Vol. 60, No. 3
INFECTION AND IMMUNITY, Mar. 1992, p. 1047-1054 0019-9567/92/031047-08$02.00/0 Copyright © 1992, American Society for Microbiology
Conservation of an Actinomyces viscosus T14V Type 1 Fimbrial Subunit Homolog among Divergent Groups of Actinomyces spp. MARIA K. YEUNG Departments of Pediatric Dentistry and Microbiology, The University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Dnive, San Antonio, Texas 78284 Received 18 October 1991/Accepted 18 December 1991
The type 1 fimbrial subunit gene of the human Actinomyces viscosus T14V was used as a DNA probe in Southern analyses to detect related DNA sequences in 16 of 30 strains ofActinomyces spp. under conditions of high stringency. The organisms with homology to the DNA probe included two human and six nonhuman A. viscosus, three human and three nonhuman A. naeslundii, and two A. bovis isolates. Homologous DNA sequences were not detected in strains of A. odontolyticus and A. israelii examined in this study. Northern (RNA) blot analysis revealed expression of a transcript from each of the A. viscosus and A. naeslundii strains and from one A. bovis strain that was comparable in size to that detected from A. viscosus T14V. Cell surface fimbriae were observed on a majority of the strains that expressed the transcript. Various degrees of cross-immunoreactivities between these strains and antibodies specific for type 1 fimbriae ofA. viscosus T14V were also observed by colony immunoassay. Thus, the data clearly demonstrate the existence in, and expression by, divergent Actinomyces groups of genomic sequences that are closely related to the type 1 fimbriae of A. viscosus T14V.
from another gram-positive bacterium, Streptococcus sanguis FW213 (18), and pilins from gram-negative organisms. Thus, whereas significant homologies exist among pilins from a wide range of gram-negative species (16, 29, 31, 33), fimbriae from Actinomyces species may have evolved from an ancestor different from that of the other bacterial fimbriae. However, the extent of genetic conservation among fimbriae of various Actinomyces species is unknown. This study was designed to examine the distribution and expression of genetic information related to A. viscosus T14VfimP (formerly the type 1 fimbrial subunit gene) among human and nonhuman strains of Actinomyces spp. The occurrence of homologous DNA sequences in a taxonomically related fimbriated strain, Corynebacterium renale (25), also was investigated. The strains with cross-hybridizing DNA sequences were assessed further for their ability to adhere to a simulated tooth surface in an in vitro assay described previously (21).
Results from studies of oral microorganisms in recent decades have shown that specific groups of bacteria are associated with different stages of plaque formation (17, 22, 30). The genusActinomyces is a member of a limited number of gram-positive bacteria that are regarded as primary colonizers of the tooth surface and that are isolated in high numbers during the early stages of plaque development (17, 37). The initial attachment by the Actinomyces spp. promotes the adherence of gram-negative bacteria in periodontal pockets and subsequent tissue inflammation and destruction (17, 28, 36). Fimbriae of human strains of Actinomyces spp. are involved in the adherence of these organisms to host tissues (3). Two distinct types of fimbriae, designated types 1 and 2, have been identified on typical strains of A. viscosus, while only one fimbrial type (type 2) is present on typical strains of A. naeslundii. The type 1 fimbriae of A. viscosus mediate bacterial adherence to the tooth surface (10). In contrast, type 2 fimbriae of A. viscosus and A. naeslundii are sites of a lactose-sensitive lectin involved in interactions of Actinomyces spp. with sialidase-treated mammalian cells (2, 6, 35) and with certain strains of Streptococcus spp. (3). The genes that encode the structural subunits of type 1 and type 2 fimbriae of A. viscosus T14V (13, 40) and the type 2 fimbriae of