Vol. 40, No. 1

INTERNATIONAL JOURNAL OF SYSTEMATIC BACTERIOLOGY, Jan. 1990, p. 92-97 0020-7713/90/010092-06$02.00/0

Taxonomy of Alkaliphilic Bacillus Strains DAGMAR FRITZE,l* JOSEF FLOSSDORF,2 A N D DIETER CLAUS' Deutsche Sammlung yon Mikroorganismen und Zellkulturen GmbH,' and Abteilung Enzymtechnologie, Gesellschaft fur Biotechnologische Forschung mbH,2 Braunschweig, Federal Republic of Germany The DNA base compositions of 78 alkaliphilic Bacillus strains were determined. These strains were grouped as follows: DNA group A, guanine-plus-cytosine(G+C) content of 34.0 to 37.5 mol% (17 strains); DNA group B, G+C content of 38.2 to 40.8 mol% (33 strains); and DNA group C, G+C content of 42.1 to 43.9 mol% (28 strains). DNA group A includes the type strain of Bacillus alcalophilus Vedder 1934. DNA-DNA hybridization studies with DNA group A strains revealed that only one strain, strain DSM 2526, exhibited a high level of DNA homology with B . alcalophilus DSM 4 S T (T = type strain). Neither strain DSM 485Tnor any other DNA group A strain is homologous to any of the Bacillus type strains with comparable base compositions. Six strains formed a distinct group containing three highly homologous strains and three strains exhibiting >50% DNA homology.

MATERIALS AND METHODS

In 1934 Vedder (37) isolated several endospore-forming bacterial strains that were able to grow at pH values up to 10 but not below pH 8.6. Since obligately alkaliphilic bacteria were not known at that time, Vedder (37) described his isolates as a new species, Bacillus alcalophilus. The restricted number of biochemical and other tests used by Vedder does not allow a conclusion about the homogeneity of this species. In the 1970s these physiologically unique bacilli became interesting because of their ability to produce unusual enzymes that are applicable in industrial and other processes due to their high pH optima (1-3, 9, 13, 14, 20, 21). To date, a large number of alkaliphilic Bacillus strains, as well as other alkaliphilic microorganisms, have been isolated. A considerable number of these have been deposited with culture collections in connection with patent applications and are designated only as Bacillus sp. Some bear species names which are not included on the Approved Lists of Bacterial Names (32) or on the validation lists in the International Journal of Systematic Bacteriology. Therefore, they have no taxonomic standing. In addition, these groups were not included in Bergey's Manual of Systematic Bacteriology (6). In 1982, Gordon and Hyde (17) investigated a large number of these alkaliphilic Bacillus strains. For characterization by standardized tests used for Bacillus strains, the alkaliphilic strains were adapted to growth on media with neutral pH values prior to examination. On the basis of the test results, Gordon and Hyde (17) grouped these organisms with the Bacillus Jirmus-Bacillus lentus complex and distinguished five subgroups. In a former publication Gordon et al. (18) had demonstrated the heterogeneity of the B . firmus-B. lentus complex, a characteristic also found in the alkaliphilic Bacillus strains. In this work we studied the extent of the guanine-pluscytosine (G +C) content variation among alkaliphilic Bacillus strains and the levels of relatedness of these strains to the type strain of B. alcalophilus as determined by DNA-DNA hybridization experiments.

