Microbial Ecology Isolation and Genetic Analysis of Haloalkaliphilic Bacteriophages in a North American Soda Lake Shereen Sabet, Weiping Chu and Sunny C. Jiang Department of Environmental Health, Science, & Policy, University of California, 1367 SE II, Irvine, CA 92697-7070, USA Received: 19 January 2006 / Accepted: 19 January 2006 / Online Publication: 6 May 2006
Abstract
Mono Lake is a meromictic, hypersaline, soda lake that harbors a diverse and abundant microbial community. A previous report documented the high viral abundance in Mono Lake, and pulsed-field gel electrophoresis analysis of viral DNA from lake water samples showed a diverse population based on a broad range of viral genome sizes. To better understand the ecology of bacteriophages and their hosts in this unique environment, water samples were collected between February 2001 and July 2004 for isolation of bacteriophages by using four indigenous bacterial hosts. Plaque assay results showed a differential seasonal expression of cultured bacteriophages. To reveal the diversity of uncultured bacteriophages, viral DNA from lake water samples was used to construct clone libraries. Sequence analysis of viral clones revealed homology to viral as well as bacterial proteins. Furthermore, dot blot DNA hybridization analyses showed that the uncultured viruses are more prevalent during most seasons, whereas the viral isolates (A7 and 72) were less prevalent, confirming the belief that uncultured viruses represent the dominant members of the community, whereas cultured isolates represent the minority species.
Introduction
Mono Lake is a closed basin, hypersaline, alkaline soda lake located in central California, USA, presently characterized by a salt concentration of 8% (80 g/L) and a pH of 10. Mono Lake is a meromictic lake that undergoes holomixis approximately once every 10 years [19], with the most recent event recorded in October of 2003 (R. Jellison, pers. comm.). The lake has a very productive microbial community [9] with a high diversity of Correspondence to: Sunny C. Jiang; E-mail:
[email protected]
DOI: 10.1007/s00248-006-9069-1
& Volume 51, 543–554 (2006) & *
bacteria [7] and some of the highest recorded viral abundance of any natural aquatic system [11], yet little is known about the viral community that resides in this extreme environment. Bacteriophages, along with their hosts, make up the largest biomass on earth, residing mostly in aquatic habitats. Viruses are believed to play a critical role in the mortality of aquatic bacteria [6], thereby affecting the microbial food web and biogeochemical processes, as well as affecting bacterial diversity by restructuring the microbial community [15, 30]. Bacteriophages can also affect their hosts via lysogeny and transduction and by mutually providing genes that enhance host survival [6, 13, 25]. Viruses and viruslike particles have been detected and isolated in hypersaline environments such as the Dead Sea, a marine solar saltern in Australia, the Great Salt Lake in Utah, and the Little Salt Pond in Yallahs, Jamaica [16, 20, 23, 31]; however, the literature is sparse regarding the ecology of these extremophilic aquatic bacteriophages. As a first step toward better understanding these viruses, especially in an extreme environment, this study aims to track the seasonal presence of Mono Lake’s representative bacteriophage community as well as uncultured phages. Fragments of viral genomes from lake waters were cloned and analyzed, and their sequences provide insight as to the nature of the viruses that reside in a hypersaline, alkaline lake. Methods Mono Lake Water Samples. Small-volume water samples (50 mL) were collected monthly by researchers at the Sierra Nevada Aquatic Research Laboratory, University of California, from Mono Lake stations 2, 6, 7, and 11 (Fig. 1) [21] between March 2003 and July 2004. A single, upper 9-m integrated water column sample was taken from all of the stations, whereas a
Springer Science+Business Media, Inc. 2006
543
544
S. SABET
ET AL.:
SEASONAL PRESENCE
OF
HYPERSALINE PHAGES
Figure 1. Map of Mono Lake showing locations of sampling stations. All stations were used for water sample collection for the duration of this study, except for stations 4, 5, and 9.
detailed vertical profile at 2, 8, 10, 12, 16, 20, 24, 28, and 35 m was sampled at station 6 using a Niskin bottle onboard a small boat. Integrated water samples were collected by lowering a 1-in.-diameter Tygon tubing weighted at one end into the water down to the 9-m depth and then lifting the weighted end out of the water and transferring the water to a bucket. Water samples were filtered through 0.22-mm pore-size syringe filters (Millex GP PES membrane; Millipore, Carrightwohill, Co. Cork, Ireland) to remove bacteria and phytoplankton, and viral fractions were shipped cold overnight to the University of California Irvine laboratory. Large-volume water samples (5–10 L) were collected at sampling stations 1, 3, 6, 8, 10, 11, and 12 from both the oxygenic (2–10 m) and anoxic (35 m) layers during field trips in February and August 2001, in August 2002, and in February and October 2003 for isolation of bacterial hosts and viruses. All samples were processed on site within 4 h of collection. Bacterial Host and Viral Isolation. Four Mono Lake bacterial hosts were isolated in February (designated M12-2c, M12-chla) and in August (designated MN112.5a, MN12-2a) of 2001 by first concentrating 10 L of water using a tangential flow filtration (TFF) system with a 30-kDa molecular weight cutoff filtration cartridge (PALL Corp., Hauppauge, NY, USA). One hundred microliters of TFF retentate was then spread onto either marine agar plates (Difco, Becton, Dickinson and Co., Sparks, MD, USA) supplemented with 1% NaCl (M12-2c and M12chla), or onto agar plates made with Mono Lake water supplemented with 0.5% peptone, which was then
sterilized by boiling. Plates were incubated for 24–72 h at room temperature and observed for colony growth. Individual colonies were picked and cultivated further in their respective media. Frozen stocks were made and stored at _80-C. M12-2c and M12-chla hosts were identified by 16S rRNA sequence analysis in a previous study (Hollibaugh, pers. comm.) and are listed under GenBank accession numbers AY730244 and AY730242, respectively. M12-2c is 99% similar to Vibrio metschnikovii and M12-chla only has a close match with a bacterium previously isolated from Mono Lake sediment in the database. In this study, the MN12-2a and MN1-12.5a bacterial isolates were also characterized by 16S rRNA sequence analysis. Those sequences are listed under the accession numbers AY856383 and AY856384, respectively. To isolate bacteriophages, 5–10 L of water samples were prefiltered through glass fiber and 0.45-mm filters (Millipore) to remove large plankton when plankton concentrations were high. The samples were then sequentially filtered via TFF, first by using a 0.22-mm pore-size filtration cartridge followed by a 30-kDa molecular weight cutoff cartridge. The first cartridge removes bacteria, whereas the second cartridge concentrates viruses in the retentate. The final retentate volume is approximately 100 mL. Plaque assays were carried out by using the top agar overlay method [1]. Briefly, 1 mL of bacterial host (described above) in 3 mL of top agar was inoculated with 1 mL of TFF-concentrated viral sample, then mixed and poured over a bottom nutrient plate. Both top and bottom agar contain the same ingredients as the bacterial
