Rare Bacterium of New Genus Isolated with Prolonged Enrichment

1 downloads 0 Views 173KB Size Report
flora was observed from the 1st phase to the 4th phase in. 84 hr. ... with conventional culture techniques:1) culture media, ..... approach using the bootstrap.
Biosci. Biotechnol. Biochem., 68 (1), 28–35, 2004

Rare Bacterium of New Genus Isolated with Prolonged Enrichment Culture Akiko H ASHIZUME,1 Ryosuke F UDOU,2 Yasuko J OJIMA,2 Ryohsuke N AKAI,1 Akira H IRAISHI,3 Akira T ABUCHI,1 Kikuo S EN,1 and Hiroshiro SHIBAI1; y 1

Department of Bioscience and Biotechnology, Faculty of Agriculture, Shinshu University, 8304, Minamiminowa, Nagano 399-4511, Japan 2 Institute of Life Science, Ajinomoto Co., Inc., Suzukicho 1-1, Kawasaki-shi, Kanagawa 210-8681, Japan 3 Department of Ecological Engineering Toyohashi University of Technology, Tenpakucho Hibarigaoka 1-1, Toyohashi-shi, Aichi 441-8580, Japan Received February 6, 2003; Accepted September 22, 2003

Dynamic change in microbial flora was monitored with an oxygen electrode. The 1st phase microorganisms, which first grew well in LB medium, were followed by the 2nd phase microorganisms, which supposedly assimilated microbial cells of the 1st phase and their metabolites. In a similar way, a change in microbial flora was observed from the 1st phase to the 4th phase in 84 hr. Based on this observation, prolonged enrichment culture was done for as long as two months to increase the ratio of existence of rare microorganisms. From these culture liquids, four slow-growing bacteria (provisionally named Shinshu-ah1, -ah2, -ah3, and -ah4), which formed scarcely visible small colonies, were isolated. Sequence analysis of their 16S rDNA showed that Shinshu-ah1 had 97% homology with Bradyrhizobium japonicum and uncultured alpha proteobacterium clone blaii 16, Shinshu-ah2 91% with Rasbo bacterium, Alpha proteobacterium 34619, Bradyrhizobium genosp. P, Afipia felis and an unidentified bacterium, Shinshu-ah3 99% with Methylobacterium mesophilicum, and Shinshu-ah4 95% with Agromyces ramosus DSM 43045. Phylogenetic study indicated that Shinshu-ah2 had a possibility to form a new family, Shinshu-ah1 a new genus, and Shinshu-ah4 a new species.

microorganisms, on one hand, there exist, in the ordinary environment, a minority of microorganisms that cannot be isolated due to their extremely low ratio of existence. Such rare microorganisms are also to be exploited for their novel microbial functions. In the field of investigational search for new biologically active substances, where more than 10,000 compounds of microbial origin accumulate, we almost cannot expect to find a new compound with practical significance by using conventional methodologies of microbial isolation and assay system. We need a new collection of novel microorganisms. Under those circumstances mentioned above, we began studies of cultivation and isolation of microorganisms belonging to the majority group that nature can cultivate but researchers cannot, and the minority group that are also difficult to find. This paper deals with discovery of several novel microorganisms belonging to the minority group that could be isolated with enrichment culture for as long as two months. In addition, only such extremely small colonies that were detected not by the naked eye but under microscopic observation were selected.

Materials and Methods Key words:

enrichment culture; rare bacteria; small colony; new genus; alpha-2 proteobacterium

It has become clear that less than 1% of microorganisms in the ordinary environment can be cultivated with conventional culture techniques:1) culture media, culture conditions, and detection of microbial colonies with the naked eye. It stimulates us, from the standpoint of applied microbiology, to develop new methodologies that allow us to cultivate and isolate uncultivable microorganisms, leading to the exploitation of their new functions. While there are a vast majority of uncultivable y

Microbial source. Soil samples were collected on the campus of the Faculty of Agriculture, Shinshu University, located in Minamiminowa-mura, Nagano, Japan. Soil samples were used as microbial sources for the enrichment culture. Similarly, the culture liquids were also used as microbial sources for isolation of rare bacteria. Culture media. Culture media used were as follows. LB consisted of 10 g/l peptone, 5 g/l yeast extract, and 5 g/l NaCl. LB was diluted 10 times and 100 times to prepare 1/10LB and 1/100LB respectively. AG(CL)

