ORIGINAL ARTICLE

Biostimulation of Estuarine Microbiota on Substrate CoatedAgar Slides: A Novel Approach to Study Diversityof Autochthonous Bdellovibrio- and Like Organisms

Ashvini Chauhan & Henry N. Williams

Received: 7 June 2007 /Accepted: 16 July 2007 /Published online: 30 October 2007# Springer Science + Business Media, LLC 2007

Abstract Characterization of Bdellovibrio- and like organ-isms (BALOs) from environmental samples involvesgrowing them in the presence of Gram-negative preybacteria and isolation of BALO plaques. This labor-intensive enrichment and isolation procedure may impedethe detection and phylogenetic characterization of unculti-vable BALOs. In this article, we describe a simple slidebiofilm assay to improve detection and characterization ofBALO microbiota. Agar spiked with biostimulants such asyeast extract (YE), casamino acids (CA), or concentratedcells of Vibrio parahaemolyticus P5 (most widely used preybacteria for isolation of halophilic BALOs) was plated ontobuffed glass slides and exposed to water samples collectedfrom Apalachicola Bay, Florida. After incubating for aweek, diversity of the biofilm bacterial community wasstudied by culture-dependent and culture-independent mo-lecular methods. The results revealed that most probablenumbers (MPNs) of BALOs and total culturable bacteriarecovered from YE agar slide were significantly higher thanthe numbers on CA- or P5-spiked agar slides. Polymerasechain reaction–restriction fragment length polymorphismfollowed by 16S rDNA sequencing of clones from differentbiostimulants resulted in identification of a plethora ofGram-negative bacteria predominantly from the alpha,gamma, delta-proteobacteria, and the Cytophaga–Flavo-bacterium–Bacteroides group. Corresponding to the higherbiomass on the YE agar slide, the BALO clone library from

YE was most diverse, consisting of Bacteriovorax spp. anda novel clade representing Peredibacter spp. Microbiotafrom all three biostimulated biofilms were exclusivelyGram-negative, and each bacterial guild represented poten-tial prey for BALOs. We propose the use of this simple yetnovel slide biofilm assay to study oligotrophic aquaticbacterial diversity which could also potentially be utilizedto isolate marine bacteria with novel traits.

Introduction

Bdellovibrio- and like organisms (BALOs) is a broadcategory of predatory bacteria that likely control bacterialmortality in nature by engaging in obligate predation of otherGram-negative bacteria in diverse environments such asoceans, rivers, soils, rhizosphere, sewage, biofilms, and evenin animal guts and gills [14, 15, 21, 22, 31, 32]. BALOs thatare halotolerant are assigned to Bacteriovoracaceae familyand freshwater BALOs to Bdellovibrionaceae family and allBALOs cluster within the delta-proteobacteria class. Bacter-iovoracaceae also consists of the estuarine BALO ribotypesand Peredibacter spp. [9, 22, 31]. However, the studies onthe diversity and phylogeny of estuarine BALOs are stilllimited, and this group remains somewhat taxonomicallyunresolved.

Most of the estuarine BALO phylogeny has stemmedfrom broth enrichment cultures in the presence of Gram-negative prey bacteria and isolation as plaque forming units(PFUs) on agar plates with prey. These PFUs are thenpurified by several rounds of subculturing to yield anaxenic BALO suspension and characterized [14, 21, 26,32]. This isolation approach is potentially biased such thatonly those BALOs are characterized which are culturable

Microb Ecol (2008) 55:640–650DOI 10.1007/s00248-007-9307-1

A. Chauhan (*) :H. N. WilliamsMarine Molecular Microbial Ecology Laboratory, EnvironmentalSciences Institute, Florida A&M University,1515, S. MLK Blvd., 305 FSHSRC,Tallahassee, FL 32307, USAe-mail: ashvini.chauhan@famu.edu

on the type prey, which, in most cases, is represented by asingle species of susceptible laboratory-maintained, Gram-negative bacteria. BALOs which are either uncultivable orthose having unidentified prey preferences will not beisolated or detected, and thus, will remain unrepresented insuch biodiversity studies.

We are interested in novel ways of exploiting moleculartools to characterize the diversity and phylogenetic distri-bution of BALOs in estuarine environments such asApalachicola Bay, Florida. This Bay is a fairly shallowsubtropical estuary in the northeastern Gulf of Mexico andconstitutes a part of the National Estuarine ResearchReserve system. Apalachicola Bay is considered relativelypristine, but recent surge in coastal developments along thesouthern side and sewage treatment plants along thenorthern side of the Bay pose a concern for futureeutrophication problems to this ecosystem. Therefore, itbecomes imperative to parameterize the microbial commu-nity and monitor changes reflective of impending eutrophi-cation in Apalachicola Bay.

Characterization of ecosystem level processes in Apa-lachicola Bay is the focus of a National Oceanic andAtmospheric Administration (NOAA)-sponsored programat Florida A&M University which emphasizes the conser-vation of aquatic reserves and monitoring the health ofestuarine ecosystems. To this end, we are studying thediversity, distribution, and role of estuarine and halophilicBALOs, which are among the least studied factorscontributing to bacterial mortality in aquatic ecosystems.BALOs may also be useful as indicators of water qualitysince as predators of Gram-negative bacteria, their numbersmay be correlated with that of coliform bacteria, Vibriospecies, and other bacterial indices of water contamination.BALOs are critical bacterial components of microbial loopsin aquatic food webs, as up to 80% of marine bacteria aresusceptible to lysis by these predatory bacteria [23].

