Effect of Mutations in the Human Immunoglobulin A1 (IgA1) Hinge on Its Susceptibility to Cleavage by Diverse Bacterial IgA1 Proteases
http://www.100md.com
感染与免疫杂志 2005年第3期
Division of Pathology and Neuroscience, University of Dundee Medical School, Dundee, United Kingdom
ABSTRACT
Components of the human immunoglobulin A1 (IgA1) hinge governing sensitivity to cleavage by bacterial IgA1 proteases were investigated. Recombinant antibodies with distinct hinge mutations were constructed from a hybrid comprised of human IgA2 bearing half of the human IgA1 hinge region. This hybrid antibody and all the mutant antibodies derived from it were resistant to cleavage by the IgA1 proteases from Streptococcus oralis and Streptococcus mitis biovar 1 strains but were cleaved to various degrees by those of Streptococcus pneumoniae, some Streptococcus sanguis strains, and the type 1 and 2 IgA1 proteases of Haemophilus influenzae, Neisseria meningitidis, and Neisseria gonorrhoeae. Remarkably, those proteases that cleave a Pro-Ser peptide bond in the wild-type IgA1 hinge were able to cleave mutant antibodies lacking a Pro-Ser peptide bond in the hinge, and those that cleave a Pro-Thr peptide bond in the wild-type IgA1 hinge were able to cleave mutant antibodies devoid of a Pro-Thr peptide bond in the hinge. Thus, the enzymes can cleave alternatives to their preferred postproline peptide bond when such a bond is unavailable. Peptide sequence analysis of a representative antibody digestion product confirmed this conclusion. The presence of a cleavable peptide bond near the CH2 end of the hinge appeared to result in greater cleavage than if the scissile bond was at the CH1 end of the hinge. Proline-to-serine substitution at residue 230 in a hinge containing potentially cleavable Pro-Ser and Pro-Thr peptide bonds increased the resistance of the antibody to cleavage by many IgA1 proteases.
INTRODUCTION
The two subclasses of human immunoglobulin A1 (IgA), IgA1 and IgA2, play key roles in protecting the mucous membranes of the respiratory, gastrointestinal, and genitourinary tracts from attack by pathogenic microorganisms and their products. While a few species of bacteria produce enzymes capable of cleaving both human IgA1 and IgA2 (16, 38), a smaller number of bacterial species produce enzymes, called IgA1 proteases, capable of cleaving only human IgA1 and not IgA2. The organisms producing IgA1 proteases include the principal causes of bacterial meningitis, Streptococcus pneumoniae, Neisseria meningitidis, and Haemophilus influenzae, and other bacterial pathogens initiating infection at mucosal surfaces of the genital tract and the oral cavity (13, 24, 38).
The action of the proteases on IgA1 results in the cleavage of the Fab arms away from the Fc region of the molecule, thereby separating the antigen recognition function from the effector function that triggers the elimination of antigens. Furthermore, the Fab fragments produced may mask relevant epitopes from the immune system and prevent binding of intact antibodies (both IgA and other isotypes) (14) and also activation of complement and Fc receptors. Thus, IgA1 proteases are thought to be virulence factors. Some IgA1 proteases may also have a role in virulence through mechanisms in addition to or distinct from that arising through IgA1 cleavage (9, 17, 29, 39). Although the narrow specificity of IgA1 proteases precludes the use of animal models as a means to prove the role of IgA1 proteases as virulence determinants, supporting evidence that they may act as virulence factors comes from observations that related but nonpathogenic bacterial species of the above genera do not produce IgA1 proteases (22, 24), that the enzymes are produced in vivo (3, 11), and that recovery from infection is associated with development of antibodies to the enzymes (4, 6, 7).
IgA1 proteases are so named because they act almost exclusively (2, 9, 17, 39) on the IgA1 isotype of humans and the higher apes (33). Human IgA2 is resistant to cleavage by IgA1 proteases because it lacks a sequence of 16 amino acids present in the hinge of human IgA1, which is the site of cleavage for all IgA1 proteases. The sequence contains a repeat of an 8-amino-acid sequence and is rich in proline, threonine, and serine. Although IgA1 proteases belong to different families, they are all postproline endopeptidases and cleave at either Pro-Ser (type 1 enzymes) or Pro-Thr (type 2 enzymes) peptide bonds (15, 19, 21, 23, 25, 26). However, the enzymes appear to be extremely specific, in that a given enzyme cleaves the specific peptide bond in only one of the duplicated 8-amino-acid sequences and not at the equivalent site in the other duplicated half of the hinge. In a previous study (37), a recombinant hybrid human IgA2-IgA1 immunoglobulin was constructed in which a sequence corresponding to one-half of the duplicated hinge region of human IgA1 was inserted into the equivalent site in IgA1 protease-resistant human IgA2. The resultant hybrid IgA2-IgA1 half-hinge molecule was found to be sensitive to cleavage by many different IgA1 proteases. In order to understand more about the amino acid requirements in the hinge of IgA1 for cleavage by IgA1 proteases with a view, ultimately, to designing IgA1 protease inhibitors, several mutated forms of the IgA2-IgA1 half-hinge hybrid were prepared, and their sensitivities to IgA1 proteases were examined.
MATERIALS AND METHODS
Primers. Primer A1H6 (5' GCGCGCGGATCCGGTCCAACCGCAGGCCC 3') contained a BamHI site (italics) and annealed 150 bp upstream of the CH1 exon of human IgA2m(1). Primer A2SEQ2 (5' TGTAGCAGCCACAGA 3') was complementary to nucleotides 873 to 887 in the CH2 exon of human IgA2m (1). Primer 224PROS (5' CTCCAGTTCCCCCAACTCCACCT 3') contained the nucleotide sequence 678 to 688 of human IgA2m(1) and that of part of the half hinge of human IgA1 (italics) with one altered base (boldface). Primer 224PROAS was the complement of primer 224PROS. Primer 230PROS (5' ACCTACCCCACCTCCACCTC 3') contained the nucleotide sequence of part of the half hinge of human IgA1 (italics) with one altered base (boldface) and that of nucleotides 689 to 695 of the CH2 exon of human IgA2m(1). Primer 230PROAS was the complement of primer 230PROS. Primer PS227S (5' CTCAACTCCATCTACCCCATCT 3') contained the nucleotide sequence of the half hinge of human IgA1 (italics) with one altered base (boldface). Primer PS227AS was the complement of primer PS227S.
Construction of mutant antibody expression vectors. The antibody expression vectors constructed, the nomenclature of the antibodies they generated, and their amino acid sequences in the hinge region are presented in Fig. 1. Plasmid pBS2 carried, downstream of the mouse VNP gene, the gene for the CH1, CH2, and CH3 domains of the chain of human IgA2m(1), with nucleotides coding for half of the hinge of human IgA1 inserted at the appropriate site, as described previously (37). The antibody expression vectors pBS10, pBS230SP, and pBS12 were constructed by PCR overlap extension mutagenesis (10), using pBS2 DNA as a template and A1H6 and A2SEQ2 as flanking primers, with 224PROS and 224PROAS as internal primers for pBS10, 230PROS and 230PROAS as internal primers for pBS230SP, and PS227S and PS227AS as internal primers for pBS12. In each case, the 920-bp PCR product, which contained a mutated form of half the IgA1 hinge region, was cleaved with BamHI and XhoI and ligated into the BamHI- and XhoI-cut site of the original IgA2m(1) expression vector (20), replacing the wild-type sequence in this region. The antibody expression vector pBS11 was made in a similar way and with the same flanking primers but with pBS10 DNA as a template and 230PROS and 230PROAS as internal primers. All the constructed expression vectors were sequenced by an ABI 377 DNA sequencer. In each case, sequencing confirmed that the half hinge of IgA1 had been modified as intended and that no PCR-generated errors in the coding regions had occurred.
Preparation of recombinant mutant hybrid IgA2-IgA1 immunoglobulins. CHO-K1 cells stably transfected previously with an appropriate mouse light chain (20) were seeded in tissue culture grade petri dishes and transfected with antibody heavy chain expression vectors using calcium phosphate as described previously (20). Positive transfectants were isolated by selection for the bacterial xanthine-guanine phosphoribosyltransferase selectable marker by growth in medium supplemented with hypoxanthine and thymidine (HT supplement; Invitrogen, Paisley, United Kingdom), xanthine (0.25 mg/ml), and mycophenolic acid (10 μg/ml). Several resistant colonies were picked, and the cell lines producing the highest yields of IgA were identified by an enzyme-linked immunosorbent assay measuring binding to the antigen NIP (3-nitro-4-hydroxy-5-iodophenylacetate), as described previously (20), before expansion into large cultures. Recombinant antibodies were purified from the supernatants of the CHO-K1 transfectant cultures by affinity chromatography on NIP-Sepharose, as described previously (20). The purified antibodies were supplemented with 0.1% sodium azide and stored in small aliquots at –20°C.
Microbial IgA1 proteases. The IgA1 proteases used were from S. pneumoniae strain SK690; Streptococcus oralis strain SK10; Streptococcus sanguis strains SK1 (ATCC 10556) (biovar 1), SK4 (biovar 2), and SK49 (biovar 4); Streptococcus mitis biovar 1 strains SK564, SK597, and SK599; N. meningitidis group B serotype 14 strain 3564 (type 1 enzyme) and group Y serotype 2c strain HF 13 (type 2 enzyme); Neisseria gonorrhoeae serogroup WI serovar 1A-2 strain 6092 (type 1 enzyme) and serogroup WII/III serovar 1B-6 strain 5489 (type 2 enzyme); and H. influenzae strain H23 (type 1 enzyme) and strain H15 (type 2 enzyme).