Bacterial strains. The Bacillus strains used in this study are listed in Table 1. Most of the strains were obtained from the National Collections of Industrial and Marine Bacteria, Ltd., Torry Research Station, Aberdeen, Scotland, the Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH, Braunschweig, Federal Republic of Germany, and the American Type Culture Collection, Rockville, Md. Seven strains were provided by K. Aunstrup, Novo Industri, Copenhagen, Denmark. One strain was obtained from A. Guffanti, Mount Sinai Medical School, New York, N.Y., and five strains came from H.-G. Triiper, Bonn, Federal Republic of Germany. The reference strains used (type strains) were obtained from the Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH. Characterization. Unless otherwise stated, the strains were characterized by using the modified methods of Gordon et al. (16). After autoclaving, the pH values of the media were adjusted to 9.7 by adding a sodium sesquicarbonate solution to a final concentration of 0.1 mollliter (2). The media used for nitrate reduction and phenylalanine deamination tests had to be acidified prior to examining the activity of the strains. Sugar utilization tests were carried out in a basal mineral medium containing (per liter) 7 g of K2HP0,, 2 g of KHZPO,, 0.1 g of MgSO, . 7H2O, 1 g of (NHJ2SO4, 5 g of NaC1, and 1 ml of a filter-sterilized vitamin solution; the vitamin stock solution contained (per liter) 20 mg of D-(+)-biotin, 20 mg of folic acid, 50 mg of thiamine hydrochloride, 50 mg of calcium D-(+)-pantothenate, 1 mg of cobalamine, 50 mg of riboflavin, 50 mg of nicotinic acid, 50 mg of p-aminobenzoic acid, and 100 mg of pyridoxine hydrochloride and was stored frozen. The pH of this medium was adjusted to 8.9 to 9.1 with NaOH. As an indicator for acid production, thymol blue was used. Growth at constant neutral and alkaline pH values was determined by using the basal mineral medium containing glucose and supplemented with sodium sesquicarbonate for achieving the alkaline pH values (defined M medium). The following additional tests were performed by using previously described methods: arginine dihydrolase test (36), Tween test (31), pullulanase test (27), urease test (33), cellulase test (34), 4-methylumbelliferone glucuronidase test (11), and diaminopimelic acid test (24). G+C ratios. About 2 x lo8 cells were harvested by

* Corresponding author. 92

VOL. 40, 1990

ALKALIPHILIC BACILLUS STRAINS

93

TABLE 1. List of bacterial strains used in this study Taxon

B . alcalophilus

"B. atcalophilus subsp. halodurans" B . circulans B . firmus B . lentus B . marinus B . pantothenticus B . sphaericus Bacillus sp.

Source andor strain designation"

DSM 4UT DSM 2526 DSM 2512 DSM 497 DSM 2513 DSM llT DSM 2528 DSM 12T Guffanti DSM ST DSM 2197T DSM 26T DSM 28T DSM 1972 NCIB 10282 to NCIB NCIB 10286 to NCIB NCIB 10296 to NCIB NCIB 10309 NCIB 10310 to NCIB NCIB 10314 NCIB 10316 to NCIB NCIB 10324 to NCIB DSM 2514 DSM 2515 ATCC 21592 DSM 2517 DSM 2518 DSM 2519 DSM 2520 DSM 2521 DSM 2522 DSM 2523 DSM 2524 DSM 2525 Aunstrup

Other culture collection designation"

ATCC 27647T NCIB 10438 ATCC 21522 ATCC 27557 ATCC 21591 ATCC 4513= NCIB 9218 ATCC 1457ST

Other designation

Reference(s) and/or comment

Vedder IT Vedder I1 221

37 17 17; Horikoshi and Ikedab

A-59

17; Horikoshi patent strain 22, 28 5, 17 4, 18 19 15, 18 30 16, 29 23, 26 17; Aunstrup et al." 17; Aunstrup et al.' 17; Aunstrup et al.' 17; Aunstrup et al." 17; Aunstrup et al.' 17; Aunstrup et al.' 17; Aunstrup et al.' 17; Aunstrup et al.' 17; Aunstrup et al.' Horikoshi and Ikedab Horikoshi and Ikedab 17; Horikoshi patent strain 17 17; Horikoshid 17; Horikoshid 17; Horikoshi patent strain R. Kenkyusho patent strain R. Kenkyusho patent strain Horikoshid Horikoshid R. Kenkyusho patent strain 17; Aunstrup et al.'

RU38 RAB

ATCC lO84OT ATCC 29841T ATCC 14576T ATCC 14577T NCIB 10281 10285 10295 10308 10313 10323 10327 ATCC 21536 ATCC 21537 ATCC ATCC ATCC ATCC ATCC ATCC ATCC ATCC ATCC

Triiper

21593 21594 21595 21596 21832 21833 31006 31007 31084

C311 C323 to C326 c334 to c343 C346 to C358 C360 C364 to C367 C369 C371 to C378 C410 to C413 0-4 Y -76 A-40 124-1 169 135 27-1 Nl N4 13 17-1 M-29 BC3, BC4, BC7, BB16, PB9, PB19, PB38(2) WN13 0 2 , 0 3 , M5, M8