The goal of this study was to investigate whetherassociations of BALOs with natural biofilms developed insitu can be utilized to study these predatory communitiesfrom environmental samples. In previous reports, BALOswere found in high numbers from aquatic biofilms or sub-merged surfaces, but not in the surrounding waters [16, 32].Recent data from our laboratory indicate that BALOs arechemotactic and respond to dense populations of bacterialprey in the environment [4]. As BALO numbers in anaquatic environment are generally very low, we propose theuse of biostimulation using substrate-incorporated agarcoated on glass slides. This approach could be applied to avariety of aquatic sources to characterize the biodiversity ofBALOs predating on endogenous prey bacteria, therebyeliminating rounds of subculturing on prey bacteria. To ourknowledge, this is the first such report.

Methods

Sampling Site Details

Samples were collected from three sites in February 2006from a pier located in the midsection of Apalachicola Bay(Latitude/longitude 29°42.128′N, 84°52.811′W) [4]. Watersamples were obtained within an area of approximately35 m2 by submerging a sterile container to an approximatedepth of 0.5 m, placed in a cooler, and transported to thelaboratory at Florida A&M University, Tallahassee. Repli-cate samples were separately filtered through 0.8-μm filtersto exclude debris and protozoan grazers [16, 21, 22, 32] andstored at 4°C. Experiments were set up within 24 h ofsampling.

Biogeochemical Parameters

At the time of sampling, selected parameters such assalinity, temperature, conductivity, pH, and dissolvedoxygen were measured at the sampling site by a PortableWater Checker U-10 probe (Horiba, Kyoto, Japan).

Slide Biofilm Assay and Biofilm Sampling

Fisherbrand premium glass microscope slides were buffedon both sides with coarse sandpaper until their glossysurface was completely buffed to aid in agar plating.These slides were then alcohol-sterilized and coatedwith 2% ultrapure low melt agar (MoBio, Solana Beach,CA) that contained either 0.01% yeast extract, casaminoacids, or concentrated cells of V. parahaemolyticus P5(final OD600 of 0.5 nm), a widely used prey to isolatehalophilic BALOs [26]. The slides were allowed to dry for30 min and were deployed into separate tubes thatcontained 50 ml of Bay samples. The tubes were incubatedwhile gently stirring on a shaker at ambient temperature.After 1 week of incubation, the biofilm-coated agar wascollected off the slides, as previously described [15], withintermittent rinsing with artificial seawater (ASW). Thebiofilm-coated low melt agar was homogenized in 5 mlof ASW by vigorous vortexing and processed for furtherexperiments.

Microscopic Analysis of Biofilm

Before harvesting of biofilms, a 1 cm2 of biofilm-coatedagar from each slide was cut with a sterile blade andmounted onto a clean glass slide. These samples werestained for 5 min with acridine orange (final concentration,5 mg/l), and a coverslip was gently placed over the biofilmfor microscopic visualization.

Diversity of Bdellovibrio- and Like Organisms (BALOs) in Apalachicola Bay 641641

Most Probable Numbers of BALOs and HeterotrophicBacteria

A subsample of the biofilm suspension was mechanicallyagitated for 10 min to homogenize the agar biofilm [15].The BALOs were quantified by the three-tube dilution mostprobable numbers (MPNs) assay, as previously described[27], with modifications. For the MPN assay, three replicate1-ml subsamples were removed from the biofilm suspen-sion and diluted to extinction by tenfold dilutions in sterileASW medium. As a prey bacterium is required to estimatenumbers of BALOs in environmental samples, each tubewas amended with a dense suspension of an overnightgrown culture of Vibrio parahaemolyticus P5 (final OD600

of 0.5 nm). Tubes that showed a reduction in absorbancedue to the activity of BALOs after 7 days of incubationwere considered positive compared to negative controltubes that were devoid of biofilm samples but containedprey cells. The presence of the predators in tubes showingpositive was confirmed by acridine orange staining.Similarly, three-tube MPN dilutions were set up in SWYEbroth for estimating total bacteria in the Bay samples; tubeswere scored positive based on growth of environmentalbacteria measured at 600 nm after a week of incubation.

Nucleic Acid Extraction and PCR Amplification

DNA from biofilm samples was extracted using theUltraClean GelSpin Kit as per the manufacturer’s instruc-tions (MoBio), except the DNA was eluted in sterilepolymerase chain reaction (PCR)-grade water for down-stream processing. Quality of the DNA was evaluated byelectrophoresis through a 0.7% agarose gel with Tris–acetate–EDTA (TAE) buffer, and concentrations of totalDNA were estimated by UV absorbance at 260 nm [24].Primers used for PCR amplification of bacterial 16S rDNAgene sequences were 27F (5′-AGAGTTTGATCMTGGCTCAG-3′) and 1492R (5′-TACGGYTACCTTGTTACGACTT-3′) [17]. All amplifications were performedin a myCycler thermocycler (Bio-Rad, Hercules, CA) usingHotStarTaq Master Mix (Qiagen, Valencia, CA). Differentdilutions of the DNA samples were denaturated at 95°Cfor 15 min, followed by 35 cycles of 94°C for 30 s, 55°Cfor 30 s, and 72°C for 30 s followed by an extensionstep of 72°C for 7 min [5]. PCR products (5 μl) wereanalyzed by electrophoresis through a 1% agarose gel inTAE buffer.

