Cellular Fatty Acid Spectra of Sporolactobacillus and Some Other Bacillus-Lactobacillus Intermediates As a Guide to Their Taxonomy

J. Gen. Appl. Microbiol., 19, 129-140 (1973)

CELLULAR FATTY ACID SPECTRA OF SPOROLACTOBACILLUS AND SOME OTHER BACILLUS-LACTOBACILLUS INTERMEDIATES AS A GUIDE TO THEIR TAXONOMY

KINJI UCHIDA AND KOYA MOGI

Central Research Laboratories, Kikkoman Shoyu Co., Ltd., Noda-shi, Chiba-ken 278, Japan

(Received September 18, 1972)

Cellular fatty acid spectra of Sporolactobacillus species and some other Bacillus-Lactobacillusintermediates were comparatively examined with special reference to their taxonomy. Bacillus coagulans, 4 strains of Sporolactobacil- lus, and other spore-forming lactic acid bacteria were found to have "iso- anteiso-type" fatty acid spectra similar to those previously reported for most species of Bacillus. Iso- or anteiso-branched odd-numbered carbon acids were predominant in the fatty acids of these organisms. Lactobacillusyamanashien- sis, a motile, diaminopimelic acid-containing, but non-spore-forming lactic acid bacterium was found, in contrast, to have "unsaturated-type" fatty acid spectrum biosynthetically different from the fatty acid spectra of the former but similar to those of L. plantarum and other lactobacilli. No transitional fatty acid pattern between the two types could be detected among the tested intermediates. Feasibility of taxonomic discrimination of Bacillus-Lactobacil- lus intermediates by examination of the cellular fatty acid spectra was dis- cussed.

In the taxonomy of lactobacilli, the relation of Lactobacillus to other taxa, particularly the borderline between genus Bacillus has been the most interesting subject to be investigated with special reference to their phylogeny (1). Numerous aberrant types sharing characters of both Lactobacillus and Bacillus have been reported (2-7). Sporolactobacillus inulinus, a catalase negative, spore-forming lactic acid bacterium isolated by KITAHARA and SUzuKI (2), is a typical one among such intermediates between the two genera. Recently, many other catalase-negative, spore-forming lactic acid fermenters closely resembling S. inulinus were isolated from rhizosphere (3, 4). However, the exact systematic position of this genus remains un- certain. DAVIS (5) critically studied the aberrant lactobacilli appearing in several literature and assumed that the genus Bacillus may be a convenient source from which lactobacilli might have originated and also that the atypical lacto

129 130 UCHIDAand MOGI VOL. 19 bacilli may represent transitional forms of an evolutionary path from the sporeformer to L, plantarum. MILLER (S, 9) studied the DNA base homology among B. coagulans, Sporolactobacillus inulinus, and L, plantarum, and observed no cross reaction indicating direct phylogenetic relationship among these organisms, in spite of their similar GC content values. The information available on the fatty acid compositions of bacteria has enabled certain phylogenetic relationship to be established. The authors have investigated fatty acid spectra or characters of the fatty acid synthetic pathways of various lactic acid bacteria in anticipation that such data might be useful in their classification (10, 11). This paper reports an investigation on the fatty acid spectra of Sporolactobacillus and some other Bacillus- Lactobacillus intermediates in relation to their taxonomy.

MATERIALS AND METHODS

Microorganisms. Bacillus coagulans P-22, B. racemilacticus M-5, B. mixolactis M-105, B, laevolacticus M-7, Sporolactobacillus racemicus M-11, S. inulinus M-20, S. laevus M-47, and Lactobacillus yamanashiensis 239 were kindly supplied by Prof. 0. Nakayama, Yamanashi University, Kofu. S. inulinus 6-11 and L, plantarum IAM 1216 were supplied by Prof. K. Kitahara, Tokyo University of Agriculture, Setagaya-ku, Tokyo. B. subtilis IAM 1523 and L, casei ATCC 7469 were obtained from the stock culture collection of our laboratories. Growth condition. All the strains of lactobacilli, sporolactobacilli, and bacilli were grown in the medium A which had been previously applied for the fatty acid analyses of lactic acid bacteria (11). Duplicate cultures were made for each strain and cells were harvested after 48 and 72 hr of growth, to insure reliability of the analysis. Under the present conditions, all the cultures for fatty acid analyses reached stationary phase of growth before 48 hr. Other conditions of the culture and harvest of cell crops were the same as previously reported (11). Analysis of cellular fatty acids. Fatty acids were extracted from alkali- hydrolyzates of whole cells, methylated, and analyzed by gas-liquid chromatography using capillary columns (butanediol succinate (BDS), 0.5 mm x45 m, and Silicon SE-30, 0.5 mm x 45 m) as previously described (11). Fatty acid compositions in the following tables were those obtained with the BDS capillary column. Detection of diaminopimelic acid. Washed cells grown on the medium A were hydrolyzed with 6 N HCl and examined by paper chromatography (12) for diaminopimelic acid. The characteristic green spots of a, -diaminopimelic acid, which faded to a permanent yellow colour, were developed by dipping the sheets into 0.1% (w/v) ninhydrin in acetone and heating at 105° for 2 min (13). 1973 Fatty Acid Spectra of Sporolactobacillus 131

