Archives of

kPnnnluoy

Arch Microbiol (1987) 148 : 305- 313

9 Springer-Verlag1987

Isolation and characterization of a carbon monoxide utilizing strain of the acetogen Peptostreptococcusproductus G. Geerligs, H. C. Aldrich*, W. Harder**, and G. Diekert Fachbereich Biologie/Mikrobiologie, Philipps-Universit/it, Karl von FrischStrage, D-3550 Marburg, Federal Republic of Germany Abstract. From sludge obtained from the sewage digester plant in Marburg-Cappel a strictly anaerobic bacterium was enriched and isolated with carbon monoxide as the sole energy source. Based on morphological and physiological characteristics the isolate was identified as a strain of Peptostreptococcus productus, which was called strain Marburg. The organism was able to grow on CO (50% at 200 kPa) as the sole energy source at a doubling time of 3 h and converted this substrate to acetate and CO2. The type strain of P. productus was not able to grow at the expense of CO. Electron microscopic investigations of strain Marburg cells revealed a cell wall which was different from that of other Gram-positive prokaryotes. D N A : D N A hybridization studies of the D N A isolated from strain Marburg and the type strain as well as some morphological and physiological properties of both strains confirmed the low degree or relatedness between the two strains. Key words: Acetogenic bacteria - Peptostreptococcus productus - Carbon monoxide utilization - Cell wall structure

Furthermore, its relatedness to strain U-1 and the type strain of P. productus was investigated. Acetogenic bacteria, which grow on CO as the sole energy source, are appropriate organisms for studies on the role of CO as the precursor for the synthesis of the carboxyl group of acetate. In order to facilitate further studies in this area some physiological properties and growth characteristics of strain Marburg are also reported here.

Material and methods Source of organisms Peptostreptococcus produetus type strain (ATCC 27340, DSM 2950) and strain U-1 (ATCC 35244, DSM 3507) were obtained from the Deutsche Sammlung yon Mikroorganismen (DSM) in G6ttingen, F R G ; Escherichia coli B strain Hayes (NCTC 10537, H I M 730-1) was kindly supplied by the Hygiene-Institut Marburg. Cultivation of organisms

Acetogenic bacteria are strictly anaerobic eubacteria, which catalyze the reduction of 2 CO2 to acetate as part of their energy metabolism (Wood 1985; Wood et al. 1986). Carbon monoxide in a bound form (rather than formate) has been proposed as an intermediate in the biosynthesis of the carboxyl group of acetate from CO2 (Diekert et al. 1984, 1985, 1986; Fuchs 1986). Therefore it is not surprising that some acetogens can grow on CO as the sole energy source (Sharak Genthner and Bryant 1982; Lynd et al. 1982; Zeikus et al. 1985). However, growth of these acetogens on CO is generally poor. Lorowitz and Bryant (1984) recently isolated an acetogen using CO as the sole energy source, which displayed fast growth rates on CO (up to 90% in the gas phase). This isolate, strain U-I, was identified mainly by its substrate spectrum as a strain of Peptostreptococcus productus. The strain was not characterized with respect to ultrastructural properties and its relatedness to other strains of P. productus. In this communication the isolation of a CO-utilizing strain of P. productus called strain Marburg is described. This organism exhibited unique ultrastructural properties.

Permanent addresses: * University of Florida, 1059 McCarty Hall, Gainesville, FL 32611, USA ** Laboratorium voor Microbiologie, Kerklaan 30, NL-9751 NN Haren, The Netherlands Offprint requests to : G. Diekert

The basal medium for the cultivation of P. productus contained 1 1 of stock solution (see below), 1 ml of vitamin solution (Wolin et al. 1963), 20 ml trace element solution (Wolin et al. 1963), 0.05% (w/v) cysteine, 0.025% (w/v) sodium sulfide, 0.2% (w/v) yeast extract and 0.2% (w/v) sodium bicarbonate. The stock solution was based on the medium described by Balch et al. (1977) and contained per 1: NH4C1, 1.0 g; KH2PO4, 0.16 g; KzHPO4 9 3 H20, 0.42 g; NazHPOr 9 2H20, 2,2 g; NaH2PO4 92 H 2 0 , 1.2 g; MgSO4 - 7H20, 0.1 g; resazurin, 2.0 rag. The pH of this stock solution was adjusted to 7.0. This solution was transferred into an anearobic chamber, where the cysteine was added. The medium was filled either into infusion bottles (125 - 2000 ml volume) in portions of 20% of the total bottle volume, or 5 ml volumes were filled into serum tubes. The bottles were closed with grey butyl rubber stoppers and screw caps, the tubes with black butyl rubber stoppers and aluminum seals, Outside the anaerobic chamber the gas phase was replaced by N2 and the bottles and tubes were autoclaved. Using a sterile syringe sodium sulfide, yeast extract, and sodium bicarbonate were added as 50-fold concentrated solutions. For growth on CO the gas phase was changed to N2:C02: CO (33% : 17% : 50% at 200 kPa). Other CO concentrations were used, where indicated; however, with the exception of > 8 3 % CO, 17% COg was in all cases added to the gas phase. For growth on He/COg the bacteria were incubated with 80% H2/20% CO2 at 200kPa in the

