Cooper, M. D., Z. A. McGee, M. H. Mulks, J. M. Koomey, and. T. L. Hindman. 1984. Attachment to ... Jack, G. W., and M. H. Richmond. 1970. A comparative study.
Vol. 56, No. 8
INFECTION AND IMMUNITY, Aug. 1988, p. 1961-1966
0019-9567/88/081961-06$02.00/0 Copyright © 1988, American Society for Microbiology
Cloning of the Gene Encoding Streptococcal Immunoglobulin A Protease and Its Expression in Escherichia coli JOANNE V. GILBERT,' ANDREW G. PLAUT,l* YOLANTA FISHMAN,2 AND ANDREW WRIGHT2 Division of Gastroenterology, Department of Medicine, Tufts-New England Medical Center Hospital,' and Department of Molecular Biology and Microbiology, Tufts University School of Medicine,2 Boston, Massachusetts 02111 Received 4 February 1988/Accepted 19 April 1988
We have identified and cloned a 6-kilobase-pair segment of chromosomal DNA from Streptococcus sanguis ATCC 10556 that encodes immunoglobulin A (IgA) protease activity when cloned into Escherichia coli. The enzyme specified by the iga gene in plasmid pJGl accumulates in the periplasm of E. coli MM294 cells and has a substrate specificity for human IgAl identical to that of native S. sanguis protease. Hybridization experiments with probes from within the encoding DNA showed no detectable homology at the nucleotide sequence level with chromosomal DNA of gram-negative bacteria that excrete IgA protease. Moreover, the S. sanguis iga gene probes showed no detectable hybridization with chromosomal DNA of S. pneumoniae, although the IgA proteases of these two streptococcal species cleaved the identical peptide bond in the human IgAl heavy-chain hinge region.
positive streptococci and gram-negative bacteria and to DNA obtained from enzyme-negative streptococci.
Immunoglobulin A (IgA) proteases are extracellular enpresent in culture filtrates of numerous human bacterial pathogens in the genera Streptococcus, Neisseria, Haemophilus, Clostridium, and Bacteroides (10, 21, 22, 26, 32). Despite their origins in a diverse group of bacteria, the IgA proteases share the feature of having human IgA as their only known substrate. Human IgAl in both serum and secretions contain a duplicated octapeptide sequence in the hinge region of their heavy (alpha) chains (35). This segment is rich in proline residues, and each IgA protease cleaves one of the peptide bonds to which proline contributes the carboxyl group. The specific bonds cleaved by enzymes from various pathogens are depicted in Fig. 1. The IgA proteases produced by the gram-negative organisms Neisseria gonorrhoeae, N. meningitidis, and Haemophilus influenzae have been studied extensively, and the iga genes encoding these proteases have been cloned (3, 17, 28). DNA sequence analysis and hybridization studies of iga genes indicate a high degree of homology, even between those from two different genera, Neisseria and Haemophilus (16; F. Grundy, unpublished observations). Streptococcus sanguis strains isolated from human dental plaque and strain ATCC 10556 were among the first bacteria found to produce IgA protease. As more enzyme-positive streptococci have been identified (15, 23), it has become clear that the IgA proteases of S. sanguis, S. mitior, and S. pneumoniae all cleave the same peptide bond in human IgAl, a single prolyl-threonyl bond in the octapeptide of the hinge duplication that lies most proximal to the amino terminus (Fig. 1). In view of this identical substrate specificity, it was predicted that these streptococcal enzymes would share common structural features. To begin an examination of these interrelationships, we have undertaken the cloning of genes specifying IgA protease from streptococci. In this paper we report the cloning and preliminary characterization of the iga gene from S. sanguis ATCC 10556 and identification of its translation product in Escherichia coli. Probes from the cloned gene have been used to explore its relationship to chromosomal DNA of other IgA proteasezymes
