Microbiol. Immunol. Vol. 36 (7), 671-681, 1992
Proposal of a New Scheme for the Serological Typing of Enterococcus faecalis Strains Shizue
MAEKAWA,*,1 Migaku YOSHIOKA,2 and Yoshiaki KUMAMOTO2 1Department of Microbiology,and 2Departmentof Urology,Sapporo
Medical College,Sapporo, Hokkaido 060, Japan (Accepted for publication, April 17, 1992)
Abstract No systematic study on serotyping of Enterococcus faecalis has been reported since 1964 when M.E. Sharpe conducted serotyping of group D streptococcus in U.K. So, we attempted to re-evaluate serotyping of E. faecalis. For this purpose, we received 42 Sharpe's strains and first examined for their biochemical characteristics as E. faecalis. Only 9 of the 42 strains were identified as E. faecalis. We raised rabbit antisera against a large number of E. faecalis strains, including the 9 Sharpe's strains, 2 strains obtained from CDC in U.S.A. and 36 strains isolated from patients hospitalized in different cities of Japan. From the results of cross-agglutination tests and absorption tests performed on these antisera using a large number of E. faecalis strains, we were able to classify 21 distinct serotype strains and to prepare 21 monospecific typing antisera by absorption of the antisera to the type strains with appropriate cross-agglutinating strains. When 832 E. faecalis strains were serotyped with the 21 typing antisera, 638 strains (76.7%) were typable. Thus, we propose a provisional scheme of 21 distinct serovars in E. faecalis.
Frequency of isolation of Enterococcusfaecalis (9) pathogens from patients with urinary tract infections is increasing in Japan as a result of frequent use of thirdgeneration cephalosporins for the treatment (5). A paper from the 'United States reported that infants hospitalized in a neonatal intensive care unit had an increased incidence of neonatal sepsis caused by E. faecalis (6), and a report on nosocomial infections in general surgery clinics and intensive care units in Germany indicated that E. .faecalis is an important nosocomial pathogen (3). In 1964, M.E.
Sharpe
reported
serological
type strains of Streptococcus.faecalis
(E. faecalis) and compared them to those of Sharpe and Shattock (1952) by precipitin reactions and reciprocal absorption tests. By this study, Sharpe classified S. faecalis into 11 different serological types (10). Since this report, no further systematic study on the serological typing of E. faecalis has been reported. Thus, we attempted to re-evaluate the serological typing of E. faecalis, and for this purpose we received 42 Sharpe's serotype strains, of which 23 had been deposited in the National Collection of Type Cultures (NCTC) of U.K. as distinct type strains, from the National Institute for Research in Dairying (NIRD) in U.K. We found 671
672
S. MAEKAWA
ET AL
however, that only 9 of these 42 strains were identified as E. faecalis. Therefore, we raised rabbit antisera to a large number of E. faecalis strains, including the 9 strains from NIRD, and performed cross-agglutination and absorption tests on each antiserum by using a large number of E. faecalis strains. From the results of these studies, we were able to identify 21 serotype strains and to prepare 21 monospecific typing antisera by absorption of the antisera to the 21 strains with appropriate crossagglutinating strains. Based on the results of the present study, we proposed to classify E. faecalis into 21 serovars. MATERIALS
AND
METHODS
Bacterial strains. E. faecalis SS498, a reference strain in Centers for Disease Control (CDC), and E. faecalis 1824-73, isolated from a wound of a patient, were received from Dr. R.R. Facklam, CDC, Atlanta, U.S.A. Enterococcussp. Sharpe's serotype strains, NIRD 1599-1610, 1616-1633, 1637-1648, 42 strains in all, were obtained from NIRD, Shinfield, Berkshire, U.K. Twenty-three of the 42 strains had been deposited as Sharpe's serotypes 1-16 and 18-24 in NCTC, Central Public Health Laboratory, London, U.K., as shown in Table 1. Japanese strains used for raising antisera were isolates from urine of patients hospitalized in different cities in Japan. Culture media. Todd-Hewitt broth (THB, Difco) was used for liquid culture and sheep blood agar plates were used as solid culture medium. Identificationof E. faecalis. Identification of E. faecalis based on 20 biochemical reactions was done by using an Api STREP 20 kit produced by API SYSTEM S.A., France (4). Preparation of rabbit antisera against E. faecalis. Antisera against various strains of E. faecalis were raised in rabbits by i.v. injection with formalin-killed bacteria. The procedures for preparation of formalin-killed bacteria, immunization of rabbits and preparation of monospecific typing sera by absorption of the antisera with cross-agglutinating strains followed essentially those prescribed for T-typing agglutination of Streptococcus pyogenes(7, 11). Agglutinationtest. On individual slide glass marked with circles, 5 pl of appropriately diluted each typing serum was placed and the bacterial suspension to be tested was added. The slide glass was gently rotated and occurrence of bacterial agglutination was checked visually for 1 min. A drop of bacterial suspension mixed with a drop of saline served as the control. RESULTS
Re-Identificationof Sharpe's SerotypingStrains by BiochemicalCharacteristicsfor E. faecalis The 42 Sharpe's serotype strains given to our laboratory by NIRD were first examined for the characteristics of E. faecalis by 20 biochemical reactions in Api STREP 20 kit. As shown in Tables 1 and 2, only 9 of the 42 strains were identified as E. faecalis. The other strains were identified as E. faecium (21 strains), E. galli-
SEROVARS
Table
1. group
Strains
OF
ENTEROCOCCUS
deposited
D streptococcus
in NIRD by
M.E.