306 headspace of thecultures. The final pH, of the medium was 6.7, if not otherwise mentioned. For growth on acids, sugars, and polyalcohols the compounds were added as 20-fold concentrated solutions; the final concentration was 50 mM [with the exception of 0.2% (w/v) casamino acids and 5 mM methanol, where indicated]; the gas phase was 80% N2/20% CO2 (200kPa). The bacteria were added as a 10% inoculum using a sterile syringe. For the first enrichment cultures sewage sludge from the digester plant in Marburg-Cappel was used. Incubation was performed at 37~ in the dark at 200 rpm on a gyratory shaker. Media for roll-tubes contained 2% (w/v) agar and 0.01% (w/v) bromoeresol green. The tubes (total volume 47 ml) contained 7 ml agar medium. They were inoculated with 0.7 ml of an enrichment culture diluted anaerobically to a cell concentration of 2 x 101 to 2 x 104 cells/ml. Acid forming colonies were detected by their yellow colour (Braun et al. 1979). They were picked using a sterile capillary tube and were then transferred to 5 ml liquid medium. For the determination of the pH optimum for growth, the phosphate buffer of the basal medium was replaced by the following buffers (25mM each): pH 6.0, morpholinoethanesulfonate (MES); pH 6.5, piperazineN,N'-bis(2-ethanesulfonate) (PIPES); pH 7.0, morpholinopropanesulfonate (MOPS); pH 7.5 and 8.0, N-(2-hydroxyethyl)-piperazine-N'-2-ethanesulfonate (HEPES). In addition, 20 mM NaC1 and 3 mM KC1 were added with the exception of PIPES buffered medium, where NaC1 was omitted, since PIPES was added as the sodium salt. Growth was determined by measuring the absorbance of the diluted culture at 578 nm in a Zeiss PL4 photometer (Zeiss, Oberkochen, FRG) using diluted medium as a blank.

Ultrastructural investigations Cells of P. productus were withdrawn from culture bottles and immediately fixed in either 1% OsO4 in pH 7.2 sodium cacodylate buffer (0.l M) for 30 rain at room temperature, or in 2.5% glutaraldehyde in the same buffer for 15 rain, followed by OsO4 treatment as before. Cells were dehydrated in a graded ethanol series, embedded in Spurr's low viscosity resin, polymerized for 1 2 - 2 4 h at 60 ~C, sectioned with a diamond knife using a LKB Ultrotome III, poststained with uranyl acetate and lead citrate, and viewed and photographed on a Philips EM-301 electron microscope. For freeze-fracturing, unfixed cells were pelleted and quick-frozen using a propane jet freezer (Mfiller et al. 1981). Replicas were prepared using a Balzers BA-360M freezefracture apparatus equipped with electron guns and a quartz crystal film thickness monitor. Deep etching was carried out for 2 min at -100~ Some thin sections were stained with silver after oxidation with periodic acid to test for the presence ofpolysaccharide in the extracellular fibrils (Erdos 1986).

Analytical and preparative methods Carbon monoxide and Hz were determined gas chromatographically using a thermal conductivity detector. Acetate was determined enzymatically as described previously (Dorn et al. 1978). Butyrate was determined gas chromatographically by flame ionization detection.