S. SABET
ET AL.:
SEASONAL PRESENCE
OF
HYPERSALINE PHAGES
host isolation and growth media. Plates were incubated at room temperature for 48 h before the number of viral plaques was counted. Selective individual plaques were picked and suspended in 300 mL of either 0.5 M Tris (pH 8.0) or MSM buffer (450 mM NaCl/50 mM MgSO4/50 mM Tris, pH 8.0). Each isolate was purified by reisolation of individual plaques four more times before it was considered to be a pure isolate. Viral lysates were harvested by plate elution using MSM buffer. Lysates were then filtered through 0.22-mm pore-size syringe filters to remove any remaining bacteria. Viral filtrates were either frozen at _80-C in a 1:1 ratio of MSM buffer and 50% sterile glycerol or stored at 4-C for further analyses. Bacterial Artificial Chromosome Cloning. To understand the diversity and dynamics of uncultured bacteriophages in Mono Lake, a bacterial artificial chromosome (BAC) clone library was constructed. First, viruses were concentrated from 240 mL of 0.22-mmfiltered lake water, collected from station 3 at 2 m in February 2003, to a final volume of 2 mL via the Centricon Plus-80 100,000 NMWL centrifugal filter (Millipore). The concentrate was then split into two equal volumes and each volume was further concentrated and washed with 0.5 M Tris–HCl (pH 8.0) via the Centricon YM 100-kDa filter unit (Millipore). To remove any potentially contaminating extraviral DNA, DNase treatment was carried out in each Centricon filter tube by adding 7 mL 0.5 M Tris–HCl (pH 8.0), 1 mL 10 enzyme reaction buffer, 2 mL (2 U) RQ1 DNase enzyme (Promega, Madison, WI, USA). Two hundred nanograms of l DNA and 150 ng of station 12 15-m (oxycline) sample were used as controls for the DNase treatment. The mixture was incubated in a 37-C incubator for at least 30 min. Following DNase digestion, viral samples were washed with 1 TE twice, incubated with 5 TE for 1 h at 4-C, followed by three more washes with 1 TE before heating at 65-C for 10 min to denature the capsid proteins and DNase. Viral DNA concentration was measured via the Turner Quantech Digital Filter Fluorometer (Barnstead International, Dubuqu, IA, USA) using the PicoGreen dsDNA Quantitation kit (Molecular Probes, Eugene, OR, USA) according to the manufacturer’s instructions. Five hundred nanograms of viral DNA sample was loaded into each of eight lanes of a 1% low-meltingpoint agarose gel (Ultrapure L.M.P. Agarose, Invitrogen, Carlsbad, CA, USA). A 5-kb DNA marker was loaded into each of the remaining two lanes on each end of the gel. The DNA sample was resolved via pulsed-field gel electrophoresis (PFGE) in the CHEF-DR II model (BioRad, Hercules, CA, USA) with the following parameters: 6 V/cm; 18 h; 1.9–7.1 second switch; running buffer final temperature 14-C. One lane containing the viral sample as well as one lane containing the 5-kb DNA marker were
545
cut from the gel and stained with SYBR Gold (Molecular Probes) to visualize the DNA bands. The rest of the gel was not stained in order to protect the viral DNA from UV damage and to increase cloning efficiency. The stained lane was then used as a guide to cut out gel slices containing the 35- to 55-kb desired band from the remainder of the unstained gel. DNA was then purified from the gel by the GELase product (Epicentre Technologies, Madison, WI, USA) according to the manufacturer’s protocol. The DNA and digested agarose were then transferred to Centricon 100-kDa filter tubes (Millipore) and washed two times with 1 TE (1000 g, 15 min). After gel extraction, DNA concentration was again measured via the fluorometer as described above. Viral DNA was ligated into the pCC1 copy control BAC vector and cells were transformed according to the manufacturer’s protocol (Epicentre). Electrocompetent transformations were carried out with the Eppendorf Electroporator model 2510. Clones were screened for size and similarity by using BglII, HindIII, and NotI restriction enzymes (Promega). Four unique clones with insert sizes ranging between 2.6 and 3.85 kb were identified. Sequencing and Analysis of BAC clones. The following BAC vector oligonucleotides were synthesized (Sigma Genosys, The Woodlands, TX, USA) and used for initial sequencing of the BAC clones: pCC1 forward sequencing primer: 50 -GGATGTGCTGCAAGGCGA TTAAGTTGG-30 ; pCC1; Reverse sequencing primer: 50 CTCGTATGTTGTGTGGAATTGTGAGC-3 0 . Primer walking was undertaken to complete the remaining length of the inserts. Sequencing was carried out with the ABI PRISM BigDye version 3.0 sequencing chemistry (Applied Biosystems, Foster City, CA, USA) by either slab gel (ABI Prism 377) or capillary (ABI Prism 3100) electrophoresis. BAC clone sequences were analyzed and contigs were assembled via the EditSeq and SeqMan programs (DNASTAR, Inc.; Madison, WI, USA). Open reading frames (ORFs) were detected by using ATG, TTG, CTG, and GTG as start codons; TGA, TAA, and TAG as stop codons; and 34 codons (approximately 100 bases) as a minimum ORF size. Homology searches were conducted against the nonredundant (nr) NCBI GenBank database and by using the blastx program; significant homology was judged by using an E value of less than 0.001, except in the case of clone 31. All BAC clone sequences were submitted to the NCBI GenBank database. Viral Probe Synthesis. Nonradioactive probes were made up of two viral isolates, A7 and 72, and two BAC clones, clone 5 and clone 76. Both phages were isolated in the summer of 2002 by using bacterial host M12-2C and MN1-12.5, respectively. The random
546
S. SABET
ET AL.:
SEASONAL PRESENCE
OF
HYPERSALINE PHAGES
primed labeling of viral genome DNA or BAC clone plasmid DNA was performed with the DIG High Prime DNA Labeling and Detection Starter Kit II (Roche, Penzberg, Germany) following the manufacturer’s protocol. Briefly, DNA was digested with the AluI restriction enzyme (Promega), denatured, and incubated with DIG-High Prime mix overnight at 37-C. Efficiency spot testing for each probe and immunological detection were carried out according to the manufacturer’s protocol (Roche). The sensitivity of each probe was measured to be at 1 pg of DNA, and all probes were used at the recommended concentration of 25 ng/mL. Probes were stored at _20-C, thawed at 42-C, and denatured at 68-C for 10 min before each use.