To whom correspondence should be addressed. Tel: +81-265-77-1611; Fax: +81-265-77-1629; E-mail: [email protected]

Novel Rare Bacteria

contained only 500 g/l supernatant of enrichment culture liquids. NB consisted of 1.67 g/l peptone, 13.33 g/l nutrient broth, and 5 g/l NaCl. Schaffer contained 8 g/l nutrient broth, 1 g/l KCl, 0.25 g/l MgSO4 7H2 O, 103 M Ca(NO3 )2 4H2 O, 105 M MnCl2 4H2 O, and 106 M FeSO4 7H2 O. Albumin consisted of 1 g/l glucose, 0.5 g/l K2 HPO4 , 0.25 g/l egg albumin, 0.2 g/l MgSO4 7H2 O, and 0.01 g/l Fe2 (SO4 )3 . YS contained 10 g/l soluble starch and 2 g/l yeast extract. R consisted of 6.67 g/l peptone, 5.33 g/l nutrient broth, 5 g/l yeast extract, 5 g/l malt extract, 5 g/l casamino acids, 2 g/l glycerol, 1 g/l MgSO4 7H2 O and 0.05 g/l Tween 80. In order to prepare a solid medium, agar powder (NACALAI TESQUE, INC.) was added at the concentration of 1.5%. So medium AG contained only 15 g/l agar powder. The media were sterilized at 121 C for 20 min. Culture condition. For solid culture, microorganisms were incubated mainly at 25 C unless otherwise noted. Prolonged enrichment culture in a shaken flask was as follows: a shaken flask (500 ml) containing 100 ml 1/ 10LB medium was inoculated with 0.05 g soil samples and cultured at 25 C for two months with shaking. In order to perform bacterial incubation under extremely low oxygen tension, an ‘‘anaerobic pack (Mitsubishi Gas Chemical Co., Inc.)’’ was used. Microbial culture in a jar-fermentor was done using a 600-ml glass fermentor (Tokyo KJM) with a control apparatus (ABLE Co., AS-04-3). The culture vessel was equipped with a membrane-coated oxygen electrode. LB medium (400 ml) was inoculated with a 0.1-g soil sample and stirred gently (initially at 150 rpm) at 25 C. The dissolved oxygen level was maintained above ‘‘zero’’ to satisfy the cells’ oxygen demand. The agitation speed was controlled according to the change in the microbial respiration rate. Aeration was not done, with the oxygen supply depending on the transfer from the free surface of the culture liquids. Sudden inactivation of cell respiration due to deficiency in assimilable

29

nutrients was monitored by watching for a sudden increase in the dissolved oxygen level. Excess foaming was suppressed by soybean oil. 16S rDNA sequencing and phylogenetic analysis. 16S rDNA fragments from the crude lysate were amplified by PCR with a universal set of primers.2) PCR products were sequenced with a SequiTherm cycle sequencing kit (Epicentre Technologies); this was followed by detection with a Pharmacia laser fluorescent DNA sequencer.3) The sequences found were compared with those retrieved from the DDBJ/EMBL/GenBank nucleotide sequence databases. A distance matrix tree was constructed by the neighbor-joining method,4) and the topology of the phylogenetic tree was built by bootstrap analysis,5) using the CLUSTAL W program.6) The 16S rDNA sequences of strains Shinshu-ah1, -ah2, and -ah4 were deposited in the DDBJ under accession numbers of AB100410, AB100409, and AB100411, respectively.

Results Dynamic aspect of change in microbial flora as detected with an oxygen electrode Microbial activity can be represented by their respiration rate, the changes of which, then can be monitored by the oxygen electrode. A small glass culture vessel containing a sterilized LB medium was inoculated with a soil sample and stirred gently (150 rpm). As shown in Fig. 1, after 7 hr lag time, microbial respiration began to be accompanied with growth. At the same time, the dissolved oxygen level markedly decreased, reflecting an increase in cell’s oxygen demand. In order to maintain the dissolved oxygen level above ‘‘zero’’, agitation speed was kept at 1,500–1,850 rpm. This value showed an activity of microbial respiration, since the high agitation speed assured sufficient oxygen supply to cope with a high rate of oxygen consumption by the cells. However, at 20 hr,

Fig. 1. Dynamic Change in Microbial Flora as Detected by Oxygen Electrode. The culture was initiated as described in Materials and Methods. The dissolved oxygen level was controlled by agitation speed according to changes in the cells’ respiration rate. Termination of cell respiration and its resumption were monitored.