Extraction and determination of quinones was performed according to the procedure of Kr6ger (1978) using cell extracts corresponding to 17.5 mg cell protein of P. productus grown on CO. The quinone extracts were dissolved in I ml ethanol and the differential spectrum of oxidized versus reduced quinones was recorded between 230 and 310 nm. The quinones were reduced by the addition of potassium borohydride. As a control quinones were extracted by the same procedure from cells of E. coli B; 5 mg of cell protein was used. Purified ubiquinone and phylloquinone (vitamin K1), 20 nmol each, were used as references. The amount of quinones was determined by the absorption differences at 280nm minus 289nm for ubiquinone and at 248 nm minus 253 nm for menaquinone in a mixture of both compounds (Kr6ger 1978). Thin layer chromatography was also performed according to Holl/inder et al. (I 977). For the extraction and determination of cytochrome b, a portion of CO-grown cells of P. productus was used corresponding to 16.5 mg cell protein. After disrupting the cells by sonification, membrane fractions of the cells were obtained by differential centrifugation. Heme IX was extracted with acetone and HC1 from these fractions (Kr6ger and Innerhofer 1976). The extracts were dried by flash evaporation and dissolved in pyridine/H20 (1:1, v:v); the pH was adjusted to 7. After eentrifugation, a differential absorption spectrum of the reduced (by Na28204) versus oxidized heme was recorded between 500 and 600 nm. Again a heine extract from E. coli B (6.5 mg cell protein) was used as a control. The heme concentration was calculated from the absorption difference at 557 nm and 540 nm. Extraction of corrinoids was performed using the method of Scherer and Sahm (1981). To 2 g of CO-grown cells (wet weight) of P. productus strain Marburg 8 ml methanol and 0.01% (w/v) KCN were added. The pH was adjusted to 6 and the suspension incubated for 10 min in a boiling water bath. After centrifugation and re-extraction of the pellet, the combined supernatants were flash evaporated at 37~C. The dried material was dissolved in HzO and applied to a QAE-Sephadex A-25 column pre-equilibrated with 50 mM tris HC1 pH 8. With the same buffer the corrinoids were eluted and applied onto a XAD-2 column (Vogelmann and Wagner 1973). The column was washed with H20 and then with 20% (v/v) methanol. The cyanocobalamins were eluted with 40% (v/v) methanol and, after flash evaporation, dissolved in H20. The corrinoid content of the solution was determined spectrophotometrically from the absorption at 361 nm (~361 = 28,060 M-1 cm-1) (Friedrich 1975). Protein was determined according to Bradford (1976) using the Bio-Rad microassay (Bio-Rad Laboratories, Mtinchen, FRG).

DNA preparation DNA was prepared by the method of Marmur (1961). The cells were harvested by centrifugation, washed in saline EDTA (0.15 M NaC1 plus 0.1 M ethylenediamine tetraacetate, pH 8) and disruptured by lysozyme, proteinase K, and sodium lauryl sulfate. DNA was isolated from the cell lysate (Marmur 1961). The purity of the preparation was analyzed by UV spectrometry and by agarose gel electrophoresis according to Maniatis et al. (1982). The purified DNA of P. productus contained DNA pieces mainly of 3 0 - 5 0 kilobase size; RNA was not detectable.

307

Determination of DNA base composition DNA base composition was calculated from the thermal denaturation profile (Marmur and Doty 1962) in standard saline/citrate buffer (0.15 M NaC1 plus 0.015 M trisodium citrate, pH 7) (De Ley et al. 1970) using an automatic recording spectrophotometer (model 250) with thermoprogrammer (model 2527) and Tm equipment from Gilford Instruments GmbH (Diisseldorf). The DNA of E. coli B with 52 tool% G + C and a T~ of 90.7~ (Gillis et al. 1970) served as reference.

D NA : DNA hybridization studies The double stranded DNA was dissolved in 1:10 diluted saline/citrate buffer (see above) and then sheared to fragments of about 1.1 x 1 0 6 daltons in a French pressure cell at 1.37 x 108 Pa (20,000 psi) and a flow rate of 1 to 2 ml/ rain. Single stranded DNA was obtained by heating the DNA for 20 rain in a boiling water bath. DNA reassociation was followed photometrically by recording the initial rate of reassociation (De Ley et al. 1970). The optimal renaturation temperature TOR was 70.3~ The degree of binding was evaluated graphically according to De Ley et al. (1970).

Results

1. Isolation of Peptostreptococcus productus strain Marburg From anaerobic sewage digester sludge obtained from the municipal waste treatment plant in Marburg-Cappel (FRG) an anaerobic, acetogenic bacterium was enriched using carbon monoxide (50% in the gas phase at 200 kPa) as the sole energy source. After only three transfers with 10% inoculum, more than 70% of the bacteria were identified as streptococci by phase contrast microscopy. After approximately 10 transfers, when the streptococci were the predominant organisms in the enrichment culture, samples from the culture were diluted and used to inoculate rolltubes containing bromocresol green indicator agar made with mineral medium supplemented with 0.2% yeast extract. The gas phase was 50% CO, 17% CO2, and 33% Nz at 200 kPa. Roll-tubes with 20% CO2 and 80% N2 were used as controls. Two to three days after inoculation acid-forming colonies were detected and used as inoculum for liquid medium. The fastest growing cultures were then again transferred to roll-tubes, and the procedure was repeated. Using this method liquid cultures were finally obtained which grew with a doubling time of approximately 3 h on carbon monoxide; an adaptation to fast growth on CO was not observed (see Lorowitz and Bryant 1984).