Prehybridization was carried out for at least 30 min, and hybridization was done overnight in a glass hybridization tube (Robbins Scientific-SciGene, Sunnyvale, CA, USA) at 43-C in a roller hybridization oven (Hybaid Mini 10; National Labnet Co., Woodbridge, NJ, USA). The membrane was washed with 0.5 SSC/0.1% SDS at 70-C. All other washes and the immunological detection steps were done at room temperature on a rocker following the manufacturer’s instructions (Roche). Membranes were exposed to Kodak Biomax MS film (New Haven, CT, USA) either for 75 min or overnight at room temperature before film development. Membranes were stripped for reprobing according to the manufacturer’s protocol (Roche).
Dot Blot Sample Preparation and Hybridization. Viruses used in the hybridization assay were concentrated by ultracentrifuging 10–20 mL of 0.22-mm-filtered water sample at 41,000 rpm with a Beckman SW41 Ti rotor for 1.5 h at 4-C. Most of the supernatant was discarded, leaving the viral pellets in õ200 mL of the original water. Then 1 mL 5 TE was added and the sample was incubated at 4-C for 1 h before being transferred to Centricon YM 100-kDa MWCO centrifugal filter units (Millipore Corporation; Billerica, MA, USA) and centrifuged at 1000 g room temperature for 20 min. The viruses were then washed two times with 100 mL 5 TE. After the second wash, 25 mL 1 TE was added to each filter unit and an invert spin was done at 500 g for 1 min to collect the viruses. Viral samples were heated at 65-C for 10 min to denature the capsid proteins and release the DNA. DNA concentration was measured via the fluorometer as described previously. A7 and 72 DNA was extracted by using the Wizard Lambda Preps DNA Purification System (Promega) and used as controls to optimize the hybridization conditions and stringency. Genomic DNA from the two bacterial hosts, M12-2c and MN1-12.5a, as well as from Aeromonas hydrophila (ATCC strain 7966) and E. coli was extracted [2] and used as controls. All control DNA concentrations were quantified spectrophotometrically with an Eppendorf BioPhotometer (Eppendorf AG, Hamburg, Germany). The DNA samples were blotted onto a positively charged nylon membrane (Magna Charge Nylon 0.45 mm, GE Osmonics, Inc., Minnetonka, MN, USA) as previously described [11] by use of the Minifold I, Microsample Filtration Manifold apparatus (Schleicher & Schuell, Keene, NH, USA). Controls were titrated onto the blots with 1, 10, 100, and 500 ng of DNA. After blotting, the nylon membrane was then rinsed for 5 min in 2 standard saline citrate (SSC), the DNA was fixed by UV cross-linking (Spectrolinker XL-1000 UV crosslinker; Spectronics Corp., Westbury, NY, USA), and the membrane was stored at _20-C until further use.
Statistical Analysis. ANOVA was performed using seasonal data of cultured bacteriophages on each host. Box and whisker graphs were plotted to represent the seasonal and between-host differences of culturable bacteriophages. All statistical analysis and graphs were created with SPSS statistical software (SPSS Inc., Chicago, IL). GenBank Accession Numbers. For bacterial hosts MN12-2a and MN1-12.5a, the accession numbers are AY856383 and AY856384, respectively. Accession numbers for each BAC clone is as follows: clone 5, AY853713; clone 31, AY853715; clone 76, AY853716; and clone 122, AY853714.
Results Mono Lake Bacteria and Phage Isolates. 16S rRNA sequence homology of the MN12-2a and MN112.5a bacterial hosts indicated that the MN12-2a host shows 98.5% homology to a Halomonas boliviensis species [24] (GenBank accession number AY245449), whereas the MN1-12.5a host is 93% homologous to a Marinospirillum alkaliphilum strain Z4, which was isolated from a soda lake in Inner Mongolia, China [36] (GenBank accession number AF275713). To determine the seasonal dynamics of culturable bacteriophages in Mono Lake, the four bacterial hosts— M12-2c, M12-chla, MN12-2a, and MN1-12.5a—were used to isolate phages from the upper 10 m of the water column (the oxygenic layer) collected during different seasons. Figure 2 shows box and whisker plots of seasonal phage counts for all four hosts. Phages infecting the M12-2c host had the lowest counts in the summer (August 2001), with a slight increase in winter (February 2003), and the highest counts in autumn (October 2003) (Fig. 2A), although overall there is no significant difference among all three seasons according to ANOVA test (P = 0.119). Interestingly, the M12-chla phages seem to be prevalent strictly during the winter season, as no phage could be cultured in either the summer or autumn
S. SABET
ET AL.:
SEASONAL PRESENCE
OF
547
HYPERSALINE PHAGES
months (Fig. 2B). In keeping with this trend, the Halomonas-like phages were far more abundant during February, whereas lower counts were observed during August and October (Fig. 2C). The ANOVA test yielded a P value of 0.003. However, the opposite result was observed for the MN1-12.5a host, which yielded an abundant number of phages in the summer season but fewer in the subsequent winter and autumn months (Fig. 2D). These data illustrate a differential seasonal pattern of phage prevalence infecting different species of Mono Lake haloalkaliphiles, suggesting that the phage community is dynamic and in a state of flux. To determine the vertical distribution of bacteriophages infecting specific bacterial hosts, water samples collected from stations 3 and 6 from the anoxic layer (35 m) during October 2003 were also used for phage plaque assays (Table 1). Phages infecting the M12-2c and MN1-12.5a hosts were isolated from this depth, but no phages of the other two hosts could be found. DNA Sequence Analysis of Cloned Viral DNA Fragments
To characterize the uncultured Mono Lake phage population, we attempted to clone whole viral DNA genomes via BAC vector cloning technology. Sample preparation entailed the removal of any possible contaminating DNA by carrying out a DNase digest of the whole viral sample before capsid denaturation. Figure 3 shows that DNase digestion was efficient at removing extraviral DNA (lane 1 vs untreated lane 2), whereas it had little or no effect on DNA encapsulated by the viral protein coat (lanes 3 and 4). Therefore, the DNase treatment ensured that the DNA sample that was subsequently purified from PFGE for cloning was indeed viral in origin. Figure 4A shows the PFGE of the viral community DNA collected from station 3, 2 m during February 2002. The 35- to 55-kb region (Fig. 4A) was targeted as the desired cloning region of interest because a relatively large amount of viral DNA was present there, which would favor cloning efficiency. The purity and size of the 35- to 55-kb DNA sample retrieved after gel purification is shown in Fig. 4B. After ligation and transformation, approximately 200 clones were screened and five clones were found to contain an insert (Fig. 4C) with insert sizes
Figure 2. Box and whisker plots of seasonal plaque counts on four
Mono Lake bacterial hosts: (A) M12-2c, (B) M12-chla, (C) MN122a, (D) MN1-12.5a. Outlier is indicated by the name of the station where water sample is assayed. Surface water was sampled from stations 1, 3, 6, 8, 10, 11, and 12 during three different seasons (*summer, from 2- to 10-m depths; **winter and autumn, from 2-m depth).
