Quality control of Saccharomyces

yeasts: differentiation of species level

and strain grouping using COX 2 gene

analysis and MALDI-TOF-MS analysis

Hiroko Kawasaki1, Shinya Ninomiya1, Kanae Teramoto2, Hiroaki Sato3, Ken-ichiro Suzuki11NITE Biological Resource Center (NBRC), National Institute of Technology and Evaluation (NITE)2JEOL Ltd., 3Resarch Institute for Environmental Management Technology, National Institute of Advanced Industrial Science and Technology (AIST)

Background

Genus Saccharomyces

The Genus Saccharomyces plays an important role in

human activities. They are used as fermenting agents

worldwide and stand out as model eukaryotic organisms in

various fields of biological science ranging from

biochemistry to genomics.

Taxonomy of Saccharomyces

The modern phylogenetic concept of Saccharomyces

based on D1/D2 domain of LSU rDNA, ITS1-5.8S-ITS2,

EF-1α, Mt Sm rDNA and COX2, restricts this genus to

eight species: Saccharomyces cerevisiae, S. bayanus,

S. cariocanus, S. kudriavzevii, S. mikatae, S.

paradoxus, S. pastorianus, S. arborilus. Some of the

Saccharomyces species were transferred to three

genera Kazachstania, Lachancea and Naumovia.

Objective

Saccharomyces species have various characterizations in strain

level

- yield of ethanol production

- ability for osmotolerance

- variety of organic compound etc.

In NBRC Yeast Collection

- 760 strains of Saccharomyces species are collected

- 551 strains are opened to the public

Because of the potential value of Saccharomyces yeast at the strain

level, discrimination and grouping of strains are needed.

In this study, we aimed to apply the high resolution biomarker for

grouping and identification of the genus Saccharomyces.

High resolution biological marker

1. Cytochrome oxidase subunit II (COX2)

- located on mitochondrial gene

- molecular phylogenetic analysis

2. Ribosomal subunit proteins

- one component of the cytoplasm ribosome

- comparative mass spectra analysis using MALDI-TOF MS

LSU rDNA D1/D2 domain:

- main current for identification in species level

- the guideline “representatives of different species tend to show

at least 1% divergence in this sequence." by Kurtzman & Robnett

(1998)

It is effective for the identification of general ascomycetos yeast,

but it is unsuitable for the identification of Saccharomyces species.

LSU rDNA D1/D2 Domain vs. COX2 gene

1 2 3 4 5 6 7 8 9

1. S. cerevisiae - 96 93 98 98 nd 92 93 91

2. S. cariocanus 99 (4) - 94 96 96 nd 91 93 92

3. S. kudriavzevii 99 (8) 99 (4) - 94 94 nd 92 92 88

4. S. mikatae 98 (12) 99 (8) 98 (10) - 98 nd 92 93 90

5. S. paradoxus 99 (5) 100 (1) 99 (5) 99 (7) - nd 91 92 90

6. S. arborilus 98 (10) 99 (3) 99 (5) 99 (9) 99 (4) - nd nd nd

7. S. bayanus 98 (11) 98 (9) 99 (7) 98 (11) 99 (8) 99 (8) - 97 89

8. S. pastorianus 98 (11) 98 (9) 99 (7) 98 (11) 99 (8) 99 (8) 100 (0) - 90

9. Lachanceakluyveri

93 (39) 93 93 92 93 93 93 93 -

The triangle under the left shows similarity (%) of LSU rDNA D1/D2 and the number in

parentheses shows the number of base substitution.

The upper right triangle shows similarity (%) of the COX2 gene.

The type strains were used.

LSU rDNA D1D2 vs COX2 gene

Reverse

Forward

NameNBRC

strainSequences

LSU rDNA blast

homology search

Cox2 blast

homology searchCluster

Saccharomyces

cerevisiae103978

Triple ‘T’ insertion

at position 106

bases

100% (585/585)

Saccharomyces sp.

AJ938046 CBS 8615

97% (597/611)

Saccharomyces

cerevisiae

AY244992 CBS 1171T

S. cerevisiae

cluster 1

All of the COX2 gene studied are single sequences.

LSU rDNA D1D2 tree

162510673106741067510676109591096010961109621889101844 1018110922109231092410956109571095810915021503160298117102640724117411084183810869108671091610917 10008100091085110852025407521957 0285113811391956199210862108631086410868108650211 1892189318941084710848109551019991685 0320027111281141117011082114209561169 10849108501094606761948106101097102130220025102970539057310481344162019611055710558105631102211023110241096810970109690250026506151097213432031101581054910550105511020001803 10998110001100111002109971099611003181518161099510990109910492180205751099310999109941060910695105521055310554105551055602591804180502631020011020021020031020041020051094710948100908531005519911995183418371831183202100258021702220203106936101010460206122517290718105148102006084708500218100803370614055502560225024302481021720181662028302890304030510559105601346143921142115226023422347235917271728183319581955195419531952195119501949237710011105147106934075110451089084902160971033405380268022102240234024510661101591661027402820292029303090317123412652106211221642252234523462375237610514105151001010012236323732343234421162133204320441226144403060308029002910253026116601051602490244023302230565026704900333096508770848104910140735105149072502090205101557100929

0.01

Sacch

aro

myces

NJ/294

strains 294 strains of Saccharomyces stocked in

NBRC are re-classified based on the LSU

rDNA D1D2 sequences.