The streptococcal strains were cultured in 2TY broth (35) (1.6% tryptone, 1% yeast extract, 0.5% sodium chloride in distilled water, pH 7) at 37°C in air containing 5% CO2, and their IgA1 proteases were concentrated and purified from the culture supernatants by fractional ammonium sulfate precipitation and subsequent dialysis against phosphate-buffered saline (PBS), pH 7.2, containing 0.1% sodium azide. The other bacteria were cultured at 37°C in air containing 5% CO2 on dialysis tubing membranes on the surfaces of appropriate solid culture media as described previously (36), and their IgA1 proteases were prepared similarly from the supernatants of the suspensions of the bacteria washed from the dialysis tubing membranes with PBS containing 0.1% sodium azide.
Digestion of recombinant hybrid IgA2-IgA1 immunoglobulins with microbial IgA1 proteases and immunoblotting. Appropriate amounts of antibody and IgA1 protease in PBS (pH 7.2) buffer containing 0.1% sodium azide in a final volume of 20 μl were incubated at 37°C for 24 h. The reactions were stopped by the addition of 10 μl of sample buffer (8 M urea, 1% sodium dodecyl sulfate, 10% glycerol, 4% -mercaptoethanol, and a trace of bromophenol blue dye in 50 mM Tris-HCl buffer, pH 6.8) and being boiled for 5 min. The reduced reaction mixtures were electrophoresed on sodium dodecyl sulfate-10% polyacrylamide gels, and the separated proteins were transferred to nitrocellulose membranes. The membranes were blocked by agitation for 30 min in 5% nonfat dry milk powder in PBS. After being thoroughly washed in PBS, the membranes were immersed in a 1-in-1,000 dilution in PBS containing 0.1% Tween 20 of either horseradish peroxidase-labeled goat antibody to human IgA (-Fab specific; Kirkegaard & Perry Laboratories, Gaithersburg, Md.) or alkaline phosphatase-labeled mouse antibody to human IgA1 (1-Fc specific; Southern Biotechnology Associates, Inc., Birmingham, Ala.) and agitated for 2 h at room temperature. After the membranes were thoroughly washed in several changes of PBS, those probed with peroxidase-labeled antibody were developed in 10 ml of 50 mM Tris-HCl buffer, pH 7.6, containing 0.3 mg of nickel chloride/ml, 10 mg of diaminobenzidine, and 60 μl of 30% hydrogen peroxide, whereas those probed with alkaline phosphatase-labeled antibody were developed in the dark in 10 ml of developing buffer (100 mM Tris-HCl, pH 9.5, containing 100 mM NaCl and 10 mM MgCl2) to which had been added 30 μl of bromochloroindolyl phosphate solution (50 mg/ml in dimethyl formamide) and 30 μl of nitroblue tetrazolium solution (100 mg/ml in 70% dimethyl formamide in water).
Protease cleavage site determination by MS. Antibody hhS224/230P was digested with H. influenzae type 1 enzyme as described above and run under reducing conditions on a 5 to 12% acrylamide gradient gel. The band corresponding to the Fab region was excised and subjected to in-gel tryptic digestion prior to matrix-assisted laser desorption ionization-time of flight mass spectrometry (MS). The MS spectrum was searched for peaks with masses corresponding to those calculated for each tryptic peptide that would be generated if one assumed cleavage could occur at any of the peptide bonds in the hinge region. Of all these possible peptide peaks, only a single one with a mass of 2,203 Da was identified. This peptide peak was extracted and subjected to liquid chromatography tandem mass spectrometry (LC MS-MS) in order to determine its amino acid sequence.
RESULTS
Analysis of the mutated antibodies. Electrophoresis of the mutated antibodies under reducing and nonreducing conditions and immunoblotting with antibodies specific for the human chain and the mouse chains revealed the presence of protein bands with masses indistinguishable from those of the recombinant IgA2-IgA1 half hinge (37) (data not shown). These bands, in turn, were virtually indistinguishable from those of recombinant wild-type human IgA2. These results indicated that the introduction of altered amino acids into the hinge region of the mutant antibodies did not appear to cause discernible effects on antibody structure or assembly.
Activities of IgA1 proteases on the mutated antibodies. The IgA1 protease activities of the enzyme preparations were demonstrated by their cleavage of recombinant wild-type human IgA1 into Fab and Fc fragments, whereas wild-type recombinant human IgA2 remained resistant to cleavage. Whenever there was cleavage of the hybrid IgA2-IgA1 half-hinge antibody and its mutant forms, the mass of the fragments generated by the different IgA1 proteases and the observed resistance of IgA2 to cleavage by them together indicated that the cleavage was taking place only in the modified or mutated IgA1 hinge and that this was due solely to IgA1 protease activity.
Cleavage by streptococcal proteases. All streptococcal IgA1 proteases cleave wild-type human IgA1 in the hinge at the Pro227-Thr228 bond (Fig. 1). Nevertheless, it was found that the IgA1 protease of S. pneumoniae SK690 cleaved antibody hhP227S, which is devoid of a Pro-Thr peptide bond in the hinge, as well as all the other mutant antibodies that contained a Pro-Thr peptide bond (Fig. 2 and Table 1). The products of each digestion were Fab fragments of two sizes. This was often seen with the S. pneumoniae IgA1 protease preparations, which are known to contain trace amounts of copurifying glycosidases (1). The size differences between the two bands were constant regardless of the antibody cleaved and appeared to be too great to be accounted for by alternative cleavage within the short half hinge present. Therefore, these fragments most likely represent different glycoforms of Fab. By contrast, the IgA1 proteases from S. mitis biovar 1 strains SK564, SK597, and SK599 and that of S. oralis strain SK10 were unable to cleave any of the antibodies (Table 1). On the other hand, the IgA1 proteases of the strains of the different S. sanguis biovars gave different cleavage results (Table 1). For example, S. sanguis strain SK1 IgA1 protease cleaved none of the mutant antibodies, whereas the IgA1 protease of S. sanguis strain SK49 (biovar 4) cleaved the hinges of all the mutant antibodies (Table 1). Antibody hhS224P appeared to be more sensitive than IgA2-IgA1 half-hinge antibody to cleavage with the IgA1 proteases of S. pneumoniae and S. sanguis (Fig. 2 and Table 1). The Fab fragments detected were of the same mass (results not shown), which suggested that cleavage in the hinge was probably always at the same peptide bond, although cleavage at an alternative neighboring peptide bond in some of the mutants could not be excluded.
Cleavage by type 1 proteases of H. influenzae, N. meningitidis, and N. gonorrhoeae. The type 1 proteases of H. influenzae, N. meningitidis, and N. gonorrhoeae all cleave wild-type human IgA1 at Pro-Ser peptide bonds, with the cleavage site for those of the last two organisms being the same and C-terminal to that of H. influenzae protease (Fig. 1). Nevertheless, all these type 1 proteases cleaved antibody hhS224/230P, which lacks a Pro-Ser peptide bond in the hinge, and also all the other mutant antibodies. Since these type 1 enzymes were able to cleave an antibody lacking the peptide bond that they preferentially cleave in wild-type IgA1, we used mass spectrometry to determine the precise peptide bond in hhS224/230P that was cleaved by the H. influenzae type 1 enzyme, as a representative of this group. The matrix-assisted laser desorption ionization-time of flight mass spectrum of the Fab fragment contained numerous peaks. We calculated the predicted masses of peptides that would be generated from the tryptic peptide encompassing the hinge region of antibody hhS224/230P if one assumed that protease cleavage could occur at any peptide bond in the hinge. Of all these possible peptide peaks, only one, with a mass of 2,203 Da, was identified in the spectrum (Fig. 3). LC MS-MS analysis of this peptide peak confirmed that the peptide had the sequence HYTNPSQDVTVPCPVPPTPP, indicating that the site cleaved by H. influenzae type 1 protease lay between Pro227 and Thr228 in antibody hhS224/230P. Thus, in the context of the IgA1 hinge, the H. influenzae type 1 enzyme is able to cleave a Pro-Thr bond if a Pro-Ser bond is unavailable. This cleavage site localization serves as an important illustration that IgA1 proteases can cleave alternatives to their preferred postproline peptide bond when the hinge sequence is constrained so that such a bond is unavailable.
An example showing the action of the type 1 IgA1 protease of N. gonorrhoeae on the mutant antibodies is presented in Fig. 4. It was noticed that the type 1 IgA1 proteases of N. gonorrhoeae and N. meningitidis appeared to cleave antibody hhS224P (Fig. 4, lane 4; Fig. 5, lane 9; and Table 1), which has a Pro-Ser peptide bond at the C-terminal end of the hinge, better than antibody hhS230P, which has a Pro-Ser peptide bond at the N-terminal end of the hinge (Fig. 4, lane 6; Fig. 5, lane 10; and Table 1). However, the type 1 IgA1 protease of H. influenzae, which normally cleaves wild-type IgA1 at a more N-terminal Pro-Ser bond, did not cleave antibody hhS230P significantly better than antibody hhS224P (Fig. 5, lanes 3 and 4). Antibody hhP227S, which has three Pro-Ser peptide bonds in the hinge, was more sensitive to these type 1 proteases than the IgA2-IgA1 half-hinge antibody, which has only two Pro-Ser peptide bonds in the hinge (Fig. 4, lane 10, and Fig. 5, lanes 6 and 12). However, antibodies hhS224P, with only one Pro-Ser peptide bond, and hhP227S appeared to be similarly sensitive (Table 1). This indicates that it is most likely the site, rather than the number, of the cleavable peptide bonds that determines the efficiency of cleavage.