Fig. 1. Gas-liquid chromatograms of the cellular fatty acids of Sporolactobacillus inulinus (A) and Lactobacillus yamanashiensis (B). Chromatographic conditions. Gas chromatograph, Hitachi K-53; detector, FID ; column, capillary, 0.5 mm X 45 m, coated with butanediol succinate (BDS) ; carrier-gas flow rate, NZ 3 ml/min for A, and 4 ml/min for B; split ratio, 1/12.5; oven temperature, 150° for A, and 160° for B. 132 UCHIDA and MOGI VOL. 19

Electron microscopy. L. yamanashiensis 239 was grown at 30° in a liquid medium containing a low concentration of glucose with the following formula (6): In 1 liter of tap water ; glucose, 0.5 g ; meat extract (Kyokuto), 5 g ; polypeptone (Daigo), 5 g ; yeast extract (Difco), 5 g ; Tween 80, 0.5 g ; MnSO4 • 4H2O, 0.1 g ; potassium citrate, 1 g ; pH 6.5. Twelve-hour and 20-hr cultured cells were harvested and washed twice gently with distilled water by centrifugation. The dried bacterial cells on grids were shadowed with chromium at an angle of 25°, and examined in a Super-Scope JEM-50 B (Japan Electron 1973 Fatty Acid Spectra of Sporolactobacillus 133

Optics Laboratories Co., Ltd.) Micrographs were taken at initial magnifica- tion of 1500.

RESULTS

The cellular fatty acid spectra of 72-hr cultured cells of the tested strains are presented in Table 1. The fatty acid compositions obtained with the 48- and 72-hr cultured cells for each strain agreed well. Table 2 indicates reproducibility in the duplicate determinations of fatty acid spectra of B, racemi- 134 UCHIDA and MOGI VOL. 19

Table 2. Reproducibility in the determinations of cellular fatty acid spectra.

.lacticus and S. laevus compared with the results of the separate experiment performed 6 months later. Major constituents of the cellular fatty acids of the typical bacillus (B. subtilis) were anteiso- and iso-branched acids of 15 or 17 carbon chain, and, among these, anteiso-pentadecanoic acid was the most predominant. Smaller amounts of even-numbered (C14, C16) iso-branched acids and straight-chain acids were also present, but unsaturated acid or cyclopropane acid was not detected. This fatty acid pattern of B. subtilis agrees well with the results appearing in other literature (14). The fatty acid spectra of B. coagulans P-22, Nakayama's spore-forming lactic acid bacteria, and all the strains of sporolactobacilli were found to be similar to that of B, subtilis. In other words, these organisms had, without exception, "Bacillus-type" cellular fatty acid spectra. In these microorganisms any trace amount of unsaturated acid or cyclopropane acid could not be detected. Sum of the anteiso-branched acids in the cellular fatty acids of S. inulinus 6-11 and also in S. inulinus M-20 reached up to 80% of the total. Fatty acid spectra of B. racemilacticus 1973 Fatty Acid Spectra of Sporolactobacillus 135