548
S. SABET
ET AL.:
SEASONAL PRESENCE
OF
HYPERSALINE PHAGES
Table 1. Plaque counts on four Mono Lake bacterial hosts using water samples collected from the anoxic layer (35 m) of stations 3 and
6 in October 2003 Plaque-forming units (pfu/L) Bacterial host Station
M12-2c
M12-chla
MN12-2a
MN1-12.5a
132 110 121 15
0 0 0 0
0 0 0 0
176 264 220 62
3 6 Average (pfu/L) SD
ranging between 3 and 5 kb. These inserts were significantly smaller than the expected whole viral genome size estimated by the PFGE DNA size markers. Restriction analysis identified four unique viral clones, designated as clones 5, 31, 76, and 122. Selective sequence homology analysis results of the inserts are shown in Table 2. The most interesting insert was clone 5, which contained the most number of ORFs (34) ranging in size from 133 to 684 bp. The average ORF size was 272 bp, whereas the median ORF size was 226 bp. Of the four clones, clone 5 ORFs showed the highest homology to phage-related sequences based on the E value and score number (Table 2). The most relevant and highest hits, shown in Table 2, were to several different species of single-strand DNA-binding protein from Chromobacterium violaceum, Nitrosomonas europaea, Streptococcus pyogenes phage 315.4, temperate phage PhiNIH1.1, and prophage LambdaSa1 phage. Holins also showed significant hits. The rest of the ORFs either returned no homology results from the database or yielded very low scores with genes of human or mouse origin (data not shown).
Clone 76 contained 14 ORFs between 123 and 654 bp, with an average length of 251 bp and a median size of 216 bp. The highest scoring match was to a hypothetical protein of Rhodobacter sphaeroides (Table 2). Other hits included sequence homology to a Mesorhizobium species, as well as to the Pas6 protein of 7Asp2, a virus that infects the eubacterial Actinoplanes genus [8]. Another
A
B M
st 3 M 200 kb 150 kb 100 kb 50 kb 35 kb 15 kb
C
Bgl II
Hind III
1
50 kb 35 kb 10 kb
Not I
1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 M kb 21
M
1
2
3
4
M
kb 5-3.5
200 150
2-1.9 1.6-1.4 0.9-0.8
100
0.5
50
10 5 Figure 3. DNase treatment of phage l control DNA and of a Mono Lake water sample before gel purification. Lane 1 contained 200 ng of l DNA digested with 1 U DNase. Lane 2 contained l DNA minus DNase. Station 12 15-m DNA sample before (lane 3) and after (lane 4) DNase treatment. Lane M contains 5-kb DNA ladder mixed with 50-kb l marker.
Figure 4. Gel purification and subsequent cloning of DNA
extracted from the station 3 2-m Mono Lake water sample. (A) Whole community viral DNA was resolved on a 1% PFGE lowmelting-point agarose gel. Arrows and brackets highlight bands of interest; the 35- to 55-kb region was cut out; lane M contains the 5-kb DNA ladder. (B) Relative purity of the 35- to 55-kb band of interest in lane 1 after gel purification; lane M contains the 5-kb DNA ladder. (C) Restriction digest analysis of five clones containing different-sized inserts. The enzyme used in each reaction is noted above the gel image. Lanes 1, clone 31; lanes 2, clone 5; lanes 3, clone 75; lanes 4, clone 76; lanes 5, clone 122; lane M = l+EcoRI+HindIII DNA marker.
3.318
2.604
31
76
14
Putative NTP-binding protein (bacteriophage phi AT3) [YP_025055.1] Hypothetical protein Rsph020849 (R. sphaeroides) [ZP_00004944.1] Hypothetical protein MBNC02002420 (Mesorhizobium sp. BNC1) [ZP_00197163.1] Pas6 (Actinoplanes phage phiAsp2) [AAT36754.1] Capsid protein gpC (virus PhiCh1) [AAM88682.1]
Single-strand DNA-binding protein (C. violaceum ATCC 12472) [AAQ59563.1] Single-strand binding protein family (N. europaea ATCC 19718) [NP_842444.1] Putative single-strand DNA-binding protein phage associated (S. pyogenes phage 315.4) [NP_665053.1] Single-strand binding protein (temperate phage PhiNIH1.1) [AAL15055.1] Prophage LambdaSa1, single-strand binding protein (Streptococcus agalactiae 2603V/R) [NP_687595.1] Prophage LambdaSo, holin, putative (Shewanella oneidensis MR-1) [NP_718538.1] Phage holin, putative (Pseudomonas putida KT2440) [NP_746013.1] Hypothetical protein BQ11480 (Bartonella quintana strain Toulouse) [YP_032690.1] Hypothetical protein Aaphi23p07 (bacteriophage Aaphi23) [NP_852729.1]
43
0.001
42
0.059
48
50
_ 2.00E 05
_ 5.00E 05
55
_ 4.00E 07
71
65
_ 5.00E 10
_ 7.00E 12
67
_ 2.00E 10
101
67
_ 2.00E 10
_ 2.00E 20
67
_ 2.00E 10
33
69
_ 7.00E 11
1.5
87
36.4/48.9
42.0/53.1
36.4/60.2
47.1/67.3
42.9/66.7
22.8/40.8
43.1/50.8
37/53.1
38/46.7
33.6/53.3
31.1/51.1
31.1/51.1
33.1/50.7
39.1/54.7
Score
E value _ 3.00E 16
CTG start; ORF 5 CTG start; ORF 5
ATG start; ORF 1 CTG start; ORF 5
Not an ORF hit; entire 3.3-kbp insert used in search ATG start; ORF 1
GTG start; ORF 6
ATG start; ORF 1
ATG start; ORF 1
ATG start; ORF 10
ATG start; ORF 10
ATG start; ORF 10
ATG start; ORF 10
ATG start; ORF 10
Comments
OF
8
34
% Identity/ Similarity
SEASONAL PRESENCE
Only ORFs with close match in the database are presented.