Cleavage by type 2 proteases of H. influenzae, N. meningitidis, and N. gonorrhoeae. The type 2 IgA1 proteases of N. meningitidis, N. gonorrhoeae, and H. influenzae all cleave wild-type IgA1 at the same Pro-Thr peptide bond, which is C-terminal to that cleaved by streptococcal IgA1 proteases (Fig. 1). However they all cleaved antibody hhP227S, which lacks a Pro-Thr peptide bond in the hinge, and also IgA2-IgA1 half hinge and hhS224P antibodies (Table 1). The activity of the type 2 IgA1 protease of H. influenzae strain H15, for example, on the mutant antibodies is presented in Fig. 6. Antibodies hhS224/230P and hhS230P were only partially cleaved by the type 2 IgA1 proteases of these bacteria, and they appeared to be generally more resistant to cleavage than the other antibodies (Table 1). A summary of the activities of the different IgA1 proteases on the mutant antibodies is presented in Table 1.
DISCUSSION
The factors determining the sensitivity of a substrate to cleavage by IgA1 proteases are not completely understood and were the subject of this investigation. The finding that unrelated substrates, like human IgA1, human chorionic gonadotropin (39), tumor necrosis factor receptor II (2), and the precursors of the IgA1 proteases of S. pneumoniae (32), the type 2 IgA1 protease of N. gonorrhoeae, and the type 1 IgA1 protease of H. influenzae (8, 27, 30), are able to be cleaved by IgA1 proteases suggests that the sequence of amino acids rather than a complex structural motif is a major determinant of sensitivity to IgA1 proteases. The observation that insertion of the hinge of human IgA1 into endoglucanase A (CenA) of Cellulomonas fimi rendered it susceptible to cleavage with both the type 1 and 2 IgA1 proteases of N. gonorrhoeae (18) appears to support this view.
On the other hand, IgA1 proteases characteristically have a narrow spectrum of activity and cleave only IgA1 from humans and apelike creatures (33) and a few other substrates (2, 17, 39). This feature, together with the finding that synthetic substrates of decapeptides with the same amino acid sequence as the human IgA1 hinge were resistant to cleavage with the type 2 protease of N. gonorrhoeae (40) suggest that in addition to a defined cleavage site sequence, other features are also necessary for sensitivity to cleavage by IgA1 proteases. These additional features most likely include a requirement for the cleavage site sequence to adopt a particular secondary structure within the context of the protein as a whole and possibly the presence of other structural features in the substrate, such as the Fc region (5). Our results with the type 2 proteases of Neisseria support this possibility. For these enzymes, the sequence motif cleaved has been described as Y-Pro-X-Pro, with cleavage occurring at the Pro-X bond, where Y represents proline (or, less satisfactorily, alanine, glycine, or threonine) and X represents threonine, serine, or alanine (28). All the mutants tested here contain such a motif in their hinges (although that of hhP227S is less satisfactory), but the antibodies hhS230P and hhS224/230P generally proved to be poor substrates for these enzymes, suggesting that the context of the Y-Pro-X-Pro motif may affect its susceptibility to cleavage.
The specific site at which bacterial IgA1 proteases cleave human wild-type IgA1 is determined by a region of the protease called the cleavage specificity determinant. Different cleavage specificity determinants permit cleavage to take place in wild-type human IgA1 at either a Pro-Ser (for a type 1 protease) or a Pro-Thr (for a type 2 protease) peptide bond, but in only one of the duplicated sequences in the hinge and not at the equivalent position in the other. Nevertheless, many IgA1 proteases, including representatives of those that cleave wild-type human IgA1 in the different duplicated halves of the hinge sequence, are still able to cleave IgA1 when only half of the IgA1 hinge is present in the hinge position of IgA1 protease-resistant IgA2, as in the IgA2-IgA1 half-hinge antibody (37). Among the IgA1 proteases that cleaved this hybrid antibody were those of S. pneumoniae and S. sanguis strains, whereas that of S. oralis failed to cleave it (37). The results presented here confirm this finding and also extend it by showing that the IgA2-IgA1 half-hinge antibody is also resistant to cleavage by the IgA1 proteases of strains of S. mitis biovar 1 and some other S. sanguis strains of different biovars. This finding is intriguing, as the IgA1 proteases of these different streptococcal species all cleave the same Pro-Thr peptide bond in wild-type human IgA1 (12, 34) and such a bond is present in the IgA2-IgA1 half-hinge antibody. The known extensive antigenic variation in the IgA1 proteases of S. mitis strains and their known closer relationship to the IgA1 proteases of S. oralis than to those of the other streptococci (31) might explain why the S. mitis IgA1 proteases behaved similarly to that of S. oralis in being unable to cleave the IgA2-IgA1 half-hinge antibody and differently from those of the other streptococci. The finding that the IgA1 proteases of S. mitis and S. oralis were also unable to cleave any of the mutant antibodies derived from the IgA2-IgA1 half-hinge antibody may indicate that it is the shortened hinge in these antibodies rather than its amino acid sequence that renders them resistant to cleavage by these IgA1 proteases. One might speculate that the C-terminal half of the IgA1 hinge normally acts as a "spacer," placing the N-terminal half of the hinge some distance away from the Fc region and allowing optimal access to the preferred Pro-Thr bond in the N-terminal half. Thus, in the hybrid antibody, the available Pro-Thr bond may be positioned too close to the Fc region of the antibody to allow efficient enzymatic activity by the protease.
The differing activities seen in this study, and also in a previous one (1), in the IgA1 proteases of different S. sanguis strains on the mutant antibodies may be a consequence of the difference in antigenicity between enzymes from strains of different biovars (34). Thus, the results support the view that the sensitivity of IgA1 to cleavage is determined not only by the presence of an appropriate cleavable peptide bond in the IgA1 hinge but also by the structures of both the IgA1 protease and the substrate upon which it acts.
The IgA1 proteases of streptococci cleave wild-type IgA1 at a Pro-Thr peptide bond (Fig. 1). Such a bond is present in the IgA2-IgA1 half-hinge antibody, whereas two are present in the mutant antibody hhS224P. This may be the reason for the finding that hhS224P is apparently more sensitive than the IgA2-IgA1 half-hinge antibody to cleavage with the IgA1 proteases of S. pneumoniae and some S. sanguis strains. Antibody hhP227S did not have a Pro-Thr site for cleavage. Thus, its sensitivity to the IgA1 proteases of S. pneumoniae and some S. sanguis strains indicates that these particular proteases are not dependent on the presence of a Pro-Thr peptide bond and that they can cleave at an alternative peptide bond. This result confirms our earlier findings with different hinge mutant IgA1 antibodies (1). The reduced sensitivity of antibodies hhS230P and hhS224/230P compared with that of antibody hhS224P to cleavage with these streptococcal IgA1 proteases, despite their very similar amino acid sequences, suggests that either the lack of serine at residue 230 or the consequent creation of a stretch of six proline residues immediately after the presumed cleavage site, or both of these things, has an inhibitory effect upon susceptibility to cleavage.
The mutant antibodies hhS224P and hhS230P were constructed so as to provide antibodies that, unlike the antibody IgA2-IgA1 half hinge from which they were derived, carried only one Pro-Ser peptide bond, either at the CH2 terminal end of the hinge, as for hhS224P, or at the CH1 terminal end, as for hhS230P. The difference between the sensitivities to cleavage of these antibodies was remarkable, for although the type 1 IgA1 proteases of N. meningitidis and N. gonorrhoeae both cleave the same Pro-Ser peptide bond in wild-type IgA1, and it is distant from the one cleaved by the type 1 protease of H. influenzae (Fig. 1), all these proteases were able to cleave antibody hhS224P more extensively than antibody hhS230P. This suggests that the availability of a cleavable peptide bond at the CH2 end of the hinge results in more efficient cleavage than when a scissile bond is present at the CH1 end.
Antibody hhS224/230P was created to provide an example of a hinge lacking Pro-Ser peptide bonds. The cleavage of this antibody by the type 1 (Pro-Ser-cleaving) proteases from H. influenzae and N. gonorrhoeae suggests that, when constrained, type 1 IgA1 proteases may be able to cleave at peptide bonds alternative to the one normally cleaved in wild-type IgA1. In this respect, they act like some streptococcal IgA1 proteases and the type 2 IgA1 proteases of N. meningitidis and H. influenzae (1).
Antibody hhP227S was created to provide an example of a half hinge lacking a Pro-Thr peptide bond. The finding that the type 2 (Pro-Thr-cleaving) IgA1 proteases of H. influenzae, N. meningitidis, and N. gonorrhoeae were able to cleave this antibody indicates that these enzymes are also similarly able to cleave a peptide bond alternative to that cleaved in wild-type IgA1. Moreover, the sensitivity of antibody hhP227S to cleavage with the Pro-Thr-cleaving IgA1 proteases of S. pneumoniae and some S. sanguis strains confirms the less stringent requirement of these streptococcal IgA1 proteases for a Pro-Thr bond in the hinge to ensure cleavage, which we reported previously (1).
Because of these findings, it might have been expected that, as all the other mutant antibodies possessed a Pro-Thr peptide bond at the same position as in the IgA2-IgA1 half-hinge antibody, they would have been cleaved by these type 2 IgA1 proteases with similar efficiencies and that this cleavage would be at least as efficient as in the cleavage of antibody hhP227S. However, as with the streptococcal IgA1 proteases, antibodies hhS230P and hhS224/230P were much more resistant than antibody hhS224P to cleavage by the Pro-Thr peptide bond-cleaving IgA1 proteases of Neisseria. This may be explained, as before for the streptococcal proteases, on the basis that the change of amino acid from serine at 230 to proline or the resultant creation of an uninterrupted sequence of six proline residues at the CH2 terminus of the hinge, or both, has a profound effect upon the susceptibility of the IgA1 hinge mutants to cleavage.