Cloning of Bacterial 16S rDNA and RFLP Analyses

Cloning of the 16S ribosomal gene was done with freshPCR amplicons which were ligated into pCRII-TOPO

cloning vector and transformed into Escherichia coliTOP10F′ cells according to the manufacturer’s instructions(Invitrogen, Carlsbad, CA). Clones were screened with X-Gal (5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside)and isopropyl-β-D-thiogalactopyranoside (IPTG) indicatorLuria–Bertani agar plates supplemented with 50 μg/ml ofkanamycin. Cloned rDNA templates were reamplified byPCR by using T7 (5′-GTAATACGACTCACTATAGGGC-3′) and SP6 (5′-ATTTAGGTGACACTATAG-3′) primershomologous to the RNA polymerase binding sites thatflank the insertion site in the vector. Restriction fragmentlength polymorphism (RFLP) analyses were conductedusing restriction enzyme HhaI and analyzed in 2% agarosegel [3]. Clone libraries were subjected to rarefaction usingaRarefactWin (version 1.3; S. Holland, Stratigraphy Lab,University of Georgia, Athens; http://www.uga.edu/∼strata/software/) to confirm that sufficient numbers of RFLPgroups were selected to represent the microbial diversity inthe clone libraries from biofilms originating from differentbiostimulants.

DNA Sequencing and Phylogenetic Analysis

Restriction profiles of the clone libraries were compared byRFLP and grouped in operational taxonomic units (OTUs).Selected clones were then sequenced at the DNA Sequenc-ing Laboratory of Florida State University with 27F primer.Chimera detection was carried out by Chimera_Checkversion 2.7 [6]. Sequences were also compared withpreviously identified sequences in the National Center ofBiotechnology Information (NCBI) database using BLAST[1] and aligned by ClustalX v. 1.8 [28]. Phylogenetic treewas generated with TREECON for Windows version 1.3b(http://bioinformatics.psb.ugent.be/software.php) usingneighbor joining with default settings [30]. Bootstrapresampling analysis for 100 replicates was performed toestimate the confidence of tree topologies.

Statistical Analyses

Samples generated from yeast extract (YE), casaminoacid (CA), and P5 libraries were subjected to web-basedUniFrac analyses (http://bmf.colorado.edu/unifrac) as de-scribed [19]. UniFrac analyses were applied to cluster thesequences from these multiple environments and testedwhich environments were significantly different using Ptest, UniFrac test, and PCA analysis. The scatter plots weregenerated with each library representing an environment,the rationale being that environments that have similarvalues of a particulate principal coordinate will appear closetogether on the PCA axis.

642 A. Chauhan, H.N. Williams

Nucleotide Sequence Accession Numbers

The partial 16S rDNA gene sequences from this study havebeen deposited in GenBank under accession numbersEF419205 to EF419239.

Results and Discussion

Biogeochemical Parameters

Selected biogeochemical parameters were noted at the timeof sampling in February 2006. For the three samples,salinity was in the mesohaline range of 8.4 ppt. The rangeof other measurements was as follows: temperature 15.5–16°C; conductivity 14.8–15.0 μS; pH 7.5–8.0, and dis-solved oxygen 50–54 mg/l.

Microscopic Analysis of Biofilm

Agar-coated slides containing either YE, CA, or V. para-haemolyticus P5 as biostimulants were observed by visualand microscopic examination to have varying degrees ofbiofilm formation after a week of incubation. The use ofYE and CA as substrates was based on our recent study,which revealed that these compounds attracted a diversepopulation and served as good sources of nutrients formicrobial communities in Apalachicola Bay [4]. Acridineorange staining of the biofilms indicated potentially live cells(stained red) and dead cell aggregates (stained green), asshown in Fig. 1A–C. In all three treatments, biofilmscontained more dead cell clumps than live cell colonies,indicating that predatory bacteria may be thriving upon thoseendogenous bacteria that adhered or grew on YE, CA, or P5.

MPNs of BALOs and Heterotrophic Bacteria

Estimates of the numbers of BALOs and total bacteria inbiofilm suspensions and estuarine water samples weredetermined by the three-tube MPN method. These resultsare shown in Table 1. Bay water contained approximately1.5×104 total bacteria and 2.1×102 BALOs. MPN analysesof biofilm samples revealed high numbers of total bacteria

c

b

Live Bacterial

a

Acridine Orange Stained Red [Live] Bacterial Communities

from YE

Acridine Orange Stained Green [Dead] Bacterial Communities from YE

Dead Bacterial Communities from CA

Live Bacterial Communities from CA

Communities from P5

[Dead BacterialCommunities from P5 Figure 1 Microscopic visualization of biofilm on agar-coated slides

containing a yeast extract (YE), b casamino acids (CA), and c Vibrioparahaemolyticus (P5) from Apalachicola Bay samples. Slides wereleft undisturbed for a week for biofilm to develop, washed, and a thinslice of agar was visualized after staining with acridine orange;biofilm cells with active synthetic pathways fluoresced red orange andinactive bacterial community fluoresced green subsequent to viewingby epifluorescent light

b

Diversity of Bdellovibrio- and Like Organisms (BALOs) in Apalachicola Bay 643643

and BALOs indicative of microbial biofilm formation andbiostimulation on all agar substrates tested (Table 1).

