SEARCH

SEARCH BY CITATION

Keywords:

  • Gammaproteobacteria;
  • anoxygenic phototrophic bacteria;
  • pufM gene

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Author contribution
  9. References

Known anoxygenic photosynthetic bacteria (APB) affiliated to Gammaproteobacteria usually use anaerobic metabolism and are restricted to oxygen-free habitats. Here, we report abundant (average of 34.5%) presence of diverse APB related to γ-like Proteobacteria in oxic oceanic surface water as indicated by the pufM gene, that encodes the M subunit of the light reaction centre complex. Thus, our sequences were most likely derived from aerobic anoxygenic phototrophs (AAnP). Two genetically distinct genotypes were revealed: one was from the oligotrophic North Pacific Ocean Gyre and the other, was from the trophic East China Sea and Bering Sea. The discovery of abundant presence of novel γ-like Proteobacterial pufM gene in the oxic seawater extends the functional ecotypes of AAnP.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Author contribution
  9. References

Anoxygenic photosynthetic bacteria (APB) have attracted much attention, because of their ability to perform photosynthesis without producing oxygen, an ancient process forming the evolutionary basis for all current types of photosynthesis (Imhoff, 1995; Xiong et al., 2000; Raymond et al., 2002). In marine environments, two anoxygenic phototrophic pathways have been revealed so far. One pathway functions only in the absence of oxygen. This predates oxygenic photosynthesis, and nowadays is restricted to anaerobic anoxygenic phototrophic bacteria (AnAnPB) that inhabit sunlight-reachable and oxygen-free habitats, which results from oxidation of the global ocean by Cyanobacterial oxygenic photosynthesis, emerging some 2100 Myr ago (Rye & Holland, 1998). The other pathway requires oxygen and is carried out by aerobic anoxygenic phototrophic bacteria (AAnPB), which were first reported only 25 years ago (Shiba et al., 1979; Shiba & Simidu, 1982). Great interest has been aroused in their wide ecological significance and their distribution in various oxic marine environments because of their capability of producing photoinduced electron transport only under aerobic conditions (Yurkov & Beatty, 1998; Kolber et al., 2000, 2001; Goericke, 2002; Allgaier et al., 2003; Rathgeber et al., 2004). The marine AAnPB identified have been found only in Alphaproteobacteria. AAnPB differ from AnAnPB with respect to their oxygen requirement in synthesizing bacteriochlorophyll a (Bchl a), and they cannot grow without oxygen even in light (Yurkov & Beatty, 1998; Rathgeber et al., 2004). Thus, it is usually considered that there is little anoxygenic photosynthesis carried out by γ-like Proteobacteria in oxic marine environments.

AAnPB and AnAnPB share a conserved photosynthetic apparatus. A 46-kb gene cluster accommodates most genes involved in bacteriochlorophyll-containing photosynthesis, including genes coding for subunits of the reaction centre complex (pufL and pufM) (Nagashima et al., 1997; Beja et al., 2002). The pufM gene has been used recently to assess the diversity of AAnPB in marine environments (Beja et al., 2002; Allgaier et al., 2003; Oz et al., 2005; Schwalbach & Fuhrman, 2005; Yutin et al., 2005), and the dominance of Roseobacter-like AAnPB has been revealed by analysis of environmental samples from the Red Sea, Mediterranean Sea and Pacific Ocean (Beja et al., 2002; Oz et al., 2005). Very recently a novel primer set for the pufM gene has been developed, and a wider diversity among marine AAnPB has been revealed including several new variants in environmental DNA samples and genomic libraries (Yutin et al., 2005). However, whether the participation of the γ-like subclass of Proteobacteria in aerobic anoxygenic photosynthesis occurs also in oxic marine environments is still unclear, because no evidence has been found apart from one γ-like Proteobacterial pufM gene sequence obtained from the BAC library, which came from c. 45 km offshore of Moss Landing, California (Beja et al., 2002).