2. Morphological characterization Phase contrast microscopy of the isolated CO-utilizing acetogens revealed elongated cocci, which occurred in pairs or short chains in the early logarithmic growth phase and as longer chains in the later phase of growth. Based on these morphological properties and on the fact that the organism utilized CO and formed acetate from this substrate, a close relatedness with the strain U-1 of Peptostreptococcus productus (Lorowitz and Bryant 1984) was assumed. Therefore the isolate was tentatively called P. produetus (strain Marburg), and the strain U-1 as well as the type strain of

P. productus were used as reference strains for the further characterization and identification of the new isolate. Electron microscopic investigations were also performed. Log phase cells of P. productus strain Marburg are ovoid, about 0.7 gm wide by 0.9-1.1 gm long. They frequently contained polyphosphate granules of various sizes (Figs. 1 A, B, 2 B). In stationary phase cells, polyglucose globules sometimes fill the cytoplasm (Fig. 1 C). The cytoplasm of late log to stationary phase ceils often includes one or two plate-like inclusions (Fig. 1 D, E). These plates insert perpendicularly upon the plasma membrane (Fig. 1 F, J). The plasma membrane shows a typical trilaminar structure approximately 10nm thick in cross section (Figs. 1A, B, 2A). The cell wall, however, exhibits a structure which we believe to be unique among prokaryotes. Exterior to the plasma membrane are two discontinuous electron dense layers comprised of peaked globular subunits about 14 nm in diameter lying adjacent to one another. These two layers are separated by a less dense, continuous layer 1 0 - 1 1 nm thick. The subunits of the outer dense layer give rise to tufts of fibrils (Fig. 2A) 4 to 5 nm in diameter which together make a fuzzy coat over the entire cell surface extending up to I00 nm from the wall (Fig. 2 A). These fibrils stain positively with the silver/methenamine stain (Erdos 1986), indicating that they are composed of polysaccharide (data not shown). They occurred in both CO utilizing strains (Marburg and U-l), but were completely absent in the type strain (data not shown). This finding was confirmed by the observation that upon centrifugation of the bacteria the COutilizers formed a slimy pellet, which in part spontaneously resuspended in the supernatant. The latter phenomenon was not observed with the type strain. Grazing sections of the two electron dense layers of the strain Marburg cell wall revealed a pattern of hexagonal units with a round central granule within each unit (Fig. 2 B, C). Freeze fractured, deep-etched preparations show one fracture plane, the PF (protoplasmic fracture) face (Chapman and Staehelin 1986), which is sometimes smooth and sometimes covered with particles (Fig. 2D). This is complementary to an EF (exoplasmic fracture) face with smooth areas corresponding to smooth PF areas, and areas with few particles and some depressions which match the rough areas on the PF face (Fig. 2 D). The deep-etched outer surface is smooth (Fig. 2E). The internal lipid plates are also particle-free (Fig. 1 D). We believe that fracture did not occur through or between the two outer discontinuous, electron-dense layers. 3. DNA characteristics The DNA was purified from the two CO-utilizing strains and from the type strain. The G + C-content was determined by thermal denaturation of the DNA. The melting temperature at which 50% of the DNA was denatured was 88.1 ___0.1~ for all three strains under the experimental conditions used. This value corresponds to a G + C-content of 45.6 + 0.2 tool%. For the type strain of P. productus a G + C-content of 45 tool% has been reported (Ezaki et al. 1983). The findings are in agreement with the assumption that the isolate is a strain of P. productus. To determine the degree of relatedness of the three strains of P. productus, the DNA: DNA homology was investigated by measuring the renaturation rate of the DNA isolated

308

Figs. 1 A - F

309 from the different strains (De Ley et al. 1970). The degree of binding between the D N A of strain Marburg and the type strain was determined to be 58% _+ 2.5% indicating a low degree of relatedness between the two strains. However, the D N A : D N A homology between strain Marburg and strain U-1 was found to be 89% _+ 5%, which is in accordance with the assumption that both CO-utilizing strains exhibit a high degree of relatedness. The latter findings are also supported by the substrate utilization patterns and the ultrastructural morphology of the three strains.