30

A. HASHIZUME et al.

the dissolved oxygen level suddenly returned to the initial value, implying that microbial respiration completely ceased due to nutritional deficiency. This phase (7–20 hr) was designated as the 1st phase, where microorganisms capable of assimilating nutrients in LB medium could proliferate. At the end of the 1st phase, it was considered that nutrients in LB medium were exhausted but organic matter consisting of microbial cells that proliferated during the 1st phase and metabolites excreted by them appeared in the medium as new nutrients. Therefore, different kinds of microorganisms were expected to begin growing using these new nutrients in the 2nd phase. Before the beginning of the 2nd phase, the agitation speed was restored to 150 rpm. As expected, shortly after, a second decrease in the dissolved oxygen began, reflecting the activation of the 2nd phase microorganisms. The dissolved oxygen level was maintained at approximately 25–30% so that its second rise could be easily detected. At 25 hr, the dissolved oxygen level again increased accompanied by the termination of the cell respiration. In a similar way, a change in microbial flora was observed up to the 4th phase in 84 hr. Culture liquids became transparent and microbial floc precipitated. Isolation of bacterial colonies from prolonged culture liquids Even though the majority of bacteria in the soil are uncultivable, as many as approximately 106 bacterial cells/gsoil formed colonies when cultivated on LBsolid medium. At first, among them, we searched for an extremely small colony, which can only be observed under the microscope (40). They were expected to be novel microorganisms, having escaped researcher’s observation. However, when transplanted again to LB medium, all of them formed visible colonies. Even though such a small-colony forming bacterium had existed at the ratio of 104 –103 , it would have been practically impossible to isolate them. The previous culture experiment in Fig. 1 suggested that the dominant microorganisms grown in the 1st phase were suitable for the cultivation in LB medium, but not in the culture system of the 2nd phase. They might even become the nutrient source for microorganisms dominantly active in the 2nd phase. Therefore, the ratio of microorganisms suitable for LB medium was expected to markedly decrease in the 2nd , 3rd and 4th phases. In order to decrease the ratio of microorganisms suitable for LB medium and increase the ratio of rare microorganisms, the culture in a shaken flask was prolonged for as long as two months. A shaken flask containing 1/10LB medium was inoculated with the soil sample, and cultured. Microorganisms living in thusprepared culture liquids were considered to have the following two growth properties: a low growth rate and requirement for nutrients in the culture liquids. Therefore, colony isolation procedures were done as follows

with particular emphasis on extremely small colony formation. Colonies were formed on AG(CL) solid medium using 0.1 ml of 105 diluted culture liquids. In 19 days, approximately 110 colonies/petri dish were observed. However, after that, small colonies continued to appear. Among them, extremely small colonies that were hardly visible with the naked eye but clearly observed under the microscope (40) were picked up. A total of 148 colonies were obtained in 8–39 days. All of the 148 colonies thus obtained were transplanted to both LB and AG(CL) media in an attempt to find a microorganism that required some nutrients in the culture liquids. All of them grew in both media to the same extent. However, growth of four colonies was extremely slow in both media: extremely small colonies scarcely visible were observed in 3 days. These microorganisms were provisionally designated as Shinshuah1, -ah2, -ah3, and -ah4 for further study. Some growth properties of four microorganisms Appearance of bacterial colonies, which were cultured on LB solid medium and observed at 10 days, was as follows. All of them were circular cones. Their colors were different: Shinshu-ah1 was white, Shinshu-ah2 transparent white (yellow after one month), Shinshu-ah3 pink, and Shinshu-ah4 white. A common characteristic was that they had extremely small colonies, whose diameter being less than 1 mm on LB medium. Sequence analysis of 16S rDNA of these bacteria, which will be shown shortly after, indicated that Shinshu-ah3 seemed identical to Methylobacterium mesophilicum. Therefore, a further detailed growth study was not done for this already known bacterium. It was considered that LB medium was not adequate for the cultivation of these bacteria (Shunshu-ah1, -ah2, and -ah4) due to insufficient or excess nutrients content. Therefore, several other media including NB, Shaeffer, Albumin, YS, R, 1/10LB and 1/100LB were examined to see whether sufficient colony formation was possible on these solid media. However, these media could not support sufficient colony formation of these bacteria, as shown in Fig. 2. Smaller colonies of these bacteria on every media examined were distinguished as compared to a large colony of E. coli. Optimum temperature for the growth of Shinshu-ah1, -ah2, and -ah4 was 25–30 C. Growth temperature of 37 C was inhibitory for the three bacteria, although the extent was different for each: the inhibition was complete for Shinshu-ah2, marked for -ah1, and slight for -ah4. Catalase activity of these bacteria was extremely weak as compared to E. coli or Corynebacterium lactofermentum (data not shown). So it was expected that they could form larger colonies by avoiding oxygen toxicity under extremely low oxygen tension. However, under such a condition, the growth of these three bacteria was markedly inhibited.