*
MATERIALS AND METHODS
Bacterial strains, media, and growth conditions. The bacteria used in this work are described in Table 1. All strains were cultivated at 37°C. Streptococci were grown in unmodified Todd-Hewitt broth (BBL Microbiology Systems, Cockeysville, Md.), with the exception of S. pneumoniae, for which the broth was supplemented with 0.5% yeast extract (Difco Laboratories, Detroit, Mich.). E. coli strains were cultivated in L broth (20). Neisseria strains were cultured in brain heart infusion (Difco) supplemented with 10% Isovitalex (BBL), and Haemophilus strains were cultured in brain heart infusion supplemented with 10 p,g of P-NAD per ml and 10 ,ug of hemin per ml (Sigma). Neisseria and Haemophilus strains were cultured in an atmosphere of 5% C02-95% air. The anaerobes Bacteroides melaninogenicus and Clostridium ramosum were grown in brain heart infusion containing 0.5% yeast extract and 5 ,ug of hemin per ml in an anaerobic chamber in an atmosphere of 5% C0210% H2-85% N2. When required, antibiotics were added to the media at the following concentrations: sodium ampicillin, 100 ,ug/ml; kanamycin sulfate, 30 ,ug/ml. General procedures. Genomic DNA was prepared from S. sanguis (5), S. pneumoniae (31), and S. mutans (25) by published methods. DNA was isolated from bacteriophage particles by the method of Berman et al. (1), and plasmid DNA was isolated by a modification of the method described by Birnboim and Doly (2). Restriction endonucleases and T4 DNA ligase were purchased from New England BioLabs, Inc., Beverly, Mass., and used as specified by the supplier. E. coli strains were transformed by the CaCl2 shock method of Berman et al. (1). Preparation of a streptococcal chromosomal DNA library. Chromosomal DNA was isolated from the IgA proteasepositive strain designated S. sanguis (Lancefield group H) ATCC 10556 (American Type Culture Collection, Rockville, Md.). After partial digestion of chromosomal DNA with the restriction endonuclease Sau3AI, the DNA was fractionated on a 10 to 40% sucrose density gradient by centrifugation at
Corresponding author. 1961
1962
INFECT. IMMUN.
GILBERT ET AL. HSanopiius influenzae 2 NLeisseria meningitidis 2 N. gonorrhoeae 2
acteroides melaninrgenicui
Clostridiui 1ramostm
IgAl
$1
H. influenme H. aeLptius
Streptoxt mdtior S. sanguis S. psemoniae
I
I
N
N. N.
meningi tidia 1 gonorrhoeae 1
1%-"--MIL--L,-ys-k,m-vai-kTo-2ser--.-i7ir-kTo-kTo-Tra-kVo-Ser-]Pro-Ser-Thr-Pro-Pro-Thr-Pro-Ser-Pro-Ser-Cys220 Z30 I'll.-
---
0
240
IgA2:A2m(1) -Cys-Pro-Val-Pro-Pro-Pro-Pro-Pro-
hinge deletion
-Cys-
IgA2:A2m(2) -Cys-Arg-Val-Pro-Pro-Pro-Pro-Pro-
hinge deletion
-Cys-
FIG. 1. Location of the peptide bonds in the human IgA heavy chain cleaved by known microbial IgA proteases. The primary sequence of the hinge region of IgAl and the two IgA2 allotypes (32) is shown. With the exception of C. ramosum, which cleaves both IgAl and IgA2 [the IgA2:A2M(1) allotype], all known IgA proteases cleave only IgAl. Enzymes of streptococcal species all cleave the bond between amino acids 227 and 228 in the heavy chain. Numbering of amino acids is from reference 35.