83 strains were obtained to become a

member of thee genera Lachancea,

Naumovia and Kazachstania and two

unknown groups.

Saccharomyces strains form a

monophyletic cluster.

Naumovia

Lachancea

Kazachstania

067619481061010971021302200251029705390573104813441620196110557105581056311022110231102410968109701096902500265061510972134320311015810549105501055110200018031099010991049218020575 10998110001100111002109971100318151816109951099610993109991099410609106951055210553105541055510556025918041805026310200110200210200310200410200510947109481009 0203020607181069360850061402431662030514392342172819541950105147108903340224101590292123421641046 1225172910200610080256102170289105602115235919581952237707510216026802450274030921062345 0210025802170222105148084703370225201803041346226017271955195110011104509710221106610282031721122346 08531005519911995183418371831183210100218055502480283105592114234718331953194910693408490538023416610293126522522375237610514105151001010012236323732343234421162133204320441226144403060308029002910253026116601051602490244023302230565026704900333096508770848104910140735105149072502090205101557100929

0.002

162510673106741067510676109591096010961109621889101844 1018110922109231092410956109571095810915021503160298117102640724117411084183810869108671091610917 10008100091085110852025407521957 0285113811391956199210862108631086410868108650211 1892189318941084710848109551019991685 0320027111281141117011082114209561169 10849108501094606761948106101097102130220025102970539057310481344162019611055710558105631102211023110241096810970109690250026506151097213432031101581054910550105511020001803 10998110001100111002109971099611003181518161099510990109910492180205751099310999109941060910695105521055310554105551055602591804180502631020011020021020031020041020051094710948100908531005519911995183418371831183202100258021702220203106936101010460206122517290718105148102006084708500218100803370614055502560225024302481021720181662028302890304030510559105601346143921142115226023422347235917271728183319581955195419531952195119501949237710011105147106934075110451089084902160971033405380268022102240234024510661101591661027402820292029303090317123412652106211221642252234523462375237610514105151001010012236323732343234421162133204320441226144403060308029002910253026116601051602490244023302230565026704900333096508770848104910140735105149072502090205101557100929

0.01

Sacch

aro

myces

NJ/294

strains S. cerevisiae

S. mikatae

S. bayanus

S. pastorianus

S. kudriavzevii

S. cariocaus

S. paradoxus

NJ/211

strains

99

92 10096

63

73

100

9963

70

Saccharomyces species are monophyletic in the LSU r DNA tree.

LSU rDNA D1D2 tree

COX2 gene tree

0213025102650573101581055010551109721102202501947105491096810969109701097110200002200297053906761048162019481961196220032027203110010100111001210557105581061011023110241343134410563 049222261802109911803109900282028303060304030803092114213321642345237602030205101591051503340209024402490267030505551069340965126523422343234423731008108902251069360490105148105149033308771051610514 18051020051099602180222105146102170853183218341991199510055195119521955023317271728 10998110031100018150259026310553105541055510609102002102003102004102006105561055210695102001 0245025605750262101418311833225210559 10947109481009 18161099510997110011100210992109941099910993180402100217025802160224105147211521162112028902900291029202932346056510155723472377022303170337123413462043204421062375101010461958024319492018026802741729072518361837195019971998053819940718 0735097110092912251226143914441045023423592363061402060248026116601661166206151954106612260162519530848084907511049025308470221085010560

0.01

S. bayanus cluster 1

S. pastorianus & S. bayanus

cluster 1

S. kudriavzevii cluster 1

S. cerevisiae cluster 1

S. cerevisiae cluster 2

S. mikatae cluster 1

S. cariocaus cluster 1

S. mikatae cluster 2

S. mikatae cluster 3

S. cerevisiae cluster 4

S. paradoxus cluster 1

S. paradoxus cluster 2

S. cerevisiae cluster 5

S. kudriavzevii cluster 2

NJ /224

strains

S. cerevisiae cluster 3

S. cariocaus cluster 2

Grouping within S. cerevisiae based on Cox2

02100217

025802160224

105147211521162112

0289029002910292029323460565101557234723770223

031703371234134620432044210623751010

1046195802431949201802680274

17290725

18361837195019971998053819940718

07350971100929

12251226

14391444

10450234

2359236306140206024802611660166116620615

1954106612260162519530848084907511049025308470221085010560

0.005

0282028303060304030803092114213321642345237602030205

1015910515

033402090244024902670305055510693409651265234223432344237310081089

022510693604901051481051490333

087710516

10514

02180222105146

102170853183218341991199510055

195119521955

023317271728

0245025605750262

101418311833225210559

S. cerevisiae

cluster 1:

alcoholic

beverage, wine,

bear,

Japanese sake,

Chinese wine

S. cerevisiae cluster 5:

alcoholic beverage,

mainly Japanese Sake,

Starter of Chinese wine

and Thai koji

S. cerevisiae cluster 2:

white wine & fermenting

must of champagne

grapes S. cerevisiae

cluster 3:

unknown

COX 2 gene is a good molecular marker for grouping

and discrimination of Saccharomyces strains.

S. cerevisiae cluster 4:

alcoholic beverage,

mainly wine

S. mikatae vs. S. cariocanus

0213025102650573101581055010551109721102202501947105491096810969109701097110200002200297053906761048162019481961196220032027203110010100111001210557105581061011023110241343134410563 049222261802109911803109900282028303060304030803092114213321642345237602030205101591051503340209024402490267030505551069340965126523422343234423731008108902251069360490105148105149033308771051610514 1805102005109960218022210514610217085318321834199119951005519511952195502331727172810998110031100018150259026310553105541055510609102002102003102004102006105561055210695102001024502560575026210141831183322521055910947109481009181610995109971100111002109921099410999109931804021002170258021602241051472115211621120289029002910292029323460565101557234723770223031703371234134620432044210623751010104619580243194920180268027417290725183618371950199719980538199407180735097110092912251226143914441045023423592363061402060248026116601661166206151954106612260162519530848084907511049025308470221085010560

0.01

S. mikatae

cluster 1

S. mikatae

NBRC 1815T

S. mikatae cluster 3

Source Country

S. mikatae cluster 1 Tree exudate

Partially decayed leaf

soil

Japan

S. mikatae cluster 2 Soil

Partially decayed leaf

Japan

S. cariocanus cluster 1 Fruit fly Brazil

LSU rDNA

D1/D2

(%)

ITS-5.8S-

ITS2

(%)

COX2

(%)

DNA-DNA

hybrid.

(%)

S.

m.

S.

c.

S.

m.

S.

c.

S.

m.

S.

c.

S.

m.

S.

c.

S. mikatae

NBRC 1815T

100 98 100 98 100 98 100 47

S. mikatae

NBRC 10994

98 98 99 98 97 97 84 48

S. mikatae

NBRC 10999

98 98 99 98 97 97 83 33

S. cariocanus

NBRC 10947T

98 100 98 100 98 100 32 100

S. cariocanus

COX2

gene tree

Saccharomyces bayanus and S. pastrianus

02130251026505731015810550105511097211022

02501947

1054910968109691097010971102000

0220029705390676104816201948196119622003202720311001010011100121055710558106101102311024

1343134410563

Group 1

Main source:

Brewer's yeast (Lager)

Mixed with S. byanus

Group 2

Main source:

Fermenting mustAle Beer

0.01

S. cerevisiae

S. cerevisiae

Saccharomyces

pastrianus

Saccharomyces

bayanus

COX2 gene tree

2. Ribosomal Subunit Protein Analysis

using MALDI-TOF MS

Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOFMS) is an emerging tool for microbe characterization and differentiation at the species and strain levels. Early reports focused on rapid identification of intact whole bacteria. Recently the MALDI-TOFMS analysis of intact yeast cells has been reported.

On the other hand, variability of data by the difference of the culture condition is problem in the method of the above-mentioned.

To solve this problem, we focus to the ribosomal subunit proteins as the biomarker for the identification of yeast.

Method

14

1. Wet cells (24 hours culture)

2. Purification of ribosomal proteinusing ultra centrifugation

4. Mass spectral analysis of protein using MALDI-TOFMS

3. Selection of bio-markerThe theory value of the observation mass calculated based on the registration amino acid

array of each ribosome subunit protein of genome sequenced S. cerevisiae NBRC 1136 was

compared with the peak on the spectrum MALDI mass of the refinement ribosome protein.

The one within 150ppm in error margin was used as biomarkers.