It becomes imperative to mention here that MPNs areknown to underestimate bacterial numbers, (more so insoils and sediments than aquatic ecosystems), as microbiotaare bound together by extrapolymeric substances, humics,and other chelating agents which results in improper mixingand dilutions required for MPN enumeration. Regardless,an interesting trend observed in the biofilm was that theMPNs of both the total bacteria and BALOs were higher onYE and lower on P5 (YE>casamino acids>P5). These dataclearly indicate that endogenous marine bacteria on nutrient-rich substrates such as YE and CA form biofilms first, andBALOs in turn respond to the established biofilm commu-nity as their “nutrient” source. The greater numbers of preyresulted in substantially greater yield of the predators,suggesting a correlation between numbers of prey andpredators. These results are also consistent with that of ourrecent study in which BALOs potentially sensed andmigrated towards native environmental prey [4]. The V.parahaemolyticus biofilm also showed similar trends, albeitat a much lower level. An endogenous bacterial consortiumstimulated with nutrients thus appears to serve as a morelucrative prey for the BALO community than the axeniclaboratory generated V. parahaemolyticus.

Acridine orange staining confirmed that the drop inoptical density of MPN tubes was due to the predationactivity of BALOs; cleared tubes showed a thriving highlymotile community of small-sized BALOs that were distin-guishable from other bacteria cells (data not shown).Although the MPN analyses were able to detect only thosepredatory bacteria that were able to attack and consume V.parahaemolyticus P5 as prey, the results show quantitativedifferences between BALOs associated with differentbiostimulants and endogenous microbiota. This was fol-lowed by molecular analyses of the bacterial community inYE, CA, and P5 biofilms.

Phylogenetic Analysis of Bacteria from Biofilm Samples

Microbial diversity of the biofilm communities that formedon YE, CA, and P5 agar surfaces was assessed by PCR-restriction fragment length polymorphism (PCR-RFLP)followed by 16S rDNA gene sequence analysis. Phylotypeswere assigned to the same OUTs based on identical orhighly similar RFLP patterns. Forty-eight clones werescreened from each library; YE was represented by 15OTUs, CA by 12 OTUs, and P5 only 7 OTUs, indicatingless diversity in this library. For all libraries, rarefactioncurves reached a plateau, indicating that sufficient numbersof clones were sequenced from each sample to represent thebiodiversity (data not shown).

Distribution of sequences within the individual clonelibraries from YE, CA, and P5 are presented in Table 2. 16SrDNA sequencing of clones from different substratesidentified Gram-negative bacteria mainly from the alpha,gamma, delta-proteobacteria, and the Cytophaga–Flavo-bacterium–Bacteroides (CFB) group. Single members werealso identified from beta and epsilon-proteobacteria andFusobacteria sp. Phylogenetic analyses from the clonelibraries are represented in Fig. 2 and discussed in detailedbelow.

α-Proteobacteria

α-Proteobacterial members are mainly phototrophic, butseveral genera such as Rhizobia can metabolize C1compounds. Hyphomonas spp., Agrobacterium spp., Mes-orhizobium spp., Rhodobacter spp., and Pseudorhodo-bacter spp., were identified from slides covered with agarcontaining YE. Roseobacter spp., Thalassospira spp.,Stappia spp., were identified from slides that containedCA. Slides that contained P5 did not harbor α-proteobac-teria. The presence of Hyphomonas spp., Rhodobacter spp.,Roseobacter spp., and Thalassospira spp. is not surprising,

Table 1 Most probable number enumerations of Bdellovibrio- and like organisms (BALOs) and total bacteria recovered from yeast extract (YE),casamino acids (CA), and Vibrio parahaemolyticus P5 biofilms from Apalachicola Bay in February 2006

Enumerationa (MPN/ml)

Biofilmb from YE Biofilm from CA Biofilm from P5

Most probable numbers ofBALOs 9.3×102 (2.0, 27.0) 4.2×102 (1.0, 13.8) 2.4×102 (0.47, 9.6)Heterotrophic bacteria 4.6×104 (1.0, 13.5) 3.6×104 (1.9, 15.8) 2.3×104 (0.5, 6.7)

aMPNs from biofilm samples were calculated by resuspension of a subsample of the scraped biofilm in sterile ASW and homogenizing thesamples before serial dilutions. Three tube MPNs were set up from each sample and BALOs were scored positive based on reduction of OD600 toat least half of that in the control (ASW + prey) and total bacteria were scored positive by increase in OD600 indicative of growth. MPN valueswere calculated from the EPA version 4.04 calculator (http://www.epa.gov/nerlcwww/other.htm).b Lower and upper limits in parentheses reflect 95% confidence interval.

644 A. Chauhan, H.N. Williams

considering that they are common colonizers of marine-submerged surfaces and aid in biofilm formation [8, 18].Agrobacterium spp. and Mesorhizobium spp. are known tofix nitrogen in plant nodules and, to our knowledge, havenot been previously reported in estuarine ecosystems.

As Agrobacterium genus is extremely heterogeneous andhas been taxonomically reclassified to new genera, such asStappia [29], it is likely that the sequences representingAgrobacterium spp. from this study belong to estuarineStappia spp. It may also be that Agrobacterium spp. andMesorhizobium spp. resulted from allochthonous depositionfrom the plume of Apalachicola River that drains freshwater into the Bay or from vegetation debris carriedinadvertently by many boats that are used by oysterharvesters in Apalachicola Bay.

β-Proteobacteria

The beta-proteobacteria group is comprised of bacteriahaving versatile degradation capacities. A single member,Acidovorax spp., was observed from this group only fromYE agar slides.