Here we report abundant presence of the diverse γ-like Proteobacterial pufM gene in a variety of marine environments, revealed when we investigated the biogeography of AAnPB on a large scale, covering tropical, subtropical, temperate and cold waters, representing trophic, subtrophic and olitotrophic waters.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Author contribution
  9. References

Sample collection and extraction of community DNA

Surface water samples (above 1 m depth) were collected from the Eastern North Pacific subtropical area, the North Pacific Ocean gyre, the temperate East China Sea and the subarctic Bering Sea (Fig. 1). Seawater was sampled using a Sea-Bird 911 plus CTD rosette assembled with Niskin bottles. Concentrations of dissolved oxygen were detected using Clark-type oxygen sensors. Chlorophyll α (Chl α) was SeaWiFS-retrieved at 9 km resolution and averaged for a period of 1 month over the corresponding investigation period. Subsamples (3–5 L) for DNA extraction were prefiltered through 200-μm mesh and subsequently filtered onto a 0.22 μm pore size filter (Pall, Gelman Sciences Inc.). These filters were immediately frozen in liquid nitrogen and then transferred to −20°C for storage. Community DNA was extracted using the hot sodium dodecyl sulfate, phenol : chloroform : isoamyl-alcohol, and ethanol precipitation extraction protocol initially described by Fuhrman et al. (1988).

image

Figure 1.  Sampling locations and sites. BS-br02: Station br02 in the Bering Sea; ECS-P6: p6 station in the East China Sea; NPG-W26, NPG-W24: w26 and W24 stations in the North Pacific Ocean Gyre; NPO-pf1: pf1 station in the North Eastern Pacific Ocean.

Download figure to PowerPoint

pufM gene clone library construction and DNA sequencing

The primer set of pufM forward (5′-TACGGSAACCTGTWCTAC-3′) and reverse (5′-CCATSGTCCAGCGCCAGAA-3′) (Achenbach et al., 2001; Beja et al., 2002) were used for PCR amplifications of partial sequences of the pufM gene (193-bp fragments). Reaction mixtures (50 μL) contained the following components: 1.5 mM MgCl2, 0.2 mM each dNTP, 0.2 μM each primer, 20 ng template DNA, and 5 U LA Taq DNA polymerase (TaKaRa, Japan). Three independent amplifications were carried out in a T3 thermocycler (Biometra Co., Germany) with the following parameters: 94°C for 4 min, followed by 30 cycles of 94°C for 1 min, 52°C for 1 min, and 72°C for 1 min, with a final extension step at 72°C for 5 min. PCR products were pooled and gel-purified using a Gel Extraction Kit (TaKaRa) according to the manufacturer's instructions. Ligation into pMD18-T vector and transformation into Escherichia coli DH5α were performed according to the product manual (TaKaRa, Code no.: D504A). Colony PCR was conducted by tooth picking the ampicillin-resistant single colonies into 25 μL PCR reaction solution for screening target inserts with pMD18-T vector primers: M13-47 (5′-CGCCAGGGTTTTCCCAGTCACGAC-3′) and RV-M (5′-AGCGGATAACAATTTCACACAGG -3′). More than 30 recombinant plasmids in each clone library were extracted and sequenced on an ABI 377A automated sequencer (Applied Biosystems) using the sequencing primer M13-47.

Phylogenetic analysis

All the nucleotide sequences of the partial pufM gene were analyzed using a blast search (http://www.ncbi.nlm.nih.gov/BLAST/) and aligned together with the highest scoring blast hits of homologous bacterial pufM sequences. Then, the partial pufM gene sequences from our study (sequences with 100% similarity were precluded), together with the reference sequences retrieved from the database and were aligned using the program clustalx 1.8 (Higgins & Sharp, 1988, 1989), and the sequence distance between each was calculated. Phylogenetic trees were constructed using the neighbor-joining method and the phylogeny inference package (3.63 edition, Joseph Felsenstein, 2004, University of Washington), were visualized with treeview software (Stirling Technologies Inc.) and edited manually. The sequences were deposited in GenBank with accession numbers DQ093226–DQ093270.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Author contribution
  9. References