Novel Rare Bacteria

31

Fig. 2. Growth of Isolated Novel Bacterial Strains on Several Kinds of Solid Medium. Shinshu-ah1, -ah2, and -ah4 were cultivated at optimum temperature of 25 C, and E. coli at 37 C. The media used were as follows. Symbols: , LB; , 1/10LB; , 1/100LB; , AG; , NB; þ, Schaeffer; , Albumin; , YS; , R.

Fig. 3. Liquid Culture of Four Strains in LB Medium as Compared to That of Escherichia coli. Symbol: , E. coli; , Shinshu-ah1; , Shinshu-ah2; , Shinshu-ah4.

All of them could grow in liquid culture, although culture time was longer and maximum growth obtained being very small as compared to Escherichia coli as shown in Fig. 3. 16S rDNA Analisis of Four Isolated Bacteria Analysis of an approximately 700–1450 base pair sequence of 16S rDNA from the above mentioned four bacteria indicated that Shinshu-ah3 seemed identical to Methylobacterium mesophilicum, however, Shinshuah1, -ah2 and -ah4 were previously undescribed bacteria: Shinshu-ah1(AB100410) with 97% sequence similarity to uncultured alpha proteobacterium, Bradyrhizobium sp. Pp3a-10, Bradyrhizobium japonicum, Bradyrhizobium sp. Pplu1-1 and Photorhizobium thompsonianum, Shinshu-ah2(AB100409) 91% to Rasbo bacterium, alpha proteobacterium 34619, unidentified bacterium, Bradyrhizobium genosp. P and Afipia felis, Shinshu-ah4(AB100411) 95% to Agromyces ramosus DSM 43045 and Curtobacterium citreum DSM 20528, 94% to Rathayibacter rathayi DSM 7485, 93% to Subtercola frigaramans and 92% to Clavibacter michi-

ganensis as shown in Table 1. The nucleotide sequence of them has been submitted to DDBJ. Phylogenetic dendrograms were constructed based on the nucleotide sequences of 16S rDNA for the strain Shinshu-ah1 and -ah2 (Fig. 4) and -ah4 (Fig. 5). Both of the strains ah1 and ah2 belonged to the order Rhizobiales and shared a common line of descent with members of the family Bradyrhizobiaceae. However, the strain ah1 could not be attributed to any known genera and formed a distinct cluster along with an uncultured alpha proteobacterium (environmental clone sequence), suggesting that it could constitute a new genus. The strain ah2 also represented a novel and deep lineage within this family, therefore, it should be allocated to a new genus or possibly to a novel higher taxon. As shown in Fig. 5, the strain ah4 was assigned to the genus Agromyces in the family Microbacteriaceae, with A. mediolanus as the closest relative species (homology: 96.3%). Since the sequence similarity level was lower than the generally recognized borderline (97%) for defining a bacterial species,7) the strain ah4 seemed to form a new species.

Discussion In an attempt to isolate novel microorganisms, extremely small colonies scarcely visible to the naked eye were selected, since visible colonies would already have been observed and identified: many microbiologists have been involved in this field for a long time. The novel bacteria in this report, which were enriched with the prolonged culture and formed small colonies, are considered to have the following two properties: a slow growth rate and inclination for poor nutritional environments. In contrast, microorganisms that have already been found and identified are those which can be isolated by the conventional technique, and so grow well in the nutritionally rich media under conventional culture conditions, forming visible colonies easily observed with the naked eye.