M-5, B, laevolacticus M-7, and B, mixolactis M-105 were almost identical to each other. It was found, in contrast, that the cellular fatty acid of the motile Lac- tobacillus yamanashiensis 239 consisted of even-numbered straight-chain saturated acids, even-numbered straight-chain mono-unsaturated acids, and cyclopropane acids ; and the most predominant of these was C19-cyclopropane (lactobacillic) acid. Detectable amount of iso- or anteiso-branched acid was not observed, even when the high resolution capillary column was used for gas chromatographic analysis. Such a fatty acid composition like this is qualitatively fully consistent with those of typical lactobacilli. Considering quantitative dif- ferences, the fatty acid pattern of L. yamanashiensis could be differentiated from that of L. plantarum by relatively smaller content of cis-vaccenic acid and higher contents of palmitic and lactobacillic acids in the former. The motility of the received culture of L. yamanashiensis 239 was pre- liminarily assured by the usual microscopic observations. Cells in the early stationary phase (48-hr culture on the medium A at 30°) and in the earlier stages were shown to be motile. Electron microscopic examination of this organism grown on the low-glucose medium revealed that this organism possessed numerous peritrichous flagella (Fig. 2). Diaminopimelic acid in L. yamanashiensis was also examined, since its occurrence in this organism had not been previously reported. Paper chromatography of the acid-hydrolyzates of the 24- and 48-hr cultured cells revealed that this motile lactobacillus contained as much diaminopimelic acid as L. plantarum or S. inulinus.

DISCUSSION

Numerous "atypical" bacilli or lactobacilli, sharing characters of both Lactobacillus and Bacillus, have been reported. It is a well known fact that the organism designated formerly as L, thermophilus was a synonym for B. coagulans. KITAHARA and SUZUKI (2) isolated a catalase-negative, nitrate- not reducing homo-lactic acid fermenter which was motile, and under, some limited conditions, produced endospores. They proposed a new taxon ; Sporo- lactobacillus as a subgenus of Lactobacillus and named their organism Sporo- lactobacillus inulinus. NAKAYAMA and YANOSHI (3) isolated many nitrate- not reducing, but catalase-positive spore-forming lactic acid fermenters from rhizosphere, and classified these organisms as B. racemilacticus, B. laevolacticus, and B. mixolactis according to the optical rotation of the lactic acids produced. Simultaneously they also isolated a number of catalase-negative, spore-forming lactic acid fermenters some of which were identical to Kitahara's organism, and the others were classified as S. laevus or S. racemicus (4, 15). From their taxonomic studies on these microorganisms, they proposed that the subgenus Sporolactobacillus should be included in genus Bacillus or family Bacillaceae rather than in genus Lactobacillus (15). 136 UCHIDA and MOGI VOL. 19

Fig. 2. Electron micrographs of Lactobacillus yamanashiensis 239 showing flagella. Cells were grown on the low-glucose (0.05%) liquid medium at 30°, harvested by centrifugation and then washed with distilled water. The grids were shadowed with chromium at an angle of 25°, and examined in a Super-Scope JEM-50 B. A : 12-hr culture, B : 20-hr culture. 1973 Fatty Acid Spectra of Sporolactobacillus 137