Proportion of γ-like Proteobacterial pufM clones

Clones that were clustered into one subclade with the pufM gene sequences from the known species affiliated to Gammaproteobacteria in the phylogenetic tree were defined as γ-like pufM gene sequences. Of the 171 pufM clones from the four sampling sites, 59 γ-like pufM gene sequences, or 34.5% of the total, were obtained. Obvious differences in the proportion of γ-like pufM gene clones were revealed at the different sampling locations. In the North Pacific Ocean gyre, the proportion of γ-like pufM clones at two stations (w24 and w26) averaged 56.1% (32 different sequences out of 57 clones) and the sequence identities varied from 76.2% to 99.5%, indicating a high abundance and high diversity in this sampling area. In the temperate East China Sea, 19 γ-like pufM clones represented by eight genotypes accounted for 59.4% of the total 32 clones observed, suggesting a high abundance but low genetic diversity. Seven γ-like pufM sequences were found from a total of 53 clones in 0 and 10 m water samples in the cold Bering Sea, whereas only one γ-like pufM gene sequence was found from the total of 29 pufM clones at station NPO-pf1 in the eastern tropical North Pacific Ocean. The concentrations of dissolved oxygen were between 238.16 and 248.43 μmol L−1, which showed that our sampling sites were oxygen rich. The nutrient conditions at our sampling sites were indicated by the Chl a concentrations (Table 1).

Table 1.   Sampling sites with information regarding γ-like pufM clones and environmental conditions
Site location (sampling time) Station (Latitude/Longitude)Total clonesγ-likeγ-like/T. (%)Group IGroup IISite temp (°C)Chl α (mg m−3)DO (μmol L−1)
  1. DO, dissolved oxygen; Chl α, chlorophyll α; –, no available data; T., total clones; temp., temperature.

North Pacific Gyre (05/12/2003)
 NPG-w26 (179.7E/23.7N)311548.49626.00.05245.26
 NPG-w24 (170.5E/21.3N)261765.413426.90.05238.16
North Pacific Ocean (25/11/2003)
 NPO-pf1 (103.6W/12.8N)2913.40128.50.21266.42
East China Sea (17/09/2003)
 ECS-P6-S (126.3E/29.0N)321959.411828.60.56278.43
Bering Sea (21/07/2003)
 BS-BR02-0 m (174.5E/56.9N)28414.30410.50.72
 BS-BR02-10m25312.01210.0

Phylogenetic analysis of γ-like Proteobacterial pufM gene

Reference sequences from different marine AAnPB isolates and related purple bacteria were retrieved from the database (GenBank+EMBL+DDBJ) to construct the phylogenetic tree. The phylogenetic tree of the pufM gene encompassed two principal clades, one containing α-3 and α-4 Proteobacteria and another containing α-1, α-2, β- and Gammaproteobacteria representatives (Fig. 2). Thus, all the γ-like Proteobacterial pufM gene sequences in this study were clustered into one subclade with the α-1 subclass, but they were grouped into two distinct groups (Group I and Group II) (Fig. 2). Group I was closest to Chromatium vinosum and Thiocaspa roseopersicina and Group II was distantly related to Thiocystis gelatinosa and resided between the Gammaproteobacteria clade and Bradyrhizobium sp. ORS278. (Fig. 2). Chromatium vinosum was the nearest relative of Group I clones retrieved from the database using blast, and this was in agreement with the phylogenic analysis. However, the nearest relatives of Group II clones were Bradyrhizobium sp. RS278, Thiocapsa roseopersicina, and Roseobacter sp. SYOP2, with lower blast scores (<97.6 bits) and sequence identities (<88%), suggesting that novel phenotypes or ecotypes of the pufM gene might exist in our sampling sites (Table 2). Most clones (91.8%) of Group I (Fig. 2a) were from the oligotrophic North Pacific Ocean Gyre, while the majority (71.4%) of the sequences from the eutrophic East China Sea and Bering Sea were clustered into Group II (Fig. 2b and Table 1).

image

Figure 2.  (a) Phylogenetic relationships of γ-like pufM gene sequences. (b) Group I. (c) Group II. Evolutionary distances were determined from an alignment of 193 nucleotide positions using the clustal w (1.8 revision) program. The trees were constructed with phylip (3.63 revision) software using the neighbor-joining algorithm. Bootstrap analysis of 100 replicates was performed. Values of >50% are shown on the nodes. The bar corresponds to base substitutions per 100 nucleotides. The pufM gene of the green nonsulphur bacterium Chloroflexus aurantiacus was used as an outgroup reference. The partial sequences of the pufM genes amplified by PCR in this study are indicated by symbols such as npg-w42xx (Table 2). Photosynthetic Alpha, Beta, and Gammaproteobacteria groups are indicated by the vertical bars to the right of the tree.