Since none of the proteases cleave wild-type IgA2, our experiments with the range of mutant IgA1 half hinges introduced into IgA2 described here allow us to conclude that many IgA1 proteases will cleave at alternative peptide bonds when presented with hinge sequences that lack the peptide bond normally cleaved in wild-type IgA1. In order to provide more detailed evidence in a representative digest, we chose to determine the precise site cleaved by the H. influenzae type 1 enzyme in antibody hhS224/230P. The site was revealed to lie between Pro227 and Thr228, showing that, in the absence of a suitably positioned Pro-Ser bond, this enzyme will cleave a Pro-Thr bond instead. This site localization provides a definitive demonstration that IgA1 proteases can cleave peptide bonds alternative to those cleaved in the wild-type IgA1 hinge if the hinge sequence is suitably constrained.
In conclusion, susceptibility to cleavage by IgA1 proteases appears to depend on the precise sequence, and hence the three-dimensional structure, of the IgA hinge, the context of the cleavage site within the hinge and within the protein as a whole, and the structure of the particular protein (as evidenced by its antigenic profile). However, it is clear that there is some flexibility in what most bacterial IgA1 proteases will accept as a substrate. Thus, many will cleave an IgA hinge lacking either the Pro-Ser or Pro-Thr bonds normally favored for cleavage. The design of reagents to inhibit IgA1 proteases is likely to represent a significant challenge. Nevertheless, the approach used here makes an important addition to our understanding of IgA1 protease-substrate interactions.
ACKNOWLEDGMENTS
We are grateful to M. Kilian for kindly donating some of the bacterial strains. We thank Douglas Lamont and Kenneth Beattie at the Fingerprints Proteomics Facility, University of Dundee, for MS analysis and for assistance with MS data interpretation.
J.M.W. thanks the Wellcome Trust for funding.
REFERENCES
1. Batten, M. R., B. W. Senior, M. Kilian, and J. M. Woof. 2003. Amino acid sequence requirements in the hinge of human immunoglobulin A1 (IgA1) for cleavage by streptococcal IgA1 proteases. Infect. Immun. 71:1462-1469.
2. Beck, S. C., and T. F. Meyer. 2000. IgA1 protease from Neisseria gonorrhoeae inhibits TNF-mediated apoptosis of human monocytic cells. FEBS Lett. 472:287-292.
3. Blake, M., K. K. Holmes, and J. Swanson. 1979. Studies on gonococcus infection. XVII. IgA1-cleaving protease in vaginal washings from women with gonorrhea. J. Infect. Dis. 139:89-92.
4. Brooks, G. F., C. J. Lammel, M. S. Blake, B. Kusecek, and M. Achtman. 1992. Antibodies against IgA protease are stimulated both by clinical disease and asymptomatic carriage of serogroup A Neisseria meningitidis. J. Infect. Dis. 166:1316-1321.
5. Chintalacharuvu, K. R., P. D. Chuang, A. Dragoman, C. Z. Fernandez, J. Qiu, A. G. Plaut, K. R. Trinh, F. A. Gala, and S. L. Morrison. 2003. Cleavage of the human immunoglobulin A1 (IgA1) hinge region by IgA1 proteases requires structures in the Fc region of IgA. Infect. Immun. 71:2563-2570.
6. Devenyi, A. G., A. G. Plaut, F. J. Grundy, and A. Wright. 1993. Post-infectious human serum antibodies inhibit IgA1 proteinases by interaction with the cleavage site specificity determinant. Mol. Immunol. 30:1243-1248.
7. Gilbert, J. V., A. G. Plaut, B. Longmaid, and M. E. Lamm. 1983. Inhibition of microbial IgA proteases by human secretory IgA and serum. Mol. Immunol. 20:1039-1049.
8. Halter, R., J. Pohlner, and T. F. Meyer. 1984. IgA protease of Neisseria gonorrhoeae: isolation and characterization of the gene and its extracellular product. EMBO J. 3:1595-1601.
9. Hauck, C. R., and T. F. Meyer. 1997. The lysosomal/phagosomal membrane protein h-lamp-1 is a target of the IgA1 protease of Neisseria gonorrhoeae. FEBS Lett. 405:86-90.
10. Horton, R. M., H. D. Hunt, S. N. Ho, J. K. Pullen, and L. R. Pease. 1989. Engineering hybrid genes without the use of restriction enzymes: gene splicing by overlap extension. Gene 77:61-68.
11. Insel, R. A., P. Z. Allen, and I. D. Berkowitz. 1982. Types and frequency of Haemophilus influenzae IgA1 proteases. Semin. Infect. Dis. 4:225-231.
12. Kilian, M., J. Mestecky, R. Kulhavy, M. Tomana, and W. T. Butler. 1980. IgA1 proteases from Haemophilus influenzae, Streptococcus pneumoniae, Neisseria meningitidis, and Streptococcus sanguis: comparative immunological studies. J. Immunol. 124:2596-2600.
13. Kilian, M., J. Reinholdt, H. Lomholt, K. Poulsen, and E. V. G. Frandsen. 1996. Biological significance of IgA1 proteases in bacterial colonization and pathogenesis: critical evaluation of experimental evidence. APMIS 104:321-338.
14. Kilian, M., and M. W. Russell. 1999. Microbial evasion of IgA functions, p. 241-251. In P. L. Ogra, J. Mestecky, M. E. Lamm, W. Strober, J. R. McGhee, and J. Bienenstock (ed.), Mucosal immunology, 2nd ed. Academic Press Inc., San Diego, Calif.
15. Kilian, M., B. Thomsen, T. E. Petersen, and H. Bleeg. 1983. Molecular biology of Haemophilus influenzae IgA1 proteases. Mol. Immunol. 20:1051-1058.
16. Kosowska, K., J. Reinholdt, L. K. Rasmussen, A. Sabat, J. Potempa, M. Kilian, and K. Poulsen. 2002. The Clostridium ramosum IgA proteinase represents a novel type of metalloendopeptidase. J. Biol. Chem. 277:11987-11994.
17. Lin, L., P. Ayala, J. Larson, M. Mulks, M. Fukuda, S. R. Carlsson, C. Enns, and M. So. 1997. The Neisseria type 2 IgA1 protease cleaves LAMP1 and promotes survival of bacteria within epithelial cells. Mol. Microbiol. 24:1083-1094.
18. Miller, P. B., H. Shen, N. R. Gilkes, D. G. Kilburn, R. C. Miller, A. G. Plaut, and R. A. J. Warren. 1992. Endoglucanase A from Cellulomonas fimi in which the hinge sequence of human IgA1 is substituted for the linker connecting its two domains is hydrolyzed by IgA proteases from Neisseria gonorrhoeae. FEMS Microbiol. Lett. 92:199-204.
19. Mortensen, S. B., and M. Kilian. 1984. Purification and characterization of an immunoglobulin A1 protease from Bacteroides melaninogenicus. Infect. Immun. 45:550-557.
20. Morton, H. C., J. D. Atkin, R. J. Owens, and J. M. Woof. 1993. Purification and characterization of chimeric human IgA1 and IgA2 expressed in COS and Chinese hamster ovary cells. J. Immunol. 151:4743-4752.
21. Mulks, M. H., S. J. Kornfeld, B. Frangione, and A. G. Plaut. 1982. Relationship between the specificity of IgA proteases and serotypes in Haemophilus influenzae. J. Infect. Dis. 146:266-274.
22. Mulks, M. H., and A. G. Plaut. 1978. IgA protease production as a characteristic distinguishing pathogenic from harmless Neisseriaceae. N. Engl. J. Med. 299:973-976.
23. Mulks, M. H., A. G. Plaut, H. A. Feldman, and B. Frangione. 1980. IgA proteases of two distinct specificities are released by Neisseria meningitidis. J. Exp. Med. 152:1442-1447.
24. Plaut, A. G. 1983. The IgA1 proteases of pathogenic bacteria. Annu. Rev. Microbiol. 37:603-622.
25. Plaut, A. G., J. V. Gilbert, M. S. Artenstein, and J. D. Capra. 1975. Neisseria gonorrhoeae and Neisseria meningitidis: extracellular enzyme cleaves human immunoglobulin A. Science 193:1103-1105.
26. Plaut, A. G., R. Wistar, and J. D. Capra. 1974. Differential susceptibility of human IgA immunoglobulins to streptococcal IgA protease. J. Clin. Investig. 54:1295-1300.
27. Pohlner, J., R. Halter, K. Beyreuther, and T. F. Meyer. 1987. Gene structure and extracellular secretion of Neisseria gonorrhoeae IgA protease. Nature 325:458-462.
28. Pohlner, J., T. Klauser, E. Kuttler, and R. Halter. 1992. Sequence-specific cleavage of protein fusions using a recombinant Neisseria type 2 IgA protease. Biotechnology 10:799-804.
29. Polissi, A., A. Pontiggia, G. Feger, M. Altieri, H. Mottl, L. Ferrai, and D. Simon. 1998. Large-scale identification of virulence genes from Streptococcus pneumoniae. Infect. Immun. 66:5620-5629.
30. Poulsen, K., J. Brandt, J. P. Hjorth, H. C. Thogersen, and M. Kilian. 1989. Cloning and sequencing of the immunoglobulin A1 protease gene (iga) of Haemophilus influenzae serotype b. Infect. Immun. 57:3097-3105.