38 38 ~

~

~~~~~

" DSM, Deutsche Sammlung von Mikroorganismen und Zellkulturen, Braunschweig, Federal Republic of Germany; ATCC, American Type Culture Collection, Rockville, Md.; NCIB, National Collection of Industrial Bacteria, Torry Research Station, Aberdeen, Scotland. K. Horikoshi and Y.Ikeda, U.S. patent 4,052,262, October 1977. K. Aunstrup, 0. Andresen, and H. Outtrup, British patent 1,243,784, August 1971. K. Horikoshi, U.S. patent 3,923,598, December 1975.

centrifugation of a sample of culture broth at 7,500 x g for 5 min. After two washes with saline-EDTA, the cells were broken by combined treatment with lysozyme and sodium lauryl sulfate and subsequently were subjected to protein degradation by proteinase K, using mainly a method developed by Marmur (25) and described in detail by Flossdorf (12). This lytic treatment yielded high-molecular-weight DNA, which was used to determine G+C contents by buoyant density centrifugation in CsCl gradients. Micrococcus Zuteus DSM 20030 (G+C content, 73.8 mol%) was routinely used as a marker. The marker was added as whole cells prior to lytic degradation. From the difFerence between the radial position of the DNA in question (rs) and the marker DNA (r,) (that is, Ar = r,-r,), together with the mean radial position of both (F) [F = (rs+rm)/2], the difference between the two G + C contents [A(G+C)] was calculated from the following equation: A(G+C) = 20.01 X F X Ar. Finally, (G+C), = (G+C), + A(G+C), (for details see reference 12). DNA-DNA hybridization. Cells were harvested from the late logarithmic phase of growth by centrifugation, washed

twice with saline-EDTA (0.15 M NaCl-O.l M EDTA, pH 8.0), and stored at -20°C until they were used. The DNA was extracted by using a slight modification of the Marmur method (10). About 2 to 3 g of wet cells, homogenized in 25 ml of saline-EDTA, was lysed by adding 10 mg of lysozyme, and the preparation was incubated at 37°C for 20 min. Subsequently, 2 ml of a 25% solution of sodium lauryl sulfate was added. After repeated precipitation and purification, the isolated DNA had a ratio of A,,, to AZ6, to A,,, of 0.43 to 0.49: 1:0.48 to 0.55. DNA-DNA hybridization was determined spectrophotometrically from the renaturation rates of the individual DNA samples and their mixtures by using the method of De Ley (8). The purified DNA solutions (A26o, approximately 1.6) were passed twice through a French pressure cell at 0.98 GPa and dialyzed against 500 volumes of 2 x SSC (Ix SSC is 0.15 M NaCl plus 0.015 M sodium citrate) at 4°C for 3 days with frequent changes of the buffer. The DNA concentration of the two samples to be hybridized was adjusted to an of 0.8 in 2 x SSC, corresponding to 40 pg of DNA per ml. A third sample was prepared containing equal amounts of both DNAs. An adenine solution

94

INT. J . SYST.BACTERIOL.

FRITZE ET AL.

TABLE 2-Continued -

TABLE 2. G+C contents of the alkaliphilic Bacillus strains and phenotypic groupings ~~~~

Taxon

~

Bacillus sp.

Strain

DSM 2519 NCIB 10318 PB 38(2) DSM 2518 NCIB 10327 NCIB 10284 M5 DSM 2528 DSM 2522 DSM 2526 M8 B . alcalophilus DSM 485= Bacillus sp. PB911 WNll WN13 DSM 2521 PB9/2 DSM 1972 NCIB 10291 NCIB 10289 NCIB 10282 03 NCIB 10299 NCIB 10286 NCIB 10287 NCIB 10288 NCIB 10290 NCIB 10314 ATCC 21592 NCIB 10294 NCIB 10300 NCIB 10296 NCIB 10292 NCIB 10293 DSM 2517 NCIB 10302 DSM 2523 02 NCIB 10283 NCIB 10285 NCIB 10303 NCIB 10298 NCIB 10297 DSM 2524 RAB N U B 10305 BC3 NCIB 10295 DSM 2520 PB19 NCIB 10307 NCIB 10308 NCIB 10316 NCIB 10313 NCIB 10312 BB16 NCIB 10304 "B. alcalophilus DSM 497 subsp. halodurans' '

G+C DNA content (mol%) group

sziu

Study of Gordon and Hydeb

34.0 35.0 35.1 35.2 35.4 35.4 35.5 35.8 36.1 36.2 36.3 36.5 36.9 36.9 37.1 37.2 37.5 38.2 38.4 38.6 38.8 38.8 39.0 39.1 39.2 39.2 39.3 39.3 39.4