On the other hand, motile lactobacilli have been reported by several authors ( 5, 6, 7 ). Interestingly most of these organisms resemble taxonomically L, plantarum more closely than any other species of Lactobacillus. L, plantarum has been considered to occupy the ultimate position within Lactobacillus for its diaminopimelic acid (16) and ribitol-teichoic acid (17 ) in the cell wall fraction, and for its distinct serological behavior (18 ). In these circumstances, DAVIS ( 5 ) concluded that the atypical lactobacilli might represent transitional forms of an evolutionary path from the spore-former to L. plantarum. Lactobacillus yamanashiensis 239, isolated from wine by NONOMuRA et al. ( 7 ), is one of these atypical lactobacilli, which is very closely related to L. plantarum except its motility. Although the motility of this strain had originarily been reported to be due to a polar flagellum ( 7 ), electron microscopic examination in the present work showed that this organism in the earlier growth phases possessed numerous peritrichous flagella as seen in other motile lactobacilli reported so far ( 6, 19). By arranging these intermediates in the order of resemblance in principal taxonomic characters (Table 3), one can find there a series of bacteria placed in every stage of continuous alteration in taxonomic features from typical bacilli to typical lactobacilli. Examination of the fatty acid spectra of these intermediates in this work showed that strains of B, coagulans, Nakayama's spore-forming lactic acid fermenters, and all the tested strains of sporolactobacilli have exclusively "Bacillus -type" or , in other words, "iso-anteiso-type" (24) fatty acid spectra, and that both L, yamanashiensis and L, plantarum have, in contrast, "Lacto- bacillus-type" or "unsaturated-type" (24) fatty acid spectra. No transitional fatty acid pattern between the two types could be detected among these taxonomical intermediates. Since all these organisms were cultured under the same conditions, the discrepancy in cellular fatty acid spectra is thought to be due solely to the difference in fatty acid synthetic pathway among these organisms. The iso-anteiso-type fatty acid composition or an enzyme system result- ing in these kinds of fatty acids is known to be distributed in a relatively wide range but in a certain limited number of bacterial taxa including Bacillus (14), Micrococcus, Staphylococcus, Sarcina (25), Streptomyces, Actinomyces, and Nocardia (26). On the other hand, the unsaturated-type has so far been found in taxonomically different groups of bacteria such as Clostridium (27), Escherichia (28), Salmonella, and other genera of Enterobacteriaceae (29), in addition to several taxa in Lactobacillaceae (11, 30, 31). These different types of fatty acids have been elucidated to be synthesized through obviously distinct synthetic pathways. The major constituent fatty acids of bacilli, odd-numbered iso-branched acids, odd-numbered anteiso-branched acids, and even-numbered iso-branched acids are synthesized through repeated addition of -CH2-CH2- groups to isovaleryl-CoA, a-methylbutyryl-CoA, and iso- 138 UCHIDA and MOGI VOL. 19

Table 3. Some taxonomic characteristics of Bacillus-Lactobacillus intermediates and their types of cellular fatty acid spectra.

butyryl-CoA, respectively (32). The synthetic mechanism of fatty acids in lactobacilli was studied by O'LEARY (33) and was shown to be similar to that of E. coll. In this mechanism, the long-chain unsaturated acids are synthesized through the elongation of already unsaturated short-chain acids, and the cyclopropane acids are synthesized by the addition of methylene groups to the double bonds of the long-chain unsaturated acids. Although some unsaturated acids are rarely formed by some limited species of bacilli under limited conditions, their biosynthetic pathway and, therefore, the position of their double bond are different from those of the unsaturated acids in lactobacilli (34). Alteration of the cellular fatty acid spectrum between these two types depending on cultural conditions has not ever been reported. UETA ct al. (24) suggested that those bacteria sharing the same fatty acid pattern or the same fatty acid synthetic pathway may belong phylogenetically to one group, in a certain sense. 1973 Fatty Acid Spectra of Sporolactobacillus 139

From the viewpoint of comparative biochemistry, at least from the ob- servation of the cellular fatty acid spectra, these Bacillus-Lactobacillus intermediates examined in the present work are obviously separated into the two phylogenetically independent groups, Bacillus group and Lactobacillus group. A relatively greater borderline is assumed to lie between spore-bearing lactic acid bacteria and motile, diaminopimelic acid-containing but not spore-forming lactobacilli. SUZUKI and KITAHARA ( 23 ) comparatively studied the guanine plus cyto- sine (GC) content in the DNA of these intermediates and some lactobacilli, and found that the GC content of B, coagulans, S. inulinus, and L. plantarum were 46.9, 39.3, and 42.9 mole %, respectively. Recently, MILLER et al. ( 8, 9 ) reported the GC values of 45.4, 47.3, and 44.5 mole %, respectively, for the same species. In their subsequent DNA hybridization experiments, however, they could observe no cross reaction indicating direct phylogenetical relationship among these organisms, in spite of their relatively similar GC values. The GC values reported for Bacillus species (22) and Lactobacillus species (8, 23, 35) generally fall within a corresponding range. It could not be said that the GC percentage is a sufficient criterion to separate lactobacilli from bacilli. The results of the present work suggest, although further investigation with additional species and strains are needed before a definite conclusion, that examination of the fatty acid spectra of bacillus-resembling "atypical lactobacilli" would enable rapid discrimination of true atypical "lactobacillus" from organisms which should rather be called atypical "bacillus".

The authors are grateful to Prof. K. Kitahara of Tokyo University of Agriculture and to Prof. 0. Nakayama of Yamanashi University for generously supplying the strains studied here and also for their valuable advices. The authors thank Mr. S. Kitahara of this laboratory for operating the electron microscope and Miss S. Sekikawa for her technical assistance. The authors also thank Dr. T. Yokotsuka, the director of their laboratories, for his support and encouragement. A preliminary report of this investigation was presented before 46th Annual Meeting of the Agricultural Chemical Society of Japan, April 1971.