Download figure to PowerPoint

Table 2.   Relative analyses and blast results of γ-like pufM gene clones
ClonesNearest relative (accession number)ScoreIdentities
Group I
 npg-jlw24-26, npg-jlw24-61, npg-jlw24-25, npg-jlw26-12, npg-jlw26-27Chromatium vinosum (AB011811)161165/193 (85%)
 npg-jlw24-60, npg-jlw24-43Chromatium vinosum (AB011811)168166/193 (86%)
 npg-jlw26-60, npg-jlw24-67, npg-jlw24-10, bs-jlbr02-10m29, ecs-jlp6s10,Chromatium vinosum (AB011811)176167/193 (86%)
 npg-jlw26-20Chromatium vinosum (D50647)178167/193 (86%)
 npg-jlw24-17Chromatium vinosum (AB011811)184168/193 (87%)
 npg-jlw26-71, npg-jlw24-69, npg-jlw26-18Chromatium vinosum (AB011811)200170/193 (88%)
 npg-jlw26-2, npg-jlw24-71, npg-jlw24-20Chromatium vinosum (AB011811)153164/193 (84%)
 npg-jlw26-5, npg-jlw26-10Chromatium vinosum (AB011811)192169/193 (87%)
Group II
 npg-jlw26-37, npg-jlw24-68, ecs-jlp6s43, bs-jlbr02-10m7, bs-jlbr02-0m45Bradyrhizobium sp. RS278 (AF182374)95.6156/192 (81%)
 npg-jlw26-4Bradyrhizobium sp. RS278 (AF182374)81.877/89 (86%)
 npo-jlpf1-0m4, ecs-jlp6s3, npg-jlw26-33Bradyrhizobium sp. RS278 (AF182374)89.778/89 (87%)
 npg-jlw26-73Thiocapsa roseopersicina (AJ544223)87.771/80 (88%)
 bs-jlbr02-0m36Thiocapsa roseopersicina (AJ544223)79.870/80 (87%)
 ecs-jlp6s11, bs-jlbr02-0m29, ecs-jlp6s5, bs-jlbr02-0m5Thiocapsa roseopersicina (AJ544223)87.771/80 (88%)
 npg-jlw26-68, ecs-jlp6s42Bradyrhizobium sp. RS278 (AF182374)97.679/89 (88%)
 npg-jlw24-64, ecs-jlp6s24Bradyrhizobium sp. RS278 (AF182374)103157/192 (81%)
 npg-jlw24-49, bs-jlbr02-10m9Chromatium vinosum (AB011811)91.782/92 (89%)
 npg-jlw26-62Roseobacter sp. SYOP2 (AY675567)83.8153/190 (80%)
 ecs-jlp6s50Roseobacter sp. SYOP2 (AY675567)194167/190 (87%)

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Author contribution
  9. References

A large proportion of γ-like Proteobacterial pufM gene clones (with highest proportion up to 65.4%) have been found unexpectedly in some oxic sampling sites (Fig. 2). Sampling sites in the subtropical North Pacific Ocean gyre are in an ultra-oligotrophic region, whereas station pf1 in the tropical Eastern North Pacific Ocean is a mesotrophic area. The cold Bering Sea is a high-nutrient region (Fujishima et al., 2001) and the East China Sea is a eutrophic marginal sea due to the nutrient input from the Yangtze River. This is in coincidence with the Bchl a concentrations. In this study, the highest proportions of γ-like Proteobacteria pufM gene clones occurred at extreme oligotrophic (65.4%) and nutrient-rich (59.4%) sampling sites, and they were grouped into two distinct ecotypes. Group I could represent oligotrophic ecotypes, because the majority of pufM gene clones were obtained from the ultra-oligotrophic Pacific Ocean gyre. However, almost all the clones from the eutrophic East China Sea and Bering Sea were clustered into group II which could be called trophic ecotype. Although the sampling site temperatures in summer varied from 10°C in the Bering Sea to 28.6°C in the East China Sea, no correlation was found between the proportions of γ-like pufM gene clones and the corresponding site temperatures (Table 1). Thus, the genetic distributions of γ-like pufM gene clones may be controlled mainly by nutrient conditions rather than temperature.