32

A. HASHIZUME et al. Table 1. Homology Relationship to Other Known Bacteria

Bacteria of this study (accession no., bp)

Homology relationship to other bacteria (accession no.)

16S rDNA length used for homology study

Shinshu-ah1 (AB100410, 926 bp)

97% homologous to Uncultured alpha proteobacterium (AJ318107) Bradyrhizobium sp. Pp3a-10 (AF514704) Bradyrhizobium japonicum (X87272) Bradyrhizobium sp. Pplu1-1 (AF384139) Photorhizobium thompsonianum (L23405)

Shinshu-ah2 (AB100409, 1466 bp)

91% homologous to Rasbo bacterium (AF007948) Alpha proteobacterium 34619 (AF288301) Unidentified bacteria (D89027) Bradyrhizobium genosp. P (Z94805)

1266 1259 864 845

Shinshu-ah3 (687 bp)

99% homologous to Methylobacterium sp. strain F 18 (D32233) Methylobacterium mesophilicum (D32225) Methylobacterium sp. KA6 (AJ400934)

683 685 646

Shinshu-ah4 (AB100411, 1003 bp)

95% homologous to Agromyces ramosus DSM 43045 (X77447) Curtobacterium citreum DSM 20528 (X77436) 94% homologous to Rathayibacter rathayi DSM 7485 (X77439) 93% homologous to Subtercola frigoramans (AF224723) 92% homologous to Clavibacter michiganensis (AF474328)

The novel microorganisms described in this paper are surely important members in the microbial ecosystem, although they are very difficult to find. In nature, a wide variety of organic compounds of animal and plant origin are constantly supplied to the soil. Microorganisms that first grow well by assimilating these organic compounds will be easily isolated with the conventional isolation technique. After the growth of such microorganisms, different kinds of ones similar to the 2nd phase microorganisms in this paper will then grow, assimilating such microorganisms or their metabolites. Furthermore, the growth of microorganisms that were activated in the 3rd and 4th phases in this paper will follow. One such example was reported by Beppu et al.,8,9) showing bacterial symbiosis. In this case, a symbiont of Symbiobacterium requires a conditional medium prepared by other bacteria. But even in the presence of the conditional medium, it forms an extremely small colony. Harayama et al. clearly showed that microbial degradation of petroleum in the ocean is carried out by a smallcolony forming bacterium.10,11) The degradation in nature is not due to large-colony forming bacteria that were easily isolated in the laboratory as petroleumassimilating, but to a small-colony forming bacterium the DNA of which rapidly increases in the petroleumcontaminated ocean. The bacterium was afterward isolated as petroleum-assimilating with the prolonged enrichment culture in the laboratory. Tsuchida et al. isolated several kinds of bacteria from activated sludge. They require a supernatant of activated sludge for their

905 903 902 902 901

926 750 902 892 908

growth, implying their growth is dependent on some metabolites of other microorganisms.12) The bacteria form large visible colonies on the supernatant-supplied medium. Furthermore, Kaeberlein succeeded in pure culture of previously uncultivated marine microorganisms in the presence of other microorganisms.13) It is to be noted that among the four bacteria isolated with our enrichment culture, three were previously undescribed new microorganisms. By applying this method we can expect to find many more novel microorganisms. A group of bacteria including the rare bacteria of this study is considered to exist, having common characteristics such as a close evolutionary relationship to each other and the difficulty in cultivating and isolating them in the laboratory. In the case of Shinshu-ah2, for example, already known bacteria of Rasbo bacterium, Bradyrhizobium and Afipia felis have 16S rDNA sequence most homologous to it. Rasbo bacterium was first isolated from blood samples of a patient with sepsis and cultivated on a monolayer of Velo cells.14) Bradyrhizobium is well known for its extremely small colony formation, and so is difficult to isolate, and several species of Afipia felis were cultivated and isolated with an amoebal co-culture procedures.15) These kinds of microorganisms mentioned above, that are to form a new microbial collection, are sure to be further investigated not only for deeper understanding of the microbial ecosystem but also for further advances in applied microbiology.

Novel Rare Bacteria

33

Fig. 4. Phylogenetic Trees Showing the Relationships of the Isolates (Shinshu-ah1 and Shinshu-ah2) and Related Microorganisms in the Order Rhizobiales Based on 16S rRNA Sequences. The partial base sequences (corresponding to 75 to 1029 in E. coli. numbering system) were used to construct the tree. Bootstrap values at the node were calculated from 1,000 resamplings. Bar, 10% estimated sequence divergence.