γ-Proteobacteria

The gamma-proteobacteria group consists of many envi-ronmentally significant bacteria and is the largest groupfound in oceans. Pseudoalteromonas spp. and Saccharo-phagus spp. were observed from YE and CA slides;Marinomonas spp. were observed from YE, CA, and P5slides. Vibrio spp. and Shewanella spp. were identifiedfrom CA and P5 slides but not from YE slides. Pseudoal-teromonas spp. are known to produce a range of extracel-lular secondary metabolites, which can even inhibit fungi,thereby rendering advantage to this guild in surfacecolonization [12]. It was interesting that the slide biofilmassay identified Saccharophagus spp. on YE and CAsubstrates. S. degradans has recently been characterized forthe ability to degrade complex plant polysaccharides such ascellulose, hemicelluloses, pectins and marine polysacchar-ides such as agar, alginate, and chitin [11]. The agar surfaceof the slides may have provided a rich nutrient source whichexplains the dominance of Saccharophagus spp. on YE andCA slide biofilms. However, agar slides that contained P5were devoid of the marine degrader Saccharophagus spp. Itis likely that P5 prevented colonization of Saccharophagusspp. Other sequences clustered mainly with Marinomonasspp., Vibrio spp., and Shewanella spp. These bacterial groupshave been consistently isolated and characterized froma variety of free-living and plankton associated marineeco-niches [4, 13, 20] and appear to be dominant membersof bacterial biofilm communities.

δ-Proteobacteria

The delta-proteobacteria group is a mixed branch of aerobicgenera such as the myxobacteria characterized by theirfruiting-body morphologies; the group also includes strictlyanaerobic genera. BALOs fall within delta-proteobacteria,and in this study, three distinct clades were identified basedon the substrate used in the slide agar assay. Sequencesoriginating from all three substrates (YE, CA, and P5)clustered with Bacteriovorax sp. JS10, which represents theunique cluster V that has, thus far, been recovered onlyfrom low salinity regions of Chesapeake Bay, PamilcoSound/Neuse River area and Apalachicola Bay, but notfrom any other marine, salt lake, and estuarine ecosystemsaround the world according to our recent study [22, 31].Eco-niches for BALOs are not only based on physico-

Table 2 Relative BALOs and potential prey phylotype abundancesfrom YE (yeast extract), CA (Casamino acids), and Vibrio para-haemolyticus P5 slide biofilms from Apalachicola Bay sampled inFebruary 2006

YEClonesa

(%)

CAClones(%)

P5Clones(%)

Closest phylogeneticrelative from NCBI

Predatory microbiotaBacteriovorax JS10 7 5 10Bacteriovorax spp. 10 7 0Peredibacter spp. 10 0 0Potential prey microbiotaPseudoalteromonas spp. 2 4 0Saccharophagus spp. 7 10 0Marinomonas spp. 2 7 4Acidovorax spp. 2 0 0Arcobacter spp. 2 2 37Hyphomonas spp. 2 0 0Agrobacterium spp. 2 0 0Mesorhizobium spp. 2 0 0Pseudorhodobacter spp. 11 0 0Rhodobacter spp. 18 0 0Bacteroidetes spp. 14 6 0Cellulophaga spp. 9 0 0Vibrio spp. 0 4 0Vibrio parahaemolyticus P5 0 0 43Shewanella spp. 0 7 2Thalassospira spp. 0 37 0Stappia spp. 0 7 0Roseobacter spp. 0 4 0Flavobacterium spp. 0 0 2Fusobacterium spp. 0 0 2

a Forty-eight clones were compared from each library. Two or morerepresentatives from each phylotype were sequenced and 16S rDNAgene sequences were compared to their phylogenetic relatives fromNCBI database.

Diversity of Bdellovibrio- and Like Organisms (BALOs) in Apalachicola Bay 645645

100

97

90

87

88

97

82

92

99

100

100

100

85

100

100

100

100

100

100

100

100

100

100

100

94

100

100

100

100

93

98

100

100

100

100

85

87

100

100

100

100

100

62

100

100

91

100

100

100

100

100

100

100

100

100

100

100

100

Peredibacter sp. clone Per1

YE-6

Peredibacter starrii

Peredibacter sp. clone Per4Bacteriovorax stolpii

YE-36

Bacteriovorax sp. F2 CA-10

Bacteriovorax sp. JS10

YE-35

P5-27

CA-2

Thalassospira lucentensis

CA-4Hyphomonas oceanitis

YE-11

Stappia sp. M8

CA-8

Agrobacterium tumefaciens

YE-31

Mesorhizobium alexandrii

YE-45

Roseobacter sp. B11

CA-13

Rhodobacter apigmentum

YE-1Pseudorhodobacter incheonensis

YE-3

Arcobacter sp. R-28314

CA-24

YE-37

P5-5

Pseudoalteromonas sp. clone CA-1B CA-43

Pseudoalteromonas sp. UL13

Pseudoalteromonas sp. BSw10016 YE-7

P5-42

Shewanella sp. TP4

CA-18

Vibrio parahaemolyticus

P5-26

Vibrio aestuarianus

CA-28

YE-33

CA-1 Saccharophagus degradans 2-40

CA-29

Marinomonas sp. BSi20470

YE-42

Marinomonas pontica strain 46-16 P5-46

Marinomonas sp. clone CA-3C CA-5

Acidovorax sp. R-25074

YE-28

Fusobacterium perfoetens

P5-33

Bacteroidetes bacterium SED4 YE-4

CA-42

Flavobacterium gelidilacus P5-40

Cellulophaga sp. CC12 YE-2

Pote

nti

al

Pre

y G

uil

ds

Pre

dato

ry G

uil

ds

10%

Arthrobacter globiformis

Figure 2 Phylogenetic tree represents partial 16S rDNA genesequences from domain Bacteria constructed with TREECON forwindows v. 1.3b using neighbor joining. A total of 48 clones werecompared from each slide biofilm library, and 16S rDNA genesequences from two representatives of each phylotype were sequenced

to construct the phylogenetic tree. Clones suffixed with YE, CA, andP5 originated from yeast extract, casamino acids, or Vibrio-coatedslides. Numbers at nodes represent bootstrap values (100 timesresampling analysis); only values >50% are presented. Arthrobacterglobiformis was used as outgroup

646 A. Chauhan, H.N. Williams

chemical selective pressures but also the bacterial commu-nity that feed BALOs. Therefore, ongoing studies in ourlaboratory focus on identification of endogenous microbialcommunities in the specific eco-niches where cluster Vsequences have been identified to correlate the distributionpatterns of cluster V to specific prey bacterial species inApalachicola Bay.