A. naeslundii WVU45 (41) have been cloned, and their DNA sequences have been determined (14, 41, 42). Although the type 1 fimbriae of A. viscosus T14V and type 2 fimbriae of A. naeslundii WVU45 possess distinct functional and immunological properties (3, 6, 8), the fimbrial subunits share 34% amino acid sequence identity with several highly conserved domains distributed in both the amino- and carboxyl-terminal ends of each protein (42). Moreover, greater than 70% sequence identity was observed between the subunits of the A. viscosus T14V and A. naeslundii WVU45 (14, 41) type 2 fimbriae that shared weak immunoreactivities (4, 6). In contrast, little or no homology was detected between Actinomyces fimbrial subunits and those of various bacterial cell surface proteins, including a fimbrial subunit
MATERUILS AND METHODS Bacterial strains and plasmids. The bacterial strains examined in the present study are summarized in Table 1. The list includes members of fiveActinomyces spp. that are placed in various clusters or taxonomical groups based on nucleic acid homologies (12, 27) and serological and/or physiological characteristics (19, 20, 24). Actinomyces strains were grown routinely in a complex medium (7) containing 0.2% glucose. For the preparation of nucleic acids, bacteria were grown in lactobacillus carrying medium (15) supplemented with 20 mM DL-threonine. The plasmid pMY3833 containing A. viscosus T14V fimP was obtained from a previous study (40). The 1.85-kb SstI-BamHI fragment that contained fimP was used as a DNA probe to screen a genomic library ofA. viscosus T14V DNA in pUC13 (38, 40). An Escherichia coli strain harboring pMY1012 was isolated that contained, in addition to fimP, flanking sequences at both the 5' and the 3' ends of the gene. Chromosomal DNA isolation and Southern analyses. Acti1047
1048
YEUNG
INFEC. IMMUN. TABLE 1. Bacterial strains
Straina
Actinomyces viscosus T14V (1+ 2+) 5519 (1+ 2-)c 5951 (1- 2+)c 147 (1- 2-)c ATCC 19246 WVU627 T6-1600
A828d R28 ATCC 27045 ATCC 27046 M3902 M4301 M4166 Actinomyces naeslundii WVU45e ATCC 19039
25f 29 N16f M4110 M4356 M4730
Host origin
Hybridization to A. viscosus T14V type 1 fimbrial gene"
Human Human Human Human Human Human Hamster Hamster Rat Dog Dog Monkey Monkey Monkey
+ + + + + + + + +
Human Human Human Human Human Monkey Monkey Monkey
Other strain designation or reference
+ + +
7 7 7 7 WVU371 (19) 19 WVU745 (7) ATCC 15987 (7) 7 A1231 (12) A1308 (12) This study This study This study
+ + + + + +
ATCC 12104 (7) WVU447 (12) W1544 (5) I (5) WVU820 (20) This study This study This study
Actinomyces israelii X522 W855 ATCC 12103 ATCC 12597 wio1 ATCC 27037
Human Human Human Human Human Human
ATCC 10048 ATCC 12102 (27) W726 (24) X373 ATCC 29322 (7) W748
Actinomyces odontolyticus X363 ATCC 17982 ATCC 29323
Human Human Human
ATCC 17929 (20)
Actinomyces bovis W827 ATCC 19012 ATCC 19013
Bovine Bovine Bovine
Corynebacterium renale
Human
7
WVU482 (20) + +
ATCC 13683 (20) A-13-S A-13-R 25
ATCC 19412 a Strains with an ATCC designation were obtained from American Type Culture Collection, Rockville, Md.; strains with numbers preceded by the letter M were isolates from monkey dental plaques obtained from J. E. Delaney, The University of Texas Health Science Center at San Antonio; all other strains were obtained from J. 0. Cisar, National Institute of Dental Research. b Determined by dot-blot hybridization of genomic DNA to fimP of A. viscosus T14V under a condition that allowed 15% nucleotide base mismatch. The + sign denotes hybridization signal at least 10 times above that of background signal, and the - sign indicates no signal, or an intensity of hybridization signal similar to that of a nonspecific background signal. c Variants of A. viscosus T14V that lack a specific type of fimbriae indicated in parentheses. d Recently classified as a member of A. viscosus serotype 1 (27). e Recently classified as a member of A. naeslundii genospecies 1 (27). f Recently classified as a member of A. naeslundii genospecies 2 (27).