REFERENCES 1) H.G. GYLLENBERGand E. EKLUND, Proc. 10th LA.M. Symp. Microbiology, p. 31 (1968). 2) K. KITAHARAand J. SUZuKI,J. Gen. Appl. Microbiol., 9, 59 (1963). 3) 0. NAKAYAMAand M. YANOSHI,J. Gen. Appl. Microbiol., 13, 139 (1967). 4) 0. NAKAYAMAand M. YANOSHI,J. Gen. Appl. Microbiol., 13, 155 (1967). 5) G.H.G. DAVIS,J. Gen. Microbiol., 34, 177 (1964). 6) M. GEMMELLand W. HODGKISS,J. Gen. Microbiol., 35, 519 (1964). 7) H. NONOMURA,T. YAMAZAKI,and Y. 0HARA, Mitt. Klosterneuburg, Ser. A, 15, 241 (1965). 8) A. MILLER III, W.E. SANDINE,and P.R. ELLIKER,J. Bacteriol., 102, 278 (1970). 140 UCHIDA and MOGI VOL. 19

9) A. MILLER III, W.E. SANDINE, and P.R. ELLIKER, Can. J. Microbial., 17, 625 (1971). 10) K. UCHIDA and K. MOGI, Agr. Biol. Chem. (Tokyo), 35, 136 (1971). 11) K. UCHIDA and K. MOGI, J. Gen. Appl. Microbiol., 18, 109 (1972). 12) H.R. PERKINS, Nature, 208, 872 (1965). 13) D.S. HOARE and E. WORK, Biochem. J., 65, 441 (1957). 14) T. KANEDA, J. Bacteriol., 93, 894 (1967). 15) 0. NAKAYAMA, Proc. 45th Meet. Agric. Chem. Soc. Japan, (1970), p. 315. 16) CS. CUMMINS and H. HARIS, J. Gen. Microbiol., 14, 583 (1956). 17) J. BADDILEY and A.L. DAVISON, J. Gen. Microbiol., 24, 295 (1961). 18) M.E. SHARP, J. Gen. Microbiol., 12, 107 (1955). 19) A.P. HARRISON, JR., and P.A. HANSEN, J. Bacteriol., 59, 444 (1950). 20) ES. BREED, E.G.D. MURRAY, and N.R. SMITH, Bergey's Manual of Determinative Bacteriology, 7th Ed. Williams and Wilkins Co., Baltimore (1957). 21) M. ANTIA, D.S. HOARE, and E. WORK, Biochem. J., 65, 448 (1957). 22) J. MARMUR and P. DOTY, J. Mol. Biol., 5, 109 (1962). 23) J. SUZUKI and K. KITAHARA, J. Gen. Appl. Microbial., 10, 305 (1964). 24) N. UETA, I. ISHIZUKA, and T. YAMAKAWA, Proc. First Int. Conf. Culture Collec- tions. (1970), p. 371. 25) I. ISHIZUKA, N. UETA, and T. YAMAKAWA,Japan. J. Exp. Med., 36, 73 (1966). 26) A. BALLIO et S. BARCELLONA, Ann. Inst. Pasteur, 114, 121 (1968). 27) C.W. MOSS and V.J. LEWIS, Appl. Microbiol., 15, 390 (1967). 28) J.D. KARKAS, H. TURLER, and E. CHARGAFF, Biochem. Biophys. Acta, 111, 96 (1965). 29) W.M. O'LEARY, The Chemistry and Metabolism of Microbial Lipids, World Publish- ing Co., New York (1967) p. 101. 30) P. MACLEOD and A. MILLER III, J. Dairy Sci., 50, 155 (1967). 31) J.H. VEERKAMP, J. Bacteriol., 108, 861 (1971). 32) T. KANEDA, J. Biol. Chem., 238, 1229 (1963). 33) W.M. O'LEARY, Biochemistry, 4, 1621 (1965). 34) A.J. FULCO, J. Biol. Chem., 244, 889 (1969). 35) F. GASSER et M. SEBALD, Ann. Inst. Pasteur, 110, 260 (1966).