26,000 rpm in an SW27 rotor (Beckman Instruments, Inc., Fullerton, Calif.) for 22 h. Fragments ranging in size from 9 to 20 kilobase pairs (kbp) were isolated and ligated (DNA/ vector ratio, 1:2) with XEMBL3 DNA which had been digested to completion with restriction endonucleases EcoRI and BamHI. The ligated DNA was packaged in vitro (6), and the resulting virions were used to produce plaques on a lawn of E. coli Q359. E. coli Q359 is lysogenic for bacteriophage P2 and is thus permissive in this case only for EMBL3 phages that carry inserts of S. sanguis DNA. Assay of IgA protease and identification of an IgA protease-
positive clone. Recombinant phage plaques in soft agar were
screened for IgA protease activity by the overlay method (11). Screening involved a 5-h incubation with unsolubilized 125I-human myeloma IgAl layered in agar over a plate containing bacterial colonies or plaques. Protease activity was localized by enzymatic release of 125I-Fab fragments captured by nitrocellulose blotting and was detected by autoradiography. IgA protease activity in suspensions of bacterial colonies, sonic extracts, or culture media was measured by using 1251-human myeloma IgAl as the substrate. Assays were conducted with 0.05 M Tris hydrochloride buffer (pH 7.5), and activity and specificity were determined by sodium
TABLE 1. Characteristics of the bacterial strains used and their observed hybridization to probe 1 Bacterial strain
E. coli K-12 MM294 MC4100 A56 Q359 MC1000
S. sanguis ATCC 10556 JF Clin A
Dl ATCC 10558 Challis S. pneumoniae 32 21 1022 1593 S. mutans 6715
C. ramosum 269 B. melaninogenicus 7846 H. influenzae ATCC 9795 ATCC 9007 N. gonorrhoeae 740 32819
Relevant characteristics
hrs hrm+ thi endA supE araDJ39 lacUJ69 rpsL B1rKl supE +80r P2 F- A(ara-leu)7697 araD139 Alac(X74) galU galK rpsL
Sanguis 1 (7) Sanguis 11 (7) Sanguis 11 (7)
Clinical isolate Clinical isolate Clinical isolate Clinical isolate
IgA proteasea
Hybridization to probe 1a
-
-
ND ND ND ND
+ + + +
+ + + +
-
+ + + +
+ +
Serotype b, protease type 1 (26) Serotype c, protease type 2 (26)
+ +
Protease type 2 Protease type 1; AHU-
+ +
a The strains are designated positive (+) or negative (-) with respect to IgA protease production and hybridization with probe 1. ND, Hybridization with probe was not undertaken.
CLONING OF STREPTOCOCCAL IgA PROTEASE GENE
VOL. 56, 1988
---
...
--P
. 1_ AEwm . I,I
0
-i=
I -
4 0, z,
x
Go
i
. /=a //a/E
I o= Z
a
x x
to0
AJG1
I
.1- zo
E:s = it 0:o --0, z Al laI w, %zil.2 x iIII I I
--co 0. -;!
I
-
'25 Ix N;
i pJG1
pBR 322
(-)(-)
probes
1963
(-)
(-) (,)(,)l
#1
#2
I
I2 E
00
x
pJG2
pBR 322 =1 Kb
FIG. 2. Partial restriction maps of the 14.2-kbp S. sanguis ATCC 10556 insertion in XJG1 and its subclones, all of which encoded active , _, vector DNA. Restriction enzymes tested but unable to cut the insert in pJG1 are BalI, BstEII, KpnI, MMl, IgA protease. Symbols: NruI, SacII, SmaI, Stul, XhoI, EagI, and MstII. The location of Tn5phoA insertions and their effect on IgA protease activity are shown: (-), not active; (+) active. Probes 1 and 2 were used in hybridization experiments, as described in the text.
dodecyl sulfate-polyacrylamide gel electrophoresis as previously described (12, 27). TnSphoA mutagenesis. MC1000 cells carrying plasmid pJG1 were infected with TnS ISSOL: :phoA (TnSphoA) (13) at a multiplicity of infection of 0.1. TnSphoA insertions in pJG1 were selected as described (13). Analysis of gene products in maxicells. The maxicell procedure was that of Sancar et al. (30), with a RecA- derivative of E. coli MC4100. Cells containing plasmids were grown in M63 medium containing 1% Casamino Acids (Difco), 1 mM MgSO4, 0.1 jig of thiamine per ml, and 0.4% glucose to as A6. of 0.7. Cell suspensions were UV irradiated at 40 J/m2 and, protected from light, incubated for 30 min at 37°C, after which cycloserine (200 ,ug/ml) was added and incubation was continued overnight. The cells were then washed in fresh M63 medium, suspended in 2.0 ml of M63 medium containing 0.4% glucose and 1 mM MgSO4, and labeled for 20 min with [35S]methionine at a concentration of 25 ,uCi/ml. 35S-labeled proteins were fractionated on sodium dodecyl sulfate-5 to 15% polyacrylamide gels and examined by autoradiography. Cell fractionation. E. coli MM294 carrying plasmid pJG1 encoding active IgA protease was grown at 37°C in L broth containing 100 ,ug of ampicillin per ml. Cells were pelleted by centrifugation, suspended in buffered sucrose, and subjected to cold osmotic shock (24), yielding a shock fluid (periplasm) and shocked cells (cytoplasm plus membrane). These fractions were assayed for IgA protease activity. P-Lactamase was assayed by using cephaloridine as substrate as described earlier (14). ,-Galactosidase production was induced by isopropyl-13-D-thiogalactopyranoside (IPTG) and assayed as described previously (20). Hybridization of streptococcal DNA to that of other bacteria. The hybridization method used was that of Southern (34). Probe DNA was isolated by electrophoresis in 1% low-melting-point agarose gels and labeled by nick translation with [32P]dATP (29). For hybridization experiments, chromosomal DNA from numerous protease-positive and -negative bacterial strains (Table 1) was used. These DNA preparations were digested with SacI and electrophoresed
on 1% agarose gels. DNA was transferred to GeneScreen Plus membranes (New England Nuclear Corp., Boston, Mass.) by capillary transfer, neutralized, and placed in a prehybridization mix containing 50% deionized formamide, 1% sodium dodecyl sulfate, 1 M NaCl, 10% dextran sulfate, and 5 x 106 cpm of probe DNA. Hybridizations were carried out at 42°C for 16 h. Wash temperatures varied from 30 to 65°C. Autoradiography was done on XAR-5 film (Eastman Kodak Co., Rochester, N.Y.).