673
FAECALIS
as serotype Sharpe
in
strains 1962
of
674
S. MAEKAWA
Table
2.
ET AL
Identification of Sharpe's serotyping strains by biochemical reactions determined in Api STREP 20 kit
Species in Enterococcus of NIRD 1605, 1641, and 1647, and species in Streptococcus of NIRD 1623 could not be identified.
narum (2 strains), E. hirae (2 strains), E. durans (1 strain), Enterococcussp. (3 strains), S. equinus(3 strains) and Streptococcus sp. (1 strain). These results were later verified by NIRD. Establishmentof MonospecificAnti-E. faecalis Typing Sera and E. faecalis SerotypeStrains We first raised rabbit antisera against two E. faecalis strains obtained from CDC and 36 E. faecalis strains isolated from patients hospitalized in different cities in Japan. Based on the results of cross-agglutination tests and reciprocal absorption tests performed on these antisera using the immunizing strains, and the results of crossagglutination tests and absorption tests on the antisera using 463 K faecalis strains
SEROVARS
OF ENTEROCOCCUS
FAECALIS
675
isolated from the other patients in Japan, we were able to classify 15 serotype strains, and to prepare 15 monospecific typing antisera by absorption of the antisera to the type strains with appropriate cross-agglutinating strains. Agglutination titers of each of these antisera against the 15 strains before and after absorption with the cross-agglutinating strains are shown in the part of Table 3 circumscribed by dotted lines. Next, 9 Sharpe's strains that were identified as E. faecalis by the present study, i.e., NCTC 8727, 8729, 8730, 8731, 8732, 8734, 8735, 8744, and 8745, were subjected to serotyping by the 15 monospecific typing antisera. NCTC 8729, 8730, 8731, 8734, 8744, and 8745 did not show agglutination reaction to any of the 15 monospecific typing antisera, whereas NCTC 8727, 8732, and 8735 showed agglutination in anti-SS498, anti-Otaru-104, and anti-Sapporo-603 monospecific typing sera, respectively. We prepared rabbit antisera against NCTC 8729, 8730, 8731, 8734, 8744, and 8745, and performed cross-agglutination tests and absorption tests on these antisera using the 6 immunizing strains and the 15 serotype strains. From the results of these experiments, we were able to classify these 6 NCTC strains as distinct serotype strains and to prepare monospecific typing antisera by absorbing the antisera with appropriate cross-agglutinating strains. Agglutination titers of these 6 antisera against the 6 immunizing strains and the 15 serotype strains before and after absorption with appropriate cross-agglutinating strains are shown in Table 3. To verify serotype of NCTC 8727, 8732, and 8735, antisera to these 3 strains were prepared. Considering the results of the agglutination reactions of these strains to the 15 monospecific typing sera, reciprocal absorption tests with the immunizing strains were performed in combination of anti-NCTC 8727 and anti-SS498, anti-NCTC 8732, and anti-Otaru-104, and anti-NCTC 8735 and anti-Sapporo-603 antisera, respectively. As shown in Table 4, NCTC 8727, 8732, and 8735 were verified to belong to serotypes of SS498, Otaru-104, and Sapporo-603, respectively. Proposalof 21 Distinct Serovarsof E. faecalis Based on the results described in the preceding section, we propose 21 distinct serovars of E. faecalis. The serovars are numbered from 1 to 21 provisionally, and type strain of each serovar is listed in Table 5. These 21 E. faecalis serotype strains, which consist of 13 strains isolated from patients in different cities in Japan, 2 strains obtained from CDC and 6 strains obtained from NIRD as NCTC Sharpe's serotype strains, were all identified as E. faecalis by Api STREP 20 kit (data not shown). DISCUSSION
In 1962, 42 group D streptococcus (S.faecalis) strains were deposited in NIRD as Sharpe's serotyping strains, and 23 of these strains were deposited in NCTC as serotype strains. Since then, no further systematic study of serological typing of E. faecalis (S. faecalis was transferred to the genus Enterococcusas E. faecalis in 1984 (9)) has been reported. Recently, the importance of E. faecalis as a pathogen in
676
S. MAEKAWA
Table 3.