6000 7000 8000 9000m/z

10000 11000 12000 13000 14000 15000 16000

x10

L39

L29

L38

S2

8-B

S3

0

S3

1

S2

8-A

L37

-AS

21

-A,B

L42

L33

-A

L36

-B

L22

-A

L31

-A

S2

0-A S

20

-B

L2

6-A

,B

S2

2-A

,BL

32

S2

4L

27

-A

S1

9-B

matrix： sinapinic acid

Range of measurement : m/z 6000-17000

S. cerevisiae NBRC1136 MALDI mass spectra

Ribosomal protein

(Teramoto et al., 2007)

Result 1: Genus level Based on the S. cerevisiae whole

genome sequence, 56 ribosome

proteins (Small subunit) and 81

ribosome proteins (Large subunit)

are presumed .

In this study, the MALDI mass

spectra were observed in the

range of m/z, and used at least 40

kinds of proteins as molecular

information.

Saccharomyces species show

clearly different molecular

patterns from other related genera,

Naumovia and Kazachstania.

Possibility to distinguish

at genus level S. bayanus

NBRC11022T

S. pastorianus

NBRC11029T

S. kudriavzevii

NBRC1802T

S. mikatae

NBRC1815T

S. cariocanus

NBRC10947T

S. paradoxus

NBRC10609T

S. cerevisiaeNBRC1136

Kazachstania unisporaNBRC215T

Naumovia dairensisNBRC211T

S. cerevisiae

NBRC10217T

7000 7800m/z

6000

L39 L29 S29B S28B S28AS30

16

Six species of Saccharomyces show the different patterns in two or more

ribosome subunit proteins. It shows the possibility to distinguish at species

level.

Differentiation among Saccharomyces species

S. pastorianus

S. bayanus

S. kudriavzevii

S. mikatae

NBRC10217

S. cerevisiaeNBRC 1136

S.

cariocanus

100

S. paradoxus

54

63

60

100

0.002

LSU

rDNA

D1/D2

tree

7000 8000 9000

S. bayanus

NBRC11022

S. pastorianus

NBRC11024

S. kudriavzevii

NBRC1802

S. mikatae

NBRC1815

S. cariocanus

NBRC10947

S. paradoxus

NBRC 10609

S. cerevisiae

NBRC1136

S. cerevisiae

NBRC10217T

m/z

S30

S28

B

S28

AS31

L38

6500 7000

L39 L29S29B

S29A

S. pastorianus

NBRC10217

S. bayanus

NBRC11022

S. kudriavzevii

NBRC1802

S. mikatae

NBRC1815

m/z

The S. mikatae-S. cariocanus immediate

group (green) can be distinguished at S24

and S30, which result corresponded to the

relationships based on the LSU rDNA

D1/D2, ITS and COX2.

Closely related strains: S. mikatae and S. cariocanus

S. mikatae

MALDI/TOFMS

tree

１００９０８０７０

NBRC 1816

NBRC 10995

NBRC 10994

NBRC 10993

NBRC 11000

NBRC 1815T

100

98

100100

98

97

99

LSU

rDNA

D1D2COX2

14600 14800 15000 15200 15400

L32 S24-

Ac

L27A

S. mikatae

NBRC10995

S. mikatae

NBRC1815

S. cariocanus

NBRC 10947T

S. mikatae

NBRC 10994

S. mikatae

NBRC 10999

m/z

The MALDI/TOF MS result corresponded

to the relationships based on the LSU

rDNA D1/D2 and COX2.

Possibility to distinguish at below

the species level

18

All of the S. cerevisiae strains have similar MALDI mass spectra.

But, some strains which have some substitutions of LSU rDNA D1D2 (1-2

bases) show the different spectra at L28 and S13.

L28 or S13 will be a biomarker for grouping?

S. cerevisiae: at below the species level

NBRC

no.

LSU rDNA

D1/D2

Similarity

1136 (Genome

strain)

-

10217 100%

0203, 0205,

2260, 10010 100%

0210 1base

substitution

1837 2base

substitution

15500 16000 16500 17000m/z

S24-AcL27A

L25S19B L28 S13

NBRC 0203

NBRC 10010

NBRC 2260

NBRC 10217

(Type strain)

NBRC 0205

NBRC 0210

NBRC 1837

NBRC 1136

(Genome strain)

Conclusions

COX2 phylogenetic analysis and ribosomal subunit mass spectra analysis

were applied to the grouping and discrimination of yeast Saccharomyces.

COX2:

The base substitution rate of the COX2 gene was larger than that of the LSU

rDNA D1/D2. More than two clusters were formed in the homogeneous group.

It show that COX2 becomes a molecular marker for grouping and

discrimination. Now we have used the COX 2 gene analysis for the quality

control of Saccharomyces group in NBRC yeast collection.

Ribosomal subunit proteins:

Several different peaks were observed in the Saccharomyces species used

by MALDI-TOF MS. This method might be useful for the identification at the

genus and species level. A few information can be used to distinguish in

homogenous group.

It will be necessary to analyze the specific ribosomal subunit proteins for

biomarker in each level. Future tasks are the promptness of the purification

of the ribosome protein and making to the data base.