3.854
Insert size (kb)
ORF hits [accession no.]
ET AL.:
5
Clone no.
No. ORFs (larger than 100 bp)
Table 2. Sequence homology analysis of the cloned Mono Lake viral DNA fragments
S. SABET HYPERSALINE PHAGES
549
550
ET AL.:
SEASONAL PRESENCE
OF
ila
identical hybridization pattern and intensity (Fig. 5B, clone 5 results; clone 76 data not shown) indicating that both clone 5- and clone 76-like phages were present in the lake during all seasons sampled. However, there was a difference overall in the intensity of hybridization among the months. Clone 5 and clone 76 probes showed the strongest hybridization to the February and March samples (Figs. 5B, C), followed by the April samples (Fig. 5B). The August samples showed relatively weak hybridization overall, and the November samples showed the weakest hybridization (Fig. 5B). These results suggest that seasonal variability of clone 5- and clone 76-like phages exists at the same stations throughout the year,
hy dr
Ae ro m on
as
12 . M 12 5a -2 c (b la nk )
N1 M
Cl
Cl
on e
76 on e Aφ 5 φ2
A Seasonal Dynamics of Cultured and Uncultured Bacteriophages by Dot Blot DNA Hybridization To determine if there is any seasonal Analysis.
variability of the Mono Lake phage community, dot blot hybridization was carried out by using genomic probes made from the cloned viral fragments, clones 5 and 76, as well as the phage isolates, A7 and 72. Figure 5A shows the hybridization results of each probe with control DNA. Clones 5 and 76 hybridized to each other as well as weakly to the genome of the M12-2c host, but they did not hybridize to either of the viral isolates, to the MN1-12.5a host, or to the negative control genomic DNA from A. hydrophila. The hybridization to the M122c host was unexpected, which implies that this bacterial host shares some sequence homology to the viral fragments or to the clone vector. However, a control probe made of empty vector alone did not hybridize with any of the control DNA including the M12-2c host (data not shown) suggesting that homology exists between the cloned viral fragments and the bacterial chromosomal DNA. Whether this is due to lysogeny or transduction is not known. The cross hybridization between clones 5 and 76 could be the result of common vector sequence within the clone because there was no sequence homology between clone 5 and 76 viral inserts as determined by sequence analysis (see results in Table 2). The phage isolates also hybridized to each other, and the A7 to the MN1-12.5a host, again unexpected results implying sequence homology; but they did not hybridize to the cloned viral fragments (Fig. 5A). The crosshybridization of these isolates was surprising because they were isolated from different hosts (A7 from M122c, a Vibrio spp.; and 72 from the MN1-12.5a, identified as a Marinospirillum spp.), which apparently do not share any similarities and are not closely related. Hybridization of clones 5 and 76 probes with Mono Lake viral community DNA collected from various stations at different seasons and depths showed nearly
HYPERSALINE PHAGES
op h
similar match was to the capsid protein of the haloalkaliphilic phage, 7Ch1. Again, as for clone 5, the majority of ORFs had no match in the database. Clone 31 contained eight ORFs, ranging in size from 108 to 153 bp, almost all of which did not show any matches with the current database. Although none of the hits met the search parameters, one of the ORFs was found to be homologous to a putative NTP-binding protein from 7AT3, a temperate phage of Lactobacillus casei (Table 2). In addition, when the entire 3.3-kb insert was used in the search, the highest scoring homology for clone 31 (E value=0.059) was to a hypothetical protein of Aa723, a temperate phage of Actinobacillus actinomycetemcomitans. Clone 122 did not share any relevant homology with the database and is not included in the final analysis and Table 2.
S. SABET
Clone 5 probe Clone 76 probe Aφ probe φ2 probe
B
st2 st6 st7 st11 August 2003 November 2003 March 2004 April 2004
C 2m 8m
12m 16m 20m 24m 28m 35m
Figure 5. Temporal and spatial distribution of specific bacteriophages among the Mono Lake viral community revealed by dot blot DNA hybridization. (A) Evaluation of the stringency and specificity of each probe used in the hybridization study using control DNA extracted from bacteria, phage, and cloned viral fragments; target DNA in each well is labeled above image, while the probe used for each blot is identified on the right side of image. (B) Hybridization of the clone 5 probe with the Mono Lake viral community DNA collected from the upper 9 m of the water column at different stations and times. One microgram of DNA was blotted from August, November, and April; whereas 500 ng of DNA was used from the March samples. (C) Hybridization of the clone 5 probe to 500 ng of DNA collected from various depths of station 6 during February and March, 2004.
S. SABET
Aφ
ET AL.:
SEASONAL PRESENCE
OF
551
HYPERSALINE PHAGES
φ2
Discussion 25 µg, 2m May 2004 1 µg, 2m 1 µg, 8m
July 2004
1 µg, 35m Figure 6. Dot blot hybridization of A7 and 72 probes with Mono Lake viral community DNA. Water samples were collected from station 6 at 2-, 8-, and 35-m depths in May and July of 2004. Amount of target DNA blotted is labeled on the right of image. Film was exposed overnight before development.
and that the phage sequences are more prevalent during the late winter and spring seasons and less prevalent during the summer and autumn months. One possible exception may be station 7, which showed weaker hybridization compared to the other stations in April (Fig. 5B). This observation may reflect microenvironmental changes at different stations during the same season. No significant variability of clones 5- and 76-like phages was observed along a water column depth profile at station 6 (Fig. 5C), despite the stratification of the water column with oxygenic and anoxic layers. Surprisingly, hybridization of the A7 and 72 probes with the same set of DNA blots under the same conditions yielded negative results (data not shown). This suggests that both A7- and 72-like phages were less abundant within the viral community than the uncultured clones 5- and 76-like phages. A 25-fold increase in the amount of viral community DNA on the blots from 1 to 25 mg and an increase in the film exposure time from 75 min to 18 h resulted in a positive hybridization signal between the A7/72 probes and the Mono Lake viral community DNA (Fig. 6). A7 hybridized most strongly to the 1 mg viral DNA collected at the 8-m depth, but much less to the 2- and 35-m depths, implying that variability exists in the vertical distribution of the viral community in Mono Lake and that the A7like virus is most prevalent at 8 m. 72 also hybridized more strongly to the sample collected from 8 m relative to the 2-m sample; however, the intensity of 72 hybridization was very low relative to A7. 72 very weakly hybridized to the 2-m sample, and is seemingly absent altogether at 35 m. In addition, the amount of target DNA was also a crucial factor since the A7 and 72 probes hybridized more intensely when 25 mg of viral community DNA was blotted—a 25-fold increase. These data suggest that A7 and, in particular, 72 represent minority populations within the Mono Lake bacteriophage community since more DNA and a longer exposure time were needed to detect their presence.