Agglutinin titers of antiserum prepared
ET AL
to each E. faecalis serotype strain against all
SEROVARS
OF
ENTEROCOCCUS
FAECALIS
of the serotype strains before and after absorption with appropriate cross-agglutinating strain(s)
677
678
Table
S. MAEKAWA 3
ET AL
continued.
Table 4.
Reciprocal absorption tests on the antisera in combination of E. faecalis strains SS498 and NCTC 8727, Sapporo-603 and NCTC 8735, and Otaru-104 and NCTC 8732
SEROVARS
Table 5.
OF
ENTEROCOCCUS
FAECALIS
Serovars of E. faecalis and type strain of each serovar to propose
679
680
S. MAEKAWA
ET AL
urinary tract infections (5), neonatal sepsis (6) and nosocomial infections (3) has been indicated. Therefore, we conducted the present study to re-evaluate the serological typing of E. faecalis. We first prepared rabbit antisera against two E. faecalis strains obtained from CDC and 36 E. faecalis strains isolated from patients hospitalized in different cities of Japan, considering that predominant serotype may somehow differ with each other in different districts or countries. From the results of absorption tests and cross-agglutination tests performed on these antisera, we were able to classify 15 distinct serotype strains, and thus we were able to prepare 15 distinct monospecific typing antisera by absorption of the antisera to the 15 serotype strains with appropriate cross-agglutinating strains. For the present study, we received 42 Sharpe's serotype strains from NIRD, and we first examined for their biochemical characteristics of E. faecalis in Api STREP 20 kit as recommended by several investigators (1, 2, 13), because the characteristics of these strains as E. faecalis have not been determined since they were deposited in NIRD and NCTC. We found that only 9 strains of the 42 strains were identified as E. faecalis, and that even 23 strains deposited in NCTC as Sharpe's serotypes 1-16, 18-24 (corresponding to Group D Sharpe and Shattock serotypes NCTC 8727, 8796, 8729-8749 in the catalogues of the National Collections of Type Cultures and Pathogenic Fungi, 1989) included strains of 4 different species of genus Enterococcusand strains of genus Streptococcus. These results, which were later verified by NIRD, might stem from the historical fact that Sharpe identified her strains as group D streptococcus but not as E. faecalis defined at present. The results of serotyping of the 9 Sharpe's strains that were identified as E. faecalis by the present study, with the 15 monospecific typing antisera, and of cross-agglutination tests and absorption tests performed on the antisera raised against the 9 strains indicated that 6 of the 9 strains were distinct in serotype from the already established 15 serotype strains, whereas the other 3 strains belonged to either one of the already established 15 serotypes. Thus, we propose a scheme of 21 distinct serovars of E. faecalis, provisionally numbered from 1 to 21, and type strain of each serovar as shown in Table 5. Since 6 NCTC strains included in our proposed serovars as type strains had been registered by M.E. Sharpe as serotype strains in 1962, Sharpe's serotype numbers for the strains may have priority. However, none of the 23 Sharpe's NCTC strains were exactly identified as E. faecalis as defined at present, so our provisionally numbered serovars of E. faecalis will be legitimate. In our scheme, serovar numbers were given to the type strains as follows: serovar 1 to SS498, a reference strain of E. faecalis in CDC; serovars 2 to 14 to the strains isolated in Japan; serovar 15 to the strain isolated in U.S.A. ; and serovars 16 to 21 to the NCTC Sharpe's strains. The serovar numbers given to NCTC 8744 and 8745 in our scheme, i.e., serovars 19 and 20, were the same as those in Sharpe's serotypes, whereas the other 4 NCTC strains were given serovar numbers different from those in Sharpe's serotypes. To evaluate the proposed serotyping of E. faecalis practically, 832 E. faecalis strains which had been isolated from 463 patients in Japan and 369 patients in
SEROVARS
OF
ENTEROCOCCUS
FAECALIS
681
U.S.A. were serotyped with the 21 monospecific typing antisera of our scheme. Among the 832 strains, 638 strains (76.7%) were typable as either one of 15 serovars. Serovars 1, 2, 4 and 7 were major serovars and occupied 72% of the typable strains, whereas serovars 16 to 21 were less than 2% (data not shown). We think that serological typing of E. faecalis with these 21 monospecific typing antisera should contribute to epidemiological study of E. faecalis infections like M-, or T-typing of S. pyogenesand serotyping of S. agalactiaein epidemiological studies (8, 12). The authors wish to thank Dr. R.R. Facklam, Division of Respiratory Disease, Center for Infectious Disease, Centers for Disease Control, U.S.A. and Dr. G. Colman, Central Public Health Laboratory, U.K. for permitting us to analyze serovars of E. faecalis strains from their laboratories. We are also deeply indebted to them for providing the E. faecalis clinical isolates stored in their laboratories. REFERENCES