4. Growth characteristics P, productus strain Marburg was routinely grown on a mineral salt medium containing 0.2% yeast extract and a gas atmosphere with 50% CO (200 kPa) as the energy source. The bacteria grew with a doubling time of near 3 h and converted CO mainly to CO2 and acetate according to the following equation (data not shown): 4 CO + 2 H 2 0 ~ 1 acetate- + 1 H + + 2 CO2 AG ~ = - 165.6 kJ/mol acetate. Trace amounts of hydrogen gas were also formed (up to 0.5% in the gas phase); butyrate was not detected. The ability to grow on CO as the sole energy source was not lost after numerous transfers of the bacteria on glucose medium. The growth rates and the lag phase were the same when the bacteria were transferred again to a medium with 50% CO as the gas phase. Pyruvate was also utilized. The cells were able to grow on H2 plus CO2 as the energy source, but growth on the latter substrates was poor compared to growth on CO. Methanol (5 mM) did not support growth as sole energy substrate, but it neither inhibited nor stimulated growth on CO when used at this concentration, The substrate spectrum of the Marburg strain was identical to that of strain U-l, whereas the type strain, under the conditions used (i.e. 50% CO, 200 kPa), was not able to utilize CO as a growth substrate, even when it was transferred repeatedly for more than 20 times in the same medium with 50% CO in the headspace (Table 1). Under these experimental conditions used, P. productus strain Marburg required yeast extract for growth on CO. Maximal growth rates were obtained at a yeast extract concentration of 0.2% (w/v) in the medium. Whether the yeast extract was required as a source for anabolistic substrates or to "detoxify" the medium from toxic trace elements is not known. No attempts were made to omit one or several trace metals from the medium. Casamino acids could not replace yeast extract as a growth stimulator. The growth rate of P. productus strain Marburg was dependent on the CO concentration in the gas phase. Optimal growth rates were observed with 5 0 - 7 0 % CO (200 kPa); at higher concentrations the growth rate decreased again (Fig. 3 A). The cell yield obtained per mol of

CO consumed reached a value of ~ 5 g cells (dry weight) only at CO concentrations of 20% or lower. This value decreased to 3.2 (50% CO) and to 1.6 (95% CO). With increasing CO concentrations the lag phase drastically increased up to nearly 60 h (95% CO) (Fig. 3 B). An explanation for this phenomenon could be that fast growth on CO is dependent on higher concentrations of CO2 (see also Lorowitz and Bryant 1984). The small increase in cell density observed in the absence of CO was obviously due to growth on yeast extract as the energy source. The bacteria grew well on CO at 30~ to 47~ optimal growth was obtained at approximately 37~ (Fig. 4A). The pH optimum of growth was pH 6 . 5 - 6 . 8 ; good growth was observed from pH 6.3 to 7.3 (Fig. 4B).

5. Other characteristics Acetogenic bacteria are known to contain high levels of corrinoids (Tanner et al. 1978), since the formation of the methyl group of acetate from COz involves at least one corrinoid enzyme (for a recent review see Fuchs 1986). Therefore the corrinoid content of P. productus strain Marburg cells was determined by extraction of the corrinoids as the dicyano-complex. After purification of the corrinoids, the concentration was determined spectrophotometrically and was 0.81 p,mol per g of cell protein. This value is in the same order as that of the corrinoid content of Acetobacterium woodii and Clostridium thermoaceticum (Tanner et al. 1978). It can be assumed that corrinoids play a similar role in the energy metabolism of P. productus to that in other acetogenic bacteria. In several acetogenic bacteria quinones and cytochromes of the b type have been detected (Gottwald et al. 1975). Therefore we tried to extract quinones and cytochrome b from CO-grown cells of P. produetus strain Marburg. Escherichia coli B ceils were used as a positive control. Neither "classical" quinones (ubiquinone and menaquinone) nor cytochrome b were extracted from the cells under the experimental conditions used. The detection limit under these conditions was 0.02 nmol/mg of cell protein for the quinones and 0.01 nmol/mg protein for cytochrome b.

Discussion In this communication the isolation and characterization of a CO-utilizing strain of Peptostreptoeoccus productus tentatively called strain Marburg are described. A COutilizing strain (U-1) of P. productus has been isolated before by Lorowitz and Bryant (1984); this strain, however, has not been characterized extensively. Therefore both strains are considered in this communication and compared to the type strain of P. productus. The type strain did not grow on CO under the cultivation conditions employed, indicating that the strains differ considerably with respect to their