RESULTS Cloning and characterization of the S. sanguis iga gene. A genomic library of S. sanguis ATCC 10556 DNA was constructed by using XEMBL3, and plaques produced by this library were screened by the overlay method for IgA protease activity. Of 3,000 plaques examined, one was found to be enzyme positive, a result consistent with the average size of the genomic fragments in the phage library. The proteasepositive plaque was purified and used for isolation of phage XJG1. DNA isolated from this phage was characterized by restriction enzyme analysis (Fig. 2), which indicated that the enzyme-positive phage contained a 14-kbp insert of streptococcal DNA. XJG1 DNA was digested with endonucleases SmaI and XhoI and ligated with plasmid pBR322 DNA which had been digested with endonucleases SalI and EcoRV. The ligated DNA was used to transform E. coli MM294, and ampicillinresistant transformants were screened for protease activity. Plasmid pJG1 (Fig. 2), obtained from a protease-positive transformant, was found to contain a 9.8-kbp insert consisting of 9.0 kbp of streptococcal DNA and a 0.8-kbp segment of the EMBL3 vector. Both XJG1 and plasmid pJG1 specified, in E. coli, IgA protease activity having identical substrate specificity to that produced by the parent S. sanguis strain (Fig. 3). Transposon mutagenesis and deletion analysis were used to localize the iga gene within plasmid pJG1. The transposon TnSphoA (19) was used for mutagenesis since its insertion into the iga gene would lead to loss of IgA protease activity.
1964
INFECT. IMMUN.
GILBERT ET AL.
HC
-
200 -
_ _-wn
"W.
.-~~~~~~~~~~AI o
Fda LC-_
_ __
04
11692.5 66 45 -
36.5 26.5 -
1 2 3 4 5 6 FIG. 3. Effect of various IgA protease preparations on human serum IgAl and IgA2 proteins examined by polyacrylamide gel electrophoresis and stained with Coomassie blue. Lanes 1, 2, and 3 are control IgAl incubated with buffer and no enzyme, XJG1 lysate, and wild-type S. sanguis ATCC 10556 streptococcal protease, respectively. Lanes 4, 5, and 6 show an IgA2 protein incubated with the' same preparations. Only IgAl is cleaved, and the digestion products with recombinant and native enzyme are the same size. The heavy chain (HC), the Fca and Fda fragments, and the light chain (LC) of the IgAl substrate are identified.