31. Poulsen, K., J. Reinholdt, C. Jespersgaard, K. Boye, T. A. Brown, M. Hauge, and M. Kilian. 1998. A comprehensive genetic study of streptococcal immunoglobulin A1 proteases: evidence for recombination within and between species. Infect. Immun. 66:181-190.
32. Poulsen, K., J. Reinholdt, and M. Kilian. 1996. Characterization of the Streptococcus pneumoniae immunoglobulin A1 protease gene (iga) and its translation product. Infect. Immun. 64:3957-3966.
33. Qiu, J., G. P. Brackee, and A. G. Plaut. 1996. Analysis of the specificity of bacterial immunoglobulin A (IgA) proteases by a comparative study of ape serum IgAs as substrates. Infect. Immun. 64:933-937.
34. Reinholdt, J., M. Tomana, S. B. Mortensen, and M. Kilian. 1990. Molecular aspects of immunoglobulin A1 degradation by oral streptococci. Infect. Immun. 58:1186-1194.
35. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
36. Senior, B. W., M. Albrechtsen, and M. A. Kerr. 1987. Proteus mirabilis strains of diverse type have IgA protease activity. J. Med. Microbiol. 24:175-180.
37. Senior, B. W., J. I. Dunlop, M. R. Batten, M. Kilian, and J. M. Woof. 2000. Cleavage of a recombinant human immunoglobulin A2 (IgA2)-IgA1 hybrid antibody by certain bacterial IgA1 proteases. Infect. Immun. 68:463-469.
38. Senior, B. W., L. M. Loomes, and M. A. Kerr. 1991. Microbial IgA proteases and virulence. Rev. Med. Microbiol. 2:200-207.
39. Senior, B. W., W. W. Stewart, C. Galloway, and M. A. Kerr. 2001. Cleavage of the hormone human chorionic gonadotropin, by the type 1 IgA1 protease of Neisseria gonorrhoeae, and its implications. J. Infect. Dis. 184:922-925.
40. Wood, S. G., and J. Burton. 1991. Synthetic peptide substrates for the immunoglobulin A1 protease from Neisseria gonorrhoeae (type 2). Infect. Immun. 59:1818-1822.(Bernard W. Senior and Jen)
ABSTRACT
Components of the human immunoglobulin A1 (IgA1) hinge governing sensitivity to cleavage by bacterial IgA1 proteases were investigated. Recombinant antibodies with distinct hinge mutations were constructed from a hybrid comprised of human IgA2 bearing half of the human IgA1 hinge region. This hybrid antibody and all the mutant antibodies derived from it were resistant to cleavage by the IgA1 proteases from Streptococcus oralis and Streptococcus mitis biovar 1 strains but were cleaved to various degrees by those of Streptococcus pneumoniae, some Streptococcus sanguis strains, and the type 1 and 2 IgA1 proteases of Haemophilus influenzae, Neisseria meningitidis, and Neisseria gonorrhoeae. Remarkably, those proteases that cleave a Pro-Ser peptide bond in the wild-type IgA1 hinge were able to cleave mutant antibodies lacking a Pro-Ser peptide bond in the hinge, and those that cleave a Pro-Thr peptide bond in the wild-type IgA1 hinge were able to cleave mutant antibodies devoid of a Pro-Thr peptide bond in the hinge. Thus, the enzymes can cleave alternatives to their preferred postproline peptide bond when such a bond is unavailable. Peptide sequence analysis of a representative antibody digestion product confirmed this conclusion. The presence of a cleavable peptide bond near the CH2 end of the hinge appeared to result in greater cleavage than if the scissile bond was at the CH1 end of the hinge. Proline-to-serine substitution at residue 230 in a hinge containing potentially cleavable Pro-Ser and Pro-Thr peptide bonds increased the resistance of the antibody to cleavage by many IgA1 proteases.
INTRODUCTION
The two subclasses of human immunoglobulin A1 (IgA), IgA1 and IgA2, play key roles in protecting the mucous membranes of the respiratory, gastrointestinal, and genitourinary tracts from attack by pathogenic microorganisms and their products. While a few species of bacteria produce enzymes capable of cleaving both human IgA1 and IgA2 (16, 38), a smaller number of bacterial species produce enzymes, called IgA1 proteases, capable of cleaving only human IgA1 and not IgA2. The organisms producing IgA1 proteases include the principal causes of bacterial meningitis, Streptococcus pneumoniae, Neisseria meningitidis, and Haemophilus influenzae, and other bacterial pathogens initiating infection at mucosal surfaces of the genital tract and the oral cavity (13, 24, 38).
The action of the proteases on IgA1 results in the cleavage of the Fab arms away from the Fc region of the molecule, thereby separating the antigen recognition function from the effector function that triggers the elimination of antigens. Furthermore, the Fab fragments produced may mask relevant epitopes from the immune system and prevent binding of intact antibodies (both IgA and other isotypes) (14) and also activation of complement and Fc receptors. Thus, IgA1 proteases are thought to be virulence factors. Some IgA1 proteases may also have a role in virulence through mechanisms in addition to or distinct from that arising through IgA1 cleavage (9, 17, 29, 39). Although the narrow specificity of IgA1 proteases precludes the use of animal models as a means to prove the role of IgA1 proteases as virulence determinants, supporting evidence that they may act as virulence factors comes from observations that related but nonpathogenic bacterial species of the above genera do not produce IgA1 proteases (22, 24), that the enzymes are produced in vivo (3, 11), and that recovery from infection is associated with development of antibodies to the enzymes (4, 6, 7).
IgA1 proteases are so named because they act almost exclusively (2, 9, 17, 39) on the IgA1 isotype of humans and the higher apes (33). Human IgA2 is resistant to cleavage by IgA1 proteases because it lacks a sequence of 16 amino acids present in the hinge of human IgA1, which is the site of cleavage for all IgA1 proteases. The sequence contains a repeat of an 8-amino-acid sequence and is rich in proline, threonine, and serine. Although IgA1 proteases belong to different families, they are all postproline endopeptidases and cleave at either Pro-Ser (type 1 enzymes) or Pro-Thr (type 2 enzymes) peptide bonds (15, 19, 21, 23, 25, 26). However, the enzymes appear to be extremely specific, in that a given enzyme cleaves the specific peptide bond in only one of the duplicated 8-amino-acid sequences and not at the equivalent site in the other duplicated half of the hinge. In a previous study (37), a recombinant hybrid human IgA2-IgA1 immunoglobulin was constructed in which a sequence corresponding to one-half of the duplicated hinge region of human IgA1 was inserted into the equivalent site in IgA1 protease-resistant human IgA2. The resultant hybrid IgA2-IgA1 half-hinge molecule was found to be sensitive to cleavage by many different IgA1 proteases. In order to understand more about the amino acid requirements in the hinge of IgA1 for cleavage by IgA1 proteases with a view, ultimately, to designing IgA1 protease inhibitors, several mutated forms of the IgA2-IgA1 half-hinge hybrid were prepared, and their sensitivities to IgA1 proteases were examined.
MATERIALS AND METHODS
Primers. Primer A1H6 (5' GCGCGCGGATCCGGTCCAACCGCAGGCCC 3') contained a BamHI site (italics) and annealed 150 bp upstream of the CH1 exon of human IgA2m(1). Primer A2SEQ2 (5' TGTAGCAGCCACAGA 3') was complementary to nucleotides 873 to 887 in the CH2 exon of human IgA2m (1). Primer 224PROS (5' CTCCAGTTCCCCCAACTCCACCT 3') contained the nucleotide sequence 678 to 688 of human IgA2m(1) and that of part of the half hinge of human IgA1 (italics) with one altered base (boldface). Primer 224PROAS was the complement of primer 224PROS. Primer 230PROS (5' ACCTACCCCACCTCCACCTC 3') contained the nucleotide sequence of part of the half hinge of human IgA1 (italics) with one altered base (boldface) and that of nucleotides 689 to 695 of the CH2 exon of human IgA2m(1). Primer 230PROAS was the complement of primer 230PROS. Primer PS227S (5' CTCAACTCCATCTACCCCATCT 3') contained the nucleotide sequence of the half hinge of human IgA1 (italics) with one altered base (boldface). Primer PS227AS was the complement of primer PS227S.
Construction of mutant antibody expression vectors. The antibody expression vectors constructed, the nomenclature of the antibodies they generated, and their amino acid sequences in the hinge region are presented in Fig. 1. Plasmid pBS2 carried, downstream of the mouse VNP gene, the gene for the CH1, CH2, and CH3 domains of the chain of human IgA2m(1), with nucleotides coding for half of the hinge of human IgA1 inserted at the appropriate site, as described previously (37). The antibody expression vectors pBS10, pBS230SP, and pBS12 were constructed by PCR overlap extension mutagenesis (10), using pBS2 DNA as a template and A1H6 and A2SEQ2 as flanking primers, with 224PROS and 224PROAS as internal primers for pBS10, 230PROS and 230PROAS as internal primers for pBS230SP, and PS227S and PS227AS as internal primers for pBS12. In each case, the 920-bp PCR product, which contained a mutated form of half the IgA1 hinge region, was cleaved with BamHI and XhoI and ligated into the BamHI- and XhoI-cut site of the original IgA2m(1) expression vector (20), replacing the wild-type sequence in this region. The antibody expression vector pBS11 was made in a similar way and with the same flanking primers but with pBS10 DNA as a template and 230PROS and 230PROAS as internal primers. All the constructed expression vectors were sequenced by an ABI 377 DNA sequencer. In each case, sequencing confirmed that the half hinge of IgA1 had been modified as intended and that no PCR-generated errors in the coding regions had occurred.