9Fig. 1 A - F . Transmission electron micrographs of Peptostreptococcusproductus strain Marburg. A Longitudinal section of a typical logphase cell showing nucleoid region (n), polyribosome clusters (r), and small polyphosphate granules (arrows). x 106,500; B cross section of typical log-phase cell showing larger polyphosphate granule (arrow). x 106,500; C stationary phase cell containing numerous globular polyglucose globules. Note typical granularity of the lead stain in the globules, x 106,500; D freeze-fractured stationary phase cell showing plate-like lipid inclusion (l). x 81,000; E thin section of stationary phase cell showing plate-like lipid inclusion (arrow). x 94,600; F freeze fracture of stationary phase cells, with plate-like lipids connecting to the plasma membrane (arrow),smooth and rough pf fracture faces~ and deep-etched smooth outer wall surface (s). x 76,500

310

Fig. 2 A - E . Transmission electron micrographs illustrating cell envelope organization in P. produetus. A High magnification view of the cell envelope illustrating plasma membrane (arrowheads), double rows of electron dense particles (paired arrows), and pointed outer row particle with polysaccharide fibrils attached (single arrow, right), x 155,000 B two cells. Cell at right contains large polyphosphate granule (p). Cell at left is seen in grazing section; two hexagonal cells of outer wall layer with central granules are seen at arrows, x 155,000; C grazing section of cell surface. Arrows indicate some of the numerous hexagonal cells, with dense central granules, in both the inner and the outer layers, x 103,200; D freeze fracture of two adjacent cells, illustrating connection of internal lipid plate to plasma membrane (arrow), convex p f face with rough and smooth areas, and concave efface with smooth and rough areas. Intramembrane particles on pf face show no obvious ordered arrangement, x 63,000; E deep-etched cell, showing particle-free area of plasma membrane (pf) and smooth outer wall surface (arrow). x 52,500

311 Table 1. Substrate spectrum of the isolate (P. productus strain Marburg) and strain U-1 of P. produetus. For growth conditions see materials and methods section Substrate

0.2

A

Strains Marburg

U-1

Glucose Fructose Lactose Galactose Maltose Xylose Saccharose Ribose Cellobiose Mannose Trehatose Arabinose Glycerol Sorbitol Inositol Mannitol Adonitol" Succinate Pyruvate Formate Citrate Tartrate Lactate Malate CO

+ + + + + + + + + + + + + + + + + + ---+

+ + + + + + + + + + + + + + + + + -+b ---+

Ha/CO2

+

+ c

Methanol Casamino acids

_ -

_ c _ c

"7" Az v

a The type strain of P. productus was reported not to grow on adonitol (Rogosa 1971) b Lorowitz and Bryant (1984) reported that strain U-1 was unable to grow on pyruvate r Data taken from Lorowitz and Bryant (1984)

physiology. The substrate spectrum of the three strains as well as the ultrastructural morphology and the D N A studies revealed a high degree of relatedness between the two COutilizing strains U - I and M a r b u r g and a low degree of relatedness between the CO utilizers and the type strain. It is, however, not recommended here to define a new species, since the 58% D N A : D N A homology between strain M a r b u r g and the type strain still justifies the classification of the isolate as P. productus. During the isolation of P. produetus more than 10 colonies were picked and transferred to liquid m e d i u m with 50% CO in the headspace. F r o m the liquid cultures only one grew at a doubling time of approximately 4 h from the beginning; the other cultures grew much slower. Lorowitz and Bryant (1984) reported that their enrichment culture required initially 5 days to deplete the CO. This was n o t the case with our enrichment cultures, which were transferred every day and consumed the CO within that period of time. Phase contrast microscopy of the liquid enrichment cultures revealed in most cases the presence of streptococci, indicating that these bacteria most p r o b a b l y were COutilizing strains of P. produetus. Only one culture was found to contain rods, but this culture grew extremely slowly on CO.

Oo

"~

o.1

o (.9

I

I

0

50

I

100

C0 concentration gas

1.5

in

the

phase (~

B .//

/ 9~/oco

54"/~c 7 / ~ , c . ~ _ _ ~

/

95./o C7

v

r-

0.5

0 0

16

32 Time

56

72

(h)

Fig. 3. A Dependency of the growth rate of P. productus strain Marburg on the initial CO concentration in the gas phase. B Growth curves of P. productus strain Marburg grown in the presence of different initial CO concentration in the gas phase. The bacteria were grown in 100 ml medium with different CO concentrations (200 kPa) in 475 ml gas phase. The rest of the gas phase consisted of N2 and/or 17-20% CO2 (at CO concentrations of more than 73% less than 17% CO2). Incubation was at 37~ and 200 rpm

The internal cellular features of the cells of the isolate resemble those of other Gram-positive cocci. The large polyglucose granules show the granular substructure typical of lead-stained glycogen and are very similar to those reported for Clostridium pasteurianum by Laishley et al. (1973). Likewise, the polyphosphate granules are typical of