The species of Gammaproteobacteria Chromatium vinosum, Thiocaspa roseopersicina and Thiocystis gelatinosa that are most similar to our pufM gene clones are typical anaerobic anoxygenic phototrophs, largely confined to marine shallow sediments or attached to decomposing organic matter (Guyoneaud et al., 1998; Imhoff et al., 1998). However, the nearest relative of most Group II clones, Bradyrhizobium sp. ORS278, is a strictly aerobic anoxygenic phototrophic bacterium belonging to the Alphaproteobacteria that has been found recently from Aeschynomene sensitiva stem nodules and is photosynthetically active during stem symbiosis (Giraud et al., 2000). This information indicates that there is no clear geographic correlation between the pufM gene containing bacteria and their nearest oxygenic photosynthetic relatives. As our samples are from oceanic oxic surface water (above 1 m depth) from a deep water column (80–4000 m at different sampling sites), it is unlikely that our environmental pufM sequences were blurred by clones coming from the anoxic sediments due to vertical mixing of the water. The above results suggest that the existence of aerobic anoxygenic photosynthesis carried out by γ-like pufM gene-containing bacteria is less likely to have originated from the same ancestor in the open ocean.

The phylogenetic tree of the pufM gene somewhat resembles previous studies but differs in some significant details (Nagashima et al., 1997; Beja et al., 2002; Yutin et al., 2005). In this study, two γ-like Proteobacteria clusters are mixed with α-1 subclass, which could be explained by the horizontal transfer of photosynthetic related genes between anoxygenic phototrophs (Nagashima et al., 1997; Beatty, 2002). Horizontal gene transfer of partial photosynthetic apparatus has been revealed by analysis of whole-genome data from five major photosynthetic prokaryotic phyla, which display a major driving force in their evolution (Raymond et al., 2002). Conjugative plasmid transfer has been detected between bacteria under simulated marine oligotrophic conditions (Goodman et al., 1993). Thus, it is speculated that the starvation is one of the important stimulatory factors for horizontal gene transfer, especially in the oligotrophic open ocean. However, in nutrient-rich areas, such as the P6 station in the East China Sea shelf water, a significant proportion of the γ-like Proteobacterial pufM gene was also found, and a small fraction appeared at the mesotrophic sampling site, station pf1, in the tropical Eastern North Pacific Ocean. It could be concluded that there are more suspended particles to which bacterial cells tend to adhere in nutrient-rich areas than in nutrient-poor areas (van Elsas & Bailey, 2002). These particles, called hot spots, support large densities of metabolically active microorganisms and accelerate the process of horizontal gene transfer (van Elsas & Bailey, 2002). Thus, it is possible that the different extreme nutrient levels stimulate the different mechanisms of horizontal gene transfer and decide the genetic and ecotype distribution of γ-like Proteobacterial pufM gene-containing bacteria. However, their achievement of aerobic phototrophic metabolism might have been selected over a long period under the oxidation processing of the ocean.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Author contribution
  9. References