Preparation of recombinant mutant hybrid IgA2-IgA1 immunoglobulins. CHO-K1 cells stably transfected previously with an appropriate mouse light chain (20) were seeded in tissue culture grade petri dishes and transfected with antibody heavy chain expression vectors using calcium phosphate as described previously (20). Positive transfectants were isolated by selection for the bacterial xanthine-guanine phosphoribosyltransferase selectable marker by growth in medium supplemented with hypoxanthine and thymidine (HT supplement; Invitrogen, Paisley, United Kingdom), xanthine (0.25 mg/ml), and mycophenolic acid (10 μg/ml). Several resistant colonies were picked, and the cell lines producing the highest yields of IgA were identified by an enzyme-linked immunosorbent assay measuring binding to the antigen NIP (3-nitro-4-hydroxy-5-iodophenylacetate), as described previously (20), before expansion into large cultures. Recombinant antibodies were purified from the supernatants of the CHO-K1 transfectant cultures by affinity chromatography on NIP-Sepharose, as described previously (20). The purified antibodies were supplemented with 0.1% sodium azide and stored in small aliquots at –20°C.
Microbial IgA1 proteases. The IgA1 proteases used were from S. pneumoniae strain SK690; Streptococcus oralis strain SK10; Streptococcus sanguis strains SK1 (ATCC 10556) (biovar 1), SK4 (biovar 2), and SK49 (biovar 4); Streptococcus mitis biovar 1 strains SK564, SK597, and SK599; N. meningitidis group B serotype 14 strain 3564 (type 1 enzyme) and group Y serotype 2c strain HF 13 (type 2 enzyme); Neisseria gonorrhoeae serogroup WI serovar 1A-2 strain 6092 (type 1 enzyme) and serogroup WII/III serovar 1B-6 strain 5489 (type 2 enzyme); and H. influenzae strain H23 (type 1 enzyme) and strain H15 (type 2 enzyme).
The streptococcal strains were cultured in 2TY broth (35) (1.6% tryptone, 1% yeast extract, 0.5% sodium chloride in distilled water, pH 7) at 37°C in air containing 5% CO2, and their IgA1 proteases were concentrated and purified from the culture supernatants by fractional ammonium sulfate precipitation and subsequent dialysis against phosphate-buffered saline (PBS), pH 7.2, containing 0.1% sodium azide. The other bacteria were cultured at 37°C in air containing 5% CO2 on dialysis tubing membranes on the surfaces of appropriate solid culture media as described previously (36), and their IgA1 proteases were prepared similarly from the supernatants of the suspensions of the bacteria washed from the dialysis tubing membranes with PBS containing 0.1% sodium azide.
Digestion of recombinant hybrid IgA2-IgA1 immunoglobulins with microbial IgA1 proteases and immunoblotting. Appropriate amounts of antibody and IgA1 protease in PBS (pH 7.2) buffer containing 0.1% sodium azide in a final volume of 20 μl were incubated at 37°C for 24 h. The reactions were stopped by the addition of 10 μl of sample buffer (8 M urea, 1% sodium dodecyl sulfate, 10% glycerol, 4% -mercaptoethanol, and a trace of bromophenol blue dye in 50 mM Tris-HCl buffer, pH 6.8) and being boiled for 5 min. The reduced reaction mixtures were electrophoresed on sodium dodecyl sulfate-10% polyacrylamide gels, and the separated proteins were transferred to nitrocellulose membranes. The membranes were blocked by agitation for 30 min in 5% nonfat dry milk powder in PBS. After being thoroughly washed in PBS, the membranes were immersed in a 1-in-1,000 dilution in PBS containing 0.1% Tween 20 of either horseradish peroxidase-labeled goat antibody to human IgA (-Fab specific; Kirkegaard & Perry Laboratories, Gaithersburg, Md.) or alkaline phosphatase-labeled mouse antibody to human IgA1 (1-Fc specific; Southern Biotechnology Associates, Inc., Birmingham, Ala.) and agitated for 2 h at room temperature. After the membranes were thoroughly washed in several changes of PBS, those probed with peroxidase-labeled antibody were developed in 10 ml of 50 mM Tris-HCl buffer, pH 7.6, containing 0.3 mg of nickel chloride/ml, 10 mg of diaminobenzidine, and 60 μl of 30% hydrogen peroxide, whereas those probed with alkaline phosphatase-labeled antibody were developed in the dark in 10 ml of developing buffer (100 mM Tris-HCl, pH 9.5, containing 100 mM NaCl and 10 mM MgCl2) to which had been added 30 μl of bromochloroindolyl phosphate solution (50 mg/ml in dimethyl formamide) and 30 μl of nitroblue tetrazolium solution (100 mg/ml in 70% dimethyl formamide in water).
Protease cleavage site determination by MS. Antibody hhS224/230P was digested with H. influenzae type 1 enzyme as described above and run under reducing conditions on a 5 to 12% acrylamide gradient gel. The band corresponding to the Fab region was excised and subjected to in-gel tryptic digestion prior to matrix-assisted laser desorption ionization-time of flight mass spectrometry (MS). The MS spectrum was searched for peaks with masses corresponding to those calculated for each tryptic peptide that would be generated if one assumed cleavage could occur at any of the peptide bonds in the hinge region. Of all these possible peptide peaks, only a single one with a mass of 2,203 Da was identified. This peptide peak was extracted and subjected to liquid chromatography tandem mass spectrometry (LC MS-MS) in order to determine its amino acid sequence.
RESULTS
Analysis of the mutated antibodies. Electrophoresis of the mutated antibodies under reducing and nonreducing conditions and immunoblotting with antibodies specific for the human chain and the mouse chains revealed the presence of protein bands with masses indistinguishable from those of the recombinant IgA2-IgA1 half hinge (37) (data not shown). These bands, in turn, were virtually indistinguishable from those of recombinant wild-type human IgA2. These results indicated that the introduction of altered amino acids into the hinge region of the mutant antibodies did not appear to cause discernible effects on antibody structure or assembly.
Activities of IgA1 proteases on the mutated antibodies. The IgA1 protease activities of the enzyme preparations were demonstrated by their cleavage of recombinant wild-type human IgA1 into Fab and Fc fragments, whereas wild-type recombinant human IgA2 remained resistant to cleavage. Whenever there was cleavage of the hybrid IgA2-IgA1 half-hinge antibody and its mutant forms, the mass of the fragments generated by the different IgA1 proteases and the observed resistance of IgA2 to cleavage by them together indicated that the cleavage was taking place only in the modified or mutated IgA1 hinge and that this was due solely to IgA1 protease activity.
Cleavage by streptococcal proteases. All streptococcal IgA1 proteases cleave wild-type human IgA1 in the hinge at the Pro227-Thr228 bond (Fig. 1). Nevertheless, it was found that the IgA1 protease of S. pneumoniae SK690 cleaved antibody hhP227S, which is devoid of a Pro-Thr peptide bond in the hinge, as well as all the other mutant antibodies that contained a Pro-Thr peptide bond (Fig. 2 and Table 1). The products of each digestion were Fab fragments of two sizes. This was often seen with the S. pneumoniae IgA1 protease preparations, which are known to contain trace amounts of copurifying glycosidases (1). The size differences between the two bands were constant regardless of the antibody cleaved and appeared to be too great to be accounted for by alternative cleavage within the short half hinge present. Therefore, these fragments most likely represent different glycoforms of Fab. By contrast, the IgA1 proteases from S. mitis biovar 1 strains SK564, SK597, and SK599 and that of S. oralis strain SK10 were unable to cleave any of the antibodies (Table 1). On the other hand, the IgA1 proteases of the strains of the different S. sanguis biovars gave different cleavage results (Table 1). For example, S. sanguis strain SK1 IgA1 protease cleaved none of the mutant antibodies, whereas the IgA1 protease of S. sanguis strain SK49 (biovar 4) cleaved the hinges of all the mutant antibodies (Table 1). Antibody hhS224P appeared to be more sensitive than IgA2-IgA1 half-hinge antibody to cleavage with the IgA1 proteases of S. pneumoniae and S. sanguis (Fig. 2 and Table 1). The Fab fragments detected were of the same mass (results not shown), which suggested that cleavage in the hinge was probably always at the same peptide bond, although cleavage at an alternative neighboring peptide bond in some of the mutants could not be excluded.
Cleavage by type 1 proteases of H. influenzae, N. meningitidis, and N. gonorrhoeae. The type 1 proteases of H. influenzae, N. meningitidis, and N. gonorrhoeae all cleave wild-type human IgA1 at Pro-Ser peptide bonds, with the cleavage site for those of the last two organisms being the same and C-terminal to that of H. influenzae protease (Fig. 1). Nevertheless, all these type 1 proteases cleaved antibody hhS224/230P, which lacks a Pro-Ser peptide bond in the hinge, and also all the other mutant antibodies. Since these type 1 enzymes were able to cleave an antibody lacking the peptide bond that they preferentially cleave in wild-type IgA1, we used mass spectrometry to determine the precise peptide bond in hhS224/230P that was cleaved by the H. influenzae type 1 enzyme, as a representative of this group. The matrix-assisted laser desorption ionization-time of flight mass spectrum of the Fab fragment contained numerous peaks. We calculated the predicted masses of peptides that would be generated from the tryptic peptide encompassing the hinge region of antibody hhS224/230P if one assumed that protease cleavage could occur at any peptide bond in the hinge. Of all these possible peptide peaks, only one, with a mass of 2,203 Da, was identified in the spectrum (Fig. 3). LC MS-MS analysis of this peptide peak confirmed that the peptide had the sequence HYTNPSQDVTVPCPVPPTPP, indicating that the site cleaved by H. influenzae type 1 protease lay between Pro227 and Thr228 in antibody hhS224/230P. Thus, in the context of the IgA1 hinge, the H. influenzae type 1 enzyme is able to cleave a Pro-Thr bond if a Pro-Ser bond is unavailable. This cleavage site localization serves as an important illustration that IgA1 proteases can cleave alternatives to their preferred postproline peptide bond when the hinge sequence is constrained so that such a bond is unavailable.