nomyces spp. genomic DNA was prepared as described previously (13). Genomic DNA (1 to 2 ,ug) was digested to completion with restriction endonucleases (GIBCO BRL Life Technologies, Inc., Gaithersburg, Md.). The digested DNA was subjected to electrophoresis in 1% agarose gels in 0.04 M Tris-acetate-0.002 M EDTA (pH 8.0) and transferred to GeneScreen (Dupont, NEN Research Products, Boston, Mass.) following a procedure recommended by the manufacturer. The DNA on the filters was prehybridized at 42°C
for 2 to 4 h in a solution containing 5 x SSC (lx SSC is 0.15 M NaCl plus 0.015 M sodium citrate), 50% formamide, 10% dextran sulfate (Pharmacia LKB Biotechnology Inc., Piscataway, N.J.), 1x Denhardt's solution (34), 1% sodium dodecyl sulfate (SDS), 1 M NaCl, 0.5% sodium PP1, and 200 ,ug of denatured DNA from herring sperm (Boehringer Mannheim Biochemicals, Indianapolis, Ind.) per ml. a-35S-dCTP (1,000 to 1,500 Ci/mmol; Dupont, NEN Research Products)-labeled DNA prepared by nick translation (34) was added, and
VOL. 60, 1992
CONSERVATION OF AN A. VISCOSUS FIMBRIAL SUBUNIT
hybridization continued at the same temperature for 18 to 20 h. The filters were washed with 0.1 x SSC-0.5% SDS at 65°C for 1 h with one change of buffer, air dried, and exposed to Kodak XAR-5 0-mat film (Eastman Kodak Co., Rochester, N.Y.). RNA isolation and Northern (RNA) analyses. Bacteria from a 100-ml culture in lactobacillus carrying medium supplemented with 20 mM DL-threonine were harvested at the mid-exponential phase of growth (optical density at 660 nm = 0.5). The cells were suspended in a solution containing 20% polyethylene glycol (molecular weight = 6,000; Sigma Chemical Co., St. Louis, Mo.), 1 mM MgCl2, and 20 mM Tris (pH 8.0) and digested with 100 mg of lysozyme (Sigma) at 37°C for 30 min. More than 90% of whole bacteria were converted to protoplasts under this condition. The protoplasts were lysed with a 3-ml ice-cold solution containing 4.2 M guanidinium thiocyanate (Fluka Chemical Corp., Ronkonkoma, N.Y.) and 1% lauryl sulfate (Sigma). Total RNA was sedimented by centrifugation on a 5.7 M CsCl-100 mM EDTA (pH 7.0) cushion as described by Sambrook et al. (34). Total RNA (10 to 15 ,ug) was denatured in a solution containing 1 M glyoxal (Sigma) and 50% dimethylsulfonate (Aldrich Chemical Co., Madison, Wis.) at 50°C for 1 h and applied to a 1% agarose gel in 10 mM sodium phosphate (pH 7.0) containing 10 mM sodium iodoacetate (Sigma). The nucleic acids were transferred to GeneScreen, and the RNA on the filters was hybridized to the DNA probe labeled with 35S-dCTP under conditions that permitted 15% nucleotide base mismatch. After hybridization, the filters were washed, in sequence, in 5x SSC-0.1% SDS at room temperature for 10 min and in 0.5 x SSC-0.5% SDS at 650C for 1 h with two changes of buffer. In vitro adherence assay. Bacteria were washed in Trisbuffered saline (TBS; 0.15 M NaCl, 0.02 M Tris-hydrochloride [pH 7.8], 0.1 mM CaCl2, 0.1 mM MgCl2, 0.02% sodium azide), and the cell density was adjusted to 109 cells per ml in the same buffer containing 2 mg of bovine serum albumin (BSA) per ml (optical density at 660 nm = 2.0 with a spectrophotometer [DU-6; Beckman Instruments, Inc., Fullerton, Calif.]). Human salivary acidic proline-rich proteins (PRPs) were a generous gift from D. A. Johnson, The University of Texas Health Science Center at San Antonio. The PRPs were purified from human parotid saliva collected from 10 people. Briefly, desalted whole saliva was brought to 45% saturation with solid ammonium sulfate. The precipitated proteins were applied to a DEAE-Sephadex A25 column (Pharmacia LKB Biotechnology) and eluted with a sodium chloride gradient (0 to 1.0 M). Two acidic protein fractions (pools I and II) both devoid of amylase activity were obtained. Proteins from pool II were used in the present study. Analyses by SDS-polyacrylamide gel electrophoresis, anionic polyacrylamide gel electrophoresis, and size exclusion high-pressure liquid column chromatography showed that pool II contained primarily the 16,300-Da protein species PRP-1, PRP-2, and parotid isoelectric-focusing variant (PIF-s) (23, 26). Latex beads (5 mg, 15.7 ,um in diameter; BDH Chemicals, Gallard-Schlesinger Chemical Corp., Carle Place, N.Y.) were coated with PRPs at 50 ,ug/ml in 0.05 M carbonate buffer (pH 9.5) at 40C overnight. Unadsorbed proteins were removed by extensive washes with TBS. The protein-coated bead suspension (50 Rl) was mixed with 50 ,ul of a washed bacterial cell suspension containing 5 x 108 cells in each well of a 96-well microtiter plate as described previously (21). A quantitative adherence