312 a n d from that: of Acetogenium kivui (Leigh et al. 1981). Uranyl acetate-stained whole mounts o f the latter revealed a hexagonal pattern very similar to the one we have seen in grazing thin sections of P. productus, although our freezefracture preparations failed to show a pattern of this type. Neither could we discern a repeating array in whole mounts stained with uranyl acetate, using the methods described by Mayer et al. (1977). It is likely that the hexagonal arrays are present in a double layer in P. productus, resulting in the twin rows of 14 nm particles which we observed. Because of this double layer, resolution of the subunits in whole m o u n t preparations might be difficult, especially considering the masking effect of the polysaccharide fibrous coating outside the wall layers. The unique nature of the wall structure in this organism suggests that additional investigation of its structure and chemistry will be profitable. Such studies are currently underway in our laboratories.

A 0.2 -3c-

0.1 O

i

i

i

20

30

40

Temperoture

a

Acknowledgements. This work was supported by a grant from the Deutsche Forschungsgemeinschaft, Bonn-Bad Godesberg and by the Fonds der Chemischen Industrie.

(~

9

References

0.16

I r" v

0.12

s s

O.OB 0

6.0

i

i

I

6.5

7.0

7.5

pH

Fig. 4. The effect of the temperature (A) and the pH (B) on the growth of P. productus on CO. The bacteria were grown in 100 ml medium with 475 ml gas phase containing 50% CO at 200 kPa. For growth conditions see materials and methods section

many prokaryotes (van Iterson, 1984 a). The plate-like lipids have been seen in Acetobacterium woodii by Mayer et al. (1977). They may represent an intracellular storage product. The cell envelope appears to be entirely unique (van Iterson, 1984b; Sleytr, 1978). Another acetogen, A. woodii, exhibits a wall layer with repeating subunits and a hexagonal array of particles in the plasma membrane (Mayer et al. 1977). We failed to find regular arrays of this type in P. productus, despite having used very similar preparation methods. The appearance o f the P. productus wall in thin section differs from that of A. woodii, which has three wall layers exterior to the plasma membrane (Mayer et al. 1977),

Balch WE, Schoberth S, Tanner RS, Wolfe RS (1977) Acetobacterium, a new genus of hydrogen oxidizing, carbondioxide reducing, anaerobic bacteria. Int J Syst Bacteriol 27: 355- 361 Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72: 248- 254 Braun M, Schoberth S, Gottschalk G (1979) Enumeration of bacteria forming acetate from Hz and COz in anaerobic habitats. Arch Microbiol 120: 201 - 204 Chapman RL, Staehelin LA (1986) Freeze-fracture (-etch) electron microscopy. In: Aldrich HC, Todd WJ (eds) Ultrastructure techniques for microorganisms. Plenum Press, New York, pp 213 240 De Ley J, Cattoir H, Reynaerts A (1970) The quantitative measurement of DNA hybridization from renaturation rates. Eur J Biochem 12:133-142 Diekert G, Hansch M, Conrad R (1984) Acetate synthesis from 2 COz in acetogenic bacteria: is carbon monoxide an intermediate? Arch Microbiol 138 : 224- 228 Diekert G, Fuchs G, Thauer RK (1985) Properties and function of carbon monoxide dehydrogenase from anaerobic bacteria. In: Poole RK, Dow CS (eds) Microbial gas metabolism. Academic Press, London Orlando San Diego New York Toronto Montreal Sydney Tokyo, pp 115-130 Diekert G, Schrader E, Harder W (1986) Energetics of CO formation and CO oxidation in cell suspensions of Acetobaeterium woodii. Arch Microbiol 144: 386- 392 Dorn M, Andreesen JR, Gottschalk G (1978) Fermentation of fumarate and g-malate by Clostridiumformieoacetieum. J Bacteriol 133 : 2 6 - 32 Erdos GW (1986) Localization of carbohydrate-containing molecules. In: Aldrich HC, Todd WJ (eds) Ultrastructure techniques for microorganisms. Plenum Press, New York, pp 399--420 Ezaki T, Yamamoto N, Ninomiya K, Suzuki S, Yabuuchi, E (1983) Transfer of Peptococcus indolicus, Peptoeoecus asaeehrarolyticus, Peptococcus prevotii, and Peptococcus magnus to the genus Peptostreptococcus and proposal of Peptostreptoeoeeus tetradius sp. nov. Int J Syst Bact 33:683-698 Friedrich W (1975) Vitamin B12 und verwandte Corrinoide. Georg Thieme Verlag, Stuttgart