TnSphoA insertions into pJG1 were selected on solid medium containing kanamycin and ampicillin. The locations of a series of TnSphoA insertions in pJG1 were analyzed by restriction analysis, and their effect on protease activity is shown in Fig. 2. TnSphoA insertions numbered 1 to 4 resulted in an IgA protease-negative phenotype, whereas insertions 5 and 6 retained protease activity. From these experiments it was concluded that the coding sequence required for activity must extend from the region near the SalI and NcoI sites, at the left end of the insert in pJG1, to or beyond the site defined by Tn5phoA insert 4. Deletion of a 3.5-kbp HpaI-NruI fragment from pJG1, giving rise to plasmid pJG2 (Fig. 2), had no effect on IgA protease activity, indicating that the coding sequence must lie to the left of these sites. In contrast, deletion of either the 3.8-kbp region between the two XbaI sites in pJG1 or the 0.6-kbp region between the two NcoI sites resulted in complete loss of IgA protease activity, indicating that these regions must lie within the essential coding sequence. Localization of recombinant IgA protease activity in E. coli. The IgA protease activity encoded by both pJG1 and pJG2 in E. coli MM294 was found primarily in the periplasm (Table 2). In both cases approximately 77% of activity was found in the periplasmic fraction; the remaining 23% was found in the fraction that contained the cytoplasm and cell envelope. No IgA protease activity was found in the growth medium. Under these conditions all ,-lactamase activity was periplasmic and all P-galactosidase activity was found in the cytoplasm and cell membrane fractions. TABLE 2. Distribution of IgA protease in cellular fractions of E. coli MM294 containing plasmids pJG1 and pJG2a % in following fraction:
Activity
Supernatant
Periplasm
Cytoplasm and cell
envelope
,-Lactamase
P-Galactosidase
IgA protease
0, 0 0, 0 0, 0
100, 100 3.4, 3.2 78.0, 76.2
0, 0 96.6, 96.8 22.0, 23.8
a Activity in each fraction is expressed as percentage of the total recovery. Paired data for each entry are from pJG1 (left) and pJG2 (right).
C D E F G FIG. 4. Autoradiography of polyacrylamide gels containing 35S-polypeptides encoded in maxicells of E. coli MC4100 by plasmids pJG1, pJG2, and their derivatives. Lanes: A, pJG2; B, pJG1; C to F, TnSphoA protease-negative mutants 4 to 1, respectively, as shown in Fig. 2; G, control MC4100 maxicells containing no plasmid. The products of pJG1 and pJG2 have molecular weights 190,000 and 200,000, respectively, based on the standards shown. Major bands in the molecular weight range 25,000 to 50,000 are encoded by Tn5phoA. A
B
Characterization of the product encoded by the streptococcal IgA protease gene. The polypeptides specified by pJG1, pJG2, and their derivatives in E. coli maxicells were labeled with [35S]methionine and analyzed as described in Materials and Methods. In addition to the proteins encoded by the pBR322 vector DNA, two unique polypeptides of 190 and 200 kilodaltons were produced in maxicells containing either pJG1 or pJG2 (Fig. 4). Neither of these proteins was produced in maxicells containing each of four IgA proteasenegative pJG1::TnSphoA derivatives. From these experiments it was concluded that the high-molecular-weight proteins represented translation products specified by the S. sanguis iga gene. Because polypeptides of this length require approximately 5.5 to 6.0 kbp of coding sequence, the iga gene in pJG2 must occupy almost all of the streptococcal DNA insert. Given the protease-positive phenotype of pJG1 and pJG2, we conclude that the 200-kilodalton polypeptide or the 190-kilodalton polypeptide, or both, must be products of the iga gene. Relationship of streptococcal DNA encoding IgA protease with DNA of other bacteria. IgA proteases are produced by many different genera of bacteria, and the various enzymes cleave at different sites in the hinge region of human IgA. To examine the relationship of the S. sanguis IgA protease with those of other genera and other species within the genus Streptococcus, we prepared two iga-specific DNA probes, a 2.4-kbp BamHI-Apal fragment and a 2.1-kbp ApaI-HpaI fragment, both internal to the iga gene (Fig. 2). The region from which each probe was obtained was judged to be from within the iga gene because TnSphoA insertions in these segments eliminated the ability to specify protease activity, and both DNA segments are present in the encoding DNA insert in plasmid pJG2. The two probes were used for Southern hybridization analysis of chromosomal DNA purified from both gram-positive and gram-negative bacteria. Both probes gave identical results; those with probe 1 are listed in Table 1, and autoradiographs are shown in Fig. 5. Hybridization was observed with chromosomal DNA of the homologous strain S. sanguis ATCC 10556 and with three other IgA protease-positive S. sanguis strains (designated Clin A, JF, and D1) under stringent hybridization conditions.
CLONING OF STREPTOCOCCAL IgA PROTEASE GENE
VOL. 56, 1988 O'
23.1
Of
9.4
-
6.5
-
4.4
-
2.3 2.0
-
-
*..