We thank Ning Hong and Yao Zhang for their assistance in sampling. Special thanks to Shaolin Shang for her assistance with providing Chl a concentrations. Professor John Hodgkiss is thanked for his help with English. This work was supported by: NSFC40576063, NSFC40232021, G2000078500, MOST2003, DF000040, 50521003, 2001CB409700 and MOE key project.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Author contribution
  9. References
  • Achenbach LA, Carey J & Madigan MT (2001) Photosynthetic and phylogenetic primers for detection of anoxygenic phototrophs in natural environments. Appl Environ Microbiol 67: 29222926.
  • Allgaier M, Uphoff H, Felske A & Wagner-Dobler I (2003) Aerobic anoxygenic photosynthesis in Roseobacter clade bacteria from diverse marine habitats. Appl Environ Microbiol 69: 50515059.
  • Beatty JT (2002) On the natural selection and evolution of the aerobic phototrophic bacteria. Photosynth Res 73: 109114.
  • Beja O, Suzuki MT, Heidelberg JF et al. (2002) Unsuspected diversity among marine aerobic anoxygenic phototrophs. Nature 415: 630633.
  • Fuhrman JA, Comeau DE, Hagstrom A & Chan AM (1988) Extraction from natural planktonic microorganisms of DNA suitable for molecular biological studies. Appl Environ Microbiol 54: 14261429.
  • Fujishima F, Ueda K, Maruo M, Nakayama E, Tokutome C, Hasegawa H, Matsui M & Sohrin Y (2001) Distribution of trace bioelements in the subarctic North Pacific Ocean and the Bering Sea (the R/V Hakuho Maru Cruise KH-97-2). J Oceanogr 57: 261273.
  • Giraud E, Hannibal L, Fardoux L, Vermeglio A & Dreyfus B (2000) Effect of Bradyrhizobium photosynthesis on stem nodulation of Aeschynomene sensitiva. Proc Nat Acad Sci USA 97: 1479514800.
  • Goericke R (2002) Bacteriochlorophyll a in the ocean: is anoxygenic bacterial photosynthesis important? Limnol Oceanogr 47: 290295.
  • Goodman AE, Hild E, Marshall KC & Hermansson M (1993) Conjugative plasmid transfer between bacteria under simulated marine oligotrophic conditions. Appl Environ Microbiol 59: 10351040.
  • Guyoneaud R, Suling J, Petri R et al. (1998) Taxonomic rearrangements of the genera Thiocapsa and Amoebobacter on the basis of 16S rDNA sequence analyses, and description of Thiolamprovum gen. nov. Int J Syst Bacteriol 48: 957964.
  • Higgins DG & Sharp PM (1988) CLUSTAL: a package for performing multiple sequence alignments on a microcomputer. Gene 73: 237244.
  • Higgins DG & Sharp PM (1989) Fast and sensitive multiple sequence alignments on a microcomputer. Comput Appl Biosci 5: 151153.
  • Imhoff JF (1995) Taxonomy and physiology of phototrophic purple bacteria and green sulfur bacteria. Anoxygenic Photosynthetic Bacteria (BlankenshipRE, MadiganMT & BauerCE, eds), p. 1. Kluwer Publishers, Dordrecht.
  • Imhoff JF, Suling J & Petri R (1998) Phylogenetic relationships among the Chromatiaceae, their taxonomic reclassification and description of the new genera Allochromatium, Halochromatium, Isochromatium, Marichromatium, Thiococcus, Thiohalocapsa and Thermochromatium. Int J Syst Bacteriol 48: 11291143.
  • Kolber ZS, Van Dover CL, Niederman RA & Falkowski PG (2000) Bacterial photosynthesis in surface waters of the open ocean. Nature 407: 177179.
  • Kolber ZS, Plumley FG, Lang AS et al. (2001) Contribution of aerobic photoheterotrophic bacteria to the carbon cycle in the ocean. Science 292: 24922495.
  • Nagashima KVP, Hiraishi A, Shimada K & Matsuura K (1997) Horizontal transfer of genes coding for the photosynthetic reaction centers of purple bacteria. J Mol Evol 45: 131136.
  • Oz A, Sabehi G, Koblizek M, Massana R & Beja O (2005) Roseobacter-like bacteria in Red and Mediterranean Sea aerobic anoxygenic photosynthetic populations. Appl Environ Microbiol 71: 344353.
  • Rathgeber C, Beatty JT & Yurkov V (2004) Aerobic phototrophic bacteria: new evidence for the diversity, ecological importance and applied potential of this previously overlooked group. Photosynth Res 81: 113128.
  • Raymond J, Zhaxybayeva O, Gogarten JP, Gerdes SY & Blankenship RE (2002) Whole-genome analysis of photosynthetic prokaryotes. Science 298: 16161620.
  • Rye R & Holland HD (1998) Paleosols and the evolution of atmospheric oxygen: a critical review. Am J Sci 298: 621672.
  • Schwalbach MS & Fuhrman JA (2005) Wide-ranging abundances of aerobic anoxygenic phototrophic bacteria in the world ocean revealed by epifluorescence microscopy and quantitative PCR. Limnol Oceanogr 50: 620628.
  • Shiba T & Simidu U (1982) Erythrobacter longus gen. nov., sp. nov., an aerobic bacterium which contains bacteriochlorophyll a. Int J Syst Bacteriol 32: 211217.
  • Shiba T, Simidu U & Taga N (1979) Distribution of aerobic bacteria which contain bacteriochlorophyll a. Appl Environ Microbiol 38: 4345.
  • Van Elsas JD & Bailey MJ (2002) The ecology of transfer of mobile genetic elements. FEMS Microbiol Ecol 42: 187197.
  • Xiong J, Fischer WM, Inoue K, Nakahara M & Bauer CE (2000) Molecular evidence for the early evolution of photosynthesis. Science 289: 17241730.
  • Yurkov VV & Beatty JT (1998) Aerobic anoxygenic phototrophic bacteria. Microbiol Mol Biol Rev 62: 695724.
  • Yutin N, Suzuki MT & Beja O (2005) Novel primers reveal wider diversity among marine aerobic anoxygenic phototrophs. Appl Environ Microbiol 71: 89588962.