1) 2) 3) 4) 5)
6)
7)
8) 9)
10) 11)
12) 13)
Bosley, G.B., Facklam, R.R., and Grossman, D. 1983. Rapid identification on enterococci. J. Clin. Microbiol. 18: 1275-1277. Facklam, R.R., and Collins, M.D. 1989. Identification of Enterococcusspecies isolated from human infections by a conventional test scheme. J. Clin. Microbiol. 27: 731-734. Holzheimer, R.G., Quoika, P., Patzmann, D., and Fussle, R. 1990. Nosocomial infections in general surgery: surveillance report from a German University Clinic. Infection 18: 219-225. Jorgensen, J.H., Crawford, S.A., and Alexander, G.A. 1983. Rapid identification of group D streptococci with the API 20S system. J. Clin. Microbiol. 17: 1096-1098. Kosakai, N., Kumamoto, Y., Hirose, Y., Tanaka, N., Hikichi, Y., Shigeta, S., Shiraiwa, Y., Yoshida, H., Ogata, M., Tazaki, H., Ili, H., Uchida, H., Kobayashi, Y., Matsuda, S., Kitagawa, R., Fujita, K., Hayashi, Y., Oguri, Y., Furusawa, T., Takeuchi, Y., Moriyama, H., Shibata, K., Yonezu, S., Takaha, M., Matsumiya, K., Tanaka, M., Yamaguchi, K., Takeda, K., Mochida, C., and Igari, J. 1990. Comparative studies on activities of antimicrobial agents against causative organisms isolated from urinary tract infections (1987). I. Susceptibility distribution. Jpn. J. Antibiot. 43: 919-953. Luginbuhl, L.M., Rotbart, H.A., Facklam, R.R., Roe, M.H., and Elliot, J.A. 1987. Neonatal enterococcal sepsis: case-control study and description of an outbreak. Pediatr. Infect. Dis. J. 6: 1022-1030. Maekawa, S., Fukuda, K., Yamauchi, T., Yamaguchi, T., Takahashi, K., and Sugawa, K. 1981. Follow-up study of pharyngeal carriers of beta-hemolytic streptococci among school children in Sapporo city during a period of 2 years and 5 months. J. Clin. Microbiol. 13: 1017-1022. Murai, T., Inazumi, Y., Ooguro, H., and Ishiguro, K. 1980. Serotype of group B streptococci isolated in Japan in recent year. Jpn. J. Infect. Dis. 54: 22-28. Schleifer, K. H., and Kilpper-Baltz, R. 1984. Transfer of Streptococcusfaecalis and Streptococcus faecium to the genus Enterococcusnom. rev. as Enterococcusfaecalis comb. nov. and Enterococcus faecium comb. nov. Int. J. Syst. Bacteriol. 34: 31-34. Sharpe, M.E. 1964. Serological types of Streptococcusfaecalis and its varieties and their cell wall type antigen. J. Gen. Microbiol. 36: 151-160. Takizawa, K., Akiyama, S., and Miyamoto, Y. 1970. Reexamination and characterization of the T-agglutination complex or pattern of Streptococcus pyogenes:preparation of anti-T factor sera. Jpn. J. Microbiol. 14: 269-277. Wilson, E., Zimmerman, R.A., and Moody, M.D. 1968. Value of T-agglutination typing of group A streptococci in epidemiologic investigations. Health Lab. Sci. 5: 199-207. You, M.S., and Facklam, R.R. 1986. New test system for identification of Aerococcus, Enterococcus, and Streptococcusspecies. J. Clin. Microbiol. 24: 607-611. (Received for publication, October 9, 1991; in revised form, April 13, 1992)
Proposal of a New Scheme for the Serological Typing of Enterococcus faecalis Strains Shizue
MAEKAWA,*,1 Migaku YOSHIOKA,2 and Yoshiaki KUMAMOTO2 1Department of Microbiology,and 2Departmentof Urology,Sapporo
Medical College,Sapporo, Hokkaido 060, Japan (Accepted for publication, April 17, 1992)
Abstract No systematic study on serotyping of Enterococcus faecalis has been reported since 1964 when M.E. Sharpe conducted serotyping of group D streptococcus in U.K. So, we attempted to re-evaluate serotyping of E. faecalis. For this purpose, we received 42 Sharpe's strains and first examined for their biochemical characteristics as E. faecalis. Only 9 of the 42 strains were identified as E. faecalis. We raised rabbit antisera against a large number of E. faecalis strains, including the 9 Sharpe's strains, 2 strains obtained from CDC in U.S.A. and 36 strains isolated from patients hospitalized in different cities of Japan. From the results of cross-agglutination tests and absorption tests performed on these antisera using a large number of E. faecalis strains, we were able to classify 21 distinct serotype strains and to prepare 21 monospecific typing antisera by absorption of the antisera to the type strains with appropriate cross-agglutinating strains. When 832 E. faecalis strains were serotyped with the 21 typing antisera, 638 strains (76.7%) were typable. Thus, we propose a provisional scheme of 21 distinct serovars in E. faecalis.
Frequency of isolation of Enterococcusfaecalis (9) pathogens from patients with urinary tract infections is increasing in Japan as a result of frequent use of thirdgeneration cephalosporins for the treatment (5). A paper from the 'United States reported that infants hospitalized in a neonatal intensive care unit had an increased incidence of neonatal sepsis caused by E. faecalis (6), and a report on nosocomial infections in general surgery clinics and intensive care units in Germany indicated that E. .faecalis is an important nosocomial pathogen (3). In 1964, M.E.
Sharpe
reported
serological
type strains of Streptococcus.faecalis
(E. faecalis) and compared them to those of Sharpe and Shattock (1952) by precipitin reactions and reciprocal absorption tests. By this study, Sharpe classified S. faecalis into 11 different serological types (10). Since this report, no further systematic study on the serological typing of E. faecalis has been reported. Thus, we attempted to re-evaluate the serological typing of E. faecalis, and for this purpose we received 42 Sharpe's serotype strains, of which 23 had been deposited in the National Collection of Type Cultures (NCTC) of U.K. as distinct type strains, from the National Institute for Research in Dairying (NIRD) in U.K. We found 671
672
S. MAEKAWA
ET AL
however, that only 9 of these 42 strains were identified as E. faecalis. Therefore, we raised rabbit antisera to a large number of E. faecalis strains, including the 9 strains from NIRD, and performed cross-agglutination and absorption tests on each antiserum by using a large number of E. faecalis strains. From the results of these studies, we were able to identify 21 serotype strains and to prepare 21 monospecific typing antisera by absorption of the antisera to the 21 strains with appropriate crossagglutinating strains. Based on the results of the present study, we proposed to classify E. faecalis into 21 serovars. MATERIALS
AND
METHODS
Bacterial strains. E. faecalis SS498, a reference strain in Centers for Disease Control (CDC), and E. faecalis 1824-73, isolated from a wound of a patient, were received from Dr. R.R. Facklam, CDC, Atlanta, U.S.A. Enterococcussp. Sharpe's serotype strains, NIRD 1599-1610, 1616-1633, 1637-1648, 42 strains in all, were obtained from NIRD, Shinfield, Berkshire, U.K. Twenty-three of the 42 strains had been deposited as Sharpe's serotypes 1-16 and 18-24 in NCTC, Central Public Health Laboratory, London, U.K., as shown in Table 1. Japanese strains used for raising antisera were isolates from urine of patients hospitalized in different cities in Japan. Culture media. Todd-Hewitt broth (THB, Difco) was used for liquid culture and sheep blood agar plates were used as solid culture medium. Identificationof E. faecalis. Identification of E. faecalis based on 20 biochemical reactions was done by using an Api STREP 20 kit produced by API SYSTEM S.A., France (4). Preparation of rabbit antisera against E. faecalis. Antisera against various strains of E. faecalis were raised in rabbits by i.v. injection with formalin-killed bacteria. The procedures for preparation of formalin-killed bacteria, immunization of rabbits and preparation of monospecific typing sera by absorption of the antisera with cross-agglutinating strains followed essentially those prescribed for T-typing agglutination of Streptococcus pyogenes(7, 11). Agglutinationtest. On individual slide glass marked with circles, 5 pl of appropriately diluted each typing serum was placed and the bacterial suspension to be tested was added. The slide glass was gently rotated and occurrence of bacterial agglutination was checked visually for 1 min. A drop of bacterial suspension mixed with a drop of saline served as the control. RESULTS
Re-Identificationof Sharpe's SerotypingStrains by BiochemicalCharacteristicsfor E. faecalis The 42 Sharpe's serotype strains given to our laboratory by NIRD were first examined for the characteristics of E. faecalis by 20 biochemical reactions in Api STREP 20 kit. As shown in Tables 1 and 2, only 9 of the 42 strains were identified as E. faecalis. The other strains were identified as E. faecium (21 strains), E. galli-
SEROVARS
Table
1. group
Strains
OF
ENTEROCOCCUS
deposited
D streptococcus
in NIRD by
M.E.
673
FAECALIS
as serotype Sharpe
in
strains 1962
of
674
S. MAEKAWA
Table
2.
ET AL
Identification of Sharpe's serotyping strains by biochemical reactions determined in Api STREP 20 kit
Species in Enterococcus of NIRD 1605, 1641, and 1647, and species in Streptococcus of NIRD 1623 could not be identified.
narum (2 strains), E. hirae (2 strains), E. durans (1 strain), Enterococcussp. (3 strains), S. equinus(3 strains) and Streptococcus sp. (1 strain). These results were later verified by NIRD. Establishmentof MonospecificAnti-E. faecalis Typing Sera and E. faecalis SerotypeStrains We first raised rabbit antisera against two E. faecalis strains obtained from CDC and 36 E. faecalis strains isolated from patients hospitalized in different cities in Japan. Based on the results of cross-agglutination tests and reciprocal absorption tests performed on these antisera using the immunizing strains, and the results of crossagglutination tests and absorption tests on the antisera using 463 K faecalis strains
SEROVARS
OF ENTEROCOCCUS
FAECALIS
675
isolated from the other patients in Japan, we were able to classify 15 serotype strains, and to prepare 15 monospecific typing antisera by absorption of the antisera to the type strains with appropriate cross-agglutinating strains. Agglutination titers of each of these antisera against the 15 strains before and after absorption with the cross-agglutinating strains are shown in the part of Table 3 circumscribed by dotted lines. Next, 9 Sharpe's strains that were identified as E. faecalis by the present study, i.e., NCTC 8727, 8729, 8730, 8731, 8732, 8734, 8735, 8744, and 8745, were subjected to serotyping by the 15 monospecific typing antisera. NCTC 8729, 8730, 8731, 8734, 8744, and 8745 did not show agglutination reaction to any of the 15 monospecific typing antisera, whereas NCTC 8727, 8732, and 8735 showed agglutination in anti-SS498, anti-Otaru-104, and anti-Sapporo-603 monospecific typing sera, respectively. We prepared rabbit antisera against NCTC 8729, 8730, 8731, 8734, 8744, and 8745, and performed cross-agglutination tests and absorption tests on these antisera using the 6 immunizing strains and the 15 serotype strains. From the results of these experiments, we were able to classify these 6 NCTC strains as distinct serotype strains and to prepare monospecific typing antisera by absorbing the antisera with appropriate cross-agglutinating strains. Agglutination titers of these 6 antisera against the 6 immunizing strains and the 15 serotype strains before and after absorption with appropriate cross-agglutinating strains are shown in Table 3. To verify serotype of NCTC 8727, 8732, and 8735, antisera to these 3 strains were prepared. Considering the results of the agglutination reactions of these strains to the 15 monospecific typing sera, reciprocal absorption tests with the immunizing strains were performed in combination of anti-NCTC 8727 and anti-SS498, anti-NCTC 8732, and anti-Otaru-104, and anti-NCTC 8735 and anti-Sapporo-603 antisera, respectively. As shown in Table 4, NCTC 8727, 8732, and 8735 were verified to belong to serotypes of SS498, Otaru-104, and Sapporo-603, respectively. Proposalof 21 Distinct Serovarsof E. faecalis Based on the results described in the preceding section, we propose 21 distinct serovars of E. faecalis. The serovars are numbered from 1 to 21 provisionally, and type strain of each serovar is listed in Table 5. These 21 E. faecalis serotype strains, which consist of 13 strains isolated from patients in different cities in Japan, 2 strains obtained from CDC and 6 strains obtained from NIRD as NCTC Sharpe's serotype strains, were all identified as E. faecalis by Api STREP 20 kit (data not shown). DISCUSSION
In 1962, 42 group D streptococcus (S.faecalis) strains were deposited in NIRD as Sharpe's serotyping strains, and 23 of these strains were deposited in NCTC as serotype strains. Since then, no further systematic study of serological typing of E. faecalis (S. faecalis was transferred to the genus Enterococcusas E. faecalis in 1984 (9)) has been reported. Recently, the importance of E. faecalis as a pathogen in
676
S. MAEKAWA
Table 3.
Agglutinin titers of antiserum prepared
ET AL
to each E. faecalis serotype strain against all
SEROVARS
OF
ENTEROCOCCUS
FAECALIS
of the serotype strains before and after absorption with appropriate cross-agglutinating strain(s)
677
678
Table
S. MAEKAWA 3
ET AL
continued.
Table 4.
Reciprocal absorption tests on the antisera in combination of E. faecalis strains SS498 and NCTC 8727, Sapporo-603 and NCTC 8735, and Otaru-104 and NCTC 8732
SEROVARS
Table 5.
OF
ENTEROCOCCUS
FAECALIS
Serovars of E. faecalis and type strain of each serovar to propose
679
680
S. MAEKAWA
ET AL
urinary tract infections (5), neonatal sepsis (6) and nosocomial infections (3) has been indicated. Therefore, we conducted the present study to re-evaluate the serological typing of E. faecalis. We first prepared rabbit antisera against two E. faecalis strains obtained from CDC and 36 E. faecalis strains isolated from patients hospitalized in different cities of Japan, considering that predominant serotype may somehow differ with each other in different districts or countries. From the results of absorption tests and cross-agglutination tests performed on these antisera, we were able to classify 15 distinct serotype strains, and thus we were able to prepare 15 distinct monospecific typing antisera by absorption of the antisera to the 15 serotype strains with appropriate cross-agglutinating strains. For the present study, we received 42 Sharpe's serotype strains from NIRD, and we first examined for their biochemical characteristics of E. faecalis in Api STREP 20 kit as recommended by several investigators (1, 2, 13), because the characteristics of these strains as E. faecalis have not been determined since they were deposited in NIRD and NCTC. We found that only 9 strains of the 42 strains were identified as E. faecalis, and that even 23 strains deposited in NCTC as Sharpe's serotypes 1-16, 18-24 (corresponding to Group D Sharpe and Shattock serotypes NCTC 8727, 8796, 8729-8749 in the catalogues of the National Collections of Type Cultures and Pathogenic Fungi, 1989) included strains of 4 different species of genus Enterococcusand strains of genus Streptococcus. These results, which were later verified by NIRD, might stem from the historical fact that Sharpe identified her strains as group D streptococcus but not as E. faecalis defined at present. The results of serotyping of the 9 Sharpe's strains that were identified as E. faecalis by the present study, with the 15 monospecific typing antisera, and of cross-agglutination tests and absorption tests performed on the antisera raised against the 9 strains indicated that 6 of the 9 strains were distinct in serotype from the already established 15 serotype strains, whereas the other 3 strains belonged to either one of the already established 15 serotypes. Thus, we propose a scheme of 21 distinct serovars of E. faecalis, provisionally numbered from 1 to 21, and type strain of each serovar as shown in Table 5. Since 6 NCTC strains included in our proposed serovars as type strains had been registered by M.E. Sharpe as serotype strains in 1962, Sharpe's serotype numbers for the strains may have priority. However, none of the 23 Sharpe's NCTC strains were exactly identified as E. faecalis as defined at present, so our provisionally numbered serovars of E. faecalis will be legitimate. In our scheme, serovar numbers were given to the type strains as follows: serovar 1 to SS498, a reference strain of E. faecalis in CDC; serovars 2 to 14 to the strains isolated in Japan; serovar 15 to the strain isolated in U.S.A. ; and serovars 16 to 21 to the NCTC Sharpe's strains. The serovar numbers given to NCTC 8744 and 8745 in our scheme, i.e., serovars 19 and 20, were the same as those in Sharpe's serotypes, whereas the other 4 NCTC strains were given serovar numbers different from those in Sharpe's serotypes. To evaluate the proposed serotyping of E. faecalis practically, 832 E. faecalis strains which had been isolated from 463 patients in Japan and 369 patients in
SEROVARS
OF
ENTEROCOCCUS
FAECALIS
681
U.S.A. were serotyped with the 21 monospecific typing antisera of our scheme. Among the 832 strains, 638 strains (76.7%) were typable as either one of 15 serovars. Serovars 1, 2, 4 and 7 were major serovars and occupied 72% of the typable strains, whereas serovars 16 to 21 were less than 2% (data not shown). We think that serological typing of E. faecalis with these 21 monospecific typing antisera should contribute to epidemiological study of E. faecalis infections like M-, or T-typing of S. pyogenesand serotyping of S. agalactiaein epidemiological studies (8, 12). The authors wish to thank Dr. R.R. Facklam, Division of Respiratory Disease, Center for Infectious Disease, Centers for Disease Control, U.S.A. and Dr. G. Colman, Central Public Health Laboratory, U.K. for permitting us to analyze serovars of E. faecalis strains from their laboratories. We are also deeply indebted to them for providing the E. faecalis clinical isolates stored in their laboratories. REFERENCES
1) 2) 3) 4) 5)
6)
7)
8) 9)
10) 11)
12) 13)
Bosley, G.B., Facklam, R.R., and Grossman, D. 1983. Rapid identification on enterococci. J. Clin. Microbiol. 18: 1275-1277. Facklam, R.R., and Collins, M.D. 1989. Identification of Enterococcusspecies isolated from human infections by a conventional test scheme. J. Clin. Microbiol. 27: 731-734. Holzheimer, R.G., Quoika, P., Patzmann, D., and Fussle, R. 1990. Nosocomial infections in general surgery: surveillance report from a German University Clinic. Infection 18: 219-225. Jorgensen, J.H., Crawford, S.A., and Alexander, G.A. 1983. Rapid identification of group D streptococci with the API 20S system. J. Clin. Microbiol. 17: 1096-1098. Kosakai, N., Kumamoto, Y., Hirose, Y., Tanaka, N., Hikichi, Y., Shigeta, S., Shiraiwa, Y., Yoshida, H., Ogata, M., Tazaki, H., Ili, H., Uchida, H., Kobayashi, Y., Matsuda, S., Kitagawa, R., Fujita, K., Hayashi, Y., Oguri, Y., Furusawa, T., Takeuchi, Y., Moriyama, H., Shibata, K., Yonezu, S., Takaha, M., Matsumiya, K., Tanaka, M., Yamaguchi, K., Takeda, K., Mochida, C., and Igari, J. 1990. Comparative studies on activities of antimicrobial agents against causative organisms isolated from urinary tract infections (1987). I. Susceptibility distribution. Jpn. J. Antibiot. 43: 919-953. Luginbuhl, L.M., Rotbart, H.A., Facklam, R.R., Roe, M.H., and Elliot, J.A. 1987. Neonatal enterococcal sepsis: case-control study and description of an outbreak. Pediatr. Infect. Dis. J. 6: 1022-1030. Maekawa, S., Fukuda, K., Yamauchi, T., Yamaguchi, T., Takahashi, K., and Sugawa, K. 1981. Follow-up study of pharyngeal carriers of beta-hemolytic streptococci among school children in Sapporo city during a period of 2 years and 5 months. J. Clin. Microbiol. 13: 1017-1022. Murai, T., Inazumi, Y., Ooguro, H., and Ishiguro, K. 1980. Serotype of group B streptococci isolated in Japan in recent year. Jpn. J. Infect. Dis. 54: 22-28. Schleifer, K. H., and Kilpper-Baltz, R. 1984. Transfer of Streptococcusfaecalis and Streptococcus faecium to the genus Enterococcusnom. rev. as Enterococcusfaecalis comb. nov. and Enterococcus faecium comb. nov. Int. J. Syst. Bacteriol. 34: 31-34. Sharpe, M.E. 1964. Serological types of Streptococcusfaecalis and its varieties and their cell wall type antigen. J. Gen. Microbiol. 36: 151-160. Takizawa, K., Akiyama, S., and Miyamoto, Y. 1970. Reexamination and characterization of the T-agglutination complex or pattern of Streptococcus pyogenes:preparation of anti-T factor sera. Jpn. J. Microbiol. 14: 269-277. Wilson, E., Zimmerman, R.A., and Moody, M.D. 1968. Value of T-agglutination typing of group A streptococci in epidemiologic investigations. Health Lab. Sci. 5: 199-207. You, M.S., and Facklam, R.R. 1986. New test system for identification of Aerococcus, Enterococcus, and Streptococcusspecies. J. Clin. Microbiol. 24: 607-611. (Received for publication, October 9, 1991; in revised form, April 13, 1992)
