, G., Vongabain, A., Cohen, S. N., 1986. The stability of Escherichia coli

gene transcripts is dependent on determinants localized to specific messenger-RNA segments. Cell 46, 245-251.

Ben Amor, K., Breeuwer, P., Verbaarschot, P., Rombouts, F. M., Akkermans, A. D. L., De Vos,

W. M. and Abee, T., 2002. Multiparametric flow cytometry and cell sorting for the assessment of viable, injured, and dead Bifidobacterium cells during bile salt stress. Applied and Environmental Microbiology 68, 5209-5216.

Ben Amor, K., Vaughan, E.E., and de Vos, W.M. 2007. Advanced molecular tools for the

identification of lactic acid bacteria. Journal of Nutrition 137, 741S-747S. Bergmaier, D., Champagne, C. P. and Lacroix, C., 2005. Growth and exopolysaccharide

production during free and immobilized cell chemostat culture of Lactobacillus rhamnosus RW-9595M. Journal of Applied Microbiology 98, 272-284.


- 130 -

Bernstein, J. A., Khodursky, A. B., Lin, P. H., Lin-Chao, S., and Cohen, S. N. (2002) Global analysis of mRNA decay and abundance in Escherichia coli at single-gene resolution using two-color fluorescent DNA microarrays. Proceedings of the National Academy of Sciences of the United States of America 99, 9697–9702.

Bezkorovainy, A., 2001. Probiotics: determinants of survival and growth in the gut. American

Journal of Clinical Nutrition 73, 399S-405S. Biavati, B., Vescovo, M., Torriani, S. and Bottazzi, V., 2000. Bifidobacteria: history, ecology,

physiology and applications. Annals of Microbiology 50, 117-131. Bibiloni, R., Zavaglia, A. G. and De Antoni, G., 2001. Enzyme-based most probable number

method for the enumeration of Bifidobacterium in dairy products. Journal of Food Protection 64, 2001-2006.

Bielecka, M. and Majkowska, A., 2000. Effect of spray drying temperature of yoghurt on the

survival of starter cultures, moisture content and sensoric properties of yoghurt powder. Food 44, 257-260.

Bin-Nun, A., Bromiker, R., Wilschanski, M., Kaplan, M., Rudensky, B., Caplan, M. and

Hammerman, C., 2005. Oral probiotics prevent necrotizing enterocolitis in very low birth weight neonates. Journal of Pediatrics. 147, 192-196.

Birmingham, P., Helm, J. M., Manner, P. A. and Tuan, R. S., 2008. Simulated joint infection

assessment by rapid detection of live bacteria with real-time reverse transcription polymerase chain reaction. Journal of Bone and Joint Surgery 90A, 602-608.

Bleve, G., Rizzotti, L., Dellaglio, F., Torriani, S., 2003. Development of reverse transcription

(RT)-PCR and real-time RT-PCR assays for rapid detection and quantification of viable yeasts and molds contaminating yogurts and pasteurized food products. Applied and Environmental Microbiology 69, 4116-4122.

Breeuwer, P. and Abee, T., 2000. Assessment of viability of microorganisms employing

fluorescence techniques. International Journal of Food Microbiology 55, 193-200. Bryan, L. E., Ohara, K. and Wong, S., 1984. Lipopolysaccharide changes in impermeability-type

aminoglycoside resistance in Pseudomonas aeruginosa. Antimicrobial Agents and Chemotherapy 26, 250-255.

Buitink, J., van den Dries, I. J., Hoekstra, F. A., Alberda, M. and Hemminga, M. A., 2000. High

critical temperature above T-g may contribute to the stability of biological systems. Biophysical Journal 79, 1119-1128.

Burin, L., Jouppila, K., Roos, Y. H., Kansikas, J. and Buera, M. P., 2004. Retention of beta-

galactosidase activity as related to Maillard reaction, lactose crystallization, collapse and glass transition in low moisture whey systems. International Dairy Journal 14, 517-525.


- 131 -

Bustin, S. A., 2000. Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays. Journal of Molecular Endocrinolology 25, 169-193.

Bustin, S. A., 2002. Quantification of mRNA using real-time reverse transcription PCR (RT-

PCR): trends and problems. Journal of Molecular Endocrinolology 29, 23–39. Cachon, R., Anterieux, P. and Diviès, C., 1998. The comparative behavior of Lactococcus lactis

in free and immobilized culture processes. Journal of Biotechnology 63, 211-218. Campbell, E. A., Korzheva, N., Mustaev A., Murakami K., Nair S., Goldfarb A., Darst S. A.,

2001. Structural mechanism for rifampicin inhibition of bacterial RNA polymerase. Cell 104, 901-912.

Canzi, E., Guglielmetti, S., Mora, D., Tamagnini, I. and Parini, C., 2005. Conditions affecting

cell surface properties of human intestinal bifidobacteria. Antonie Van Leeuwenhoek International Journal of General and Molecular Microbiology 88, 207-219.

Caplan, M. S., Miller-Catchpole, R., Kaup, S., Russell, T., Lickerman, M., Amer, M., Xiao, Y.

and Thomson, R., Jr., 1999. Bifidobacterial supplementation reduces the incidence of necrotizing enterocolitis in a neonatal rat model. Gastroenterology 117, 577-583.

Carvalho, A. S., Silva, J., Ho, P., Teixeira, P., Malcata, F. X. and Gibbs, P., 2004. Relevant

factors for the preparation of freeze-dried lactic acid bacteria. International Dairy Journal 14, 835-847.

Castagliuolo, I., Galeazzi, F., Ferrari, S., Elli, M., Brun, P., Cavaggioni, A., Tormen, D.,

Sturniolo, G. C., Morelli, L. and Palu, G., 2005. Beneficial effect of auto-aggregating Lactobacillus crispatus on experimentally induced colitis in mice. FEMS Immunology and Medical Microbiology 43, 197-204.

Castro, H. P., Teixeira, P. M. and Kirby, R., 1995. Storage of lyophilized cultures of

Lactobacillus bulgaricus under different relative humidities and atmospheres. Applied Microbiology and Biotechnology 44, 172-176.

Castro, H. P., Teixeira, P. M. and Kirby, R., 1996. Changes in the cell membrane of

Lactobacillus bulgaricus during storage following freeze-drying. Biotechnology Letters 18, 99-104.

Cesena, C., Morelli, L., Alander, M., Siljander, T., Tuomola, E., Salminen, S., Mattila-Sandholm,

T., Vilpponen-Salmela, T., von Wright, A., 2001. Lactobacillus crispatus and its nonaggregating mutant in human colonization trials. Journal of Dairy Science 84, 1001-1010.

Champagne, C. P., Mondou, F., Raymond, Y. and Roy, D., 1996. Effect of polymers and storage

temperature on the stability of freeze-dried lactic acid bacteria. Food Research International 29, 555-562.


- 132 -

Champagne, C. P., Gardner, N. J. and Roy, D., 2005. Challenges in the addition of probiotic cultures to foods. Critical Reviews in Food Science and Nutrition 45, 61-84.

Champagne, C. P. and Fustier, P., 2007. Microencapsulation for the improved delivery of

bioactive compounds into foods. Current Opinion in Biotechnology 18, 184-190. Chavez, B. E. and Ledeboer, A. M., 2007. Drying of probiotics: Optimization of formulation and

process to enhance storage survival. Drying Technology 25, 1193-1201. Cheikhyoussef, A., Pogori, N., Chen, W. and Zhang, H., 2008. Antimicrobial proteinaceous

compounds obtained from bifidobacteria: From production to their application. International Journal of Food Microbiology 125, 215-222.

Chen, H. C., Teplitski, M., Robinson, J. B., Rolfe, B. G. and Bauer, W. D., 2003. Proteomic

analysis of wild-type Sinorhizobium meliloti responses to N-acyl homoserine lactone quorum-sensing signals and the transition to stationary phase. Journal of Bacteriology 185, 5029-5036.

Chouraqui, J.P., Van Egroo, L.D. and Fichot, M.C., 2004. Acidified milk formula supplemented

with Bifidobacterium lactis: Impact on infant diarrhea in residential care settings. Journal of Pediatric Gastroenterology and Nutrition 38, 288-292.

Cinquin, C., Le Blay, G., Fliss, I. and Lacroix, C., 2004. Immobilization of infant fecal

microbiota and utilization in an in vitro colonic fermentation model. Microbial Ecology 48, 128-138.

Coakley, M., Banni, S., Johnson, M. C., Mills, S., Devery, R., Fitzgerald, G., Ross, R. P. and

Stanton, C., 2009. Inhibitory Effect of Conjugated alpha-Linolenic Acid from Bifidobacteria of Intestinal Origin on SW480 Cancer Cells. Lipids 44, 249-256.

Collado, M. C., Moreno, Y., Cobo, J. M. and Hernandez, M., 2006. Microbiological evaluation

and molecular characterization of bifidobacteria strains in commercial fermented milks. European Food Research and Technology 222, 112-117.