A A A A A A A A A A A A A A A A A B B B B B B B B B B B B

5 6 5 5 5 5 6 6 6 5 6 5 6 5 5 6 6 5 2 2 2 6 1 1 2 2 1 2 1

B . lentus I Intermediate Intermediate Intermediate Intermediate Intermediate

39.4 39.4 39.5 39.5 39.5 39.5 39.5 39.6 39.6 39.6 39.6 39.6 39.7 39.7 39.7 39.7 40.2 40.3 40.4 40.4 40.8 42.1 42.2 42.3 42.5 42.5 42.5 42.5 42.5

B B B B B B B B B B B B B B B B B B B B B C C C C C C C C

1 1 1 1 2 1 1 2 1 1 1 1 1 1 1 1 1 1 1 1 2 4 4 4 4 4 4 4 4

B .firmus B . firmus B . firmus B . firmus B . lentus I B . firmus B . firmus

Intermediate Intermediate Intermediate Intermediate

Intermediate B . lentus I B . lentus I B . lentus I B . lentus I B . firmus B . lentus I B . lentus I B. lentus I B . firmus B . lentus I B . firmus

B .firmus B . firmus B . firmus B . firmus B . firmus

~

Phenotypic group

Phenotypic group Taxon

~~

Strain

Bacillus sp.

NCIB 10306 NCIB 10301 NCIB 10322 NCIB 10325 BC4 NCIB 10321 NCIB 10309 NCIB 10310 NCIB 10311 NCIB 10324 NCIB 10323 "B. alcalophilus DSM 2513 subsp. halodurans" Bacillus sp. NCIB 10320 DSM 2514 NCIB 10319 NCIB 10317 NCIB 2525 DSM 2515 NCIB 10326 DSM 2512 a

Continued

s$ia

Study of Gordon and Hydeb

42.6 42.6 42.6 42.6 42.7 42.7 42.8 42.8 42.8 42.8 42.8 42.8

C C C C C C C C C C C C

4 4 4 4 4 4 3 4 3 4 4 4

B . lentus 111 B . lentus 111 B . lentus I11 B . lentus I11 Intermediate B . lentus I11 B . lentus I1 B . lentus 111 B . lentus I1 B . lentus I11 B . lentus I11 Intermediate

43 .O 43.2 43.3 43.3 43.5 43.6 43.9 43.9

C C C C C C C C

4 3 3 3 3 3 3 3

B . lentus I11 B . lentus I1 B . lentus I1 B. lentus I11 B . lentus I1

See Table 3. See reference 17.

having an A,,, of 0.8 in 2X SSC was used as a blank to allow correction for electronic disturbances. Values were determined by using a model 2600 UV spectrophotometer system (Gilford Instrument Laboratories, Inc.) and a model 2527 thermoprogrammer (Gilford) equipped with a thermoelectronic control. The DNA solutions were denatured by raising the temperature to 98°C for 15 min. The thermoprogrammer was then adjusted to the optimum renaturation temperature (TOR), which was calculated from the following equation given by De Ley (7): TOR = [0.51(G+C content)] + 47. The cooling process took 2 to 2.5 min. Then the decrease in A,,, at the constant optimal renaturation temperature was plotted for 40 min with a model 7225A plotter (Hewlett-Packard Co.). The degree of binding (%D) was calculated from the following equation given by De Ley (7): %D =

4v'M - v'A - v'B

2 x B . firmus Intermediate B . firmus B . firmus Intermediate B . lentus I11 B . lentus I11 B . lentus I11 B. lentus I11 B . lentus I11 Intermediate B . lentus I11 B . lentus I11

G+C DNA content (mol%) group

VZiTiTi x

loo

Each hybridization experiment was carried out two to four times. TABLE 3. Comparison of the phenotypic properties of the six phenotypic groups ~~~~

Group

z;ii?i

Colony color

1 2 3 4 5 6

22 9 9 19 10 9

Bright yellow (7)" Dull yellow Cream white Cream white Cream white (1) Cream white (4)

Growth Deamination of Reduction of at 50"c phenylalanine N03 to No,

-

+ +

-

+

-

-

+

-

+

The numbers in parentheses are the numbers of strains having different colony colors.

VOL. 40, 1990

ALKALIPHILIC BACILLUS STRAINS

95

TABLE 4. Phenotypic properties of DNA group A strains and four Bacillus species type strainsa Reaction of strain: Property

DSM DSM Mg pB9/1 pB9/2 pB38(2) NCIB NCIB NCIB DSM M5 Wl3 DSM DSM DSM DSM DSM DSM DSM DSM llT 4UT 2526 10284 10327 10318 2528 2521 2522 2518 2519 26T 1297* 9T

Growth at: 10°C + 15°C + 26°C + 40°C + 45°C 47°C Growth in: NaCl(5%) NaCl (10%) NaCl(20%) Defined M medium at pH 9.7 Defined M medium at pH 7.0 Acid produced from: Mannose Melibiose Mannitol Raiiinose Sorbitol Xylose Glucose Hydrolysis of: Casein + + Gelatin Tween 20 + Tween 40 + Tween 60 Hippurate Pullulan + Cellulose Production of urease Reduction of NO, to NO2 Splitting of 4-MUGc + Presence of Dpmd

b

+

*

+ + +

+ + +

-

+ + +

+ + + + + + -

-

*

-

-

-

-

+ + +

-

-

+ +

-

-

-

-

+ + +

+

+

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

+ + +

+

-

+ + +

+ -

+ + + +

+ + + + + + + + + + + + + + + +

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

+ ++ + + + +-

+ + ++ ++ -+ + 2 + + + + + + + + + +

+ + + -

-

-

-

-

+

+

+

+

-

-

+

-

+

-

+

+ + + + + -

+

+

-

+

-

-

-

-

+

-

-

-

-

-

-

-

-

+ + - + - + + + + + + - - - + + * + + + - - + + - - - - - + + + + + + + + + - + -

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

+

+

+

+

The following reactions were positive for all strains: catalase activity, oxidase activity, hydrolysis of Tween 80, and utilization of citrate. The following reactions were negative for all strains: growth at 50 and 6"C, deamination of phenylalanine,acid production from arabinose, formation of crystalline dextrins, lysis by KOH, aminopeptidase activity, arginine dihydrolase activity, and lecithinase activity. +, Positive reaction; -, negative reaction; k, variable reaction. 4-MUCi, 4-Methylumbelliferone glucuronide. Presence of diaminopimelic acid (Dpm) in the cell walls.

RESULTS A total of 64 alkaliphilic Bacillus strains which were included in the studies of Gordon and Hyde (17), as well as 14 other strains, were investigated in this study (Table 1). The DNA base compositions of the 78 strains ranged from 34.0 to 43.9 mol% G+C (Table 2), with the strains grouped in the following three G+C content ranges: 17 strains (DNA group A) with G+C contents ranging from 34.0 to 37.5 mol%; 33 strains (DNA group B) with G+C contents ranging from 38.2 to 40.8 mol%; and 28 strains (DNA group C) with G+C contents ranging from 42.1 to 43.9 mol%. Relying on the profound physiological characterization of the strains by Gordon and Hyde (17), we chose some meaningful characteristics for our tentative differentiation under alkaline conditions. These characteristics were found by Gordon and Hyde to be present or absent in nearly 100% of the strains in certain groups under neutral conditions. Under alkaline conditions these characteristics divided the strains into six groups (Table 3). The six phenotypic groups

correlated with the three G+C content clusters (DNA groups A, B , and C) (Table 2). DNA group A was formed by phenotypic groups 5 and 6, DNA group B was formed by phenotypic groups 1 and 2, and DNA group C was formed by phenotypic groups 3 and 4. Except for 13 strains out of the set of 64, the six phenotypic groups confirmed the arrangement of Gordon and Hyde (17) into B.firrnus, B. lentus I, B . lentus 11, B . lentus 111, and intermediates. For strains BC3, PB19, DSM 2513, BC4, BB16, DSM2519, NCIB10318, DSM2528, PB9/1, and PB9/2 (strains of the intermediate group) the assignment to our phenotypic groups 1, 2, 4, 5, and 6 was supported by the G + C values (Table 2). Three other strains (strains NCIB10286, NCIB10326, and DSM1972) could be assigned to closely neighboring groups. As the type strain of B . alcalophilus, strain DSM 485, was found in DNA group A, this group was characterized phenotypically and genotypically in more detail. A marked diversity in phenotypic (Table 4) and genotypic (Table 5 ) characteristics was found within this DNA group. As deter-

96

INT. J. SYST.BACTERIOL.

FRITZE ET GL.

TABLE 5. DNA-DNA hybridization values among DNA group A strains and between DNA group A strains and Bacillus type strains having G+C contents in the DNA group A G+C content rangen % Hybridization with DNA from strain:

Strain

DSM 2526 DSM 485T M8 PB9/2 PB9/1 PB38(2) NCIB 10284 NCIB 10327 NCIB 10318 DSM 2528 M5 WN13 DSM 2521 DSM 2522 DSM 2518 DSM 2519

7iy 87 100

M8 -b

7 100

PB9/1

NCIB 10327

10318

-

-

-

-

16 32 92 100

48 19

25 22

36 27

WN13

DSM 2521

DSM 2522

DSM 2518

DSM 2519

DSM 26=

DSM 1297T

DSM ST

DSM

-

-

-

-

-

-

-

-

-

15 28

29 45

23 ND“

-

28 31

-

40 34

-

24 18

-

21 16

-

33 20

-

16 16

22 16

17 29

9

16

1 28

10

45 34 11 36 23 8 13 15 32 ND 29 25 19

M5

-

-

-

-

-

26 85 96 100

30

32

5

29 16 -

-

-

-

54 100

56 72 100

52 55 57 100

-

24 25 14 100

-

-

ND ND 5 ND 100

28 ND 19 28 ND 100

-

-

-

-

44 14 28 36 10 ND 46 100

16 4 6 30 27 ND 39 22 100

18 16 17 10 28 ND 11 18 14

23 38 22 18 15 ND 10 12 16

-

llT

16 11 11 10 17 8 34 ND 31 6 21

a The G+C contents of strains DSM 26T, DSM 1297T, DSM 9T, and DSM llTare 36.8, 38.0,36.4, and 35.4 mol%, respectively. For the G+C contents of the other strains, see Table 2. The G+C contents of the type strains were determined by Fahmy et al. (10). -, Value not determined because of high level of homology to other strains. ND, Not determined.

’

mined by DNA-DNA reassociation experiments, only the following three homology groups were distinguished: strains DSM 485T (T = type strain) and DSM 2526; strains PB9/1 and PB9/2; and strains PB38 (2), NCIB10327, and NCIBl0284. Three other strains (strains NCIB10318, M5, and DSM2528) were assigned loosely to the last group. All of the other strains showed no significant levels of DNA-DNA homology with each other. N o close relationship was revealed to any of the Bacillus type strains examined for comparison (Bacillus circulans DSM llT,Bacillus pantothenticus DSM 26T, B . lentus DSM 9T, and Bacillus marinus DSM 1297T)whose DNAs had similar G+C ratios (Table 5). DISCUSSION Since the study of Vedder no profound approach to the taxonomy of alkaliphilic Bacillus strains was made until the study of Gordon and Hyde in 1982 (17). These authors placed the strains loosely in the B. Jirmus-B. lentus complex. According to a broad range of physiological and biochemical properties, they found a spectrum of strains which at one end resembled the species B . firmus and at the other end resembled the species B . lentus, with several variations in between (B. lentus I, B. lentus 11, B . lentus 111, and intermediates). Gordon and Hyde assigned the type strain of B . alcalophilus strain DSM 485, to the intermediate group. Our investigations of G+C ratios revealed that the “complex” of alkaliphilic Bacillus strains can be divided into separate G+C content clusters. In addition, these clusters have been shown to correlate with certain physiological properties that are stable under neutral and alkaline conditions. Therefore, these properties may be used as valid characteristics for distinguishing among groups of alkaliphilic Bacillus strains. Through DNA-DNA reassociation studies with the strains of one of the G+C content groups (DNA group A), it has been shown that a variety of species may be found within all groups. As a consequence, the concept of strains forming a “continuum” (18) instead of separate clusters cannot be maintained. Of the 78 akaliphilic strains which we studied, only 2 could be assigned genetically to the species B . alcalophilus Vedder

1934 although isolation methods similar to those of Vedder were used for most other strains. These two were strains DSM 485T and DSM 2526, both of which were isolated by Vedder (37). This might indicate that B . alcalophilus is a rare species. That it is indeed a species on its own is shown by the low DNA-DNA homology values (