Several previous studies have reported the isolation and phylogenic characterization of the Mono Lake bacterial community [7, 14, 17, 18, 22, 27, 32]; however, very little is known of the Mono Lake viral community. It is not surprising that four of the Mono Lake bacterial isolates that have successfully served as phage hosts resemble culturable species from similar environments elsewhere in the world. H. boliviensis strain LC1, the closest match for MN12-2a, is a moderately halophilic, psychrophilic, alkali-tolerant bacterium isolated from a soil sample around lake Laguna Colorada in Bolivia [24]. M. alkaliphilum Z4 strain, the match for MN1-12.5a, was isolated from Haoji soda lake in the Inner Mongolia region of China and optimally grows at 2% NaCl, pH 9.5, and 37-C [36]. M12-2c was identified as a Vibrio spp., perhaps the most commonly isolated species from the brackish environment [5, 28]. However, M12-chla only resembles a bacterium previously isolated from Mono Lake. A model describing the effect of viral infection on the maintenance of bacterial diversity was put forth by Wommack and Colwell [35] in which specific host subpopulations that underwent fast growth and reached high concentrations were selectively killed by their viruses. This model takes into account the presence of multiple phage–host systems in the same environment, each with its own independent cycle of bacterial host growth and reduction, and is generally known as the Bkill-the-winner^ theory. The observation of seasonal dynamics of culturable bacteriophages in Mono Lake partially supports this theory. For example, phages infecting host M12-chla were found only during the winter season when their host was initially isolated, suggesting the winter condition favors the growth of the host and, subsequently, the proliferation of its phages. Strong seasonal variability of MN12-2a and MN1-12.5a phage concentration (ranging from 50 to over 8000 pfu/L) was also observed, suggesting the active interaction between these phages and their hosts. Future investigations that incorporate the temporal dynamics of specific host populations may provide additional insights to support this theory. However, phages infecting M12-2c did not display any temporal patterns that support the kill-the-winner theory. They were present at constantly low levels (100–200 pfu/L) at all seasons and depths sampled (Fig. 2 and Table 1), which suggests that the kill-thewinner theory may not apply to all host–phage systems. Temperate interaction between virus and host, where virus and host concentrations reach equilibrium, may also be an important mode of phage and host interaction in the aquatic environment [12]. All plaques observed and counted on their specific host lawn were clear or semiclear in this study, suggesting lytic infection. Turbid plaques resembling those formed
552
by temperate coliphage on their host lawn have not been observed or confirmed with host–phage systems isolated from Mono Lake. However, we cannot completely rule out the presence of temperate phage among culturable phages found in Mono Lake. For example, although both A7 and 72 form clear plaques on their host lawns, we also found cross-hybridization between the A7 genome and its host, MN1-12.5a. Future genetic studies focusing on identification of the homologous DNA sequence between this phage and its host will shed light on the question of Blytic^ or Blysogenic^ of this specific host– phage system. Assuming a portion of culturable phages observed in Mono Lake is temperate, then the increase in specific phage density could reflect an event of lysogenic induction that changes the host bacterial population from steady state to the state of Bemergency response^ under environmental stress conditions. Such an observation was noted for halophages isolated in a Jamaican hypersaline lagoon. When the pond was in an evaporative state with a robust bacterial density and high salinity (5 M [Cl_]), relatively few phages were isolated, whereas when the pool was diluted by fresh water rains and the halophilic bacterial population was destroyed due to reduced salinity (1.3 M [Cl_]), there was a much higher phage yield [31]. It was suggested that bacterial populations may escape viral predation by exploiting an environmental niche (in this case, high salinity). Hosts would maintain a stable relationship with their viral parasites within a robust population under ideal host conditions, but when environmental conditions were no longer ideal and the host population was Bdoomed^ to death, viruses would then act as scavengers by becoming active lytic phages [31]. Therefore, the seasonal variability of phages in Mono Lake may be indicative of the removal of certain predominant bacterial populations as a result of environmental changes that produce a denser host population (kill-the-winner theory); or it may be due to environmental changes that trigger lysogens to become lytic. The absence of a more defined seasonal expression of phages infecting host M12-2c may be due to a continuous, stable presence of the M12-2c host in Mono Lake year-round, or simply to a chronic, nonlytic infection. Our attempt at cloning entire bacteriophage genomes has not been successful in spite of the promise of the BAC vector technology [4, 26]. This failure may be due to the expression of toxic phage genes in the E. coli host and/or the possibility that the lake viruses were composed of circular DNA, or DNA that is otherwise modified. The smaller than expected size inserts may be due to the shearing of the viral genomes during the cloning procedure. It was also interesting to observe that two of the five clones containing viral fragments were identical based on initial sequencing, implying some viral fragments are more clonable than others due to cloning
S. SABET
ET AL.:
SEASONAL PRESENCE
OF
HYPERSALINE PHAGES