Collado, M. C., Meriluoto, J. and Salminen, S., 2007. Measurement of aggregation properties

between probiotics and pathogens: In vitro evaluation of different methods. Journal of Microbiological Methods 71, 71-74.

Corcoran, B. M., Ross, R. P., Fitzgerald, G. F. and Stanton, C., 2004. Comparative survival of

probiotic lactobacilli spray-dried in the presence of prebiotic substances. Journal of Applied Microbiology 96, 1024-1039.

Corcoran, B.M., Stanton, C., Fitzgerald, G., and Ross, R.P., 2008. Life under stress: The

probiotic stress response and how it may be manipulated. Current Pharmaceutical Design 14, 1382-1399.


- 133 -

Costa, E., Usall, J., Teixido, N., Garcia, N. and Vinas, I., 2000. Effect of protective agents, rehydration media and initial cell concentration on viability of Pantoea agglomerans strain CPA-2 subjected to freeze-drying. Journal of Applied Microbiology 89, 793-800.

Costerton, J. W., Stewart, P. S. and Greenberg, E. P., 1999. Bacterial biofilms: A common cause

of persistent infections. Science 284, 1318-1322. Coutard, F., Pommepuy, M., Loaec, S. and Hervio-Heath, D., 2005. mRNA detection by reverse

transcription-PCR for monitoring viability and potential virulence in a pathogenic strain of Vibrio parahaemolyticus in viable but nonculturable state. Journal of Applied Microbiology 98, 951-961.

Coutard, F., Lozach, S., Pommepuy, M., Hervio-Heath, D., 2007. Real-time reverse transcription-

PCR for transcriptional expression analysis of virulence and housekeeping genes in viable but nonculturable Vibrio parahaemolyticus after recovery of culturability. Applied and Environmental Microbiology 73, 5183-5189.

Crociani, F., Biavati, B., Alessandrini, A., Chiarini, C. and Scardovi, V., 1996. Bifidobacterium

inopinatum sp. nov and Bifidobacterium denticolens sp nov, two new species isolated from human dental caries. International Journal of Systematic Bacteriology 46, 564-571.

Crowe, J. H., Crowe, L. M., Carpenter, J. F. and Wistrom, C. A., 1987. Stabilization of dry

phospholipid-bilayers and proteins by sugars. Biochemistry Journal 242, 1-10. Daley, R.J., 1977. Direct epifluorescence enumeration of native aquatic bacteria: uses,

limitations, and comparative accuracy. In: Costerton J. W., Colwell. R. R., Native Aquatic Bacteria: Enumeration, Activity and Ecology. American Society for Testing and Materials, Philadelphia, pp. 29–45.

Davey, M. E. and O'Toole, G. A., 2000. Microbial biofilms: from ecology to molecular genetics.

Microbiology and Molecular Biology Reviews 64, 847-867. Del Re, B., Sgorbati, B., Miglioli, M. and Palenzona, D., 2000. Adhesion, autoaggregation and

hydrophobicity of 13 strains of Bifidobacterium longum. Letters in Applied Microbiology 31, 438-442.

de Man, J. D., Rogosa, M., Sharpe, M. E., 1960. A medium for the cultivation of Lactobacilli.

Journal of. Applied Bacteriology 23, 130–135. Denich, T. J., Beaudette, L. A., Lee, H. and Trevors, J. T., 2003. Effect of selected environmental

and physico-chemical factors on bacterial cytoplasmic membranes. Journal of Microbiological Methods 52, 149-182.

De Schryver, P., Crab, R., Defoirdt, T., Boon, N. and Verstraete, W., 2008. The basics of bio-

flocs technology: The added value for aquaculture. Aquaculture 277, 125-137.


- 134 -

De Smet, I., De Boever, P. and Verstraete, W., 1998. Cholesterol lowering in pigs through enhanced bacterial bile salt hydrolase activity. British Journal of Nutrition 79, 185-194.

Desmond, C., Stanton, C., Fitzgerald, G. F., Collins, K. and Ross, R. P., 2001. Environmental

adaptation of probiotic lactobacilli towards improvement of performance during spray drying. International Dairy Journal 11, 801-808.

Desmond, C., Ross, R. P., O'Callaghan, E., Fitzgerald, G. and Stanton, C., 2002. Improved

survival of Lactobacillus paracasei NFBC 338 in spray-dried powders containing gum acacia. Journal of Applied Microbiology 93, 1003-1011.

de Vrese, M., Winkler, P., Rautenberg, P., Harder, T., Noah, C., Laue, C., Ott, S., Hampe, J.,

Schreiber, S., Heller, K., and Schrezenmeir, J., 2005. Effect of Lactobacillus gasseri PA 16/8, Bifidobacterium longum SP 07/3, B. bifidum MF 20/5 on common cold episodes: a double blind, randomized, controlled trial. Clinical Nutrition 24, 481-491.

de Vrese, M. and Schrezenmeir, J.: Probiotics, Prebiotics, and Synbiotics. In Food

Biotechnology. Edited by; 2008:1-66. Advances in Biochemical Engineering / Biotechnology, vol 111.]

Ding, W. K. and Shah, N. P., 2007. Acid, bile, and heat tolerance of free and microencapsulated

probiotic bacteria. Journal of Food Science 72, M446-M450. Ding, W. K. and Shah, N. P., 2009. Effect of Various Encapsulating Materials on the Stability of

Probiotic Bacteria. Journal of Food Science 74, M100-M107. Diviès, C. and Siess, M. H., 1976. Study on L-malic acid catabolism by Lactobacillus casei cells

immobilized into polyacrylamide-gel lattice. Annales De Microbiologie B127, 525-539. Doleyres, Y., Paquin, C., LeRoy, M. and Lacroix, C., 2002a. Bifidobacterium longum ATCC

15707 cell production during free- and immobilized-cell cultures in MRS-whey permeate medium. Applied Microbiology and Biotechnology 60, 168-173.

Doleyres, Y., Fliss, I. and Lacroix, C., 2002b. Quantitative determination of the spatial

distribution of pure- and mixed-strain immobilized cells in gel beads by immunofluorescence. Applied Microbiology and Biotechnology 59, 297-302.

Doleyres, Y., Fliss, I. and Lacroix, C., 2004a. Increased stress tolerance of Bifidobacterium

longum and Lactococcus lactis produced during continuous mixed-strain immobilized-cell fermentation. Journal of Applied Microbiology 97, 527-539.

Doleyres, Y., Fliss, I. and Lacroix, C., 2004b. Continuous production of mixed lactic starters

containing probiotics using immobilized cell technology. Biotechnology Progress 20, 145-150.

Doleyres, Y. and Lacroix, C., 2005. Technologies with free and immobilised cells for probiotic

bifidobacteria production and protection. International Dairy Journal 15, 973-988.


- 135 -

Edwards, C. A. and Parrett, A. M., 2002. Intestinal flora during the first months of life: new perspectives. British Journal of Nutrition 88, S11 – S18.

Egert, M., de Graaf, A. A., Smidt, H., de Vos, W. M. and Venema, K., 2006. Beyond diversity:

functional microbiomics of the human colon. Trends in Microbiology 14, 86-91. Espina, F. and Packard, V. S., 1979. Survival of Lactobacillus acidophilus in a spray –drying

process. Journal of Food Protection 42, 149-152. Falk, P. G., Hooper, L. V., Midtvedt, T. and Gordon, J. I., 1998. Creating and maintaining the

gastrointestinal ecosystem: What we know and need to know from gnotobiology. Microbiology and Molecular Biology Reviews 62, 1157-+.

Fanaro, S., Chierici, R., Guerrini, P. and Vigi, V., 2003. Intestinal microflora in early infancy:

composition and development. Acta Paediatrica 92, 48-55. FAO/WHO (2001) Evaluation of health and nutritional properties of powder milk with live lactic

acid bacteria. Report from FAO/WHO Expert Consultation, 1–4 October 2001, Cordoba, Argentina.

Favaro-Trindale, C. S. and Grosso, C. R. F., 2002. Microencapsulation of L-acidophilus (La-05)

and B-lactis (Bb-12) and evaluation of their survival at the pH values of the stomach and in bile. Journal of Microencapsulation 19, 485-494.

Freudig, B., Hogekamp, S. and Schubert, H., 1999. Dispersion of powders in liquids in a stirred

vessel. Chemical Engineerin and Processing 38, 525-532. Friedrich, U., Lenke, J., 2006. Improved enumeration of lactic acid bacteria in mesophilic dairy

starter cultures by using multiplex quantitative real-time PCR and flow cytometry-fluorescence in situ hybridization. Applied and Environmental Microbiology 72, 4163-4171.

Fukushima, Y., Kawata, Y., Hara, H., Terada, A. and Mitsuoka, T., 1998. Effect of a probiotic

formula on intestinal immunoglobulin A production in healthy children. International Journal of Food Microbiology 42, 39-44.