Acknowledgments We would like to express our sincere thanks to M. Tokura for his precious advice and excellent technical assistance.

References 1)

Amann, R. I., Ludwig, W., and Schleifer, K., Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiol. Rev., 59, 143–169 (1995).

2)

3)

4)

5)

Lane, D. J., 16S/23S sequencing. In ‘‘Nucleic Acid Techniques in Bacterial Systematics’’, eds. Stackebrandt, E., and Goodfellow, M., Wiley, pp. 115–175 (1991). Iizuka, T., Yamanaka, S., Nishiyama, T., and Hiraishi, A., Isolation and phylogenetic analysis of aerobic copiotrophic ultra-microbacteria from urban soil. J. Gen. Appl. Microbiol., 44, 75–84 (1998). Saitou, H., and Nei, M., The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol., 4, 406–425 (1987). Felsenstein, J., Confidence limits on phylogenesis: an approach using the bootstrap. Evolution., 39, 783–791 (1985).

34

A. HASHIZUME et al.

Fig. 5. Phylogenetic Trees Showing the Relationships of the Strain Shinshu-4 and Representative Bacteria in the Family Microbacteriaceae Based on 16S rRNA Sequences. The partial base sequences (corresponding to 46 to 1029 in E. coli. numbering system) were used to construct the tree. Bootstrap values at the node were calculated from 1,000 resamplings. Bar, 1% estimated sequence divergence.

6)

7)

8)

Thompson, J. D., Higgins, D. G., and Gibson, T. J., CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res., 22, 4673–4680 (1994). Stackebrandt, E., and Goebel, B. M., Taxonomic note: a place for DNA-DNA reassociation and 16S rRNA sequence analysis in the present species definition of bacteriology. Int. J. Syst. Bacteriol., 44, 846–849 (1994). Ohno, M., Okano, I., Watsuji, T., Kakinuma, T., Ueda,

9)

K., and Beppu, T., Establishing the independent culture of a strictly symbiotic bacterium Symbiobacterium thermophilum from its supporting Bacillus strain. Biosci. Biotechnol. Biochem., 63, 1083–1090 (1999). Ueda, K., Ohno, M., Yamamoto, K., Nara, H., Mori, Y., Shimada, M., Hayashi, M., Oida, H., Terashima, Y., Nagata, M., and Beppu, T., Distribution and diversity of symbiotic thermophiles, Symbiobacterium thermophilum and related bacteria, in natural environments. Appl. Environ. Microbiol., 67, 3779–3784 (2001).

Novel Rare Bacteria

10)

11)

12)

13)

Sugiura, K., Ishihara, M., Shimauchi, T., and Harayama, S., Physicochemical properties and biodegradation of crude oil, Environ. Sci. Technol., 31, 45–51 (1997). Harayama, S., Kishira, H., Kasai, Y., and Shutsubo, K., Petroleum biodegradation in marine environments. J. Mol. Microbiol. Biotechnol., 1, 63–70 (1999). Tsuchida, T., Manome, A., Kanagawa, T., Hosoya, H., Kamagata, T., Tani, Y., and Kurane, R., Novel isolation method for unculturable bacteria from bioconsortia. Program and Abstracts of Fifth International Symposium on Environmental Biotechnology. O5-5 (2000). Kaeberlein, T., Lewis, K., and Epstein, S. S., Isolating ‘‘uncultivable’’ microorganisms in pure culture in a

14)

15)

35

simulated natural environment. Science, 296, 1127–1129 (2002). Blomqvist, G., Wesslen, L., Pahlson, C., Hjelm, E., Pettersson, B., Nikkila, T., Allard, U., Svensson, O., Uhlen, M., Morein, B., and Friman, G., Phylogenetic placement and characterization of a new alpha-2 proteobacterium isolated from a patient with sepsis. J. Clin. Microbiol., 35, 1988–1995 (1997). Scola, B., Barrassi, L., and Raoult, D., Isolation of new Fastidious alpha proteobacteria and Afipia felis from hospital water supplies by direct plating and amoebal coculture procedures. FEMS Microbiol. Ecol., 34, 129–137 (2000).