bias. We are confident the cloned inserts are of viral origin based on the follow reasons: (1) The Centricon-80 with 100,000 NMWL does not favor the recovery of dissolved extracellular DNA; (2) Dissolved DNA forms a smear on PFGE gels rather than DNA bands with defined sizes, therefore, extracting DNA from a defined-size DNA band does not favor recovery of any significant amount of dissolved DNA; and (3) DNase treatment before PFGE efficiently removed dissolved DNA from the samples. Sequence homology analysis indicated that the uncultured Mono Lake phages contain both viral and bacterial genes (Table 2). The largest ORF exhibited relatively high homology to single-strand DNA-binding protein from several different bacterial and viral species (Table 2). Single-strand DNA-binding protein, a helix destabilizer, has high affinity to single-stranded DNA and is involved in replication, recombination, and repair. It is not unreasonable to think that viruses that harbor this gene would employ it in either their own replication or perhaps contribute to the function of their host’s genome. Other sequence hits were to phage holin, Pas6 protein of 7Asp2, and capsid protein from 7Ch1. Along with lysin, holin is part of a dual protein system in lytic double-stranded DNA phages that forms a pore in the bacterial cell’s membrane allowing lysin access to the cell wall to degrade the peptidoglycan layer ultimately for phage release. Pas6 is a hypothetical protein believed to be involved in head morphogenesis, and is similar to bacteriophage SPP1 Gp7 (accession number X89721). Bacteriophage 7Ch1 is a haloalkaliphilic phage that infects Natrialba magadii, a bacterium isolated from an alkaline East African soda lake [29]. Interestingly, 7Ch1 contains segmented nucleic acid composed of both DNA and RNA molecules [34]. The same ORF that showed high homology to Pas6 and 7Ch1 also showed high homology to a hypothetical protein from a Mesorhizobium species. ORFs from clone 31 did not reflect any high homology to current database sequences. In light of this observation, a search was conducted using the sequence of the entire 3.3-kb insert, and a homology hit with an E value of 0.059 was returned to a hypothetical protein from bacteriophage Aa723, isolated from an oral bacterium. This temperate phage expresses a 44-kb linear DNA molecule that has 1.6-kb terminal redundancy signifying circular permutations [33]. Another finding from clone 31 is the sequence hit to a putative NTPbinding protein from 7AT3, a temperate bacteriophage isolated from an anaerobic lactic-acid-forming bacterium. The fact that clone 31 did not express any genes with strong homologies to current database sequences implies that clone 31 ORFs are unique and belong to a phage that is quite distinct from any that has been described thus far. This is not surprising because as much as 74%
S. SABET
ET AL.:
SEASONAL PRESENCE
OF
553
HYPERSALINE PHAGES
of metagenomic sequences do not have hits to current database sequences [3], indicating that a large percentage of environmental viruses are still undescribed and unlike any of their laboratory isolate cousins. Hybridization results obtained by using probes made of cultured phage isolates and uncultured viral clones in this study led to the conclusion, long suspected, that laboratory isolates do not accurately represent the dominant or majority species in the environment, and that the uncultured members of the microbial community are those that truly reflect the dominant, majority groups. The phages represented by clones 5 and 76 were present at a fairly constant level year-round. Perhaps their hosts are more similar in seasonal abundance to the M12-2c host, which yielded phages year-round, and less like the other hosts, which only yielded phages seasonally (Fig. 2). An interesting observation of the clone 5 and clone 76 seasonal hybridization pattern is the relatively weak hybridization to the November 2004 DNA sample (Fig. 5B). This may relate to the physical mixing event that occurred in November 2003. The last time the lake underwent holomixis (lake turnover) was between October 22 and November 14, 2003. The November water sample used in this study was collected on November 14, the last day of holomixis (R. Jellison, pers. comm.). It is likely the microbial community was affected during the holomixis such that the abundance of the phage population as represented by clones 5 and 76 was reduced compared to when the lake was stratified. This could explain the relatively weaker hybridization intensity of these clones to the November sample. Another possibility is that the phages as represented by clones 5 and 76 do indeed express a seasonal profile and are simply not as abundant during the autumn months. Such a result would confirm our findings (Fig. 2) of seasonal variability among phage isolates. The A7 and 72 probes hybridized with each other’s phage DNA (Fig. 5A), indicating that sequence homology exists between these two viruses although they were isolated from very distinct bacterial hosts. This result contradicts the previous finding that only phages infecting related host bacteria share genome similarities [10]. The results here suggest that bacteriophages from the same environment may share some genetic homology as it relates to increased survivability and/or better infection in their shared respective habitat, even though they infect diverse hosts. Environmental viruses might then have more in common with each other on the basis of their geographical distribution, which may in turn lead to a way to classify them based on their shared Bgeographical^ genes, in addition to their host range. In summary, the findings in this report contribute to a better understanding of the seasonal population dynamics of both phage isolates and uncultured members of the Mono Lake bacteriophage community.
Furthermore, we have illustrated a more detailed image of the genetic makeup of these phages. They contain typical phage genes that encode for holins and capsid proteins, as well as for single-strand DNA-binding proteins. It is very likely that the phage community contains temperate phages as well as lytic phages, linear DNA-containing viruses as well as viruses that contain circular permutations.
Acknowledgments
We thank Dr. Grieg Steward and Dr. Robert Jellison for their technical expertise and their collaboration in field sampling. We also thank Sandra Roll, Kimberly Rose, and other members of the SNARL team for collecting water samples during this study. We thank Sam Choi for his help with statistical analysis and anonymous reviewers for their suggestions for improving this manuscript. This study was supported by NSF awards DEB-01-30528, DEB-01-29174, and DEB-01-29160 to S.C.J., G.F.S., and R.J., respectively.