A second clade was also observed from YE and CAbiofilms consisting of sequences clustering within thehalophilic group of the Bacteriovorax spp. [31]. Surpris-ingly, a third clade was also identified, but only from theYE biofilm community. Sequences from this clade werenovel such that they clustered within Peredibacter spp., afreshwater isolate. Peredibacter-like sequences from thisstudy shared a similarity index of 90–94% with otherPeredibacter spp. and may potentially be phylogeneticrelatives of other freshwater ecotypes [9].

Most studies to detect BALOs have been based on brothenrichments in the presence of V. parahaemolyticus P5, orsome other Gram-negative prey bacterium, and by isolationas PFUs on double-layered agar plates. These PFUs arethen purified, and samples are characterized for thebiodiversity of BALOs. This isolation approach is inher-ently biased, as only P5-specific, and cultivable BALOs areisolated, and those having other prey preferences will notform plaques. Therefore, results from this study furthersupport the need to have culture-independent tools to studyBALO diversity.

The YE agar slide biofilms resulted in the maximumdiversity of bacteria nearly all of which could be potentialprey for BALOs (Table 1). This may be the reason for thetwo to four times higher MPN of BALOs and the higherdiversity (Table 2 and Fig. 2) recovered from the YEbiofilm as compared to the CA and P5 libraries. Bdellovi-brio spp., Bacteriovorax spp., and a novel clade clusteringwith Peredibacter spp., were identified from YE slidebiofilms. These BALOs likely had several prey species toselect from among the bacterial biofilm guilds on YEslides. On the other hand, the P5 slides not only harboredlesser numbers of BALOs (Table 2), but enriched for asingle clade, Bacteriovorax spp., which may prefer Vibriospp. over other bacteria [21, 22, 26, 31].

ɛ-Proteobacteria

The epsilon-proteobacteria group consists of few genera,and in this study, only Arcobacter spp. was identified fromall biofilms, but numbers were considerably higher from theP5-coated slide. Arcobacter has been previously reported tobe associated with free-living and planktonic bacteria froma variety of marine sources [20], and it may be thatArcobacter spp. share a synergistic relationship with V.

parahaemolyticus P5, accounting for the higher numbersdetected in the clone libraries from P5 biofilms (Table 2).

Fusobacteria

A single member was observed from this group clusteringwith Fusobacterium spp. detected only in the P5 biofilm.This was rather surprising, as Fusobacterium spp. is anobligate anaerobe [10], and there was no effort made insetting up the experiment to accommodate anaerobicconditions.

CFB Group

The CFB group is considered to be of special relevance foraquatic environments. They are predominant marine het-erotrophic bacterioplanktonic guilds on marine snow orfree-living in nutrient-rich micro-niches [2]. The diverseenzymatic profiles of Bacteroides spp. indicate that they arepotential “specialists” in the degradation of high-molecular-weight compounds in marine pools of dissolved organicmatter and particulate organic matter [7]. Therefore, theCFB group is thought to play a significant role in themarine carbon cycle by responding to high-molecular-weight substrates.

PCA Analyses of 16S rDNA Sequences

Statistical analyses was performed with UniFrac to deter-mine which of the environments represented by YE, CA,and P5 were significantly different and selected for differentmicrobiota based on 16S rDNA profiling and eventualsegregation by principal coordinate analysis (Fig. 3). BeforeUniFrac analysis, the 16S rRNA sequences were groupedbased on their origination environments, i.e., YE (yeastextract), CA (casamino acids), and V. parahaemolyticus P5.Principal coordinates axis results are shown in Fig. 3; thesequences that came from YE, CA, or P5 separated out onthree different axis of the PCA, indicating that theseenvironments selected for different bacteria to some extent.Based on the p test values, YE and CA clones were highlysignificant (P<0.001) when compared with clones from P5,which were marginally significant (P<0.02). PCA axis 1explained 21.2% of the variability, and PCA axis 2explained 14.09%, with a cumulative percentage of35.29%, which potentially separated sequences based ontheir origin (Fig. 3). Phylogenetic and statistical resultsindicate that all three biostimulants selected for differentmicrobial communities probably based on abilities ofendogenous microbiota to utilize biostimulants as carbonand energy sources. It may also be that biofilms turned outto be different due to the synergistic and antagonistic

Diversity of Bdellovibrio- and Like Organisms (BALOs) in Apalachicola Bay 647647

influences of primary colonizers, similar to the biofilmscharacterized from natural environments.