1049
assay using [3H]thymidine-labeled bacterial cells as described by Gibbons et al. (21) also was performed. Colony immunoassay. The immunological properties of the Actinomyces strains were examined by following a colony immunoassay protocol as described previously (40). Nitrocellulose filter strips containing bacteria (107 cells per spot) were incubated with different antibodies directed against A. viscosus T14V whole bacteria and fimbriae prepared from this strain. The specificities of the different antibodies have been described previously (5, 8). Electron microscopy. Bacteria were washed in TBS and suspended in the same buffer supplemented with 1% BSA. A small droplet (2 to 5 ,ul) of the bacterial suspension was placed on a copper grid coated with a thin film of 2% Formvar. Excess liquid was removed, and a drop of 1% phosphotungstic acid was immediately applied for 10 s. The drop of phosphotungstic acid was removed, and the dried film was examined with a JEOL 1200 EX microscope (JEOL, Ltd., Tokyo, Japan). RESULTS Detection of DNA sequences homologous to fimP among Actinomyces strains. Genomic DNA from Actinomyces strains was examined initially by dot-blot hybridization for the presence of DNA sequences related to the A. viscosus T14V type 1 fimbrial subunit gene, fimP. The probe was the 1.85-kb SstI-BamHI fragment from pMY3833 (40) that contained fimP. Hybridization at various conditions of stringency was achieved by adjusting the percentage of formamide in the reaction solutions (34). Chromosomal DNA from 30 Actinomyces strains was tested, and 16 DNAs produced strong hybridization signals under conditions that permitted 30% nucleotide base mismatch. These included genomic DNA from 8 of 10A. viscosus, 6 of 8A. naeslundii, and 2 of 3 A. bovis strains. Hybridization signals of comparable intensities were obtained when the formamide concentration was at a level that permitted 15% mismatch (Table 1). Under conditions that allowed only 10% mismatch, hybridization signals detected in human A. viscosus and human A. naeslundii strains remained comparable to those observed under conditions permitting 15% mismatch, while those detected in rodent A. viscosus and monkey A. viscosus and A. naeslundii were reduced by four- and onefold, respectively. Hybridization with DNA from A. bovis strains was not detected under the conditions described. DNA sequences homologous tofimP were not detected in any of the A. odontolyticus and A. israelii strains at a low signal-tonoise ratio that allowed 70% mismatch (39). Similarly, shared homology between Actinomyces strains and C. renale was not detected (Table 1). Thus, A. viscosus T14V fimP is highly conserved among A. viscosus and A. naeslundii from a wide range of host origins, as well as in certain
strains of A. bovis.
Restriction fragment length polymorphism associated with
fimP. Five DNA fragments (Fig. 1, probes A through E) that
were internal to and flanking A. viscosus T14V fimP were used to hybridize to BamHI-digested genomic DNA from each of the 16 Actinomyces strains selected by dot-blot analysis (Table 1). As shown in Fig. 2, panel I, probe B (containingfimP) hybridized to two bands of approximately 2 and 0.4 kb in all the rodent A. viscosus strains and a single fragment of approximately 7 kb in all the monkeyA. viscosus and A. naeslundii isolates. In contrast, the DNA bands recognized by probe B in each of the human A. viscosus strains were approximately 3 or 7 kb, those in human A.
1050
YEUNG
INFECT. IMMUN. Subunit Coding Region ;N
0
cl; _ .n
Hindill Sstl
Sall
Pstl
BamHI
BamHl
_ oL-;.
_"I,.i>
C~ 3: C-; E
B
".-
0.
0
L
1-
I-
pUC13
A
11,
-
:-
z
INFECTION AND IMMUNITY, Mar. 1992, p. 1047-1054 0019-9567/92/031047-08$02.00/0 Copyright © 1992, American Society for Microbiology
Conservation of an Actinomyces viscosus T14V Type 1 Fimbrial Subunit Homolog among Divergent Groups of Actinomyces spp. MARIA K. YEUNG Departments of Pediatric Dentistry and Microbiology, The University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Dnive, San Antonio, Texas 78284 Received 18 October 1991/Accepted 18 December 1991
The type 1 fimbrial subunit gene of the human Actinomyces viscosus T14V was used as a DNA probe in Southern analyses to detect related DNA sequences in 16 of 30 strains ofActinomyces spp. under conditions of high stringency. The organisms with homology to the DNA probe included two human and six nonhuman A. viscosus, three human and three nonhuman A. naeslundii, and two A. bovis isolates. Homologous DNA sequences were not detected in strains of A. odontolyticus and A. israelii examined in this study. Northern (RNA) blot analysis revealed expression of a transcript from each of the A. viscosus and A. naeslundii strains and from one A. bovis strain that was comparable in size to that detected from A. viscosus T14V. Cell surface fimbriae were observed on a majority of the strains that expressed the transcript. Various degrees of cross-immunoreactivities between these strains and antibodies specific for type 1 fimbriae ofA. viscosus T14V were also observed by colony immunoassay. Thus, the data clearly demonstrate the existence in, and expression by, divergent Actinomyces groups of genomic sequences that are closely related to the type 1 fimbriae of A. viscosus T14V.
from another gram-positive bacterium, Streptococcus sanguis FW213 (18), and pilins from gram-negative organisms. Thus, whereas significant homologies exist among pilins from a wide range of gram-negative species (16, 29, 31, 33), fimbriae from Actinomyces species may have evolved from an ancestor different from that of the other bacterial fimbriae. However, the extent of genetic conservation among fimbriae of various Actinomyces species is unknown. This study was designed to examine the distribution and expression of genetic information related to A. viscosus T14VfimP (formerly the type 1 fimbrial subunit gene) among human and nonhuman strains of Actinomyces spp. The occurrence of homologous DNA sequences in a taxonomically related fimbriated strain, Corynebacterium renale (25), also was investigated. The strains with cross-hybridizing DNA sequences were assessed further for their ability to adhere to a simulated tooth surface in an in vitro assay described previously (21).
Results from studies of oral microorganisms in recent decades have shown that specific groups of bacteria are associated with different stages of plaque formation (17, 22, 30). The genusActinomyces is a member of a limited number of gram-positive bacteria that are regarded as primary colonizers of the tooth surface and that are isolated in high numbers during the early stages of plaque development (17, 37). The initial attachment by the Actinomyces spp. promotes the adherence of gram-negative bacteria in periodontal pockets and subsequent tissue inflammation and destruction (17, 28, 36). Fimbriae of human strains of Actinomyces spp. are involved in the adherence of these organisms to host tissues (3). Two distinct types of fimbriae, designated types 1 and 2, have been identified on typical strains of A. viscosus, while only one fimbrial type (type 2) is present on typical strains of A. naeslundii. The type 1 fimbriae of A. viscosus mediate bacterial adherence to the tooth surface (10). In contrast, type 2 fimbriae of A. viscosus and A. naeslundii are sites of a lactose-sensitive lectin involved in interactions of Actinomyces spp. with sialidase-treated mammalian cells (2, 6, 35) and with certain strains of Streptococcus spp. (3). The genes that encode the structural subunits of type 1 and type 2 fimbriae of A. viscosus T14V (13, 40) and the type 2 fimbriae of A. naeslundii WVU45 (41) have been cloned, and their DNA sequences have been determined (14, 41, 42). Although the type 1 fimbriae of A. viscosus T14V and type 2 fimbriae of A. naeslundii WVU45 possess distinct functional and immunological properties (3, 6, 8), the fimbrial subunits share 34% amino acid sequence identity with several highly conserved domains distributed in both the amino- and carboxyl-terminal ends of each protein (42). Moreover, greater than 70% sequence identity was observed between the subunits of the A. viscosus T14V and A. naeslundii WVU45 (14, 41) type 2 fimbriae that shared weak immunoreactivities (4, 6). In contrast, little or no homology was detected between Actinomyces fimbrial subunits and those of various bacterial cell surface proteins, including a fimbrial subunit
MATERUILS AND METHODS Bacterial strains and plasmids. The bacterial strains examined in the present study are summarized in Table 1. The list includes members of fiveActinomyces spp. that are placed in various clusters or taxonomical groups based on nucleic acid homologies (12, 27) and serological and/or physiological characteristics (19, 20, 24). Actinomyces strains were grown routinely in a complex medium (7) containing 0.2% glucose. For the preparation of nucleic acids, bacteria were grown in lactobacillus carrying medium (15) supplemented with 20 mM DL-threonine. The plasmid pMY3833 containing A. viscosus T14V fimP was obtained from a previous study (40). The 1.85-kb SstI-BamHI fragment that contained fimP was used as a DNA probe to screen a genomic library ofA. viscosus T14V DNA in pUC13 (38, 40). An Escherichia coli strain harboring pMY1012 was isolated that contained, in addition to fimP, flanking sequences at both the 5' and the 3' ends of the gene. Chromosomal DNA isolation and Southern analyses. Acti1047
1048
YEUNG
INFEC. IMMUN. TABLE 1. Bacterial strains
Straina
Actinomyces viscosus T14V (1+ 2+) 5519 (1+ 2-)c 5951 (1- 2+)c 147 (1- 2-)c ATCC 19246 WVU627 T6-1600
A828d R28 ATCC 27045 ATCC 27046 M3902 M4301 M4166 Actinomyces naeslundii WVU45e ATCC 19039
25f 29 N16f M4110 M4356 M4730
Host origin
Hybridization to A. viscosus T14V type 1 fimbrial gene"
Human Human Human Human Human Human Hamster Hamster Rat Dog Dog Monkey Monkey Monkey
+ + + + + + + + +
Human Human Human Human Human Monkey Monkey Monkey
Other strain designation or reference
+ + +
7 7 7 7 WVU371 (19) 19 WVU745 (7) ATCC 15987 (7) 7 A1231 (12) A1308 (12) This study This study This study
+ + + + + +
ATCC 12104 (7) WVU447 (12) W1544 (5) I (5) WVU820 (20) This study This study This study
Actinomyces israelii X522 W855 ATCC 12103 ATCC 12597 wio1 ATCC 27037
Human Human Human Human Human Human
ATCC 10048 ATCC 12102 (27) W726 (24) X373 ATCC 29322 (7) W748
Actinomyces odontolyticus X363 ATCC 17982 ATCC 29323
Human Human Human
ATCC 17929 (20)
Actinomyces bovis W827 ATCC 19012 ATCC 19013
Bovine Bovine Bovine
Corynebacterium renale
Human
7
WVU482 (20) + +
ATCC 13683 (20) A-13-S A-13-R 25
ATCC 19412 a Strains with an ATCC designation were obtained from American Type Culture Collection, Rockville, Md.; strains with numbers preceded by the letter M were isolates from monkey dental plaques obtained from J. E. Delaney, The University of Texas Health Science Center at San Antonio; all other strains were obtained from J. 0. Cisar, National Institute of Dental Research. b Determined by dot-blot hybridization of genomic DNA to fimP of A. viscosus T14V under a condition that allowed 15% nucleotide base mismatch. The + sign denotes hybridization signal at least 10 times above that of background signal, and the - sign indicates no signal, or an intensity of hybridization signal similar to that of a nonspecific background signal. c Variants of A. viscosus T14V that lack a specific type of fimbriae indicated in parentheses. d Recently classified as a member of A. viscosus serotype 1 (27). e Recently classified as a member of A. naeslundii genospecies 1 (27). f Recently classified as a member of A. naeslundii genospecies 2 (27).
nomyces spp. genomic DNA was prepared as described previously (13). Genomic DNA (1 to 2 ,ug) was digested to completion with restriction endonucleases (GIBCO BRL Life Technologies, Inc., Gaithersburg, Md.). The digested DNA was subjected to electrophoresis in 1% agarose gels in 0.04 M Tris-acetate-0.002 M EDTA (pH 8.0) and transferred to GeneScreen (Dupont, NEN Research Products, Boston, Mass.) following a procedure recommended by the manufacturer. The DNA on the filters was prehybridized at 42°C
for 2 to 4 h in a solution containing 5 x SSC (lx SSC is 0.15 M NaCl plus 0.015 M sodium citrate), 50% formamide, 10% dextran sulfate (Pharmacia LKB Biotechnology Inc., Piscataway, N.J.), 1x Denhardt's solution (34), 1% sodium dodecyl sulfate (SDS), 1 M NaCl, 0.5% sodium PP1, and 200 ,ug of denatured DNA from herring sperm (Boehringer Mannheim Biochemicals, Indianapolis, Ind.) per ml. a-35S-dCTP (1,000 to 1,500 Ci/mmol; Dupont, NEN Research Products)-labeled DNA prepared by nick translation (34) was added, and
VOL. 60, 1992
CONSERVATION OF AN A. VISCOSUS FIMBRIAL SUBUNIT
hybridization continued at the same temperature for 18 to 20 h. The filters were washed with 0.1 x SSC-0.5% SDS at 65°C for 1 h with one change of buffer, air dried, and exposed to Kodak XAR-5 0-mat film (Eastman Kodak Co., Rochester, N.Y.). RNA isolation and Northern (RNA) analyses. Bacteria from a 100-ml culture in lactobacillus carrying medium supplemented with 20 mM DL-threonine were harvested at the mid-exponential phase of growth (optical density at 660 nm = 0.5). The cells were suspended in a solution containing 20% polyethylene glycol (molecular weight = 6,000; Sigma Chemical Co., St. Louis, Mo.), 1 mM MgCl2, and 20 mM Tris (pH 8.0) and digested with 100 mg of lysozyme (Sigma) at 37°C for 30 min. More than 90% of whole bacteria were converted to protoplasts under this condition. The protoplasts were lysed with a 3-ml ice-cold solution containing 4.2 M guanidinium thiocyanate (Fluka Chemical Corp., Ronkonkoma, N.Y.) and 1% lauryl sulfate (Sigma). Total RNA was sedimented by centrifugation on a 5.7 M CsCl-100 mM EDTA (pH 7.0) cushion as described by Sambrook et al. (34). Total RNA (10 to 15 ,ug) was denatured in a solution containing 1 M glyoxal (Sigma) and 50% dimethylsulfonate (Aldrich Chemical Co., Madison, Wis.) at 50°C for 1 h and applied to a 1% agarose gel in 10 mM sodium phosphate (pH 7.0) containing 10 mM sodium iodoacetate (Sigma). The nucleic acids were transferred to GeneScreen, and the RNA on the filters was hybridized to the DNA probe labeled with 35S-dCTP under conditions that permitted 15% nucleotide base mismatch. After hybridization, the filters were washed, in sequence, in 5x SSC-0.1% SDS at room temperature for 10 min and in 0.5 x SSC-0.5% SDS at 650C for 1 h with two changes of buffer. In vitro adherence assay. Bacteria were washed in Trisbuffered saline (TBS; 0.15 M NaCl, 0.02 M Tris-hydrochloride [pH 7.8], 0.1 mM CaCl2, 0.1 mM MgCl2, 0.02% sodium azide), and the cell density was adjusted to 109 cells per ml in the same buffer containing 2 mg of bovine serum albumin (BSA) per ml (optical density at 660 nm = 2.0 with a spectrophotometer [DU-6; Beckman Instruments, Inc., Fullerton, Calif.]). Human salivary acidic proline-rich proteins (PRPs) were a generous gift from D. A. Johnson, The University of Texas Health Science Center at San Antonio. The PRPs were purified from human parotid saliva collected from 10 people. Briefly, desalted whole saliva was brought to 45% saturation with solid ammonium sulfate. The precipitated proteins were applied to a DEAE-Sephadex A25 column (Pharmacia LKB Biotechnology) and eluted with a sodium chloride gradient (0 to 1.0 M). Two acidic protein fractions (pools I and II) both devoid of amylase activity were obtained. Proteins from pool II were used in the present study. Analyses by SDS-polyacrylamide gel electrophoresis, anionic polyacrylamide gel electrophoresis, and size exclusion high-pressure liquid column chromatography showed that pool II contained primarily the 16,300-Da protein species PRP-1, PRP-2, and parotid isoelectric-focusing variant (PIF-s) (23, 26). Latex beads (5 mg, 15.7 ,um in diameter; BDH Chemicals, Gallard-Schlesinger Chemical Corp., Carle Place, N.Y.) were coated with PRPs at 50 ,ug/ml in 0.05 M carbonate buffer (pH 9.5) at 40C overnight. Unadsorbed proteins were removed by extensive washes with TBS. The protein-coated bead suspension (50 Rl) was mixed with 50 ,ul of a washed bacterial cell suspension containing 5 x 108 cells in each well of a 96-well microtiter plate as described previously (21). A quantitative adherence
1049
assay using [3H]thymidine-labeled bacterial cells as described by Gibbons et al. (21) also was performed. Colony immunoassay. The immunological properties of the Actinomyces strains were examined by following a colony immunoassay protocol as described previously (40). Nitrocellulose filter strips containing bacteria (107 cells per spot) were incubated with different antibodies directed against A. viscosus T14V whole bacteria and fimbriae prepared from this strain. The specificities of the different antibodies have been described previously (5, 8). Electron microscopy. Bacteria were washed in TBS and suspended in the same buffer supplemented with 1% BSA. A small droplet (2 to 5 ,ul) of the bacterial suspension was placed on a copper grid coated with a thin film of 2% Formvar. Excess liquid was removed, and a drop of 1% phosphotungstic acid was immediately applied for 10 s. The drop of phosphotungstic acid was removed, and the dried film was examined with a JEOL 1200 EX microscope (JEOL, Ltd., Tokyo, Japan). RESULTS Detection of DNA sequences homologous to fimP among Actinomyces strains. Genomic DNA from Actinomyces strains was examined initially by dot-blot hybridization for the presence of DNA sequences related to the A. viscosus T14V type 1 fimbrial subunit gene, fimP. The probe was the 1.85-kb SstI-BamHI fragment from pMY3833 (40) that contained fimP. Hybridization at various conditions of stringency was achieved by adjusting the percentage of formamide in the reaction solutions (34). Chromosomal DNA from 30 Actinomyces strains was tested, and 16 DNAs produced strong hybridization signals under conditions that permitted 30% nucleotide base mismatch. These included genomic DNA from 8 of 10A. viscosus, 6 of 8A. naeslundii, and 2 of 3 A. bovis strains. Hybridization signals of comparable intensities were obtained when the formamide concentration was at a level that permitted 15% mismatch (Table 1). Under conditions that allowed only 10% mismatch, hybridization signals detected in human A. viscosus and human A. naeslundii strains remained comparable to those observed under conditions permitting 15% mismatch, while those detected in rodent A. viscosus and monkey A. viscosus and A. naeslundii were reduced by four- and onefold, respectively. Hybridization with DNA from A. bovis strains was not detected under the conditions described. DNA sequences homologous tofimP were not detected in any of the A. odontolyticus and A. israelii strains at a low signal-tonoise ratio that allowed 70% mismatch (39). Similarly, shared homology between Actinomyces strains and C. renale was not detected (Table 1). Thus, A. viscosus T14V fimP is highly conserved among A. viscosus and A. naeslundii from a wide range of host origins, as well as in certain
strains of A. bovis.
Restriction fragment length polymorphism associated with
fimP. Five DNA fragments (Fig. 1, probes A through E) that
were internal to and flanking A. viscosus T14V fimP were used to hybridize to BamHI-digested genomic DNA from each of the 16 Actinomyces strains selected by dot-blot analysis (Table 1). As shown in Fig. 2, panel I, probe B (containingfimP) hybridized to two bands of approximately 2 and 0.4 kb in all the rodent A. viscosus strains and a single fragment of approximately 7 kb in all the monkeyA. viscosus and A. naeslundii isolates. In contrast, the DNA bands recognized by probe B in each of the human A. viscosus strains were approximately 3 or 7 kb, those in human A.
1050
YEUNG
INFECT. IMMUN. Subunit Coding Region ;N
0
cl; _ .n
Hindill Sstl
Sall
Pstl
BamHI
BamHl
_ oL-;.
_"I,.i>
C~ 3: C-; E
B
".-
0.
0
L
1-
I-
pUC13
A
11,
-
:-
z