References 1. Adams, MH (1959) Bacteriophages. Interscience Publishers, New York 2. Ausubel, FM (1987) Current Protocols in Molecular Biology. Greene Publishing Associates and Wiley, Brooklyn, NY 3. Breitbart, M, Salamon, P, Andresen, B, Mahaffy, JM, Segall, AM, Mead, D, Azam, F, Rohwer, F (2002) Genomic analysis of uncultured marine viral communities. Proc Natl Acad Sci USA 99: 14250–14255 4. Choi, S, Wing, RA (2003) The Construction of Bacterial Artificial Chromosome (BAC) Libraries 2003. Clemson University Genomics Institute (CUGI), Clemson, SC 5. Eilers, H, Pernthaler, J, Glockner, FO, Amann, R (2000) Culturability and in situ abundance of pelagic bacteria from the North Sea. Appl Environ Microbiol 66: 3044–3051 6. Fuhrman, JA (1999) Marine viruses and their biogeochemical and ecological effects. Nature 399: 541–548 7. Humayoun, SB, Bano, N, Hollibaugh, JT (2003) Depth distribution of microbial diversity in Mono Lake, a meromictic soda lake in California. Appl Environ Microbiol 69: 1030–1042 8. Jarling, M, Bartkowiak, K, Pape, H, Meinhardt, F (2004) The genome of phiAsp2, an Actinoplanes infecting phage. Virus Genes 29: 117–129 9. Jellison, R, Melack, JM (1993) Algal photosynthetic activity and its response to meromixis in hypersaline Mono Lake, California. Limnol Oceanogr 38: 818–837 10. Jiang, S, Fu, W, Chu, W, Fuhrman, JA (2003) The vertical distribution and diversity of marine bacteriophage at a station off Southern California. Microb Ecol 45: 399–410 11. Jiang, S, Steward, G, Jellison, R, Chu, W, Choi, S (2004) Abundance, distribution, and diversity of viruses in alkaline, hypersaline Mono Lake, California. Microb Ecol 47: 9–17 12. Jiang, SC, Paul, JH (1998) Significance of lysogeny in the marine environment: studies with isolates and a model of lysogenic phage production. Microb Ecol 35: 235–243
554
13. Mann, NH, Cook, A, Millard, A, Bailey, S, Clokie, M (2003) Marine ecosystems: bacterial photosynthesis genes in a virus. Nature 424: 741 14. Mason, DT (1967) Limnology of Mono Lake, California. Univ Calif Publ Zool 83: 1–110 15. Middelboe, M (2000) Bacterial growth rate and marine virus–host dynamics. Microb Ecol 40: 114–124 16. Nuttall, SD, Dyall-Smith, ML (1993) HF1 and HF2: novel bacteriophages of halophilic archaea. Virology 197: 678–684 17. Oremland, RS, Hoeft, SE, Santini, JM, Bano, N, Hollibaugh, RA, Hollibaugh, JT (2002) Anaerobic oxidation of arsenite in Mono Lake water and by a facultative, arsenite-oxidizing chemoautotroph, strain MLHE-1. Appl Environ Microbiol 68: 4795–4802 18. Oremland, RS, Miller, LG, Culbertson, CW, Robinson, SW, Smith, RL, Lovley, D, Whiticar, ML, King, GM, Kene, RP, Iversen, N, Sargent, M (1993) Aspects of the biogeochemistry of methane in Mono Lake and the Mono Basin of California. In: Oremland, RS (Ed.) Biogeochemistry of Global Change. Chapman & Hall, New York 19. Oren, A (2002) Halophilic microorganisms and their environments. In: Seckbach, J (Ed.) Cellular Origin and Life in Extreme Habitats, Vol. 5. Kluwer Academic Publishers, Boston, pp 497–498 20. Oren, A, Bratbak, G, Heldal, M (1997) Occurrence of virus-like particles in the Dead Sea. Extremophiles 1: 143–149 21. Pelagos Corporation (1987) A bathymetric and geologic survey of Mono Lake, California. Project report, San Diego, CA 22. Pikuta, EV, Hoover, RB, Bej, AK, Marsic, D, Detkova, EN, Whitman, WB, Krader, P (2003) Tindallia californiensis sp. nov., a new anaerobic, haloalkaliphilic, spore-forming acetogen isolated from Mono Lake in California. Extremophiles 7: 327–334 23. Post, FJ (1981) Microbiology of the Great Salt Lake north arm. Hydrobiologia 81: 59–69 24. Quillaguaman, J, Hatti-Kaul, R, Mattiasson, B, Alvarez, MT, Delgado, O (2004) Halomonas boliviensis sp. nov., an alkalitolerant, moderate halophile isolated from soil around a Bolivian hypersaline lake. Int J Syst Evol Microbiol 54: 721–725 25. Seaton, GG, Hurry, VM, Rohozinski, J (1996) Novel amplification of non-photochemical chlorophyll fluorescence quenching following viral infection in Chlorella. FEBS Lett 389: 319–323
S. SABET
ET AL.:
SEASONAL PRESENCE
OF
HYPERSALINE PHAGES
26. Shizuya, H, Birren, B, Kim, U, Mancino, V, Slepak, T, Tachiiri, Y, Simon, M (1992) Cloning and stable maintenance of 300-kilobasepair fragments of human DNA in Escherichia coli using an Ffactor-based vector. Proc Natl Acad Sci USA 89: 8794–8797 27. Switzer Blum, J, Burns Bindi, A, Buzzelli, J, Stolz, JF, Oremland, RS (1998) Bacillus arsenicoselenatis, sp. nov., and Bacillus selenitireducens, sp. nov.: two haloalkaliphiles from Mono Lake, California that respire oxyanions of selenium and arsenic. Arch Microbiol 171: 19–30 28. Thompson, FL, Iida, T, Swings, J (2004) Biodiversity of vibrios. Microbiol Mol Biol Rev 68: 403–431 29. Tindall, BJ, Ross, HNM, Grant, WD (1984) Natronobacterium gen nov. and Natronococcus gen. nov., two new genera of haloalkaliphilic archaebacteria. Syst Appl Microbiol 5: 41–57 30. van Hannen, EJ, Zwart, G, van Agterveld, MP, Gons, HJ, Ebert, J, Laanbroek, HJ (1999) Changes in bacterial and eukaryotic community structure after mass lysis of filamentous cyanobacteria associated with viruses. Appl Environ Microbiol 65: 795–801 31. Wais, AC, Daniels, LL (1985) Populations of bacteriophage infecting Halobacterium in a transient brine pool. FEMS Microbiol Ecol 31: 323–326 32. Ward, BB, Martino, DP, Diaz, MC, Joye, SB (2000) Analysis of ammonia-oxidizing bacteria from hypersaline Mono Lake, California, on the basis of 16S rRNA sequences. Appl Environ Microbiol 66: 2873–2881 33. Willi, K, Meyer, J (1998) DNA analysis of temperate bacteriophage Aa(phi)23 isolated from Actinobacillus actinomycetemcomitans. Mol Gen Genet 258: 323–325 34. Witte, A, Baranyi, U, Klein, R, Sulzner, M, Luo, C, Wanner, G, Kruger, DH, Lubitz, W (1997) Characterization of Natronobacterium magadii phage phi Ch1, a unique archaeal phage containing DNA and RNA. Mol Microbiol 23: 603–616 35. Wommack, KE, Colwell, RR (2000) Virioplankton: viruses in aquatic ecosystems. Microbiol Mol Biol Rev 64: 69–114 36. Zhang, W, Xue, Y, Ma, Y, Grant, WD, Ventosa, A, Zhou, P (2002) Marinospirillum alkaliphilum sp. nov., a new alkaliphilic helical bacterium from Haoji soda lake in Inner Mongolia Autonomous Region of China. Extremophiles 6: 33–37