Ecological Significance

Biofilms represent an environmentally important niche forthe ecology and survival of BALOs due to being a majorand continuous source of nutrients in the form of flourish-ing prey bacterial communities that are protected in theextra polymeric substance layer(s). It does not come as asurprise that BALOs were found in high numbers fromaquatic biofilms, but not in the surrounding waters, inprevious reports [15, 32]. We recently identified chemotaxis[4] as one of the mechanisms by which BALOs sense andrespond to high populations of bacterial prey, as those that

may be found in biofilms as opposed to the low numbersfound in the water column. This may be one of the reasonsfor the high abundance and diversity of the predatorsobserved in YE agar slide biofilm. Another very likelyreason for higher BALO diversity from YE biofilms couldpotentially be explained on the basis of studies by Sanchez-Amat and Torrella [25]. Their study yielded clues on theidentification of yeast extract as most potent substance ascompared to sugars, ammonia, amino acids, and organicamines in inducing rupture of bdelloplast, the resting stageof Bdellovibrios inside the host periplasmic space. There-fore, YE may be one of the triggers which facilitates thelytic factor such that bdelloplasts in the sample ruptured inthe presence of YE-coated agar slides, which eventually ledto the proliferation of BALO guilds and their identification.

Figure 3 Shown are the first two principal coordinates from a PCAafter sequences that originated from yeast extract, casamino acids, andVibrio parahaemolyticus P5 were subjected to web-based UniFracanalyses. Percentages represented in the axis labels are percentages of

variation that are explained by the principal coordinates. PCA analysesindicate that three substrates selectively stimulated the predator–preyguilds from Apalachicola Bay water sample

648 A. Chauhan, H.N. Williams

One week of sample incubation with nutrient-enrichedagar surfaces may have introduced some bias associatedwith bacterial growth rather than biostimulation alone. Tocircumvent this problem, the biostimulants were spiked inlower concentrations (0.01%). In this way, the substrateswould have been depleted rapidly, giving just enoughnutrient source for the endogenous microbiota to increase innumbers and feed the BALO community, resulting in theirproliferation, detection, and ease in identification. Thisappeared to be the case, as bacterial MPNs, even after aweek of incubation, were found to be low (approximately104 cells/ml). Therefore, most likely, identified BALOdiversity was restricted to those that were able to grow onthe endogenous susceptible bacterial population as opposedto single prey species selected by the investigator.

Davidov et al. [9] have recently designed and reportedthe use of Bdellovibrio-, Bacteriovorax-, and Peredibacter-specific primers targeting the conserved 16S rDNA regionfrom these BALO groups. These authors found certainBALO guilds that were identified only by a PCR-DGGEapproach employing group-specific primers, and only asmall number of enriched BALO ribotypes were alsodetected by culture-dependent approach. We believe thatthe slide agar biofilm assay reported here combines theadvantages of a culture with a molecular-based approach tocharacterize microbial biodiversity, especially from oligo-trophic aquatic ecosystems where the bacterial biomassmay be low. This technique could also be very useful in theisolation of novel marine oligotrophic bacteria harboringunique traits such as antiobiotic production or those havingunique battery of hydrolytic enzymes, etc.

However, the present study did not examine seasonalvariations, tidal cycles, or triggers of biofilm promoterssuch as the synergistic and antagonistic affects of endog-enous bacterial communities. If these variations are takeninto account, the result may likely identify differences usingdifferent biostimulants in BALOs and bacterial diversity.

Collectively, our results clearly show that estuarineBALOs respond well to a biofilm community, and thisapproach can be exploited to study BALO biodiversity innatural environments, bypassing the rounds of labor inten-sive subculturing with laboratory-maintained single prey.Studies are underway in our laboratory involving the use ofyeast-extract-coated agar surfaces to be deployed in aquaticecosystems to study field-based predator–prey interactions.

Acknowledgments Funding for this study was provided by theNational Science Foundation grant OCE0455276 and the NationalOceanic and Atmospheric Administration grant NA17AE1624. Theauthors wish to thank the anonymous reviewers for their very helpfulcritique.

References

1. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990)Basic local alignment search tool. J Mol Biol 215:403–410

2. Bauer M, Kube M, Teeling H, Richter M, Lombardot T, Allers E,Wurdemann CA, Quast C, Kuhl H, Knaust F, Woebken D,Bischof K, Mussmann M, Choudhuri JV, Meyer F, Reinhardt R,Amann RI, Glockner FO. (2006). Whole genome analysis of themarine Bacteroidetes ‘Gramella forsetii’ reveals adaptations todegradation of polymeric organic matter. Environ Microbiol 8(12):2201–2213

3. Chauhan A, Ogram A (2006) Fatty acid-oxidizing consortia alonga nutrient gradient in the Florida Everglades. Appl EnvironMicrobiol 72:2400–2406

4. Chauhan A, Williams HN (2006) Response of bdellovibrio andlike organisms (BALOs) to the migration of naturally occurringbacteria to chemo attractants. Curr Microbiol 53(6):516–522

5. Chauhan A., Ogram A, Reddy KR (2004) Syntrophic-methano-genic associations along a nutrient gradient in the FloridaEverglades. Appl Environ Microbiol 70:3475–3484

6. Cole JR, Chai B, Marsh TL, Farris RJ, Wang Q, Kulam SA,Chandra S, McGarrell DM, Schmidt TM, Garrity GM, Tiedje JM(2003) The Ribosomal Database Project (RDP-II): previewing anew autoaligner that allows regular updates and the newprokaryotic taxonomy. Nucleic Acids Res 31:442–443

7. Cottrell MT, Kirchman DL (2000). Natural assemblages of marineproteobacteria and members of the Cytophaga-Flavobacter clusterconsuming low- and high-molecular-weight dissolved organicmatter. Appl Environ Microbiol 66(4):1692–1697

8. Dang H, Lovell CR (2002) Numerical dominance and phylotypediversity of marine Rhodobacter species during early colonizationof submerged surfaces in coastal marine waters as determined by16S ribosomal DNA sequence analysis and fluorescence in situhybridization. Appl Environ Microbiol 68(2):496–504