0
0.36
-
A B C D E
B
FIG. 5. (Left) Southern blot analysis of chromosomal DNA of various bacterial species (Table 1) with probe 1 from the IgA protease-encoding region of S. sanguis ATCC 10556 streptococcal DNA (Fig. 2). The numbers on the vertical axis are size in kilobase pairs. Lanes: A to D; streptococcal strains Clin A, Dl, JF, and ATCC 10556, respectively, all of which hybridized to the probe; E, S. sanguis ATCC 10558, which did not hybridize. Nonhybridizing bacterial DNAs of other species listed in Table 1 are not shown. (Right) Higher exposure time of gel in the left panel to better illustrate the weak hybridization in lane B.
The probe failed to hybridize with two S. sanguis strains that did not produce IgA protease, S. sanguis ATCC 10558 and S. sanguis Challis. DNA from S. pneumoniae failed to hybridize with either probe, even under conditions of low stringency, despite these clinical isolates having IgA protease activity indistinguishable in specificity from that of the S. sanguis strain. No hybridization was detected with chromosomal DNA of S. mutans, which is IgA protease negative, and none was observed with Haemophilus, Neisseria, Clostridium, and Bacteroides strains, all of which are IgA protease positive.
DISCUSSION In these experiments we have identified and cloned a segment of DNA encoding IgA protease activity from chromosomal DNA of S. sanguis ATCC 10556. Through the use of transposon and deletion mutagenesis of the encoding DNA expressed in E. coli, we have determined that the IgA protease gene (iga) is approximately 6 kbp long. The IgA protease activity expressed by the S. sanguis iga gene cloned in E. coli has a substrate specificity identical to that of the native enzyme arising from the S. sanguis parent strain. The enzyme specified by plasmid pJG1 is localized in the periplasm of E. coli MM294 cells, indicating that the IgA protease contains a signal sequence that allows efficient secretion through the cytoplasmic membrane into the periplasm. It is noteworthy that the IgA protease genes of N. gonorrhoeae and H. influenzae encode enzymes that are secreted into the growth medium when cloned in E. coli (9, 17), suggesting that these proteases carry additional determinants that allow their secretion through the outer membranes of gram-negative organisms. The S. sanguis iga gene gave rise to two polypeptides, 190 and 200 kilodaltons in size, in E. coli maxicells containing plasmid pJG1 or pJG2. It seems likely that one or both of these high-molecular-mass products are the enzyme protein, because they are not expressed by protease-negative TnSphoA insertion mutants of these plasmids. Although the relationship between the two products is unclear, it is most
1965
likely that the smaller polypeptide is a processed or degraded form of the larger protein or that transcription of the iga gene can begin at more than one site. We are currently studying these possibilities. Hybridization experiments with two probes from within the S. sanguis IgA protease-encoding DNA showed the predicted homology with chromosomal DNA of other protease-positive bacteria classified as S. sanguis by the criteria of Facklam (7). Interestingly, even under conditions of low stringency the chromosomal DNA of S. pneumoniae failed to hybridize with either probe, although both the S. sanguis and S. pneumoniae IgA proteases cleave the identical peptide bond in the human IgAl heavy-chain hinge region. This result indicates that there are substantial differences between the iga genes of these closely related species, at least at the nucleotide sequence level. These differences are somewhat surprising in view of earlier results with gramnegative bacteria. For example, the iga genes encoding IgA proteases within a given genus, e.g., Haemophilus or Neisseria, are homologous even if they encode proteases of different (type 1 and type 2) substrate specificity (16, 28). Furthermore, the iga genes of two different genera, again Haemophilus and Neisseria (16; Grundy, unpublished), show significant homology by Southern hybridization and DNA sequence analysis. Cloning of the pneumococcal protease gene and direct comparison of its nucleotide sequence with that of the S. sanguis iga gene are needed to clarify the relationship between the enzymes of the two species. It is possible that a more precise assignment of a given streptococcal species to various taxonomic groups, or a redefinition of the groups themselves, could be advanced by the use of iga gene probes. It is not yet clear how the IgA proteases contribute to the onset or perpetuation of infectious processes. S. sanguis species are abundant in the oral cavity, and it has been shown that they are among the earliest microorganisms to adhere to and colonize the proteinaceous acquired pellicle on newly cleaned dental enamel surfaces (18, 33). In addition to S. sanguis, oral streptococci such as S. mitior produce IgA proteases, but it is noteworthy that S. mutans, a decidedly cariogenic species invariably found in human dental plaque, is IgA protease negative. Although the exact role of S. sanguis in caries pathogenesis remains to be determined, this species may establish an environment suitable for secondary colonization by other bacteria. Naturally occurring dental caries is a bacterial disease that cannot be attributed to a single microbial species, and this has complicated the development of rational antimicrobial therapy. The goal of caries control will require an understanding about how plaque bacteria interact with one another and with the salivary antibody important in controlling infections of nonshedding tooth surfaces. The cloning of the iga gene from S. sanguis and its capacity of expression in E. coli offer the opportunity to study the specific biochemical events of streptococcal-IgA interaction. The cloned gene will also allow construction of IgA protease-negative mutants that can be compared with the otherwise isogenic parent strain. Haemophilus and Neisseria iga genes have been used to construct mutants of this type in order to examine the role of these enzymes in pathogenesis (4, 8). Finally, these mutants may be helpful in analyzing the role of IgA protease in determining the virulence of streptococcal strains and their involvement in the development of dental disease.
