DIAGN MICROBIOLINFECTDIS 1992;15:281-285
281
Genetic Heterogeneity Among Nutritionally Deficient Streptococci Daniel S. Stein and Claudia R. Libertin
The nutritionally deficient (variant) streptococci (NDS) share the auxotrophic characteristic of requiring pyridoxal or thiol group supplementation for growth. The deoxyribonucleic acid relatedness of these organisms among themselves is unknown. Improved speciation of NDS would lead to a better knowledge of their pathogenesis and possible insight into improved clinical management. Therefore, DNA-DNA hybridization and biotyping of 23 nutritionally deficient streptococci were performed. Biochemical testing using the API Rapid Strept Iden-
tification method revealed that the organisms in this study were characterized among three broad biotype groups. Only one strain was nontypeable. DNA-DNA hybridization among the nutritionally deficient streptococci that we compared revealed genetic heterogeneity. Only four (17%) of 23 isolates were highly homologous; all were of biotypes 2 and 3. Reference viridans streptococcal strains had minimal homology to the NDS strains. The data indicate that the NDS are genetically heterogeneous.
INTRODUCTION
mortality. Taxonomic relatedness among nutritionally deficient streptococci among themselves and to other streptococci could lead to insight into improved antibiotic management of NDS-infected patients. The taxonomic relationship of the nutritionally deficient or variant streptococci among themselves and to other streptococci is unclear. An early taxonomic proposal that was based on patterns of sugar fermentation and the composition of the cell wall suggested that the NDS were largely mutant subspecies of Streptococcus mitior (Roberts et al., 1979). Serologic testing has shown that there are two major serologic groups that are minimally cross-reactive to other viridans streptococci (Van de Rijn et al., 1979). Biochemical testing using the API Rapid Strept Identification method has given evidence for three broad biotype groups (Bouvet et al., 1985). Penicillin-binding protein (PBP) studies have classified biotype-1 strains as having similar PBP patterns that were termed PBPgroup I (Bouvet et al., 1989). The biotype2 and 3 strains had PBP patterns termed PBP group II. A small number of strains were not classified into either a biotype or a PBP group. Another identifiable feature of NDS is the presence of a hot-acid-extractable chromophore that was also found in -85% of S. mitis and S. sanguis II (Bouvet et al., 1981 and 1985; Stein and Libertin, 1989a). Recently, the characterization of NDS into two species was made based on DNA relatedness studies (Bouvet et al., 1989). In this study, we determine the taxonomic relationship
The nutritionally deficient streptococci (NDS) were first described in 1961 as fastidious Gram-positive cocci requiring the secretions of other bacteria or thiol group supplementation for growth (Frenkel et al., 1961). Since that time, many other descriptive names have been used for this group of organisms: satelliting; sulfhydryl, pyridoxal, or vitamin-B6 requiring; and nutritionally variant (Roberts et al., 1979; Ruoff, 1991). Nutritionally deficient streptococci have been responsible for serious infections including sepsis and endocarditis in humans (Stein and Nelson, 1987; Stein and Libertin, 1989b; Ruoff, 1991). The response to therapy of endocarditis caused by these organisms appears to be worse than for viridans streptococcal endocarditis (Bisno et al., 1989; Roberts et al., 1979; Stein and Nelson, 1987). Improved speciation could lead to a better understanding of the pathogenesis of these organisms and identification of possible predictors of morbidity and From the Division of AIDS, Medical Branch, NIAID (D.S.S.), Rockville,Maryland;and Departments of Medicine and Pathology,LoyolaUniversityof Chicago (C.R.L.), Maywood, Illinois, USA. Address reprint requests to Dr. C.R. Libertin, Departments of Medicine and Pathology,LoyolaUniversityof Chicago, 2160 South First Avenue, Maywood,IL 60153, USA. Received 4 February 1991; revised and accepted 19 August 1991. © 1992 Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010 0732-8893/92/$5.00
282
among a group of nutritionally deficient streptococci and to reference strain of viridans streptococci by a D N A - D N A hybridization method.
MATERIALS AND METHODS Organisms The nutritionally deficient (variant) streptococci were obtained from our laboratory collection and from kind donations from Walter Wilson, MD, Mayo Clinic, Rochester, MN, and Roberta Carey, PhD, St. Francis Hospital, Evanston, IL. All of the organisms were blood isolates from patients with endocarditis. The organisms were checked for their inability to grow in the absence of pyridoxal supplementation by subculturing onto 5% sheep blood-tryptic soy agar (SBA) plates (BBL Microbiology Systems, Cockeysville, MD) and incubating at 37°C for 72 hr. All isolates were tested for satelliting growth around a Staphylococcus aureus streak on SBA plates and the ability to produce a hot-acid-extractable chromophore when done as previously described (Stein and Libertin, 1989a). Cultures were maintained on brain-heart infusion anaerobic (BHI ANA) plates (BBL Microbiology Systems) or on pyridoxal-supplemented 5% SBA. The reference strains, S. mitis ATCC 15911, S. sanguis II ATCC 15909, and S. mitis CDC 1178 were obtained from the American Type Culture Collection or the Centers for Disease Control (Atlanta, GA).
Biotyping The Rapid Strept ID system (API Analytics, Plainsfield, NY) was used as recommended by the manufacturer. This system consists of strips that contain 20 cupules. Each cupule has a desiccated substrate for the following tests: acetoin production (VogesProskauer test); hydrolysis of esculin; the presence of activity for the following enzymes: ~-glucosidase, pyrrolidonylarylamidase (PYR), o~-galactosidase (cxgal), ~-glucuronidase (~-GUR), ~-galactosidase (~gaS), alkaline phosphatase, leucine arylamidase, and arginine dihydrolase; and the anaerobic production of acid from ribose, r-arabinose, mannitol, sorbitol, lactose, trehalose, inulin, raffinose, starch, and glycogen. The following modifications in the methods suggested by the manufacturer were done: organisms from two BHI ANA plates incubated at 37°C for 24 hr were used for the inoculum and the first nine tests of the strip were read at 4 hr (as recommended by manufacturer) with the remaining tests and esculin reduction read at 24 hr. The organisms were suspended in 2 ml of sterile 100 ~g/ml pyridoxal HC1 solution, which was used to inoculate the manufacturer's supplied media.
D.S. Stein and C.R. Libertin
Biotyping was done according to the proposal of Bouvet et al. (1985). Biotyping I was defined as positive results in the pyrrolidonylarylamidase, a-galactosidase, ~-galactosidase tests, and the production of acid from trehalose; biotype II was defined as positive results in PYR and f~-glucuronidase tests with variable production of acid from inulin; and biotype III was defined as a positive PYR test with variable production of acid from inulin.
DNA Preparation DNA was extracted from organisms grown in T o d d Hewitt broth supplemented with 10 ~g/ml pyridoxal HCI (Sigma Chemical Company, St. Louis, MO) (Maurer, 1986; Stein and Libertin, 1989b). After overnight growth, the suspended cellular pellet underwent lysis at 37°C for 2 hr with 4 mg/ml of lysozyme followed by alkaline sodium dodecyl sulfate. The DNA released was phenol--chloroform extracted and then precipitated with cold ethanol. It was then suspended and treated with pronase K and RNAse A (Boehringer-Mannheim), phenol-chloroform extracted once again, and precipitated with cold ethanol. The DNA was solubilized in TRIS-EDTA buffer to be stored at - 20 ° or - 70°C until used. Purity and yields were addressed spectrophotometrically (Spectronic 1001; Baush and Lomb) at 260 and 280 nm.
Hybridizations A dot-blot hybridizations were performed in a 96well manifold (Schleicher and Scheull, Keene, NH) (Shields et al., 1986). All organisms were tested simultaneously on the same filter; hybridizations were done in triplicate. Serial dilutions of N a O H denatured DNA were applied to a nitrocellulose filter in a vacuum manifold and rinsed with 1 M ammonium acetate. Prior to baking at 80°C, filters were first washed with 2 x SSC, then 0.2 x Denhardt's solution in 3 x SSC, and air dried. Denhardt's solution is 0.1% (wt/vol) of each of the following: Ficoll, polyvinylpyrrolidone, and bovine serum albumin. Chromosomal DNA probes were labeled with 32p by nick translation. Hybridizations were conducted in 50% deionized formamide, I x Denhardt's solution, 50 ~g/ml herring sperm DNA, 0.1% SDS, and 4 x SSC, at 37°C or 50°C for 12-24 hr. Posthybridization washes were conducted at 60°C and consisted of three 30-min washes in I x Denhardt's solution, 50 ~g/ml herring sperm DNA, 0.1% SDS in 4 x SSC followed by one 30-min wash in I x SSC. Additional hybridizations with probes 1178 and 20559 were done at 50°C in the same hybridization solution as above except for the addition of 10 ~g/ml poly A and the use of 3 x SSC instead of 4 x SSC. Posthybridization wash
Nutritionally Deficient Streptococci
solutions were done at 65°C and were as noted above except that 3 x SSC replaced 4 x SSC. Exposure of the dot-blot hybridizations were made at -70°C on Kodak X-Omat film. Microdensitometry was performed on a Biorad model 620 Vido Densitometer. Sensitivity strips of serial dilutions of the probes revealed detection of DNA at a threshold of 0.8 ng. All DNA data were expressed as a percentage of the control probe with correction for the amount of DNA in each blot. The calculated Tm, the temperature at which 50% of the DNA base pairs would be hydrogen bonded, would be 62°C using the above conditions and assuming no base-pair mismatches (Davis et al., 1986). The dot-blot hybridization conditions were therefore optimal (Tm-25°C) at 37°C (Johnson, 1986).
283
with a low degree of hybridization (42.3%) to the viridans streptococci S. sanguis ATCC 15909 and S. mitis ATCC 15911. The 20559 probe demonstrated homology of >60% for three strains; 7548 (90%), LM8601 (76.0%), and MB (71.4%). At the higher stringency conditions (50°C), no significant difference in the percent of homology among all strains (data not shown) was noted. Strains 7548 and MB were classified as biotype 3, whereas LM8601 and 20554 were classified as biotype 2. Neither of the two viridans streptococci (S. sanguis ATCC 15909, S. mitis CDC 1178) had significant homology to any of the strains tested.
DISCUSSION RESULTS
All organisms in the nutritionally deficient streptococci (NDS) group were pyridoxal-requiring bacteria as determined by their absence of growth without pyridoxal but ability to have satelliting growth around Staphylococcus aureus, and identification of chromophore after acid extraction. In addition, one organism CCGA, was a ~-hemolytic group-A antigen-positive streptococci that met the prime criteria for NDS, but was chromophore negative (Stein and Libertin, 1989a). The organisms' biotypes are listed in Table 1. Of the 23 strains, 11 (47.8%) were biotype 1, four (17.4%) biotype 2, and seven (30.4%) biotype 3. Only three organisms from the group of nutritionally deficient streptococci had an atypical biochemical profile. According to the scheme proposed by Bouvet et al. (1989), all biotype definitions included a positive test for pyrrolidonylarylamidase. However, one of these isolates (AB8611) was negative for the pyrrolidonylarylamidase test despite four separate trials of testing and prolongation of the incubation period to 24 hr. AB8611 was therefore designated nontypeable. Minor variations were made in assigning biotypes for two other organisms. Strain BA8212 had a positive result in the alkaline phosphatase (PAL) test; and strain LR8604 best fit a biotype-3 classification though the results did not affect biotype identification. PAL was positive for CCGA, BA8212, and AB8611-LU; ADH was positive for LR8604, CCGA and 20559; and acid was produced from mannitol for CCGA, BC840, and AB8611-LU. These results were not reported in a prior biochemical survey of different strains (Bouvet et al., 1985). Marked genetic heterogeneity among the NDS was found. The chromosomal DNA probes for the biotype 1 (6995) and biotype 2 (20559) did not significantly hybridize to each other. The 6995 probe had little homology to any of the 21 NDS strains tested
Our results indicate genetic heterogeneity among the strains of nutritionally deficient streptococci that we compared. Only four strains of NDS among the biotypes 2 and 3 were highly homologous to each other as determined by the membrane-bound method of D N A - D N A hybridization, a method used in other studies of taxonomic classification (Coykendall et al., 1975 and 1987; Ezaki et al., 1988; Roberts et al., 1987; Welborn et al., 1983). The NDS strains 20559 was 90% homologous to the NDS 7548, 76% homologous to the NDS LM8601, and 71.4% homologous to the NDS strain MB. These strains were isolated at four different institutions, thereby making it highly unlikely that they share any epidemiologic link. These four strains are of the same species by both the classic taxonomic criteria of being >70% homologous to the control strain (Brenner, 1973) and by the more recent taxonomic recommendations (Wayne et al., 1987). The majority of the NDS strains that we tested did not meet the classification scheme that is composed of two species, S. defectivus spp. nov. and S. adjacens spp. nov., proposed by Bouvet et al. (1989). Our results indicate that some NDS in biotypes 2 and 3, which would be designated S. adjacens spp. nov., are related at the species level, but not all of them. Seven of the strains in biotypes 2 and 3 did not show a high degree of homology. Only one of these strains (20559) was atypical biochemically in that it hydrolyzed arginine which was unlike the proposed species S. adjacens spp. nov.; all others were biochemically similar. In addition, we were unable to demonstrate a homologous relationship among NDS in biotype 1, S. defectivus spp. nov. Among these 11 strains, none were homologous to each other despite being the same biotype. Among the NDS in this study, marked genetic heterogeneity existed. Possibly, a group of unrelated strains exist that meet the biotyping requirements for the S. defectivus spp. nov. characterization that represent a heretofore previously undescribed species or
284
TABLE I
D.S. Stein a n d C.R. Libertin
Percent D N A - D N A H o m o l o g y A m o n g Nutritionally Deficient Streptococci Percent Hybridizationa to Probe DNA from
Biotype 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 3 3 3 3 3 3 3 NT
NDS Chromosomal DNBA
20559
1178
6995
15909
93634 16316 19363 8060 BA8212b JV8611 BC7729 SS8706 BC7901 MC8 6995 20559 MC4 LM8601 CCGAc L61 7548 8246 CL8707 BC840-2 LR8604~ MB AB8611 S. sanguis ATCC 15909 S. mitis ATCC 15911 S. mitis CDC 1178
1.0 1.3 0.1 0.2 6.9 0.3 0.2 1.3 0.5 0.3 27.2 100.0 12.0 76.0 2.3 4.5 90.0 6.0 35.5 6.1 2.8 71.4 3.3 6.9 12.5 8.4
0.4 0.3 ND ND 3.3 0.2 ND 0.1 ND ND 45.0 -1.2 18.9 2.2 0.4 26.8 0.9 5.6 0.2 2.9 18.4 3.1 6.7 12.9 100.0
3.3 1.7 2.6 1.0 2.1 0.4 0.9 ND ND 1.7 100.0 4.9 1.8 8.0 ND 1.2 12.9 2.9 3.0 ND 8.0 8.9 5.4 42.3 42.3 --
0.6 0.5 ND ND 0.2 0.1 ND ND 0.2 0.3 20.5 0.2 0.6 1.7 0.2 0.5 5.1 0.6 0.6 0.2 1.0 1.9 1.1 100.0 30.8 --
NT, not typeable because pyrrolidonylarylamidase test was negative; NA, not applicable; ND, not detectable; and --, redundant study. See Materials and Methods for explanation of test procedures and biotype assignment. aData expressed as the percentage of the DNA control. Data shown are for 37°C hybridization (optimal conditions). All organisms' DNAs were hybridized simultaneously on the same filter; all results were reproduced in triplicate. Data represent one experiment. bPositve alkaline phosphatase test. cPositive alkaline phosphatase and arginine dehydrolase tests for production of acid from trehalose and chromophore test were negative. e~-Galactosidase test were positive.
a variant of the p r o p o s e d type strain. A direct comparison of the NDS strains to those used by Bouvet et al. (1989) w o u l d be necessary to explore this possibility further. The absence of significant genetic relatedness a m o n g NDS a n d viridans streptococci in this s t u d y is consistent with the w o r k b y Bouvet et al. (1989). In fact, others have s h o w n that within the viridans group of streptococci, organisms speciated biochemically do not hybridize as expected to viridans streptococci of the same species (Ezaki et al., 1988). The n u m b e r of viridans streptococcal strains that were c o m p a r e d to the NDS in this s t u d y a n d so far to date are probably too small to conclude that NDS are not h o m o l o g o u s
to some species of viridans streptococci. As noted, the p r o p o s e d taxonomical relationship a m o n g viridans streptococci remain unclear; therefore, the relation° ship of NDS to the viridans g r o u p of streptococci must be done with caution. Further investigations into these relationships m a y yield a better u n d e r s t a n d i n g of the p a t h o g e n e s i s of these o r g a n i s m s a n d lead to insights into i m p r o v e d clinical m a n a g e m e n t of patients infected with these organisms. The authors thank Drs. Roberta Carey and Walter Wilson for their kind gifts of NDS. This research was supported in part by the estate of William G. Potts.
Nutritionally Deficient Streptococci
285
REFERENCES Bisno, AL, Dismukes WE, Durack DT, et al. (1989). Antimicrobial treatment of infectious endocarditis due to viridans streptococci, enterococci, and staphylococci. JAMA 261:1471-1477. Bouvet A, Grimond F, Grimond PAD (1989) Streptococcus defectivities sp. nov. and Streptococcus adjacens sp. nov., nutritionally variant streptococci from human clinical specimens. Int J Syst Bacteriol 39:290-294. Bouvet A, van de Rijn I, McCarty M (1981) Nutritionally variant streptococci from patients with endocarditis: growth parameters in a semisynthetic medium and demonstration of a chromophore. J Bacteriol 146:10751082. Bouvet, A, Villeroy R, Cheng F, et al. (1985) Characterization of nutritionally variant streptococci by biochemical tests and penicillin binding proteins. J Clin Microbiol 22:1030-1034. Brenner DJ (1973) Deoxyribonucleic acid reassociation in the taxonomy of enteric bacteria. Int J System Bacteriol 23:298-307. Coykendall AL, Specht PA (1975) DNA base sequence homologies among strains of Streptococcus sanguis. J Gen Microbiol 91:92-98. Coykendall AL, Wesbecher PM, Gustafson KB (1987) Streptococcus milleri, Streptococcus constellatus, and Streptococcus intermedius are later synonyms of Streptococcus anginosus. Int J Syst Bacteriol 37:222-228. Davis LG, Dibner MD, Battey JF (1986) Methods in Molecular Biology. New York: Elsevier. Ezaki T, Hashimoto Y, Takeuchi N, et al. (1988) Simple genetic method to identify viridans streptococci by colorimetric dot hybridization and fluorometric hybridization in microdilution wells. J CLin Microbiol 26:17081713. Frenkel A, Hirsch W (1961) Spontaneous development of L forms of streptococci requiring secretions of other bacteria or sulphydryl compounds for normal growth. Nature 191:728-730. Johnson JL (1986) Nucleic acid in bacterial classification. In Bergey'sManual of Systematic Bacteriology. Ed, NR Krieg
and JG Hott. Baltimore, MD: Williams and Wilkins, pp 8-11. Maurer R (1986) Laboratory Procedure Manual Cold Spring Harbor Symposium 1986. New York: Cold Spring Harbor. Roberts RB, Krieger AG, Schiller NL, Gross KC (1979) Viridans streptococcal endocarditis: the role of various species including pyridoxal dependent streptococci. Rev Infect Dis 1:955-965. Roberts MC, McMillan C, Coyle MB (1987) Whole chromosomal DNA probes for rapid identification of Mycobacterium tuberculosis and Mycobacterium avium complex. J Clin Microbiol 25:1239-1243. Ruoff KL (1991) Nutritionally variant streptococci. Clin Microbiol Rev 4:184-190. Shields MS, Kline BC, Tam JE (1986) A rapid method for the quantitative measurement of gene dosage: mini-F plasmid concentration as a function of cell growth rate. J Microbiol Methods 6:33-46. Stein DS, Libertin CR (1989a) A double-blinded comparative evaluation of three media for chromophore testing with viridans and nutritionally variant (deficient) streptococci. Am J Clin Pathol 91:589-593. Stein DS, Libertin CR (1989b) Molecular analysis of viridans and nutritionally deficient (variant) streptococci causing sequential episodes of endocarditis in a patient. Am J Clin Pathol 91:620-624. Stein DS, Nelson KE (1987) Endocarditis due to nutritionally deficient streptococci: a therapeutic dilemma. Rev Infect Dis 9:908-916. Van de Rijn I, George M, Bouvet A, Roberts RB (1979) Enzyme linked immunosorbent assay for the detection of antibodies to nutritionally variant streptococci in patients with endocarditis. J Infect Dis 1:955-965. Wayne L, Brenner DJ, Colwell P (1987) Report of the Ad Hoc Committee on Reconciliation of Approaches to Bacterial Systematics. Int J Syst Bacteriol 37:463-464. Welborn PP, Hadley WK, Newbrun E, Yajko DM (1983) Characterization of strains of viridans streptococci by deoxyribonucleic acid hybridization and physiological tests. Int J Syst Bacteriol 33:293-299.
281
Genetic Heterogeneity Among Nutritionally Deficient Streptococci Daniel S. Stein and Claudia R. Libertin
The nutritionally deficient (variant) streptococci (NDS) share the auxotrophic characteristic of requiring pyridoxal or thiol group supplementation for growth. The deoxyribonucleic acid relatedness of these organisms among themselves is unknown. Improved speciation of NDS would lead to a better knowledge of their pathogenesis and possible insight into improved clinical management. Therefore, DNA-DNA hybridization and biotyping of 23 nutritionally deficient streptococci were performed. Biochemical testing using the API Rapid Strept Iden-
tification method revealed that the organisms in this study were characterized among three broad biotype groups. Only one strain was nontypeable. DNA-DNA hybridization among the nutritionally deficient streptococci that we compared revealed genetic heterogeneity. Only four (17%) of 23 isolates were highly homologous; all were of biotypes 2 and 3. Reference viridans streptococcal strains had minimal homology to the NDS strains. The data indicate that the NDS are genetically heterogeneous.
INTRODUCTION
mortality. Taxonomic relatedness among nutritionally deficient streptococci among themselves and to other streptococci could lead to insight into improved antibiotic management of NDS-infected patients. The taxonomic relationship of the nutritionally deficient or variant streptococci among themselves and to other streptococci is unclear. An early taxonomic proposal that was based on patterns of sugar fermentation and the composition of the cell wall suggested that the NDS were largely mutant subspecies of Streptococcus mitior (Roberts et al., 1979). Serologic testing has shown that there are two major serologic groups that are minimally cross-reactive to other viridans streptococci (Van de Rijn et al., 1979). Biochemical testing using the API Rapid Strept Identification method has given evidence for three broad biotype groups (Bouvet et al., 1985). Penicillin-binding protein (PBP) studies have classified biotype-1 strains as having similar PBP patterns that were termed PBPgroup I (Bouvet et al., 1989). The biotype2 and 3 strains had PBP patterns termed PBP group II. A small number of strains were not classified into either a biotype or a PBP group. Another identifiable feature of NDS is the presence of a hot-acid-extractable chromophore that was also found in -85% of S. mitis and S. sanguis II (Bouvet et al., 1981 and 1985; Stein and Libertin, 1989a). Recently, the characterization of NDS into two species was made based on DNA relatedness studies (Bouvet et al., 1989). In this study, we determine the taxonomic relationship
The nutritionally deficient streptococci (NDS) were first described in 1961 as fastidious Gram-positive cocci requiring the secretions of other bacteria or thiol group supplementation for growth (Frenkel et al., 1961). Since that time, many other descriptive names have been used for this group of organisms: satelliting; sulfhydryl, pyridoxal, or vitamin-B6 requiring; and nutritionally variant (Roberts et al., 1979; Ruoff, 1991). Nutritionally deficient streptococci have been responsible for serious infections including sepsis and endocarditis in humans (Stein and Nelson, 1987; Stein and Libertin, 1989b; Ruoff, 1991). The response to therapy of endocarditis caused by these organisms appears to be worse than for viridans streptococcal endocarditis (Bisno et al., 1989; Roberts et al., 1979; Stein and Nelson, 1987). Improved speciation could lead to a better understanding of the pathogenesis of these organisms and identification of possible predictors of morbidity and From the Division of AIDS, Medical Branch, NIAID (D.S.S.), Rockville,Maryland;and Departments of Medicine and Pathology,LoyolaUniversityof Chicago (C.R.L.), Maywood, Illinois, USA. Address reprint requests to Dr. C.R. Libertin, Departments of Medicine and Pathology,LoyolaUniversityof Chicago, 2160 South First Avenue, Maywood,IL 60153, USA. Received 4 February 1991; revised and accepted 19 August 1991. © 1992 Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010 0732-8893/92/$5.00
282
among a group of nutritionally deficient streptococci and to reference strain of viridans streptococci by a D N A - D N A hybridization method.
MATERIALS AND METHODS Organisms The nutritionally deficient (variant) streptococci were obtained from our laboratory collection and from kind donations from Walter Wilson, MD, Mayo Clinic, Rochester, MN, and Roberta Carey, PhD, St. Francis Hospital, Evanston, IL. All of the organisms were blood isolates from patients with endocarditis. The organisms were checked for their inability to grow in the absence of pyridoxal supplementation by subculturing onto 5% sheep blood-tryptic soy agar (SBA) plates (BBL Microbiology Systems, Cockeysville, MD) and incubating at 37°C for 72 hr. All isolates were tested for satelliting growth around a Staphylococcus aureus streak on SBA plates and the ability to produce a hot-acid-extractable chromophore when done as previously described (Stein and Libertin, 1989a). Cultures were maintained on brain-heart infusion anaerobic (BHI ANA) plates (BBL Microbiology Systems) or on pyridoxal-supplemented 5% SBA. The reference strains, S. mitis ATCC 15911, S. sanguis II ATCC 15909, and S. mitis CDC 1178 were obtained from the American Type Culture Collection or the Centers for Disease Control (Atlanta, GA).
Biotyping The Rapid Strept ID system (API Analytics, Plainsfield, NY) was used as recommended by the manufacturer. This system consists of strips that contain 20 cupules. Each cupule has a desiccated substrate for the following tests: acetoin production (VogesProskauer test); hydrolysis of esculin; the presence of activity for the following enzymes: ~-glucosidase, pyrrolidonylarylamidase (PYR), o~-galactosidase (cxgal), ~-glucuronidase (~-GUR), ~-galactosidase (~gaS), alkaline phosphatase, leucine arylamidase, and arginine dihydrolase; and the anaerobic production of acid from ribose, r-arabinose, mannitol, sorbitol, lactose, trehalose, inulin, raffinose, starch, and glycogen. The following modifications in the methods suggested by the manufacturer were done: organisms from two BHI ANA plates incubated at 37°C for 24 hr were used for the inoculum and the first nine tests of the strip were read at 4 hr (as recommended by manufacturer) with the remaining tests and esculin reduction read at 24 hr. The organisms were suspended in 2 ml of sterile 100 ~g/ml pyridoxal HC1 solution, which was used to inoculate the manufacturer's supplied media.
D.S. Stein and C.R. Libertin
Biotyping was done according to the proposal of Bouvet et al. (1985). Biotyping I was defined as positive results in the pyrrolidonylarylamidase, a-galactosidase, ~-galactosidase tests, and the production of acid from trehalose; biotype II was defined as positive results in PYR and f~-glucuronidase tests with variable production of acid from inulin; and biotype III was defined as a positive PYR test with variable production of acid from inulin.
DNA Preparation DNA was extracted from organisms grown in T o d d Hewitt broth supplemented with 10 ~g/ml pyridoxal HCI (Sigma Chemical Company, St. Louis, MO) (Maurer, 1986; Stein and Libertin, 1989b). After overnight growth, the suspended cellular pellet underwent lysis at 37°C for 2 hr with 4 mg/ml of lysozyme followed by alkaline sodium dodecyl sulfate. The DNA released was phenol--chloroform extracted and then precipitated with cold ethanol. It was then suspended and treated with pronase K and RNAse A (Boehringer-Mannheim), phenol-chloroform extracted once again, and precipitated with cold ethanol. The DNA was solubilized in TRIS-EDTA buffer to be stored at - 20 ° or - 70°C until used. Purity and yields were addressed spectrophotometrically (Spectronic 1001; Baush and Lomb) at 260 and 280 nm.
Hybridizations A dot-blot hybridizations were performed in a 96well manifold (Schleicher and Scheull, Keene, NH) (Shields et al., 1986). All organisms were tested simultaneously on the same filter; hybridizations were done in triplicate. Serial dilutions of N a O H denatured DNA were applied to a nitrocellulose filter in a vacuum manifold and rinsed with 1 M ammonium acetate. Prior to baking at 80°C, filters were first washed with 2 x SSC, then 0.2 x Denhardt's solution in 3 x SSC, and air dried. Denhardt's solution is 0.1% (wt/vol) of each of the following: Ficoll, polyvinylpyrrolidone, and bovine serum albumin. Chromosomal DNA probes were labeled with 32p by nick translation. Hybridizations were conducted in 50% deionized formamide, I x Denhardt's solution, 50 ~g/ml herring sperm DNA, 0.1% SDS, and 4 x SSC, at 37°C or 50°C for 12-24 hr. Posthybridization washes were conducted at 60°C and consisted of three 30-min washes in I x Denhardt's solution, 50 ~g/ml herring sperm DNA, 0.1% SDS in 4 x SSC followed by one 30-min wash in I x SSC. Additional hybridizations with probes 1178 and 20559 were done at 50°C in the same hybridization solution as above except for the addition of 10 ~g/ml poly A and the use of 3 x SSC instead of 4 x SSC. Posthybridization wash
Nutritionally Deficient Streptococci
solutions were done at 65°C and were as noted above except that 3 x SSC replaced 4 x SSC. Exposure of the dot-blot hybridizations were made at -70°C on Kodak X-Omat film. Microdensitometry was performed on a Biorad model 620 Vido Densitometer. Sensitivity strips of serial dilutions of the probes revealed detection of DNA at a threshold of 0.8 ng. All DNA data were expressed as a percentage of the control probe with correction for the amount of DNA in each blot. The calculated Tm, the temperature at which 50% of the DNA base pairs would be hydrogen bonded, would be 62°C using the above conditions and assuming no base-pair mismatches (Davis et al., 1986). The dot-blot hybridization conditions were therefore optimal (Tm-25°C) at 37°C (Johnson, 1986).
283
with a low degree of hybridization (42.3%) to the viridans streptococci S. sanguis ATCC 15909 and S. mitis ATCC 15911. The 20559 probe demonstrated homology of >60% for three strains; 7548 (90%), LM8601 (76.0%), and MB (71.4%). At the higher stringency conditions (50°C), no significant difference in the percent of homology among all strains (data not shown) was noted. Strains 7548 and MB were classified as biotype 3, whereas LM8601 and 20554 were classified as biotype 2. Neither of the two viridans streptococci (S. sanguis ATCC 15909, S. mitis CDC 1178) had significant homology to any of the strains tested.
DISCUSSION RESULTS
All organisms in the nutritionally deficient streptococci (NDS) group were pyridoxal-requiring bacteria as determined by their absence of growth without pyridoxal but ability to have satelliting growth around Staphylococcus aureus, and identification of chromophore after acid extraction. In addition, one organism CCGA, was a ~-hemolytic group-A antigen-positive streptococci that met the prime criteria for NDS, but was chromophore negative (Stein and Libertin, 1989a). The organisms' biotypes are listed in Table 1. Of the 23 strains, 11 (47.8%) were biotype 1, four (17.4%) biotype 2, and seven (30.4%) biotype 3. Only three organisms from the group of nutritionally deficient streptococci had an atypical biochemical profile. According to the scheme proposed by Bouvet et al. (1989), all biotype definitions included a positive test for pyrrolidonylarylamidase. However, one of these isolates (AB8611) was negative for the pyrrolidonylarylamidase test despite four separate trials of testing and prolongation of the incubation period to 24 hr. AB8611 was therefore designated nontypeable. Minor variations were made in assigning biotypes for two other organisms. Strain BA8212 had a positive result in the alkaline phosphatase (PAL) test; and strain LR8604 best fit a biotype-3 classification though the results did not affect biotype identification. PAL was positive for CCGA, BA8212, and AB8611-LU; ADH was positive for LR8604, CCGA and 20559; and acid was produced from mannitol for CCGA, BC840, and AB8611-LU. These results were not reported in a prior biochemical survey of different strains (Bouvet et al., 1985). Marked genetic heterogeneity among the NDS was found. The chromosomal DNA probes for the biotype 1 (6995) and biotype 2 (20559) did not significantly hybridize to each other. The 6995 probe had little homology to any of the 21 NDS strains tested
Our results indicate genetic heterogeneity among the strains of nutritionally deficient streptococci that we compared. Only four strains of NDS among the biotypes 2 and 3 were highly homologous to each other as determined by the membrane-bound method of D N A - D N A hybridization, a method used in other studies of taxonomic classification (Coykendall et al., 1975 and 1987; Ezaki et al., 1988; Roberts et al., 1987; Welborn et al., 1983). The NDS strains 20559 was 90% homologous to the NDS 7548, 76% homologous to the NDS LM8601, and 71.4% homologous to the NDS strain MB. These strains were isolated at four different institutions, thereby making it highly unlikely that they share any epidemiologic link. These four strains are of the same species by both the classic taxonomic criteria of being >70% homologous to the control strain (Brenner, 1973) and by the more recent taxonomic recommendations (Wayne et al., 1987). The majority of the NDS strains that we tested did not meet the classification scheme that is composed of two species, S. defectivus spp. nov. and S. adjacens spp. nov., proposed by Bouvet et al. (1989). Our results indicate that some NDS in biotypes 2 and 3, which would be designated S. adjacens spp. nov., are related at the species level, but not all of them. Seven of the strains in biotypes 2 and 3 did not show a high degree of homology. Only one of these strains (20559) was atypical biochemically in that it hydrolyzed arginine which was unlike the proposed species S. adjacens spp. nov.; all others were biochemically similar. In addition, we were unable to demonstrate a homologous relationship among NDS in biotype 1, S. defectivus spp. nov. Among these 11 strains, none were homologous to each other despite being the same biotype. Among the NDS in this study, marked genetic heterogeneity existed. Possibly, a group of unrelated strains exist that meet the biotyping requirements for the S. defectivus spp. nov. characterization that represent a heretofore previously undescribed species or
284
TABLE I
D.S. Stein a n d C.R. Libertin
Percent D N A - D N A H o m o l o g y A m o n g Nutritionally Deficient Streptococci Percent Hybridizationa to Probe DNA from
Biotype 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 3 3 3 3 3 3 3 NT
NDS Chromosomal DNBA
20559
1178
6995
15909
93634 16316 19363 8060 BA8212b JV8611 BC7729 SS8706 BC7901 MC8 6995 20559 MC4 LM8601 CCGAc L61 7548 8246 CL8707 BC840-2 LR8604~ MB AB8611 S. sanguis ATCC 15909 S. mitis ATCC 15911 S. mitis CDC 1178
1.0 1.3 0.1 0.2 6.9 0.3 0.2 1.3 0.5 0.3 27.2 100.0 12.0 76.0 2.3 4.5 90.0 6.0 35.5 6.1 2.8 71.4 3.3 6.9 12.5 8.4
0.4 0.3 ND ND 3.3 0.2 ND 0.1 ND ND 45.0 -1.2 18.9 2.2 0.4 26.8 0.9 5.6 0.2 2.9 18.4 3.1 6.7 12.9 100.0
3.3 1.7 2.6 1.0 2.1 0.4 0.9 ND ND 1.7 100.0 4.9 1.8 8.0 ND 1.2 12.9 2.9 3.0 ND 8.0 8.9 5.4 42.3 42.3 --
0.6 0.5 ND ND 0.2 0.1 ND ND 0.2 0.3 20.5 0.2 0.6 1.7 0.2 0.5 5.1 0.6 0.6 0.2 1.0 1.9 1.1 100.0 30.8 --
NT, not typeable because pyrrolidonylarylamidase test was negative; NA, not applicable; ND, not detectable; and --, redundant study. See Materials and Methods for explanation of test procedures and biotype assignment. aData expressed as the percentage of the DNA control. Data shown are for 37°C hybridization (optimal conditions). All organisms' DNAs were hybridized simultaneously on the same filter; all results were reproduced in triplicate. Data represent one experiment. bPositve alkaline phosphatase test. cPositive alkaline phosphatase and arginine dehydrolase tests for production of acid from trehalose and chromophore test were negative. e~-Galactosidase test were positive.
a variant of the p r o p o s e d type strain. A direct comparison of the NDS strains to those used by Bouvet et al. (1989) w o u l d be necessary to explore this possibility further. The absence of significant genetic relatedness a m o n g NDS a n d viridans streptococci in this s t u d y is consistent with the w o r k b y Bouvet et al. (1989). In fact, others have s h o w n that within the viridans group of streptococci, organisms speciated biochemically do not hybridize as expected to viridans streptococci of the same species (Ezaki et al., 1988). The n u m b e r of viridans streptococcal strains that were c o m p a r e d to the NDS in this s t u d y a n d so far to date are probably too small to conclude that NDS are not h o m o l o g o u s
to some species of viridans streptococci. As noted, the p r o p o s e d taxonomical relationship a m o n g viridans streptococci remain unclear; therefore, the relation° ship of NDS to the viridans g r o u p of streptococci must be done with caution. Further investigations into these relationships m a y yield a better u n d e r s t a n d i n g of the p a t h o g e n e s i s of these o r g a n i s m s a n d lead to insights into i m p r o v e d clinical m a n a g e m e n t of patients infected with these organisms. The authors thank Drs. Roberta Carey and Walter Wilson for their kind gifts of NDS. This research was supported in part by the estate of William G. Potts.
Nutritionally Deficient Streptococci
285
REFERENCES Bisno, AL, Dismukes WE, Durack DT, et al. (1989). Antimicrobial treatment of infectious endocarditis due to viridans streptococci, enterococci, and staphylococci. JAMA 261:1471-1477. Bouvet A, Grimond F, Grimond PAD (1989) Streptococcus defectivities sp. nov. and Streptococcus adjacens sp. nov., nutritionally variant streptococci from human clinical specimens. Int J Syst Bacteriol 39:290-294. Bouvet A, van de Rijn I, McCarty M (1981) Nutritionally variant streptococci from patients with endocarditis: growth parameters in a semisynthetic medium and demonstration of a chromophore. J Bacteriol 146:10751082. Bouvet, A, Villeroy R, Cheng F, et al. (1985) Characterization of nutritionally variant streptococci by biochemical tests and penicillin binding proteins. J Clin Microbiol 22:1030-1034. Brenner DJ (1973) Deoxyribonucleic acid reassociation in the taxonomy of enteric bacteria. Int J System Bacteriol 23:298-307. Coykendall AL, Specht PA (1975) DNA base sequence homologies among strains of Streptococcus sanguis. J Gen Microbiol 91:92-98. Coykendall AL, Wesbecher PM, Gustafson KB (1987) Streptococcus milleri, Streptococcus constellatus, and Streptococcus intermedius are later synonyms of Streptococcus anginosus. Int J Syst Bacteriol 37:222-228. Davis LG, Dibner MD, Battey JF (1986) Methods in Molecular Biology. New York: Elsevier. Ezaki T, Hashimoto Y, Takeuchi N, et al. (1988) Simple genetic method to identify viridans streptococci by colorimetric dot hybridization and fluorometric hybridization in microdilution wells. J CLin Microbiol 26:17081713. Frenkel A, Hirsch W (1961) Spontaneous development of L forms of streptococci requiring secretions of other bacteria or sulphydryl compounds for normal growth. Nature 191:728-730. Johnson JL (1986) Nucleic acid in bacterial classification. In Bergey'sManual of Systematic Bacteriology. Ed, NR Krieg
and JG Hott. Baltimore, MD: Williams and Wilkins, pp 8-11. Maurer R (1986) Laboratory Procedure Manual Cold Spring Harbor Symposium 1986. New York: Cold Spring Harbor. Roberts RB, Krieger AG, Schiller NL, Gross KC (1979) Viridans streptococcal endocarditis: the role of various species including pyridoxal dependent streptococci. Rev Infect Dis 1:955-965. Roberts MC, McMillan C, Coyle MB (1987) Whole chromosomal DNA probes for rapid identification of Mycobacterium tuberculosis and Mycobacterium avium complex. J Clin Microbiol 25:1239-1243. Ruoff KL (1991) Nutritionally variant streptococci. Clin Microbiol Rev 4:184-190. Shields MS, Kline BC, Tam JE (1986) A rapid method for the quantitative measurement of gene dosage: mini-F plasmid concentration as a function of cell growth rate. J Microbiol Methods 6:33-46. Stein DS, Libertin CR (1989a) A double-blinded comparative evaluation of three media for chromophore testing with viridans and nutritionally variant (deficient) streptococci. Am J Clin Pathol 91:589-593. Stein DS, Libertin CR (1989b) Molecular analysis of viridans and nutritionally deficient (variant) streptococci causing sequential episodes of endocarditis in a patient. Am J Clin Pathol 91:620-624. Stein DS, Nelson KE (1987) Endocarditis due to nutritionally deficient streptococci: a therapeutic dilemma. Rev Infect Dis 9:908-916. Van de Rijn I, George M, Bouvet A, Roberts RB (1979) Enzyme linked immunosorbent assay for the detection of antibodies to nutritionally variant streptococci in patients with endocarditis. J Infect Dis 1:955-965. Wayne L, Brenner DJ, Colwell P (1987) Report of the Ad Hoc Committee on Reconciliation of Approaches to Bacterial Systematics. Int J Syst Bacteriol 37:463-464. Welborn PP, Hadley WK, Newbrun E, Yajko DM (1983) Characterization of strains of viridans streptococci by deoxyribonucleic acid hybridization and physiological tests. Int J Syst Bacteriol 33:293-299.