9. Davidov Y, Friedjung A, Jurkevitch E (2006) Structure analysis ofa soil community of predatory bacteria using culture-dependentand culture-independent methods reveals a hitherto undetecteddiversity of Bdellovibrio- and -like organisms. Environ Microbiol8(9):1667–1673

10. Dworkin M (2005) Prokaryotic life cycles. In Dworkin M et al(eds) The prokaryotes: an evolving electronic resource for themicrobiological community, 3rd edn. Online BETA-Release 3.20(12/31/)

11. Ensor L, Stotz SK, Weiner RM (1999) Expression of multipleinsoluble complex polysaccharide degrading enzyme systems by amarine bacterium. J Ind Microbiol Biotech 23:123–126

12. Franks A, Egan S, Holmstrom C, James S, Lappin-Scott H,Kjelleberg S (2006) Inhibition of fungal colonization by Pseu-doalteromonas tunicata provides a competitive advantage duringsurface colonization. Appl Environ Microbiol 72(9):6079–6087

13. Jensen S, Bergh O, Enger O, Hjeltnes B (2002) Use of PCR-RFLPfor genotyping 16S rRNA and characterizing bacteria culturedfrom halibut fry. Can J Microbiol 48(5):379–386

14. Jurkevitch E (2000) The genus Bdellovibrio. In: Dworkin M et al(eds) The prokaryotes: an evolving electronic resource for themicrobiological community, 3rd edition, release 3.7. November 2,2001, Springer, New York http://link.springer-ny.com/link/service/books/10125/

15. Kelley JI, Turng B, Williams HN, Baer ML (1997) Effects oftemperature, salinity, and substrate on the colonization of surfacesin situ by aquatic Bdellovibrios. Appl Environ Microbiol 63(1):84–90

Diversity of Bdellovibrio- and Like Organisms (BALOs) in Apalachicola Bay 649649

16. Kelley JI, Williams HN (1992) Bdellovibrios in Callinectussapidus, the blue crab. Appl Environ Microbiol 58:1408–1410

17. Lane DJ (1991) 16S/23S rRNA sequencing. In: Stackebrandt E,Goodfellow M (eds) Nucleic acid techniques in bacterialsystematics. Wiley, New York, pp 115–175

18. Langille SE, Weiner RM (1998) Spatial and temporal depositionof hyphomonas strain VP-6 capsules involved in biofilmformation. Appl Environ Microbiol 64(8):2906–2913

19. Lozupone C, Knight R (2005) UniFrac: a new phylogeneticmethod for comparing microbial communities. Appl EnvironMicrobiol 71(12):8228–8235 (Dec)

20. Maugeri TL, Carbone M, Fera MT, Irrera GP, Gugliandolo C(2004) Distribution of potentially pathogenic bacteria as freeliving and plankton associated in a marine coastal zone. J ApplMicrobiol 97(2):354–361

21. Pineiro SA, Sahaniuk GE, Romberg E, Williams HN (2004)Predation pattern and phylogenetic analysis of Bdellovibrio-naceae from the Great Salt Lake, Utah. Curr Microbiol 48:113–117

22. Pineiro SA, Stine OC, Chauhan A, Steyert SR, Smith R, WilliamsHN (2007) Global survey of diversity among environmental salt-water Bacteriovoracaceae. Environ Microbiol 9(10):2441–2450(Oct)

23. Rice TD, Williams HN, Turng BF (1998) Susceptibility ofbacteria in estuarine environments to autochthonous Bdellovi-brios. Microb Ecol 35:256–264

24. Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: alaboratory manual, 2nd edn. Cold Spring Harbor LaboratoryPress, Cold Spring Harbor

25. Sanchez-Amat A, Torrella F (1990) Formation of stable bdello-plasts as a starvation–survival strategy of marine Bdellovibrios.Appl Environ Microbiol 56(9):2717–2725

26. Schoeffield AJ, Williams HN (1990) Efficiencies of recovery ofBdellovibrios from brackish-water environments by using variousbacterial species as prey. Appl Environ Microbiol 56:230–236

27. Sutton DC, Besant PJ (1994). Ecology and characteristics ofbdellovibrios from three tropical marine habitats. Mar Biol119:313–320

28. Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, HigginsDG (1997) The ClustalX windows interface: flexible strategies formultiple sequence alignment aided by quality analysis tools.Nucleic Acids Res 24:4876–4882

29. Uchino Y, Hirata A, Yokota A, Sugiyama J (1998) Reclassificationof marine Agrobacterium species: proposals of Stappia stellulatagen. nov., comb. nov., Stappia aggregata sp. nov., nom. rev.,Ruegeria atlantica gen. nov., comb. nov., Ruegeria gelatinovoracomb. nov., Ruegeria algicola comb. nov., and Ahrensia kieliensegen. nov., sp. nov., nom. rev. J Gen Appl Microbiol 44(3):201–210

30. Van de Peer Y, De Wachter R (1994) TREECON for Windows: asoftware package for the construction and drawing of evolutionarytrees for the Microsoft Windows environment. Comput ApplicBiosci 10:569–570

31. Williams HN, Pineiro S (2006) Ecology of the predatoryBdellovibrio and like organisms. In: Jurkevitch E (ed) Predatoryprokaryotes. Springer, Berlin, pp 214–244

32. Williams HN, Scheoffield AJ, Guether D, Kelley J, Shah D,Falker WA (1995) Recovery of bdellovibrios from submergedsurfaces and other aquatic habitats. Microb Ecol 29:39–48

650 A. Chauhan, H.N. Williams