1966
GILBERT ET AL.
INFECT. IMMUN.
ACKNOWLEDGMENTS We thank Frank Grundy, Martin Taubman, and Anne Kane for their valuable contribution to our research and Carol Nichols for editorial and typing assistance. The research was supported by Public Health Service grants DE 07257 and Al 20337 from the National Institutes of Health and by the Center for Gastroenterology Research on Absorptive and Secretory Processes (grant P30-AM 39428 from the National Institutes of Health).
immunoglobulin Al protease-producing streptococci in the human oral cavity and pharynx. Infect. Immun. 31:868-873. 16. Koomey, J. M., and S. Falkow. 1984. Nucleotide sequence homology between immunoglobulin Al protease genes of Neisseria gonorrhoeae, Neisseria meningitidis, and Haemophilus influenzae. Infect. Immun. 43:101-107. 17. Koomey, J. M., R. E. Gill, and S. Falkow. 1982. Genetic and biochemical analysis of gonococcal IgAl protease: cloning in Escherichia coli and construction of mutants of gonococci that fail to produce the activity. Proc. Natl. Acad. Sci. USA 79:
LITERATURE CITED 1. Berman, M. L., L. W. Enquist, and T. J. Silhavy. 1984. Experiments with gene fusions. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 2. Birnboim, H., and J. Doly. 1979. A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7:1513-1523. 3. Bricker, J., M. H. Mulks, A. G. Plaut, E. R. Moxon, and A. Wright. 1983. IgAl proteases of Haemophilus influenzae: cloning and characterization in E. coli K-12. Proc. Natl. Acad. Sci. USA 80:2681-2685. 4. Cooper, M. D., Z. A. McGee, M. H. Mulks, J. M. Koomey, and T. L. Hindman. 1984. Attachment to and invasion of human Fallopian tube mucosa by an IgAl protease-deficient mutant of Neisseria gonorrhoeae and its wild-type parent. J. Infect. Dis. 150:737-744. 5. Coykendall, A. L. 1970. Base composition of deoxyribonucleic acid isolated from cariogenic streptococci. Arch. Oral Biol. 15: 365-368. 6. Enquist, L., and N. Sternberg. 1979. In vitro packaging of DAM vectors and their use in cloning DNA fragments. Methods Enzymol. 68:281-298. 7. Facklam, R. R. 1977. Physiological differentiation of viridans streptococci. J. Clin. Mircrobiol. 5:184-201. 8. Farley, M. M., D. S. Stephens, M. H. Mulks, M. D. Cooper, J. V. Bricker, S. S. Mirra, and A. Wright. 1986. Pathogenesis of IgAl protease-producing and -nonproducing Haemophilus influenzae in human nasopharyngeal organ cultures. J. Infect. Dis. 154:752-759. 9. Fishman, Y., J. Bricker, J. V. Gilbert, A. G. Plaut, and A. Wright. 1985. Cloning of type 1 immunoglobulin Al protease from Neisseria gonorrhoeae and secretion of the enzyme from Escherichia coli, p. 164-168. In G. K. Schoolnik (ed.), The pathogenic neisseriae. American Society for Mircrobiology, Washington, D.C. 10. Frandsen, E. V. G., J. Reinholdt, and M. Kilian. 1987. Enzymatic and antigenic characterization of immunoglobulin Al proteases from Bacteroides and Capnocytophaga spp. Infect. Immun. 55:631-638. 11. Gilbert, J. V., and A. G. Plaut. 1983. Detection of IgA protease activity among multiple bacterial colonies. J. Immunol. Methods 57:247-251. 