313 Fuchs G (1986) C O 2 fixation in acetogenic bacteria: variations on a theme. FEMS Microbiol Rev 39:181 - 2 1 3 Gillis M, De Ley J, De Cleene M (1970) The determination of molecular weight of bacterial genome DNA from renaturation rates. Eur J Biochem 12:143-153 Gottwald M, Andreesen JR, Le Gall J, Ljungdahl LG (1975) Presence of cytochrome and menaquinone in Clostridium formicoaceticum and Clostridium thermoaceticum. J Bacteriol 122: 325- 328 Hollfinder R, Wolf G, Mannheim W (1977) Lipoquinones of some bacteria and mycoplasmas, with consideration of their physiological significance. Ant v Leeuw 43:177 - 185 Iterson W van (1984a) Inner structures of bacteria. Van Nostrand Reinhold Co., New York, p 501 Iterson W van (1984b) Outer structures of bacteria. Van Nostrand Reinhold Co., New York Kr6ger A (1978) Determination of contents and redox states of ubiquinone and menaquinone. In: Methods in enzymology, vol LIII. Academic Press, London New York, pp 579- 591 Kr6ger A, Innerhofer A (1976) The function of the b cytochromes in the electron transport from formate to fumarate of Vibrio succinogenes. Eur J Biochem 69 :497 - 506 Laishley EJ, MacAlister TJ, Clements I, Young C (1973) Isolation and morphology of native intracellular polyglucose molecules from Clostridium pasteurianum. Can J Microbiol 19 : 991 - 994 Leigh JA, Mayer F, Wolfe RS (1981) Acetogenium kivui, a new thermophilic hydrogen-oxidizing, acetogenic bacterium. Arch Microbiol 129 : 2 7 5 - 280 Lorowitz WH, Bryant MP (1984) Peptostreptococcus productus strain that grows rapidly with CO as the energy source. Appl Environ Microbiol 47 :961 - 964 Lynd L, Kerby R, Zeikus JG (1982) Carbon monoxide metabolism of the methylotrophic acidogen Butyribacterium methylotrophicum. J Bacteriol 149:255-263 Maniatis T, Fritsch ET, Sambrook J (1982) Molecular cloning (A laboratory manual). Cold Spring Harbour Laboratory Marmur J (1961) A procedure for the isolation of deoxyribonucleic acid from microorganisms. J Mol Biol 3 : 2 0 8 - 218

Marmur J, Doty P (1962) Determination of the base composition of deoxyribonucleic acid from its thermal denaturation temperature. J Mol Biochem 5:109-118 Mayer F, Lurz R, Schoberth S (1977) Electron microscope investigation of the hydrogen-oxidizing acetate-forming anaerobic bacterium Acetobacterium woodii. Arch Microbiol 115 : 2 0 7 213 Mfiller M, Meister N, Moor H (1981) Freezing in a propane jet and its application in freeze-fracturing. Mikroskopie 36:129-140 Rogasa M (1971) Peptococcaceae, a new family to include the Grampositive, anaerobic cocci of the genera Peptococcus, Peptostreptococcus, and Ruminococcus. Int J Syst Bact 21:234-237 Scherer P, Sahm H (1981) Effects of trace elements and vitamins on the growth of Methanosarcina barkeri. Acta Biotechnol 1 : 5 7 65 Sharak Genthner BR, Bryant MP (1982) Growth of Eubaeterium limosum with CO as the energy source. Appl Environ Microbiol 43 : 7 0 - 74 Sleytr UB (1978) Regular arrays of macromolecules on bacterial cell walls: Structure, chemistry, assembly, and function. Internat Rev Cytol 53 : 1 - 64 Tanner RS, Wolfe RS, Ljungdahl LG (1978) Tetrahydrofolate enzyme levels in Acetobacterium woodii and their implication in the synthesis of acetate from CO2. J Bacteriol 134:668-670 Vogelmann H, Wagner F (1973) Elutionschromatographische Trennungen von Corrinoiden an dem unpolaren Adsorbens Amberlite XAD-2. J Chromatogr 76: 359-- 379 Wolin EA, Wolin M J, Wolfe RS (1963) Formation of methane by bacterial extracts. J Biol Chem 238:2882-2886 Wood HG (1985) Then and now. Ann Rev Biochem 54:1-41 Wood HG, Ragsdale SW, Pezacka E (1986) The acetyl-CoA pathway of autotrophic growth. FEMS Microbiol Rev 39: 3 4 5 - 362 Zeikus JG, Kerby R, Krzycki JA (1985) Single-Carbon chemistry of acetogenic and methanogenic bacteria. Science 227:11671173 Received March 30, 1987/Accepted May 22, 1987