An example showing the action of the type 1 IgA1 protease of N. gonorrhoeae on the mutant antibodies is presented in Fig. 4. It was noticed that the type 1 IgA1 proteases of N. gonorrhoeae and N. meningitidis appeared to cleave antibody hhS224P (Fig. 4, lane 4; Fig. 5, lane 9; and Table 1), which has a Pro-Ser peptide bond at the C-terminal end of the hinge, better than antibody hhS230P, which has a Pro-Ser peptide bond at the N-terminal end of the hinge (Fig. 4, lane 6; Fig. 5, lane 10; and Table 1). However, the type 1 IgA1 protease of H. influenzae, which normally cleaves wild-type IgA1 at a more N-terminal Pro-Ser bond, did not cleave antibody hhS230P significantly better than antibody hhS224P (Fig. 5, lanes 3 and 4). Antibody hhP227S, which has three Pro-Ser peptide bonds in the hinge, was more sensitive to these type 1 proteases than the IgA2-IgA1 half-hinge antibody, which has only two Pro-Ser peptide bonds in the hinge (Fig. 4, lane 10, and Fig. 5, lanes 6 and 12). However, antibodies hhS224P, with only one Pro-Ser peptide bond, and hhP227S appeared to be similarly sensitive (Table 1). This indicates that it is most likely the site, rather than the number, of the cleavable peptide bonds that determines the efficiency of cleavage.
Cleavage by type 2 proteases of H. influenzae, N. meningitidis, and N. gonorrhoeae. The type 2 IgA1 proteases of N. meningitidis, N. gonorrhoeae, and H. influenzae all cleave wild-type IgA1 at the same Pro-Thr peptide bond, which is C-terminal to that cleaved by streptococcal IgA1 proteases (Fig. 1). However they all cleaved antibody hhP227S, which lacks a Pro-Thr peptide bond in the hinge, and also IgA2-IgA1 half hinge and hhS224P antibodies (Table 1). The activity of the type 2 IgA1 protease of H. influenzae strain H15, for example, on the mutant antibodies is presented in Fig. 6. Antibodies hhS224/230P and hhS230P were only partially cleaved by the type 2 IgA1 proteases of these bacteria, and they appeared to be generally more resistant to cleavage than the other antibodies (Table 1). A summary of the activities of the different IgA1 proteases on the mutant antibodies is presented in Table 1.
DISCUSSION
The factors determining the sensitivity of a substrate to cleavage by IgA1 proteases are not completely understood and were the subject of this investigation. The finding that unrelated substrates, like human IgA1, human chorionic gonadotropin (39), tumor necrosis factor receptor II (2), and the precursors of the IgA1 proteases of S. pneumoniae (32), the type 2 IgA1 protease of N. gonorrhoeae, and the type 1 IgA1 protease of H. influenzae (8, 27, 30), are able to be cleaved by IgA1 proteases suggests that the sequence of amino acids rather than a complex structural motif is a major determinant of sensitivity to IgA1 proteases. The observation that insertion of the hinge of human IgA1 into endoglucanase A (CenA) of Cellulomonas fimi rendered it susceptible to cleavage with both the type 1 and 2 IgA1 proteases of N. gonorrhoeae (18) appears to support this view.
On the other hand, IgA1 proteases characteristically have a narrow spectrum of activity and cleave only IgA1 from humans and apelike creatures (33) and a few other substrates (2, 17, 39). This feature, together with the finding that synthetic substrates of decapeptides with the same amino acid sequence as the human IgA1 hinge were resistant to cleavage with the type 2 protease of N. gonorrhoeae (40) suggest that in addition to a defined cleavage site sequence, other features are also necessary for sensitivity to cleavage by IgA1 proteases. These additional features most likely include a requirement for the cleavage site sequence to adopt a particular secondary structure within the context of the protein as a whole and possibly the presence of other structural features in the substrate, such as the Fc region (5). Our results with the type 2 proteases of Neisseria support this possibility. For these enzymes, the sequence motif cleaved has been described as Y-Pro-X-Pro, with cleavage occurring at the Pro-X bond, where Y represents proline (or, less satisfactorily, alanine, glycine, or threonine) and X represents threonine, serine, or alanine (28). All the mutants tested here contain such a motif in their hinges (although that of hhP227S is less satisfactory), but the antibodies hhS230P and hhS224/230P generally proved to be poor substrates for these enzymes, suggesting that the context of the Y-Pro-X-Pro motif may affect its susceptibility to cleavage.
The specific site at which bacterial IgA1 proteases cleave human wild-type IgA1 is determined by a region of the protease called the cleavage specificity determinant. Different cleavage specificity determinants permit cleavage to take place in wild-type human IgA1 at either a Pro-Ser (for a type 1 protease) or a Pro-Thr (for a type 2 protease) peptide bond, but in only one of the duplicated sequences in the hinge and not at the equivalent position in the other. Nevertheless, many IgA1 proteases, including representatives of those that cleave wild-type human IgA1 in the different duplicated halves of the hinge sequence, are still able to cleave IgA1 when only half of the IgA1 hinge is present in the hinge position of IgA1 protease-resistant IgA2, as in the IgA2-IgA1 half-hinge antibody (37). Among the IgA1 proteases that cleaved this hybrid antibody were those of S. pneumoniae and S. sanguis strains, whereas that of S. oralis failed to cleave it (37). The results presented here confirm this finding and also extend it by showing that the IgA2-IgA1 half-hinge antibody is also resistant to cleavage by the IgA1 proteases of strains of S. mitis biovar 1 and some other S. sanguis strains of different biovars. This finding is intriguing, as the IgA1 proteases of these different streptococcal species all cleave the same Pro-Thr peptide bond in wild-type human IgA1 (12, 34) and such a bond is present in the IgA2-IgA1 half-hinge antibody. The known extensive antigenic variation in the IgA1 proteases of S. mitis strains and their known closer relationship to the IgA1 proteases of S. oralis than to those of the other streptococci (31) might explain why the S. mitis IgA1 proteases behaved similarly to that of S. oralis in being unable to cleave the IgA2-IgA1 half-hinge antibody and differently from those of the other streptococci. The finding that the IgA1 proteases of S. mitis and S. oralis were also unable to cleave any of the mutant antibodies derived from the IgA2-IgA1 half-hinge antibody may indicate that it is the shortened hinge in these antibodies rather than its amino acid sequence that renders them resistant to cleavage by these IgA1 proteases. One might speculate that the C-terminal half of the IgA1 hinge normally acts as a "spacer," placing the N-terminal half of the hinge some distance away from the Fc region and allowing optimal access to the preferred Pro-Thr bond in the N-terminal half. Thus, in the hybrid antibody, the available Pro-Thr bond may be positioned too close to the Fc region of the antibody to allow efficient enzymatic activity by the protease.
The differing activities seen in this study, and also in a previous one (1), in the IgA1 proteases of different S. sanguis strains on the mutant antibodies may be a consequence of the difference in antigenicity between enzymes from strains of different biovars (34). Thus, the results support the view that the sensitivity of IgA1 to cleavage is determined not only by the presence of an appropriate cleavable peptide bond in the IgA1 hinge but also by the structures of both the IgA1 protease and the substrate upon which it acts.
The IgA1 proteases of streptococci cleave wild-type IgA1 at a Pro-Thr peptide bond (Fig. 1). Such a bond is present in the IgA2-IgA1 half-hinge antibody, whereas two are present in the mutant antibody hhS224P. This may be the reason for the finding that hhS224P is apparently more sensitive than the IgA2-IgA1 half-hinge antibody to cleavage with the IgA1 proteases of S. pneumoniae and some S. sanguis strains. Antibody hhP227S did not have a Pro-Thr site for cleavage. Thus, its sensitivity to the IgA1 proteases of S. pneumoniae and some S. sanguis strains indicates that these particular proteases are not dependent on the presence of a Pro-Thr peptide bond and that they can cleave at an alternative peptide bond. This result confirms our earlier findings with different hinge mutant IgA1 antibodies (1). The reduced sensitivity of antibodies hhS230P and hhS224/230P compared with that of antibody hhS224P to cleavage with these streptococcal IgA1 proteases, despite their very similar amino acid sequences, suggests that either the lack of serine at residue 230 or the consequent creation of a stretch of six proline residues immediately after the presumed cleavage site, or both of these things, has an inhibitory effect upon susceptibility to cleavage.
The mutant antibodies hhS224P and hhS230P were constructed so as to provide antibodies that, unlike the antibody IgA2-IgA1 half hinge from which they were derived, carried only one Pro-Ser peptide bond, either at the CH2 terminal end of the hinge, as for hhS224P, or at the CH1 terminal end, as for hhS230P. The difference between the sensitivities to cleavage of these antibodies was remarkable, for although the type 1 IgA1 proteases of N. meningitidis and N. gonorrhoeae both cleave the same Pro-Ser peptide bond in wild-type IgA1, and it is distant from the one cleaved by the type 1 protease of H. influenzae (Fig. 1), all these proteases were able to cleave antibody hhS224P more extensively than antibody hhS230P. This suggests that the availability of a cleavable peptide bond at the CH2 end of the hinge results in more efficient cleavage than when a scissile bond is present at the CH1 end.