12. Gilbert, J. V., A. G. PlaUt, B. Longmaid, and M. E. Lamm. 1983. Inhibition of microbial IgA proteases by human secretory IgA and serum. J. Mol. Immunol. 20:1039-1049. 13. Gutierrez, C., J. Barondess, C. Manoil, and J. Beckwith. 1987. The use of transposon TnphoA to detect genes for cell envelope proteins subject to a common regulatory stimulus. J. Mol. Biol. 195:289-297. 14. Jack, G. W., and M. H. Richmond. 1970. A comparative study of eight distinct beta-lactamases synthesized by gram-negative bacteria. J. Gen. Microbiol. 61:43-61. 15. Kilian, M., and K. Holmgren. 1981. Ecology and nature of
7881-7885. 18. Loesche, W. J. 1986. Role of Streptococcus mutans in human dental decay. Microbiol. Rev. 50:353-380. 19. Manoil, C., and J. Beckwith. 1985. TnphoA: a transposon probe for protein export signals. Proc. Natl. Acad. Sci. USA 82:81298133. 20. Miller, J. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 21. Mortensen, S. B., and M. Kilian. 1984. Purification and characterization of an immunoglobulin Al protease from Bacteroides melaninogenicus. Infect. Immun. 45:550-557. 22. Mulks, M. H. 1985. Microbial IgA proteases p. 81-104. In I. A. Holder (ed.), Bacterial enzymes and virulence. CRC Press, Inc., Boca Raton, Fla. 23. Mulks, M. H., S. Kornfeld, and A. G. Plaut. 1980. Specific proteolysis of human IgA by Streptococcus pneumoniae and Haemophilus influenzae. J. Infect. Dis. 141:450-456. 24. Neu, H.C., and L. A. Heppel. 1965. The release of enzymes from E. coli by osmotic shock and during the formation of spheroplasts. J. Biol. Chem. 240:3685-3692. 25. Perry, D., L. J. Nilsen, and H. W. Kuramitsu. 1985. Mapping of a cloned glycosyltransferase gene in Streptococcus mutans. Infect. Immun. 50:130-135. 26. Plaut, A. G. 1983. The IgAl proteases of pathogenic bacteria. Annu. Rev. Microbiol. 37:603-622. 27. Plaut, A. G., J. V. Gilbert, G. Leger, and M. Blumenstein. 1985. IgAl protease cleaves heavy chains independently in dimeric human IgAl. J. Mol. Immunol. 7:821-826. 28. Pohlner, J., R. Halter, K. Beyreuther, and T. F. Meyer. 1987. Gene structure and extracellular secretion of Neisseria gonorrhoeae IgA protease. Nature (London) 325:458-462. 29. Rigby, P. W. J., M. Dieckmann, C. Rhodes and P. Berg. 1977. Labeling deoxyribonucleic acid to high specific activity in vitro by nick translation with DNA polymerase I. J. Mol. Biol. 113: 237-251. 30. Sancar, A., A. M. Hack, and W. D. Rupp. 1979. Simple method for identification of plasmid-coded proteins. J. Bacteriol. 137: 692-693. 31. Saunders, C. W., and W. R. Guild. 1980. Properties and transforming activities of two plasmids in Streptococcus pneumoniae. Mol. Gen. Genet. 180:573-578. 32. Senda, S., F. Fujiyama, T. Ushijima, K. Hodohara, T. Bamba, S. Hosoda, and K. Kobayashi. 1985. Clostridium ramosum, an IgA protease-producing species and its ecology in the human intestinal tract. Microbiol. Immunol. 29:1019-1028. 33. Shaw, J. H. 1987. Causes and control of dental caries. N. Engl. J. Med. 317:996-1004. 34. Southern, E. M. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98:503-517. 35. Tsuzukida, Y., C. C. Wang, and F. W. Putnam. 1979. Structure of the A2m (1) allotype of human IgA-a recombinant molecule. Proc. Natl. Acad. Sci. USA 76:1104-1108.