Fux, C. A., Costerton, J. W., Stewart, P. S. and Stoodley, P., 2005. Survival strategies of

infectious biofilms. Trends in Microbiology 13, 34-40. Gagnon, M., Kheadr, E. E., Dabour, N., Richard, D. and Fliss, I., 2006. Effect of Bifidobacterium

thermacidophilum probiotic feeding on enterohemorrhagic Escherichia coli O157 : H7 infection in BALB/c mice. International Journal of Food Microbiology 111, 26-33.

Gardiner, G. E., O'Sullivan, E., Kelly, J., Auty, M. A. E., Fitzgerald, G. F., Collins, J. K., Ross,

R. P. and Stanton, C., 2000. Comparative survival rates of human-derived probiotic Lactobacillus paracasei and L-salivarius strains during heat treatment and spray drying. Applied and Environmental Microbiology 66, 2605-2612.


- 136 -

Gasser, F., 1994. Safety of lactic acid bacteria and their occurrence in human clinical infections. Bulletin De L Institut Pasteur 92, 45-67.

Gevers, D., Huys, G. and Swings, J., 2001. Applicability of rep-PCR fingerprinting for

identification of Lactobacillus species. FEMS Microbiological Letters 205, 31-36. Ghigo, J. M., 2003. Are there biofilm-specific physiological pathways beyond a reasonable

doubt? Research in Microbiology 154, 1-8. Gionchetti, P., Rizzello, F., Helwig, U., Venturi, A., Lammers, K. M., Brigidi, P., Vitali, B.,

Poggioli, G., Miglioli, M. and Campieri, M., 2003. Prophylaxis of pouchitis onset with probiotic therapy: a double-blind, placebo-controlled trial. Gastroenterology 124, 1202-1209.

Goldenberger, D., Kunzli, A., Vogt, P., Zbinden, R. and Altwegg, M., 1997. Molecular diagnosis

of bacterial endocarditis by broad-range PCR amplification and direct sequencing. Journal of Clinical Microbiology 35, 2733-2739

Gonzalez-Escalona, N., Hammack, T. S., Russell, M., Jacobson, A. P., De Jesus, A. J., Brown, E.

W., Lampel, K. A., 2009. Detection of live Salmonella spp. cells in produce by a TaqMan(R) based quantitative reverse transcriptase real-time PCR (qRT-PCR) targeting invA mRNA. Applied and Environmental Microbiology 75, 3714-3720.

Grattepanche, F., Audet, P. and Lacroix, C., 2007. Enhancement of functional characteristics of

mixed lactic culture producing nisin Z and exopolysaccharides during continuous prefermentation of milk with immobilized cells. Journal of Dairy Science 90, 5361-5373.

Grattepanche, F. and Lacroix, C. (in press). Production of high quality probiotics using novel

fermentation and stabilization technologies. In D. Bagchi, F. C. Lau and D. K. Ghosh (Eds.), Biotechnology in Functional Foods and Nutraceuticals. Taylor & Francis Group. Abingdon, UK.

Green, S. L., 1978. Case-report – Fatal anaerobic pulmonary infection due to Bifidobacterium

eriksonii. Postgraduate Medicine 63, 187-192. Guarner, F. and Malagelada, J. R., 2003. Gut flora in health and disease. Lancet 361, 512-519. Guarner, F., Bourdet-Sicard, R., Brandtzaeg, P., Gill, H. S., McGuirk, P., van Eden, W.,

Versalovic, J., Weinstock, J. V. and Rook, G. A. W., 2006. Mechanisms of disease: the hygiene hypothesis revisited. Nature Clinical Practice Gastroenterology & Hepatology 3, 275-284.

Gueimonde, M., Tolkko, S., Korpimaki, T. and Salminen, S., 2004. New real-time quantitative

PCR procedure for quantification of bifidobacteria in human fecal samples. Applied and Environmental Microbiology 70, 4165-4169.


- 137 -

Gueimonde, M., Garrigues, C., van Sinderen, D., de los Reyes-Gavilan, C.G., and Margolles, A., 2009. Bile-inducible efflux transporter from Bifidobacterium longum NCC2705, conferring bile resistance. Applied Environmental Microbiology 75, 3153-3160.

Guillot, A., Obis, D. and Mistou, M. Y., 2000. Fatty acid membrane composition and activation

of glycine-betaine transport in Lactococcus lactis subjected to osmotic stress. International Journal of Food Microbiology 55, 47-51.

Hambraeus, G., von Wachenfeldt, C., Hederstedt, L., 2003. Genome-wide survey of mRNA half-

lives in Bacillus subtilis identifies extremely stable mRNAs. Molecular Genetics and Genomics 269, 706-714.

Hansen, L. T., Allan-Wojtas, P. M., Jin, Y. L. and Paulson, A. T., 2002. Survival of Ca-alginate

microencapsulated Bifidobacterium spp. in milk and simulated gastrointestinal conditions. Food Microbiology 19, 35-45.

Harmsen, H. J. M., Wildeboer-Veloo, A. C. M., Raangs, G. C., Wagendorp, A. A., Klijn, N.,

Bindels, J. G. and Welling, G. W., 2000. Analysis of intestinal flora development in breast-fed and formula-fed infants by using molecular identification and detection methods. Journal of Pediatric Gastroenterology and Nutrition 30, 61-67.

Hebuterne, X., 2003. Gut changes attributed to ageing: effects on intestinal microflora. Current

Opinion in Clinical Nutrition and Metabolic Care 6, 49-54. Heenan, C. N., Adams, M. C., Hosken, R. W. and Fleet, G. H., 2002. Growth medium for

culturing probiotic bacteria for applications in vegetarian food products. Food Science and Technology 35, 171-176.

Heipieper, H. J., Keweloh, H. and Rehm, H. J., 1991. Influence of phenols on growth and

membrane-permeability of free and immobilized Escherichia coli. Applied and Environmental Microbiology 57, 1213-1217.

Hellyer, T. J., DesJardin, L. E., Hehman, G. L., Cave, M. D., Eisenach, K. D., 1999. Quantitative

analysis of mRNA as a marker for viability of Mycobacterium tuberculosis. Journal of Clinical Microbiology 37, 290-295.

Hewitt, C. J., Nebe-Von-Caron, G., 2001. An industrial application of multiparameter flow

cytometry: Assessment of cell physiological state and its application to the study of microbial fermentations. Cytometry 44, 179-187.

Higgins, M. J. and Novak, J. T., 1997. Characterization of exocellular protein and its role in

bioflocculation. Journal of Environmental Engineering 123, 479-485. Hokamp, K., Roche, F. M., Acab, M., Rousseau, M. E., Kuo, B., Goode, D., Aeschliman, D.,

Bryan, J., Babiuk, L. A., Hancock, R. E., et al., 2004. ArrayPipe: a flexible processing pipeline for microarray data. Nucleic Acids Research 32, W457-459.


- 138 -

Hooper, L. V., Xu, J., Falk, P. G., Midtvedt, T. and Gordon, J. I., 1999. A molecular sensor that allows a gut commensal to control its nutrient foundation in a competitive ecosystem. Proceedings of the National Academy of Sciences of the United States of America 96, 9833-9838.

Hooper, L. V., Wong, M. H., Thelin, A., Hansson, L., Falk, P. C. and Gordon, J. I., 2001.

Molecular analysis of commensal host-microbial relations hips in the intestine. Science 291, 881-884.

Hooper, L. V., Midtvedt, T. and Gordon, J. I., 2002. How host-microbial interactions shape the

nutrient environment of the mammalian intestine. Annual Review of Nutrition 22, 283-307.

Hopkins, M. J. and MacFarlane, G. T., 2002. Changes in predominant bacterial populations in

human faeces with age and with Clostridium difficile infection. Journal of Medical Microbiology 51, 448-454.

Hoskisson, P. A. and Hobbs, G., 2005. Continuous culture - making a comeback? Microbiology

151, 3153-3159. Hubalek, Z., 2003. Protectants used in the cryopreservation of microorganisms. Cryobiology 46,

205-229. Hundt, S., Zaigler, A., Lange, C., Soppa, J., Klug, G., 2007. Global analysis of mRNA decay in

Halobacterium salinarum NRC-1 at single-gene resolution using DNA Microarrays. Journal of Bacteriology 189, 6936-6944.

Huys, G., Vancanneyt, M., D'Haene, K., Vankerckhoven, V., Goossens, H. and Swings, J., 2006.

Accuracy of species identity of commercial bacterial cultures intended for probiotic or nutritional use. Research in Microbiology 157, 803-810.

Ibrahim, S. A. and Bezkorovainy, A., 1994. Growth promoting factors fro Bifidobacterium

longum. Journal of Food Science 59, 189-191. Ibrahim, S. A., Hassan, O. A., Salameh, M. M. and Shahbazi, A., 2005. Effect of media

composition and incubation temperatures on autoaggregation behavior of bifidobacteria. Milk Science International 60, 127-129.

Ishibashi, N. and Shimamura, S., 1993. Bifidobacteria: Research and Development in Japan.

Food Technology 47, 129-134. Isolauri, E., Arvola, T., Sutas, Y., Moilanen, E. and Salminen, S., 2000. Probiotics in the

management of atopic eczema. Clinical and Experimental Allergy 30, 1604-1610. Jacobsen, C. S., Holben, W. E., 2007. Quantification of mRNA in Salmonella sp. seeded soil and

chicken manure using magnetic capture hybridization RT-PCR. Journal of Microbiological Methods 69, 315-321.


- 139 -

Jiang, T., Mustapha, A. and Savaiano, D. A., 1996. Improvement of lactose digestion in humans by ingestion of unfermented milk containing Bifidobacterium longum. Journal of Dairy Science 79, 750-757.

Jirku, V., 1999. Whole cell immobilization as a means of enhancing ethanol tolerance. Journal of

Industrial Microbiology and Biotechnology. 22, 147-151. Johnson, J.A.C. and Etzel, M.R., 1993. Inactivation of lactic acid bacteria during spray drying.

In: Barbosa-Cánovas, G.V. and Okos, M.R. (ed.) Food Dehydration. American Institute of Chemical Engineers, New York, USA, pp. 98-107.

Johnson, J. A. C. and Etzel, M. R., 1995. Properties of Lactobacillus helveticus CNRZ-32

attenuated by spray drying, freeze drying, or freezing. Journal of Dairy Science 78, 761-768.

Johnson, D. R., Lee, P. K. H., Holmes, V. F., Alvarez-Cohen, L., 2005. An internal reference

technique for accurately quantifying specific mRNAs by real-time PCR with application to the tceA reductive dehalogenase gene. Applied and Environmental Microbiology 71, 3866-3871.

Josephson, K. L., Gerba, C. P. and Pepper, I. L., 1993. Polymerase chain-reaction detection of

nonviable bacterial pathogens. Applied and Environmental Microbiology 59, 3513-3515. Jouenne, T., Tresse, O. and Junter, G. A., 1994. Agar-entrapped bacteria as an in-vitro model of

biofilms and their susceptibility to antibiotics. FEMS Microbiology Letters 119, 237-242. Junter, G. A. and Jouenne, T., 2004. Immobilized viable microbial cells: from the process to the

proteome... or the cart before the horse. Biotechnology Advances 22, 633-658. Kastner, S., Perreten, V., Bleuler, H., Hugenschmidt, G., Lacroix, C. and Meile, L., 2006.

Antibiotic susceptibility patterns and resistance genes of starter cultures and probiotic bacteria used in food. Systematic and Applied Microbiology 29, 145-155.

Keer, J. T., Birch, L., 2003. Molecular methods for the assessment of bacterial viability. Journal

of Microbiological Methods 53, 175-183. Kheadr, E., Dabour, N., Le Lay, C., Lacroix, C. and Fliss, I., 2007. Antibiotic susceptibility

profile of bifidobacteria as affected by oxgall, acid, and hydrogen peroxide stress. Antimicrobial Agents and Chemotherapy 51, 169-174.

Kim, S. S. and Bhowmik, S. R., 1990. Survival of lactic acid bacteria during spray drying of

plain yogurt. Journal of Food Science 55, 1008-1010. Kim, T. B., Song, S. H., Kang, S. C. and Oh, D. K., 2003. Quantitative comparison of lactose and

glucose utilization in Bifidobacterium longum cultures. Biotechnology Progress 19, 672-675.


- 140 -

King, V. A. E. and Su, J. T., 1993. Dehydration of Lactobacillus acidophilus. Process Biochemistry 28, 47-52.

Klein, S. M., Elmer, G. W., McFarland, L. V., Surawicz, C. M. and Levy, R. H., 1993. Recovery

and elimination of the biotherapeutic agent, Saccharomyces boulardii, in healthy human volunteers. Pharmaceutical Research 10, 1615-1619.

Klein, G., Pack, A., Bonaparte, C. and Reuter, G., 1998. Taxonomy and physiology of probiotic

lactic acid bacteria. International Journal of Food Microbiology 41, 103-125. Klijn, A., Mercenier, A. and Arigoni, F., 2005. Lessons from the genomes of bifidobacteria.

FEMS Microbiology Reviews 29, 491-509. Knorr, D., 1998. Technology aspects related to microorganisms in functional foods. Trends in

Food Science and Technology 9, 295-306. Koch SR: Effects of fermentation conditions on viability, physiological and technological

characteristics of autolytic dried direct vat set lactic starter cultures. PhD Dissertation ETH Nr.16909. Zurich: ETH Zurich; 2006

Kolter, R., 1999. Growth in studying the cessation of growth. Journal of Bacteriology 181, 697-

699. Koop, H. M., Valentijnbenz, M., Amerongen, A. V. N., Roukema, P. A. and Degraaff, J., 1989.

Aggregation of 27 oral bacteria by human whole saliva – Influence of culture medium, calcium, and bacterial-cell concentration, and interference by autoaggregation. Antonie Van Leeuwenhoek Journal of Microbiology 55, 277-290.

Kos, B., Suskovic, J., Vukovic, S., Simpraga, M., Frece, J. and Matosic, S., 2003. Adhesion and

aggregation ability of probiotic strain Lactobacillus acidophilus M92. Journal of Applied Microbiology 94, 981-987.

Krasaekoopt, W., Bhandari, B. and Deeth, H., 2003. Evaluation of encapsulation techniques of

probiotics for yoghurt. International Dairy Journal 13, 3-13. Krisch, J. and Szajani, B., 1997. Ethanol and acetic acid tolerance in free and immobilized cells

of Saccharomyces cerevisiae and Acetobacter aceti. Biotechnology Letters 19, 525-528. Krishnan, S., Gowda, L. R., Misra, M. C. and Karanth, N. G., 2001. Physiological and

morphological changes in immobilized L-plantarum NCIM 2084 cells during repeated batch fermentation for production of lactic acid. Food Biotechnology. 15, 193-202.

Kullen, M. J., Amann, M. M., O’ Shaughnessy, M. J., O’ Sullivan, D. J., Busta, F. F. and Brady,

L. J., 1996. Differentiation of ingested and endogenous bifidobacteria by DNA fingerprinting demonstrates the survival of an unmodified strain in the gastrointestinal tract of humans. Journal of Nutrition 127, 89-94.


- 141 -

Kullen, M. J. and Klaenhammer, T. R., 1999. Identification of the pH-inducible, proton-translocating F1F0-ATPase (atpBEFHAGDC) operon of Lactobacillus acidophilus by differential display: gene structure, cloning and characterization. Molecular Microbiology 33, 1152-1161.

Lacroix, C., Grattepanche, F., Doleyres, Y., & Bergmaier, D. (2005). Immobilized cell

technology for the dairy industry. In V. Nevidovic, and R. Willaert (Eds.), Cell immobilization biotechnology. focus in biotechnology. Springer: Heidelberg and Berlin, Germany.

Lacroix, C. and Yidirim, S., 2007. Fermentation technologies for the production of probiotics

with high viability and functionality. Current Opinion in Biotechnology 18, 176-183. Lahtinen, S. J., Ouwehand, A. C., Reinikainen, J. P., Korpela, J. M., Sandholm, J. and Salminen,

S. J., 2006. Intrinsic properties of so-called dormant probiotic bacteria, determined by flow cytometric viability assays. Applied and Environmental Microbiology 72, 5132-5134.

Lahtinen, S. J., Ahokoski, H., Reinikainen, J. P., Gueimonde, M., Nurmi, J., Ouwehand, A. C.,

Salminen, S. J., 2008. Degradation of 16S rRNA and attributes of viability of viable but nonculturable probiotic bacteria. Letters in Applied Microbiology 46, 693-698.

Lamboley, L., Lacroix, C., Champagne, C. P. and Vuillemard, J. C., 1997. Continuous mixed

strain mesophilic lactic starter production in supplemented whey permeate medium using immobilized cell technology. Biotechnology and Bioengineering 56, 502-516.

Lamboley, L., Lacroix, C., Artignan, J. M., Champagne, C. P. and Vuillemard, J. C., 1999. Long-

term mechanical and biological stability of an immobilized cell reactor for continuous mixed-strain mesophilic lactic starter production in whey permeate. Biotechnology Progress 15, 646-654.

Lammers, K. M., Brigidi, P., Vitali, B., Gionchetti, P., Rizzello, F., Caramelli, E., Matteuzzi, D.

and Campieri, M., 2003. Immunomodulatory effects of probiotic bacteria DNA: IL-1 and IL-10 response in human peripheral blood mononuclear cells. FEMS Immunology and Medical Microbiology 38, 165-172.

Leahy, S. C., Higgins, D. G., Fitzgerald, G. F. and van Sinderen, D., 2005. Getting better with

bifidobacteria. Journal of Applied Microbiology 98, 1303-1315. Lee, K. Y. and Heo, T. R., 2000. Survival of Bifidobacterium longum immobilized in calcium

alginate beads in simulated gastric juices and bile salt solution. Applied and Environmental Microbiology 66, 869-873.

Lee, H. Y., Park, J. H., Seik, S. H., Cho, S. A., Baek, M. W., Kim, D. J. and Lee, Y. H., 2004.

Dietary intake of various lactic acid bacteria suppresses type 2 helper T cell production in antigen-primed mice splenocyte. Journal of Microbiology and Biotechnology 14, 167-170.


- 142 -

Lee, J.H., Karamychev, V.N., Kozyavkin, S.A., Mills, D., Pavlov, A.R., Pavlova, N.V., Polouchine, N.N., Richardson, P.M., Shakhova, V.V., Slesarev, A.I., Weimer, B. and O'Sullivan, D.J., 2008. Comparative genomic analysis of the gut bacterium Bifidobacterium longum reveals loci susceptible to deletion during pure culture growth. BMC Genomics 9, 247.

Leslie, S. B., Israeli, E., Lighthart, B., Crowe, J. H. and Crowe, L. M., 1995. Trehalose and

sucrose protect both membranes and proteins in intact bacteria during drying. Applied and Environmental Microbiology 61, 3592-3597.

Ley, R. E., Peterson, D. A. and Gordon, J. I., 2006. Ecological and evolutionary forces shaping

microbial diversity in the human intestine. Cell 124, 837-848. Leyer, G. J., Li, S. G., Mubasher, M. E., Reifer, C. and Ouwehand, A. C., 2009. Probiotic effects

on cold and influenza-like symptom incidence and duration in children. Pediatrics 124, E172-E179.

Lian, W. C., Hsiao, H. C. and Chou, C. C., 2002. Survival of bifidobacteria after spray-drying.

International Journal of Food Microbiology 74, 79-86. Lilly, D. M. and Stillwel.Rh, 1965. Probiotics – Growth promoting factors produced by

microorganisms. Science 147, 747- 748. Lindner, J. D. D., Canchaya, C., Zhang, Z., Neviani, E., Fitzgerald, G. F., van Sinderen, D. and

Ventura, M., 2006. Exploiting Bifidobacterium genomes: The molecular basis of stress response. International Journal of Food Microbiology120, 13-24.

MacConaill, L.E., Butler, D., O'Connell-Motherway, M., Fitzgerald, G.F. and van Sinderen, D.,

2003. Identification of two-component regulatory systems in Bifidobacterium infantis by functional complementation and degenerate PCR approaches. Applied and Environmental Microbiology 69:4219-4226.

Mangin, I., Bourget, N., Bouhnik, Y., Bisetti, N., Simonet, J. M., Decaris, B., 1994. Identification

of Bifidobacterium strains by rRNA gene restriction patterns. Applied and Environmental Microbiology 60, 1451-1458.

Marteau, P., Pochart, P., Dore, J., Bera-Maillet, C., Bernalier, A. and Corthier, G., 2001.

Comparative study of bacterial groups within the human cecal and fecal microbiota. Applied and Environmental Microbiology 67, 4939-4942.

Masco, L., Huys, G., Gevers, D., Verbrugghen, L. and Swings, J., 2003. Identification of

Bifidobacterium species using rep-PCR fingerprinting. Systematic and Applied Microbiology 26, 557-563.

Masco, L., Huys, G., De Brandt, E., Temmerman, R. and Swings, J., 2005. Culture-dependent

and culture-independent qualitative analysis of probiotic products claimed to contain bifidobacteria. International Journal of Food Microbiology 102, 221-230.


- 143 -

Masco, L., Vanhoutte, T., Temmerman, R., Swings, J. and Huys, G., 2007. Evaluation of real-time PCR targeting the 16S rRNA and recA genes for the enumeration of bifidobacteria in probiotic products. International Journal of Food Microbiology 113, 351-357.

Masson, F., Lacroix, C. and Paquin, C., 1994. Direct measurement of pH profiles in gel beads

immobilizing Lactobacillus helveticus using a pH sensitive microelectrode. Biotechnology Techniques 8, 551-556.

Matsuda, K., Tsuji, H., Asahara, T., Kado, Y. and Nomoto, K., 2007. Sensitive quantitative

detection of commensal bacteria by rRNA-targeted reverse transcription-PCR. Applied and Environmental Microbiology 73, 32-39.

Matsuki, T., Watanabe, K., Fujimoto, J., Takada, T. and Tanaka, R., 2004. Use of 16S rRNA

gene-targeted group-specific primers for real-time PCR analysis of predominant bacteria in human feces. Applied and Environmental Microbiology 70, 7220-7228.

Mättö, J., Alakomi, H.-L., Vaari, A., Virkajärvi, K. and Saarela, M., 2006. Influence of

processing conditions on Bifidobacterium animalis subsp lactis functionality with a special focus on acid tolerance and factors affecting it. International Dairy Journal 16, 1029-1037.

Maukonen, J., Alakomi, H. L., Nohynek, L., Hallamaa, K., Leppamaki, S., Mättö, J. and Saarela,

M., 2006. Suitability of the fluorescent techniques for the enumeration of probiotic bacteria in commercial non-dairy drinks and in pharmaceutical products. Food Research International 39, 22-32.

Mauriello, G., Aponte, M., Andolfi, R., Moschetti, G. and Villani, F., 1999. Spray-drying of

bacteriocin-producing lactic acid bacteria. Journal of Food Protection 62, 773-777. Maus, J. E. and Ingham, S. C., 2003. Employment of stressful conditions during culture

production to enhance subsequent cold- and acid-tolerance of bifidobacteria. Journal of Applied Microbiology 95, 146-154.

Medina, M., Izquierdo, E., Ennahar, S. and Sanz, Y., 2007. Differential immunomodulatory

properties of Bifidobacterium longum strains: relevance to probiotic selection and clinical applications. Clinical and Experimental Immunology 150, 531-538.

Meng, X. C., Stanton, C., Fitzgerald, G. F., Daly, C. and Ross, R. P., 2008. Anhydrobiotics: The

challenges of drying probiotic cultures. Food Chemistry 106, 1406-1416. Modesto, M., Biavati, B. and Mattarelli, P., 2006. Occurrence of the family Bifidobacteriaceae in

human dental caries and plaque. Caries Research 40, 271-276. Moorman, M. A., Thelemann, C. A., Zhou, S. Y., Pestka, J. J., Linz, J. E. and Ryser, E. T., 2008.

Altered Hydrophobicity and membrane composition in stress-adapted Listeria innocua. Journal of Food Protection 71, 182-185.


- 144 -

Morgan, C.A., Herman, N., White, P.A. and Vesey, G., 2006. Preservation of micro-organisms by drying; A review. Journal of Microbiological Methods 66, 183-193.

Mozzetti, V. Novel technological approaches to en chance stress tolerance of Bifidobacterium

longum NCC2705 cells using continuous cultures. PhD Dissertation ETH Nr.18419. Zurich: ETH Zurich; 2009.

Nagawa, M., Nakabayashi, A. and Fujino, S., 1988. Preparation of the bifidus milk powder.

Journal of Dairy Science 71, 1777-1782. National Committee for Clinical Laboratory Standards (1991) Antimicrobial susceptibility

testing. 3rd edn. Villanova, PA: National Committee for Clinical Laboratory Standards. Nebe-von-Caron, G., Stephens, P. J., Hewitt, C. J., Powell, J. R. and Badley, R. A., 2000.

Analysis of bacterial function by multi-colour fluorescence flow cytometry and single cell sorting. Journal of Microbiological Methods 42, 97-114.

Nielsen, D. S., Moller, P. L., Rosenfeldt, V., Paerregaard, A., Michaelsen, K. F. and Jakobsen,

M., 2003. Case study of the distribution of mucosa-associated Bifidobacterium species, Lactobacillus species, and other lactic acid bacteria in the human colon. Applied and Environmental Microbiology 69, 7545-7548.

Norton, S., Lacroix, C. and Vuillemard, J. C., 1994. Kinetic –study of continuous whey permeate

fermentation by immobilized Lactobacillus helveticus for lactic-acid production. Enzyme and Microbial Technology 16, 457-466.

O'Mahony, L., McCarthy, J., Kelly, P., Hurley, G., Luo, F. Y., Chen, K. S., O'Sullivan, G. C.,

Kiely, B., Collins, J. K., Shanahan, F., et al., 2005. Lactobacillus and Bifidobacterium in irritable bowel syndrome: Symptom responses and relationship to cytokine profiles. Gastroenterology 128, 541-551.

O'Riordan, K., Andrews, D., Buckle, K. and Conway, P., 2001. Evaluation of microencapsulation

of a Bifidobacterium strain with starch as an approach to prolonging viability during storage. Journal of Applied Microbiology 91, 1059-1066.

Oliver, A. E., Crowe, L. M. and Crowe, J. H., 1997. Methods for dehydration-tolerance:

Depression of the phase transition temperature in dry membranes and carbohydrate vitrification. Seed Science Research 8, 211-221.

Ouellette, V., Chevalier, P. and Lacroix, C., 1994. Continuous fermentation of a supplemented

milk with immobilized Bifidobacterium infantis. Biotechnology Techniques 8, 45-50. Ouwehand, A. C., Salminen, S. J., 1998. The health effects of cultured milk products with viable

and non-viable bacteria. International Dairy Journal 8, 749-758. Ouwehand, A. C., Salminen, S. and Isolauri, E., 2002. Probiotics: an overview of beneficial

effects. Antonie Van Leeuwenhoek 82, 279-289.


- 145 -

Ouwehand, A. C., Kurvinen, T. and Rissanen, P., 2004. Use of a probiotic Bifidobacterium in a dry food matrix, an in vivo study. International Journal of Food Microbiology 95, 103-106.

Parche, S., Beleut, M., Rezzonico, E., Jacobs, D., Arigoni, F., Titgemeyer, F. and Jankovic, I.,

2006. Lactose-over-glucose preference in Bifidobacterium longum NCC2705: glcP, encoding a glucose transporter, is subject to lactose repression. Journal of Bacteriology 188, 1260-1265.

Parche, S., Amon, J., Jankovic, I., Rezzonico, E., Beleut, M., Barutcu, H., Schendel, I., Eddy,

M.P., Burkovski, A., Arigoni, F., and Titgemeyer, F., 2007. Sugar transport systems of Bifidobacterium longum NCC2705. Journal of Molecular Microbiology and Biotechnology 12, 9-19.

Paul, E., Fages, J., Blanc, P., Goma, G. and Pareilleux, A., 1993. Survival of alginate-entrapped

cells of Azospirillum lipoferum during dehydration and storage in relation to water properties. Applied Microbiology and Biotechnology 40, 34-39.

Perez, P. F., Minnaard, Y., Disalvo, E. A. and De Antoni, G. L., 1998. Surface properties of

bifidobacterial strains of human origin. Applied and Environmental Microbiology 64, 21-26

Petschow, B. W. and Talbott, R. D., 1990. Growth promotion of Bifidobacterium species by

whey and casein fractions from human and bovine-milk. Journal of Clinical Microbiology 28, 287-292.

Picot, A. and Lacroix, C., 2004. Encapsulation of bifidobacteria in whey protein-based

microcapsules and survival in simulated gastrointestinal conditions and in yoghurt. International Dairy Journal 14, 505-515.

Poch, M. and Bezkorovainy, A., 1991. Bovine-milk κ-casein trypsin digest is a growth enhancer

for the genus Bifidobacterium. Journal of Agricultural and Food Chemistry 39, 73-77. Poirier, I., Maréchal, P. A., Richard, S. and Gervais, P., 1999. Saccharomyces cerevisiae viability

is strongly dependant on rehydration kinetics and the temperature of dried cells. Journal of Applied Microbiology 86, 87-92.

Prestrelski, S. J., Tedeschi, N., Arakawa, T. and Carpenter, J. F., 1993. Dehydration-induced

conformational transitions in proteins and their inhibition by stabilizers. Biophysical Journal 65, 661-671.

Prévost, H., Diviès, C. and Rousseau, E., 1985. Continuous yogurt production with Lactobacillus

bulgaricus and Streptococcus thermophilus entrapped in calcium alginate. Biotechnology Letters 7, 247-252.


- 146 -

Prévost, H. and Diviès, C., 1987. Fresh fermented cheese production with continuous pre-fermented milk by a mixed culture of mesophilic lactic streptococci entrapped in calcium alginate. Biotechnology Letters 9, 789-794.

Prévost, H. and Diviès, C., 1992. Cream fermentation by a mixed culture of lactococci entrapped

in 2-layer calcium alginate gel beads. Biotechnology Letters 14, 583-588. Price, C. E., Reid, S. J., Driessen, A. J. M. and Abratt, V. R., 2006. The Bifidobacterium longum

NCIMB 702259(T) ctr gene codes for a novel cholate transporter. Applied and Environmental Microbiology 72, 923-926.

Rahman, M. M., Kim, W. S., Kumura, H. and Shimazaki, K., 2008. Autoaggregation and surface

hydrophobicity of bifidobacteria. World Journal of Microbiology and Biotechnology 24, 1593-1598.

Rajilic-Stojanovic, M., Smidt, H. and de Vos, W. M., 2007. Diversity of the human

gastrointestinal tract microbiota revisited. Environmental Microbiology 9, 2125-2136. Rautava, S., Salminen, S., and Isolauri, E. 2009. Specific probiotics in reducing the risk of acute

infections in infancy - a randomised, double-blind, placebo-controlled study. British Journal of Nutrition 101, 1722-1726.

Ray, B., Jezeski, J. J. and Busta, F. F., 1971. Effect of rehydration on recovery, repair, and

growth of injured freeze-dried Salmonella anatum. Applied Microbiology 22, 184-&. Redon, E., Loubiere, P., Cocaign-Bousquet, M., 2005. Role of mRNA stability during genome-

wide adaptation of Lactococcus lactis to carbon starvation. Journal of Biological Chemistry 280, 36380-36385.

Reid, G., 2008. Probiotics and prebiotics - Progress and challenges. International Dairy Journal

18, 969-975. Reilly, S. S. and Gilliland, S. E., 1999. Bifidobacterium longum survival during frozen and

refrigerated storage as related to pH during growth. Journal of Food Science 64, 714-718. Resta-Lenert, S., Barrett, K. E., 2003. Live probiotics protect intestinal epithelial cells from the

effects of infection with enteroinvasive Escherichia coli (EIEC). Gut 52, 988-997. Rezzonico, E., Lariani, S., Barretto, C., Cuanoud, G., Giliberti, G., Delley, M., Arigoni, F. and

Pessi, G., 2007. Global transcriptome analysis of the heat shock response of Bifidobacterium longum. FEMS Microbiology Letters 271, 136-145.

Riveros, B., Ferrer, J. and Borquez, R., 2009. Spray Drying of a Vaginal Probiotic Strain of

Lactobacillus acidophilus. Drying Technology 27, 123-132.


- 147 -

Ross, R. P., Desmond, C., Fitzgerald, G. F. and Stanton, C., 2005. Overcoming the technological hurdles in the development of probiotic foods. Journal of Applied Microbiology 98, 1410-1417.

Rossetti, L. and Giraffa, G., 2005. Rapid identification of dairy lactic acid bacteria by M13-

generated, RAPD-PCR fingerprint databases. Journal of Microbiological Methods 63, 135-144.

Ruas-Madiedo, P., Hernandez-Barranco, A., Margolles, A. and de los Reyes-Gavilan, C. G.,

2005. A bile salt-resistant derivative of Bifidobacterium animalis has an altered fermentation pattern when grown on glucose and maltose. Applied and Environmental Microbiology 71, 6564-6570.

Ruiz, L., Sanchez, B., Ruas-Madiedo, P., de los Reyes-Gavilan, C. G. and Margolles, A., 2007.

Cell envelope changes in Bifidobacterium animalis ssp. lactis as a response to bile. Fems Microbiology Letters 274, 316-322.

Ruiz, L., Sanchez, B., de los Reyes-Gavilan, C. G., Gueimonde, M. and Margolles, A., 2009.

Coculture of Bifidobacterium longum and Bifidobacterium breve alters their protein expression profiles and enzymatic activities. International Journal of Food Microbiology 133, 148-153

Saarela, M., Lahteenmaki, L., Crittenden, R., Salminen, S. and Mattila-Sandholm, T., 2002. Gut

bacteria and health foods--the European perspective. Int J Food Microbiol 78, 99-117. Saarela, M., Rantala, M., Hallamaa, K., Nohynek, L., Virkajärvi, I. and Mättö, J., 2004.

Stationary-phase acid and heat treatments for improvement of the viability of probiotic lactobacilli and bifidobacteria. Journal of Applied Microbiology 96, 1205-1214.

Saarela, M., Virkajärvi, I., Alakomi, H. L., Sigvart-Mattila, P. and Mättö, J., 2006. Stability and

functionality of freeze-dried probiotic Bifidobacterium cells during storage in juice and milk. International Dairy Journal 16, 1477-1482.

Salminen, S., Ouwehand, A., Benno, Y. and Lee, Y. K., 1999. Probiotics: how should they be

defined? Trends in Food Science and Technology 10, 107-110. Salminen, M. K., Tynkkynen, S., Rautelin, H., Saxelin, M., Vaara, M., Ruutu, P., Sarna, S.,

Valtonen, V. and Jarvinen, A., 2002. Lactobacillus bacteremia during a rapid increase in Probiotic use of Lactobacillus rhamnosus GG in Finland. Clinical Infectious Diseases 35, 1155-1160.

Sanchez, B., Champomier-Verges, M. C., Anglade, P., Baraige, F., Reyes-Gavilan, C. G. D.,

Margolles, A. and Zagorec, M., 2005. Proteomic analysis of global changes in protein expression during bile salt exposure of Bifidobacterium longum NCIMB 8809. Journal of Bacteriology 187, 5799-5808.


- 148 -

Sanchez, B., Champomier-Verges, M. C., Stuer-Lauridsen, B., Ruas-Madiedo, P., Anglade, P., Baraige, F., de los Reyes-Gavilan, C. G., Johansen, E., Zagorec, M. and Margolles, A., 2007. Adaptation and response of Bifidobacterium animalis subsp. lactis to bile: a proteomic and physiological approach. Applied and Environmental Microbiology 73, 6757-6767.

Sanchez, B., Ruiz, L., de los Reyes-Gavilan, C. G. and Margolles, A., 2008. Proteomics of stress

response in Bifidobacterium. Frontiers in Bioscience 13, 6905-6919. Santivarangkna, C., Kulozik, U. and Foerst, P., 2007. Alternative drying processes for the

industrial preservation of lactic acid starter cultures. Biotechnology Progress 23, 302-315. Santivarangkna, C., Kulozik, U. and Foerst, P., 2008. Inactivation mechanisms of lactic acid

starter cultures preserved by drying processes. Journal of Applied Microbiology 105, 1-13.

Savijoki, K., Suokko, A., Palva, A., Valmu, L., Kalkkinen, N. and Varmanen, P., 2005. Effect of

heat-shock and bile salts on protein synthesis of Bifidobacterium longum revealed by [S-35]methionine labelling and two-dimensional gel electrophoresis. FEMS Microbiology Letters 248, 207-215.

Saxelin, M., 2008. Probiotic formulations and applications, the current probiotics market, and

changes in the marketplace: A European perspective. Clinical Infectious Diseases 46, S76-S79.

Schachtsiek, M., Hammes, W. P. and Hertel, C., 2004. Characterization of Lactobacillus

coryniformis DSM 20001(T) surface protein Cpf mediating coaggregation with and aggregation among pathogens. Applied and Environmental Microbiology 70, 7078-7085.

Schell, M. A., Karmirantzou, M., Snel, B., Vilanova, D., Berger, B., Pessi, G., Zwahlen, M. C.,

Desiere, F., Bork, P., Delley, M., Pridmore, R.D. and Arigoni, F., 2002. The genome sequence of Bifidobacterium longum reflects its adaptation to the human gastrointestinal tract. Proceedings of the National Academy of Sciences of the United States of America 99, 14422-14427.

Schembri, M. A., Kjaergaard, K. and Klemm, P., 2003. Global gene expression in Escherichia

coli biofilms. Molecular Microbiology 48, 253-267. Schepers, A. W., Thibault, J. and Lacroix, C., 2006. Continuous lactic acid production in whey

permeate/yeast extract medium with immobilized Lactobacillus helveticus in a two-stage process: Model and experiments. Enzyme and Microbial Technology 38, 324-337.

Schmidt, G. and Zink, R., 2000. Basic features of the stress response in three species of

bifidobacteria: B. longum, B. adolescentis, and B. breve. International Journal of Food Microbiology 55, 41-45.


- 149 -

Sela, D.A., Chapman, J., Adeuya, A., Kim, J.H., Chen, F., Whitehead, T.R., Lapidus, A., Rokhsar, D.S., Lebrilla, C.B., German, J.B., Price, N.P., Richardson, P.M., and Mills, D.A., 2008. The genome sequence of Bifidobacterium longum subsp. infantis reveals adaptations for milk utilization within the infant microbiome. Proceedings of the National Academy of Sciences 105, 18964-18969.

Shah, N. P. and Ravula, R. R., 2000. Microencapsulation of probiotic bacteria and their survival

in frozen fermented dairy desserts. Australian Journal of Dairy Technology 55, 139-144. Shahidi, F. and Han, X. Q., 1993. Encapsulation of food ingredients. Critical Reviews in Food

Science and Nutrition 33, 501-547. Shanahan, F., 2002. The host-microbe interface within the gut. Best Practice and Research in

Clinical Gastroenterology 16, 915-931. Sheehan, V. M., Sleator, R. D., Fitzgerald, G. F. and Hill, C., 2006. Heterologous expression of

BetL, a betaine uptake system, enhances the stress tolerance of Lactobacillus salivarius UCC118. Applied and Environmental Microbiology 72, 2170-2177.

Sheridan, G. E. C., Masters, C. I., Shallcross, J. A. and Mackey, B. M., 1998. Detection of

mRNA by reverse transcription PCR as an indicator of viability in Escherichia coli cells. Applied and Environmental Microbiology 64, 1313-1318.

Shimakawa, Y., Matsubara, S., Yuki, N., Ikeda, M. and Ishikawa, F., 2003. Evaluation of

Bifidobacterium breve strain Yakult-fermented soymilk as a probiotic food. International Journal of Food Microbiology 81, 131-136.

Silva, A. M., Bambirra, E. A., Oliveira, A. L., Souza, P. P., Gomes, D. A., Vieira, E. C. and

Nicoli, J. R., 1999. Protective effect of bifidus milk on the experimental infection with Salmonella enteritidis subsp. typhimurium in conventional and gnotobiotic mice. Journal of Applied Microbiology 86, 331-336.

Silva, J., Carvalho, A. S., Teixeira, P. and Gibbs, P. A., 2002. Bacteriocin production by spray-

dried lactic acid bacteria. Letters in Applied Microbiology 34, 77-81. Silva, J., Carvalho, A. S., Pereira, H., Teixeira, P. and Gibbs, P. A., 2004. Induction of stress

tolerance in Lactobacillus delbrueckii ssp bulgaricus by the addition of sucrose to the growth medium. Journal of Dairy Research 71, 121-125.

Simpson, P. J., Stanton, C., Fitzgerald, G. F. and Ross, R. P., 2005. Intrinsic tolerance of

Bifidobacterium species to heat and oxygen and survival following spray drying and storage. Journal of Applied Microbiology 99, 493-501.

Snydman, D. R., 2008. The safety of probiotics. Clinical Infectious Disease 46, 104–111.


- 150 -

Sodini, I., Boquien, C. Y., Corrieu, G. and Lacroix, C., 1997. Use of an immobilized cell bioreactor for the continuous inoculation of milk in fresh cheese manufacturing. Journal of Industrial Microbiology and Biotechnology 18, 56-61.

Soejima, T., Iida, K. I., Qin, T., Taniai, H., Seki, M., Yoshida, S. I., 2008. Method to detect only

live bacteria during PCR amplification. Journal of Clinical Microbiology 46, 2305-2313. Stanley, N. R. and Lazazzera, B. A., 2004. Environmental signals and regulatory pathways that

influence biofilm formation. Molecular Microbiology 52, 917-924. Stevens, M. J. A., Wiersma, A., de Vos, W. M., Kuipers, O. P., Smid, E. J., Molenaar, D.,

Kleerebezem, M., 2008. Improvement of Lactobacillus plantarum aerobic growth as directed by comprehensive transcriptome analysis. Applied and Environmental Microbiology 74, 4776-4778.

Sturme, M. H. J., Kleerebezem, M., Nakayama, J., Akkermans, A. D., Vaughan, E. E. and de

Vos, W. M., 2002. Cell to cell communication by autoinducing peptides in gram-positive bacteria. Antonie Van Leeuwenhoek 81, 233-243.

Su, P., Henriksson, A., Tandianus, J. E., Park, J. H., Foong, F. and Dunn, N. W., 2005. Detection

and quantification of Bifidobacterium lactis LAFTI (R) B94 in human faecal samples from a consumption trial. FEMS Microbiology Letters 244, 99-103.

Sugimoto, S., Abdullah Al, M. and Sonomoto, K., 2008. Molecular Chaperones in Lactic Acid

Bacteria: Physiological Consequences and Biochemical Properties. Journal of Bioscience and Bioengineering 106, 324-336.

Sultana, K., Godward, G., Reynolds, N., Arumugaswamy, R., Peiris, P. and Kailasapathy, K.,

2000. Encapsulation of probiotic bacteria with alginate-starch and evaluation of survival in simulated gastrointestinal conditions and in yoghurt. International Journal of Food Microbiology 62, 47-55.

Sun, W. R. and Griffiths, M. W., 2000. Survival of bifidobacteria in yogurt and simulated gastric

juice following immobilization in gellan-xanthan beads. International Journal of Food Microbiology 61, 17-25.

Sunny-Roberts, E. O. and Knorr, D., 2009. The protective effect of monosodium glutamate on

survival of Lactobacillus rhamnosus GG and Lactobacillus rhamnosus E-97800 (E800) strains during spray-drying and storage in trehalose-containing powders. International Dairy Journal 19, 209-214.

Takayama, K., Kjelleberg, S. A., 2000. The role of RNA stability during bacterial stress

responses and starvation. Environmental Microbiology 2, 355-365. Talwalkar, A. and Kailasapathy, K., 2003. Effect of microencapsulation on oxygen toxicity in

probiotic bacteria. Australian Journal of Dairy Technology 58, 36-39.


- 151 -

Tamime, A. Y., Marshall, V. M. E. and Robinson, R. K., 1995. Microbiological and technological aspects of milks fermented by bifidobacteria. Journal of Dairy Research 62, 151-187.

Tannock, G. W., 1999. Analysis of the intestinal microflora: a renaissance. Antonie Van

Leeuwenhoek 76, 265-278. Teixeira, P. C., Castro, M. H. and Kirby, R. M., 1995a. Death kinetics of Lactobacillus

bulgaricus in spray drying process. Journal of Food Protection 58, 934-936. Teixeira, P. C., Castro, M. H., Malcata, F. X. and Kirby, R. M., 1995b. Survival of Lactobacillus

delbrueckii ssp. bulgaricus following spray drying. Journal of Dairy Science 78, 1025-1031.

Teixeira, P., Castro, H. and Kirby, R., 1996. Evidence of membrane lipid oxidation of spray-

dried Lactobacillus bulgaricus during storage. Letters in Applied Microbiology 22, 34-38. To, B. C. S. and Etzel, M. R., 1997. Survival of Brevibacterium linens (ATCC 9174) after spray

drying, freeze drying, or freezing. Journal of Food Science 62, 167-170. TolkerNielsen, T., Larsen, M. H., Kyed, H., Molin, S., 1997. Effects of stress treatments on the

detection of Salmonella typhimurium by in situ hybridization. International Journal of Food Microbiology 35, 251-258.

Trauth, E., Lemaitre, J. P., Rojas, C., Diviès, C. and Cachon, R. M., 2001. Resistance of

immobilized lactic acid bacteria to the inhibitory effect of quaternary ammonium sanitizers. Food Science and Technology 34, 239-243.

US Patent 2003/0109025. Particles containing coated living microorganisms, and method for

producing same. Van der Meulen, R., Adriany, T., Verbrugghe, K. and De Vuyst, L., 2006. Kinetic analysis of

bifidobacterial metabolism reveals a minor role for succinic acid in the regeneration of NAD(+) through its growth-associated production. Applied and Environmental Microbiology 72, 5204-5210.

Vaughan, E. E., de Vries, M. C., Zoetendal, E. G., Ben-Amor, K., Akkermans, A. D. L. and de

Vos, W. M., 2002. The intestinal LABs. Antonie Van Leeuwenhoek 82, 341-352. Vaughan, E. E., Heilig, H. G., Ben-Amor, K. and de Vos, W. M., 2005. Diversity, vitality and

activities of intestinal lactic acid bacteria and bifidobacteria assessed by molecular approaches. FEMS Microbiology Reviews 29, 477-490.

Vega, C. and Roos, Y. H., 2006. Invited review: Spray-dried dairy and dairy-like - emulsions

compositional considerations. Journal of Dairy Science 89, 383-401.


- 152 -

Ventling, B. L. and Mistry, V. V., 1993. Growth characteristics of bifidobacteria in ultrafiltered milk. Journal of Dairy Sciences 76, 962-971.

Ventura, M., Canchaya, C., Del Casale, A., Dellaglio, F., Neviani, E., Fitzgerald, G. F. and van

Sinderen, D., 2006. Analysis of bifidobacterial evolution using a multilocus approach. International Journal of Systematic and Evolutionary Microbiology 56, 2783-2792.

Ventura, M., O'Connell-Motherway, M., Leahy, S., Moreno-Munoz, J. A., Fitzgerald, G. F. and

van Sinderen, D., 2007. From bacterial genome to functionality; case bifidobacteria. International Journal of Food Microbiology 120, 2-12.

Vincent, D., Roy, D., Mondou, F. and Dery, C., 1998. Characterization of bifidobacteria by

random DNA amplification. International Journal of Food Microbiology 43, 185-193. Vitali, B., Candela, M., Matteuzzi, D. and Brigidi, P., 2003. Quantitative detection of probiotic

Bifidobacterium strains in bacterial mixtures by using real-time PCR. Systematic and Applied Microbiology 26, 269-276.

Wallner, G., Erhart, R., Amann, R., 1995. Flow cytometric analysis of activated-sludge with

ribosomal-RNA-targeted probes. Applied and Environmental Microbiology 61, 1859-1866.

Wang, Y. C., Yu, R. C. and Chou, C. C., 2004. Viability of lactic acid bacteria and bifidobacteria

in fermented soymilk after drying, subsequent rehydration and storage. International Journal of Food Microbiology 93, 209-217.

Weizman, Z., Asli, G. and Alsheikh, A., 2005. Effect of a probiotic infant formula on infections

in child care centers: comparison of two probiotic agents. Pediatrics 115, 5-9. Whiteley, M., Bangera, M. G., Bumgarner, R. E., Parsek, M. R., Teitzel, G. M., Lory, S. and

Greenberg, E. P., 2001. Gene expression in Pseudomonas aeruginosa biofilms. Nature 413, 860-864.

Whitman, W. B., Coleman, D. C. and Wiebe, W. J., 1998. Prokaryotes: The unseen majority.

Proceedings of the National Academy of Sciences of the United States of America 95, 6578-6583.

Wilson, D. N. and Nierhaus, K. H., 2005. Ribosomal proteins in the spotlight. Critical Reviews in

Biochemistry and Molecular Biology 40, 243-267. Wolber, P. K., Collins, P. J., Lucas, A. B., De Witte, A. and Shannon, K. W., 2006. The Agilent

in situ-synthesized microarray platform. Methods in Enzymology 410, 28-57. Wollowski, I., Rechkemmer, G. and Pool-Zobel, B. L., 2001. Protective role of probiotics and

prebiotics in colon cancer. American Journal of Clinical Nutrition, 73, 451–455.


- 153 -

Woodmansey, E. J., McMurdo, M. E. T., Macfarlane, G. T. and Macfarlane, S., 2004. Comparison of compositions and metabolic activities of fecal microbiotas in young adults and in antibiotic-treated and non-antibiotic-treated elderly subjects. Applied and Environmental Microbiology 70, 6113-6122.

Wouters, J. A., Rombouts, F. M., de Vos, W. M., Kuipers, O. P. and Abee, T., 1999. Cold shock

proteins and low-temperature response of Streptococcus thermophilus CNRZ302. Applied and Environmental Microbiology 65, 4436-4442.

Zamora, L. M., Carretero, C. and Pares, D., 2006. Comparative survival rates of lactic acid

bacteria isolated from blood, following spray-drying and freeze-drying. Food Science and Technology International 12, 77-84.

Zavaglia, A. G., Kociubinski, G., Perez, P., Disalvo, E. and De Antoni, G., 2002. Effect of bile

on the lipid composition and surface properties of bifidobacteria. Journal of Applied Microbiology 93, 794-799.

Zayed, G. and Roos, Y. H., 2004. Influence of trehalose and moisture content on survival of

Lactobacillus salivarius subjected to freeze-drying and storage. Process Biochemistry. 39, 1081-1086.

Zoetendal, E. G., Collier, C. T., Koike, S., Mackie, R. I. and Gaskins, H. R., 2004. Molecular

ecological analysis of the gastrointestinal microbiota: a review. Journal of Nutrition 134, 465-472.

Zoetendal, E. G., Vaughan, E. E. and de Vos, W. M., 2006. A microbial world within us.

Molecular Microbiology 59, 1639-1650.

Curriculum vitae

- 154 -

Acknowledgements My thanks go to Christophe Lacroix, Leo Meile, Enea Rezzonico and Bernard Berger for the

supervision during this project. Thank you also Paul Ross who accepted to co-examine this thesis

and provide his comments. A special thank to Franck Grattepanche for patient supervision and

proof reading my manuscripts and also for sharing taste of (most) of my music library. Many

thanks to my diploma students Rafael Benz, who spend many hours beside continuous culture

and Claudia Baggenstos for precise work and endurance during spray drying and storage

experiments. My thanks go also to Valeria Mozzetti, for good cooperation during 3 ½ years, I

wish you and your young family all the best for the future.

Many thanks to past and present colleagues of the Laboratory of Food Biotechnology who

contributed with discussions, knowledge or only by mere presence to this work, especially to

Mark Stevens, Alfonso Die and of course to Frau Venakis: Danke Kiki, für die Geduld mit mir!

A heartfelt thanks goes to my family and friends, especially to my mother Sybille who supported

me always in every situation. A very special thanks to my father, Alois, who left us so

unexpectedly 3 days after I started my PhD project and with whom I would have liked to share

good and bad times during the last four years.

At the very end I want to express my deepest thanks to my wife, Maria, for your love and

confidence and patience to share your life with a “rab lat”.

This study was carried out thanks to the financial support of the Commission of Technology and

Innovation of Switzerland (CTI-Project Nr. 75272 LSPP-LS), Nestlé (Switzerland) and the

Walter Hochstrasser Foundation.

Documents

Rights / License: Research Collection In Copyright - Non ...492/eth-492-02.pdfAccepted on the recommendation of Prof. Dr. Christophe Lacroix, examiner Prof. Dr. Leo Meile, co-examiner