Antibody hhS224/230P was created to provide an example of a hinge lacking Pro-Ser peptide bonds. The cleavage of this antibody by the type 1 (Pro-Ser-cleaving) proteases from H. influenzae and N. gonorrhoeae suggests that, when constrained, type 1 IgA1 proteases may be able to cleave at peptide bonds alternative to the one normally cleaved in wild-type IgA1. In this respect, they act like some streptococcal IgA1 proteases and the type 2 IgA1 proteases of N. meningitidis and H. influenzae (1).
Antibody hhP227S was created to provide an example of a half hinge lacking a Pro-Thr peptide bond. The finding that the type 2 (Pro-Thr-cleaving) IgA1 proteases of H. influenzae, N. meningitidis, and N. gonorrhoeae were able to cleave this antibody indicates that these enzymes are also similarly able to cleave a peptide bond alternative to that cleaved in wild-type IgA1. Moreover, the sensitivity of antibody hhP227S to cleavage with the Pro-Thr-cleaving IgA1 proteases of S. pneumoniae and some S. sanguis strains confirms the less stringent requirement of these streptococcal IgA1 proteases for a Pro-Thr bond in the hinge to ensure cleavage, which we reported previously (1).
Because of these findings, it might have been expected that, as all the other mutant antibodies possessed a Pro-Thr peptide bond at the same position as in the IgA2-IgA1 half-hinge antibody, they would have been cleaved by these type 2 IgA1 proteases with similar efficiencies and that this cleavage would be at least as efficient as in the cleavage of antibody hhP227S. However, as with the streptococcal IgA1 proteases, antibodies hhS230P and hhS224/230P were much more resistant than antibody hhS224P to cleavage by the Pro-Thr peptide bond-cleaving IgA1 proteases of Neisseria. This may be explained, as before for the streptococcal proteases, on the basis that the change of amino acid from serine at 230 to proline or the resultant creation of an uninterrupted sequence of six proline residues at the CH2 terminus of the hinge, or both, has a profound effect upon the susceptibility of the IgA1 hinge mutants to cleavage.
Since none of the proteases cleave wild-type IgA2, our experiments with the range of mutant IgA1 half hinges introduced into IgA2 described here allow us to conclude that many IgA1 proteases will cleave at alternative peptide bonds when presented with hinge sequences that lack the peptide bond normally cleaved in wild-type IgA1. In order to provide more detailed evidence in a representative digest, we chose to determine the precise site cleaved by the H. influenzae type 1 enzyme in antibody hhS224/230P. The site was revealed to lie between Pro227 and Thr228, showing that, in the absence of a suitably positioned Pro-Ser bond, this enzyme will cleave a Pro-Thr bond instead. This site localization provides a definitive demonstration that IgA1 proteases can cleave peptide bonds alternative to those cleaved in the wild-type IgA1 hinge if the hinge sequence is suitably constrained.
In conclusion, susceptibility to cleavage by IgA1 proteases appears to depend on the precise sequence, and hence the three-dimensional structure, of the IgA hinge, the context of the cleavage site within the hinge and within the protein as a whole, and the structure of the particular protein (as evidenced by its antigenic profile). However, it is clear that there is some flexibility in what most bacterial IgA1 proteases will accept as a substrate. Thus, many will cleave an IgA hinge lacking either the Pro-Ser or Pro-Thr bonds normally favored for cleavage. The design of reagents to inhibit IgA1 proteases is likely to represent a significant challenge. Nevertheless, the approach used here makes an important addition to our understanding of IgA1 protease-substrate interactions.
ACKNOWLEDGMENTS
We are grateful to M. Kilian for kindly donating some of the bacterial strains. We thank Douglas Lamont and Kenneth Beattie at the Fingerprints Proteomics Facility, University of Dundee, for MS analysis and for assistance with MS data interpretation.
J.M.W. thanks the Wellcome Trust for funding.
REFERENCES
1. Batten, M. R., B. W. Senior, M. Kilian, and J. M. Woof. 2003. Amino acid sequence requirements in the hinge of human immunoglobulin A1 (IgA1) for cleavage by streptococcal IgA1 proteases. Infect. Immun. 71:1462-1469.
2. Beck, S. C., and T. F. Meyer. 2000. IgA1 protease from Neisseria gonorrhoeae inhibits TNF-mediated apoptosis of human monocytic cells. FEBS Lett. 472:287-292.
3. Blake, M., K. K. Holmes, and J. Swanson. 1979. Studies on gonococcus infection. XVII. IgA1-cleaving protease in vaginal washings from women with gonorrhea. J. Infect. Dis. 139:89-92.
4. Brooks, G. F., C. J. Lammel, M. S. Blake, B. Kusecek, and M. Achtman. 1992. Antibodies against IgA protease are stimulated both by clinical disease and asymptomatic carriage of serogroup A Neisseria meningitidis. J. Infect. Dis. 166:1316-1321.
5. Chintalacharuvu, K. R., P. D. Chuang, A. Dragoman, C. Z. Fernandez, J. Qiu, A. G. Plaut, K. R. Trinh, F. A. Gala, and S. L. Morrison. 2003. Cleavage of the human immunoglobulin A1 (IgA1) hinge region by IgA1 proteases requires structures in the Fc region of IgA. Infect. Immun. 71:2563-2570.
6. Devenyi, A. G., A. G. Plaut, F. J. Grundy, and A. Wright. 1993. Post-infectious human serum antibodies inhibit IgA1 proteinases by interaction with the cleavage site specificity determinant. Mol. Immunol. 30:1243-1248.
7. Gilbert, J. V., A. G. Plaut, B. Longmaid, and M. E. Lamm. 1983. Inhibition of microbial IgA proteases by human secretory IgA and serum. Mol. Immunol. 20:1039-1049.
8. Halter, R., J. Pohlner, and T. F. Meyer. 1984. IgA protease of Neisseria gonorrhoeae: isolation and characterization of the gene and its extracellular product. EMBO J. 3:1595-1601.
9. Hauck, C. R., and T. F. Meyer. 1997. The lysosomal/phagosomal membrane protein h-lamp-1 is a target of the IgA1 protease of Neisseria gonorrhoeae. FEBS Lett. 405:86-90.
10. Horton, R. M., H. D. Hunt, S. N. Ho, J. K. Pullen, and L. R. Pease. 1989. Engineering hybrid genes without the use of restriction enzymes: gene splicing by overlap extension. Gene 77:61-68.
11. Insel, R. A., P. Z. Allen, and I. D. Berkowitz. 1982. Types and frequency of Haemophilus influenzae IgA1 proteases. Semin. Infect. Dis. 4:225-231.
12. Kilian, M., J. Mestecky, R. Kulhavy, M. Tomana, and W. T. Butler. 1980. IgA1 proteases from Haemophilus influenzae, Streptococcus pneumoniae, Neisseria meningitidis, and Streptococcus sanguis: comparative immunological studies. J. Immunol. 124:2596-2600.
13. Kilian, M., J. Reinholdt, H. Lomholt, K. Poulsen, and E. V. G. Frandsen. 1996. Biological significance of IgA1 proteases in bacterial colonization and pathogenesis: critical evaluation of experimental evidence. APMIS 104:321-338.
14. Kilian, M., and M. W. Russell. 1999. Microbial evasion of IgA functions, p. 241-251. In P. L. Ogra, J. Mestecky, M. E. Lamm, W. Strober, J. R. McGhee, and J. Bienenstock (ed.), Mucosal immunology, 2nd ed. Academic Press Inc., San Diego, Calif.
15. Kilian, M., B. Thomsen, T. E. Petersen, and H. Bleeg. 1983. Molecular biology of Haemophilus influenzae IgA1 proteases. Mol. Immunol. 20:1051-1058.
16. Kosowska, K., J. Reinholdt, L. K. Rasmussen, A. Sabat, J. Potempa, M. Kilian, and K. Poulsen. 2002. The Clostridium ramosum IgA proteinase represents a novel type of metalloendopeptidase. J. Biol. Chem. 277:11987-11994.
17. Lin, L., P. Ayala, J. Larson, M. Mulks, M. Fukuda, S. R. Carlsson, C. Enns, and M. So. 1997. The Neisseria type 2 IgA1 protease cleaves LAMP1 and promotes survival of bacteria within epithelial cells. Mol. Microbiol. 24:1083-1094.
18. Miller, P. B., H. Shen, N. R. Gilkes, D. G. Kilburn, R. C. Miller, A. G. Plaut, and R. A. J. Warren. 1992. Endoglucanase A from Cellulomonas fimi in which the hinge sequence of human IgA1 is substituted for the linker connecting its two domains is hydrolyzed by IgA proteases from Neisseria gonorrhoeae. FEMS Microbiol. Lett. 92:199-204.
19. Mortensen, S. B., and M. Kilian. 1984. Purification and characterization of an immunoglobulin A1 protease from Bacteroides melaninogenicus. Infect. Immun. 45:550-557.
20. Morton, H. C., J. D. Atkin, R. J. Owens, and J. M. Woof. 1993. Purification and characterization of chimeric human IgA1 and IgA2 expressed in COS and Chinese hamster ovary cells. J. Immunol. 151:4743-4752.
21. Mulks, M. H., S. J. Kornfeld, B. Frangione, and A. G. