Contagious Equine Metritis: Characterization of Small and ...

Bull. Equine Res. Inst. No. 24, 23-32 (1987)

Contagious Equine Metritis: Characterization of Small and

Large Colonial Variants of Taylorella equigenitalis

lsolated from a Laboratory Strain

MaSanobu KAMADA, 1) Tohru ANZAI, 1) Takuml KANEMARU,1)

Ryuichi WADA 2) and Takeshi KUMANOMIDO2)

An experiment was carried out to clarify morphological, serological and bacteriological

characteristics of small and large colonial variants of Taylorella equigenitalis isolated from a laboratory strain. During subculture small colonial variants occurred spontaneously

by a reversion from small to large colonies in vitro. They were convex, smooth, friable and colorless with entire edges and grew slower on EGA than the large colonial variants

to reach a diameter of 0.7 mm or less at 7 days of incubation. On the contrary, the large

colonial variants were convex, smooth, viscid and grayish-white to light brown with entire

edges and grew relatively rapid and 3 to 5 mm in diameter at 7 days. Both colonial variants were visible on EGA at 3 days of incubation. When observed by the stereoscopic micro-

scope, the small and large colonial variants were classified into the transparent and opaque

phenotypes in opacity. No considerable difference was in biochemical characteristics be-tween both colonial variants. Besides, these variants had a close antigenic relationship

in the cross complement fixation and indirect hemagglutination tests.

Key words : Taylorella equigenitalis, Haemophilus equigenitalis, colonial variation, colonial

morphology, GEM

Introduction

Contagious equine metritis (CEM) is a contagious venereal disease of horses characterized by acute endometritis, show-ing clinical signs, such as vaginal exudate, and congestion and edema of the cervix and vagina. 1) It is caused by a micro-aerophilic, Gram-negative and nonfer-mentative coccobacillus or bacillus, which was newly classified as Taylorella equi-

genitalis. 2, 3)In general, changes in the colonial

morphology of bacteria are often indica-

tive of changes in antigenicity4-6) and in

virulence. 4, 7) Up to date, there have

been some reports on colonial variations

of T. equigenitalis in vitro ,8_10) but few

reports on changes in its antigenicity or

its virulence. In their previous paper, 11)

the authors described that small colonial

variants isolated from a laboratory strain

were much weaker in pathogenicity for

mares than large colonial variants. As a

result, it was indicated that changes in

colonial morphology were closely assoc-

iated with changes in virulence of T.

equigenitalis, as well as a number of other

pathogens.

Received for publication May 31, 1987.1) Epizootic Research Station, Equine Research Institute, Japan

Racing Association. 1400-4, Shiba, Kokubunji-machi, Shimotsuga-

gun, Tochigi 32904 JAPAN2) Equine Research Institute. Japan Racing Association.

- 23-

Kamada, M. et al.

The present study was carried out to

determine whether small colonial variants

of T. equigenitalis showing changes in

virulence for mares were associated with

changes in antigenicity. Its aims were to

classify these variants morphologically

according to the description of Hitchcock

et al.lo) and to know their bacteriological

characteristics. This paper deals with

various characteristics obtained from cul-

tural, morphological, biochemical and

serological tests.

Materials and Methods

Isolation history of bacteria and strains used. The To-89 strain 12) of T. equigenitalis was isolated from a mare with metritis in the Hidaka district of Hokkaido during the 1980 breeding season. After the 3rd sub-culture on Eugon chocolate agar (ECA) in an atmosphere containing 10% CO2, it revealed the presence of numerous colonial phenotypes. Small and large colonial variants, called the To-895 and To-89L strains, respectively, were picked up and then subcultured to ensure a single colony type. Small colonial var-iants were initially isolated at a low frequency and grew slower on ECA than any other colonial variant. They reached 0.7 mm or less in diameter at 7 days of incubation. On the contrary, large colo-nial variants 3 to 5 mm in diameter were isolated at a high frequency at 7 days of incubation. Both colonial variants were visible on ECA within 3 days after cultivation. After the 8th subculture, they were used for cultural, morphological,

biochemical and serological examinations.Cultural and morphological examinations.

Growth conditions were tested with ECA

(BBL) supplemented with glucose or

without. Moreover, X and V factor

requirements for growth were tested with

X and V discs (Nissui) on ECA and

brain heart infusion agar (Nissui). These

cultures were incubated at 37•Ž in an

atmosphere containing 10% CO2.

Colonial morphology was examined

with cultures on ECA and clear typing

medium, which was a modification of

the medium described by James and

Swanson. 13) Namely, each liter of med-

ium contained 3.75 g of trypticase pep-

tone (BBL), 7.5 g of thiotone peptone

(BBL), 4 g of K2HPO4 (anhydrous) 1 g

of KH2PO4 (crystalline), 5 g of NaGI, 2

g of soluble starch (BBL), and 10 g of

Noble agar (Difco). The medium was

autoclaved at 121 •Ž for 15 min. Ten

milliliters of Iso Vital X (BBL) and 2

ml of factors X and V containing 1 mg

of hemin (Sigma) per ml, 1 mg of L-

histidine per ml, and 0.3 mg of NAD

(Sigma) per ml were thoroughly mixed

into a cooled (50•Ž) medium before

pouring on plates. Plates had been stored

at 4•Ž until use. Cultures were incubated

in an atmosphere containing 10% CO2

at 37•Ž for 96 to 120 h. Morphological

observation on colonies was carried out

by a stereoscopic microscope equipped

with a substage reflecting mirror with a

diffusing surface and with coaxial vertical

illumination (SZH-151; Olympus Opti-

cal Co., Ltd., Tokyo).

Biochemical examination. The presence

of oxidase was determined by Kovac's

method. 14) The production of catalase

was assessed with growth removed from

an agar culture and placed in a drop of

3% H2O2 on a glass slide. Motility was

tested with growth by the hanging drop method,. The porphyrin test was carried

- 24-

Characterization of Colony Variants of CEMO

out by the method of Killian. 15) Most

of the biochemical tests were generally

performed with media described by Edw-ards and Ewing 16) according to the

previous reports. 2)

Fig. 1. Reversion from small to large colonial variants of the To-89S strain of T.

equigenitalis on Eugon chocolate agar at 8 days of incubation. Large colonies

looking like white powder

Cellular enzyme test. The presence of

97 specific cellular enzymes was tested

by means of API ZYM and additional

eight experimental ZYM galleries con-

sisting of ZYM II and AP 1-7 (API

System, S. A. ). Each strain was cultured

on an agar plate at 35•Ž for two days,

and the growth suspended in distilled

water (turbidity: No. 6 on the McFarland

scale). Two drops of the suspension were

incubated into each of the cupules in

the test galleries, which were incubated

at 35•Ž for 4 h. After the incubation, one

drop of each of API reagents A and B

was added to each of the cupules. The

results were read according to the direc-

tions of the manufacturer. Color grades

3+, 4+, and 5+ were interpreted as

positive.

Streptomycin sensitivity test. Two strains

of T. equigenitalis were examined for

susceptibility to streptomycin by the disk

diffusion method 17) with commercial discs

(50 ƒÊg per disc) produced by the Showa

Yakuhin Kanko Co., Ltd., Tokyo. Before

incubation at 37•Ž for 48 to 72 h, discs

were placed on the agar upon which a

lawn of bacteria had been streaked by a

bending glass rod. Resistance was scored

on the basis of inhibition of colonial

growth within a 5-mm radius of the discs.

Serological tests. After incubated on

ECA for 4 days, the To-89S and To-89L

strains were suspended in PBS and

centrifuged at 2000 rpm for 10 min.

- 25-

Kamada, M. et al.

After washed three times with PBS, the

sediment was suspended in PBS contain-

ing 0.2% formalin solution and stored at

4•Ž until use. It was used as an inocula

for immunization or as an antigen for

the complement fixation (CF) and in-direct hemagglutination (IHA) tests.

Fig. 2. Transparent colony of the To-89S

strain grown on clear typing medium,

as observed under the stereoscopic

microscope with reflection light

Fig. 3. Transparent colonies of the To-89S

strain grown on Eugon chocolate agar,

as observed under the stereoscopic mi-

croscope with coaxial vertical illumina-

tion

Fig. 4. Opaque colony of the To-89L strain

grown on clear typing medium, as ob-

served under the stereoscopic micros-

cope with reflection light

Fig. 5. Opaque colonies of the To-89L strain

grown on Eugon chocolate agar, as

observed under the stereoscopic micros-

cope with coaxial vertical illumination

Rabbits were injected intramuscularly five times with the above-mentioned inocula ranging from 109 to 1010 colony

- 26-

Characterization of Colony Variants of CEMO

forming units per ml at 3- to 4-day

intervals. They bled 2 weeks after the

last injection. Before the CF and IHA

tests, they were inactivated at 56•Ž for

30 min.

Table 1. Reversion of the To-89 strain of

T. equigenitalis from small to large

colonies in vitro during subculture

Time in days of incubation

Remarks. * Small and large colonies.** Small colonies alone .

The antigen was adjusted and both

tests were performed in the same manner

as described by Kamada et al. 18) and

Sahu et al. 19) The CF titer was expressed

as the reciprocal of the highest serum

dilution that had reduced hemolysis to

50% or less. The IHA titer was expressed

as the reciprocal of the highest serum

dilution that had agglutinated sheep

erythrocytes to 50% or over.

Results

Cultural and morphological characteristics. During subculture the To-89S strain of T. equigenitalis showed a reversion from small to large colonies in vitro (Table 1). Namely, after the 5th subculture, large colonies grew spontaneously on ECA at a low frequency after 6 days of incubation

(Fig. 1). When growth of the organism on ECA without glucose was comparable

to that on ECA with glucose, it had a

tendency to show a reversion from small

to large colonies at a very low frequency

during the prolonged cultivation.

Table 2. Colony size of the To-89S and To-

89L strains of T. equigenitalis at the

8th passage level grown on Eugon

chocolate agar

Time in days of incubation

Remark. The figure indicates the colony size (mm in diameter).

Table 3. Characteristics of the To-89S and

To-89L strains of T. equigenitalis

Remarks.* Negative . ** Positive. *** Arabinose, fructose,

glucose, inulin, lactose, maltose, mannitol, mannose,

raffinose, rhamnose, saccharose, salicin, sorbitol,

starch, trehalose, and xylose.

The To-89S strain grew slower on

ECA than large colonial variants to

reach 0.7 mm or less in diameter at 7

days of incubation, whereas the To-89L

- 27-

Kamada, M, et al.

strain grew relatively rapid to reach 3 to 5 mm in diameter (Table 2). Both colonial variants were visible on ECA within 3 days after cultivation. When observed by the stereoscopic microscope, small colonial variants were transparent with only a few translucent foci (Figs. 2 and 3) and large colonial variants opaque

(Figs. 4 and 5). The To-89S strain was classified as a transparent phenotype in opacity and the To-89L strain as an opaque phenotype.

Table 4. Enzymes of the To-89S and To-89L

strains of T, equigenitalis detected

commonly in the cellular enzyme test.

Table 5. Antigenic relationship between the To-89S and To-89L strains of T. equigenitalis by cross complement fi-xation (CF) and indirect hemagglu-tination (IHA) tests

Remarks.* The CF titer is expressed as the reciprocal of

the highest serum dilution that had reduced hemo-lysis to 50% or less. ** The IHA titer is exp-ressed as the reciprocal of the highest serum dilu-tion that had agglutinated sheep erythrocytes to 50% or over.

Biological and biochemical characteristics.

The To-89S and To-89L strains were

Gram-negative, nonmotile and capsulated

coccobacilli or short rods. The To-89L strain was convex, smooth, viscid and

grayish-white to light brown with entire edges, whereas the To-89S strain was convex, smooth, friable and colorless with entire edges.

In the conventional biochemical tests, no significant difference was observed in biochemical characteristics between both strains. Namely, they were positive only in catalase, oxidase and phosphatase tests, being negative in most of the tests

(Table 3). They were also positive for the porphyrin test and resistant to strepto-mycin. When cellular enzymes were ex-amined, many enzymes, such as phos-

phoamidase, glutamate transpeptidase, and arylaminidases of amino acids and oligopeptidases, were detected (Table 7).

Antigenic characteristics. In the CF and IHA tests, there have been few significant differences in antigenicity between the To-89S and To-89L strains of T, equi-

genitalis (Table 5). Namely, they were found to have a close antigenic relation-ship in both tests.

Discussion

Hitchcock et al. 10) described that all the colonial variants of T. equigenitalis

grown on clear typing medium could be classified into one of three groups based on opacity phenotype: transparent, o-

paque, and intermediate. In their report, transparent variants with a few opaque foci were classified as a transparent

phenotype. The small colonial variants used in the present experiment were transparent with only a few translucent foci on ECA and clear typing medium when observed by the stereoscopic micro-scope equipped with a substage reflecting

- 28-

Characterization of Colony Variants of CEMO

mirror with a diffusing surface and

equipped with coaxial vertical illumina-

tion. So the To-895 strain of T. equigenitalis

should be classified into the transparent

phenotype described by those authors. 10) On the other hand, the large colonial

variants used in this study were opaque

under the stereoscopic microscope with

reflection light. When observed by the

stereoscopic microscope with coaxial ver-

tical illumination, they were opaque with

a few translucent foci. According to

Hitchcock et al., 10) the opaque pheno-

type was expressed as opaque colonies in

opacity under the stereoscopic microscope

with reflection light or transmitted light.

Therefore, the To-89L strain should be

classified into the opaque phenotype de-

scribed by them. 10)

In general, variation of bacteria show-

ing the opaque or transparent appearance

of the colony can be induced by the

presence or absence of a capsule, 10) outer membrane protein 5, 20, 21) and other sur-

face appendages.7, 22, 23) These changes in

morphological properties often reflect

changes in virulence, 7, 21) and antigenic

properties. 20) On the other hand, morph-ological observations of T. equigenitalis

showed the presence of a fibrillar layer

on the outer membrane of various variants

or the presence of intracellular strands

comprised in a bleb-forming outer mem-

brane like pill of Neisseria gonorrhoeae. 10)

In another previous paper, 11) the authors

indicated that changes in colonial mor-

phology were closely associated with changes in virulence in T. equigenitalis, as

well as in a number of pathoens.

Besides, it was suggested that small

colonial variants might probably be close-

ly associated with changes in such adhesive

properties of bacteria as pili or outer membrane proteins because of rapid

elimination of them from the uterus and

cervix. The results in this study, however,

indicated that small and large colonial

variants showing an apparent difference

in virulence for mares had a close anti-

genic relationship, since no considerable difference was in serological and bio-

chemical characteristics between them.

Therefore, it is thought that changes in

the colonial morphology may probably

reflect changes in antigenic quantity, as

well as in antigenic quality. In future, it

will be necessary to carry out a molecular

microbiological investigation of colonial

variations of T. equigenitalis. Such inves-

tigation has been performed in the

studies of N, gonorrhoeae. 5, 6, 20, 21) It will

bring about success to the antigenic

analysis of colonial variations of T. equi-

genitalis.Sahu and Weber 8) and Sahu et al. 9)

described six coloniall variants of T.

equigenitalis: five phenotypes isolated from

clinical specimens and observed at 14

days of incubation or later, and a sixth

slow-growing phenotype grown in ECA

plates at 5 days of incubation or later. The tiny colonies of the six phenotypes

were 0.25 mm or less in diameter after 10

days of incubation, showing a reversion

from tiny to large colonies in vivo. At

present, it is unknown whether their tiny colonies exhibit opacity variation when

grown on ECA or clear typing medium or whether large colonies occur when

the tiny ones are grown on ECA during

a prolonged cultivation.

On the other hand, the To-895 strain

of T. equigenitalis, transparent phenotype

in opacity, presented the following char-

- 29-

Kamada, M. et al.

acteristics. Namely, small colonial var-

iants were convex, smooth, friable and

colorless with entire edges, grew slower

on ECA than the large colonial variants

to reach 0.7 mm or less in diameter at 7

days of incubation, and were visible on

ECA within 3 days after cultivation.

The bacteria exhibited a reversion spon-

taneously from small to large colonies in

vitro, but not in vivo. It is possible, how-

ever, that the small colonial variants

used in this study may give rise to large

colonies in vivo when inoculated at a

rather large inoculum size into experi-

mental mares because of the occurrence

of the above-mentioned reversion in vitro.

Therefore, it is of great interest in a

future study to determine whether the

two small colonial variants described

above are classified into the same pheno-

type, because they resemble in various

characteristics, except a slow-growing

phenotype.

Observation on the colonial morpho-logy of T. equigenitalis grown on ECA revealed the presence of small colonial variants in 15 of 33 primary isolations from mares with GEM (unpublished data). To date, pure isolates of small colonial variants have been obtained directly from 3 cases in the field. When these isolates belonging to the transparent

phenotype were subcultured on ECA, large colonial variants grew within 8 days after cultivation. These results indicate that colonial variations of T. equigenitalis exhibiting changes in virulence occur in vitro and in vivo and that they have been spread among mares in the field. Since GEM broke out in Japan for the first

time during the 1980 breeding season, 12)

it is still difficult to detect carrier mares

of the disease clinically, because there is

an increase in number of naturally in-

fected mares presenting mild or no clinical

signs of the disease. The occurrence of

colonial variations exhibiting changes in

virulence might have been responsible

for the gradual decrease in number of

naturally infected horses showing typical

clinical signs of GEM in the field.

It depends upon the development of a

diagnostic method available for the de-

tection of carrier mares that any measure

taken for the prevention of GEM finally

leads to a success in the control or erad-

ication of the disease in Japan and other

counties. Usually, the final diagnosis of

the disease is carried out by the isolation

of etiological bacteria from the genital

tract of mares and stallions. The small

colonial variants have no characteristics

in colonial morphology, except that they

are transparent and of small phenotype.

It is very difficult to differentiate them

from other bacteria showing almost the

same colonial appearance at the time of

primary isolation. In addition, the pre-sence of such slow-growing strains of T.

equigenitalis as described in the previous

reports17, 24) makes bacterial differentiation

rather difficult. Therefore, the serological

reactions, including the CF and IHA

tests and enzyme linked immunosorbent

assay, may be available as additional

means for the diagnosis of the disease.

The results of the present cross CF and

IHA tests indicated a close antigenic

relationship between small and large

colonial variants of T. equigenitalis.

- 30-

Characterization of Colony Variants of CEMO

Acknowledgments

The authors are grateful to Drs. T. Hirasawa, A. Amada and Y. Fukunaga, of their institute, for val-uable advice given to the present work, and to Miss Inoue for technical assistance. Thanks are also due to the members of the Hidaka Agricultural Mutual Aid Association who extended a helping hand in collect-ing specimens, and to Dr. K. Hirose and the staff members of the Hidaka Livestock Hygiene Service Center, Prefecture of Hokkaido, for cooperation for this work.

Literature Cited

1) Crowhurst, R. C. 1977. Genital infection in

mares. Vet. Rec. 100, 476.

2) Sugimoto, C., Isayama, Y., Sakazaki, R. and

Kuramochi, S. 1983. Transfer of Haemophilus equi-

genitalis Taylor et al. 1978 to the genus Taylorella

gen. nov. as Taylorella equigenitalis com. nov.

Current Microbiol. 9, 155-162.

3) International Committee on Systematic Bacte-

riology. 1984. Validation of the publication of new

names and new combinations previously effectively

published outside the IJSB. Int. J. Syst. Bacteriol.

34, 503-504.

4) Cooper, M. L., Keller, H. M. and Watlers,

E. W. 1957. Microscopic characteristics of col-

onies of Shigella flexneria 2a and 2b and their anti-

genic composition, mouse virulence and immuno-

genicity. J. Imrnunol. 78, 160-171.

5) Walstad, D. L., Guyman, L. F, and Sparling,

P. F. 1977. Altered outer membrane protein in

different colonial types of Neisseria gonorrhoeae. J.

Bacteriol. 129, 1623-1627.

6) Swanson, J. 1978. Studies on gonococcus infec-

tion. XIV. Cell wall protein differences among

color/opacity colony variants of Neisseria gonor-

rhoeae. Infect. Immun. 21, 292-302.

7) Kellogg, E. S., Cohen, I. R., Non, L. C.,

Schroeter, A. L, and Reising, G. 1968. Neisseria

gonorrhoeae. ‡U. Colonial variation and pathogen-

icily during 35 months in vitro. J. Bacteriol. 96,

596-605.

8) Sahu, S. P. and Weber, S. 1982. Contagious

equine metritis: Effect of intraauterine inoculation

of tiny colony forms in pony mares. Vet. Rec. 110,

250-251.

9) Sahu, S. P., Wool, S. and Breese, S. S. 1982.

Observation on the morphology of contagious

equine metritis bacterial colonies isolated from

infected pony mares. Am. J. Vet. Res. 43, 796-

800.

10) Hitchcock, P. J., Brown, T. M., Corwin, D.,

Hayes, S. F., Olszewski, A. and Todd, W. 1985.

Morphology of three strains of contagious equine

metritis organism. Infect. Immun. 48, 94-108.

11) Kanemaru, T., Kamada, M., Anzai, T, and

Kumanomido, T. 1987. Contagious equine me-

tritis: The pathogenicity for mares of small and

large colonial variants of Taylorella equigenitalis

isolated from a laboratory strain. 5th Inter. Conf.

Equine Inf. Dis. (accepted for publication).

12) Kamada, M., Akiyama, Y., Oda, T, and Fuku-

zawa, Y. 1981. Contagious equine metritis: Iso-

lation of Haemophilus equigenitalis from horses with

endometritis in Japan. Jpn. J. Vet. Sci. 43, 565-

568.

13) James, J. F, and Swanson, J. 1978. Studies on

Gonococcus infection. X‡V. Occurrence of color/

opacity colonial variants in clinical cultures.

Infect. Immun. 19, 332-340.

14) Kovacs, N. 1956. Identification of Pseudomonas

pyocyanae by the oxidase reaction. Nature (London)

178, 703.

15) Killian, M. 1974. A rapid method for the

differentiation of Haemophilus strains : The por-

phyrin test. Acta Pathol. Microbiol. Scand. 82, 835-

842.

16) Edwards, P. R. and Ewing, W. H. 1972. Iden-

tification of Enterobacteriaceae. Minneapolis: Bur-

gess.

17) Kamada, M., Kumanomido, T., Anzai, T.,

Kanemaru, T., Senba, H. and Ohishi, H. 1986.

Isolation and drug susceptibility of streptomycin

sensitive Taylorella equigenitalis from mares with

metritis and infertility in Japan. Bull. Equine Res.

Inst. 23, 55-61.

18) Kamada, M., Oda, T., Fukuzawa, Y., Ohishi,

H., Wada, R., Fukunaga, Y. and Kumanomido,

T. 1983. Contagious equine metritis : A survey on

complement fixing antibod against Haemophilus

equigenitalis in light horses in Japan. Bull. Equine

Res. Inst. 20, 119-125.

19) Sahu, S. P., Rommel, F. A., Fales, W. H.,

Hamdy, F. M., Swerczek, T. W., Youngquist,

R. S. and Bryans, J. T. 1983. Evaluation of var-

ious serotests to detect antibodies in ponies and

horses infected with contagious equine metritis

bactetia. Am. J. Vet. Res. 44, 1405-1409.

20) Swanson, J. 1978. Colony opacity and protein ‡U

. Compositions of gonococci. Infect. Immun. 37,

359-368.

21) Lambden, P. R., Heckels, J. E., James, L. T.

and Watt, P. J. 1979. Variation in surface protein

composition associated with virulence properties

in opacity types of Neisseria gonorrhoeae. J. Gen.

Microbiol. 114, 305-312.

22) Swanson, J. and Barrera, O. 1983. Gonococcal

pilus subunit size heterogeneity correlates with

transitions in colony piliation phenotype, not with

changes in colony opacity. J. Exp. Med. 143, 886-

906.

23) Salit, I. E., Blake, M, and Gotschlich, E. C.

1980. Intra-strain heterogeneity of gonococcul pili

is related to opacity colony variance. J. Exp.

-31-

Kamada, M. et al.

Med. 151, 716-725.24) Ward, J., Hourigan, M., McGuirk, j. and

Gogarty, A. 1984. Incubation times for primary

isolation of the contagious equine metritis organism.

Vet. Rec. 114, 298.

馬伝染性 子宮炎:保 存株か ら分離 されたTaylorella equigenitalisの 小型 および大型集落 変異株

の性 状:鎌 田正信1)・安斉 了1)・兼丸卓 美1)・和 田隆一・2)・熊埜 御堂 毅2)(1)日本 中央 競馬会競走馬総

合研究所栃 木支所, 2)日本中央競馬会競走馬総合研究所)一 保存株 か ら分離 されたT. equigenitalzs

の小型お よび大型集落変異株について形態学的,血 清学的,細 菌学的性状を調べた.小 型集落変異株 は

継代 中にご大 型集落を散 発的に産 生 した.こ の現象は ブ ドウ糖無添加培地 よ り添加培地で高頻 度で認め

られ る傾 向を示 した.小 型集落変異株は隆起 した平滑 な粘稠性 の少 ない無 色透 明な円形 集落 を形成 し,

発育速度は大型集落変異株にご比べて遅 く,培 養7日 目で直径0.7mm以 下 であった.一 方,大 型集落

変異 株は隆起 した平滑な粘稠性の強い灰 白色か ら淡褐色不透 明な円形集落 を形成 し,発 育速度が早 く,

培 養7日 目で直径3か ら5mmで あ った.実 体顕微鏡に よる集落形態 の観 察で,小 型 お よび大 型集落

変異株は透明度変異においてそれぞれ透 明,不 透 明変異株 と分類 された.な お,今 回の実験 において

両株間での生化学性状 の差は認め られず,交 叉補体結合 お よび間接血球凝集 試験 において両株は類似

した抗原性を有 す ることが明 らか とな った.

- 32-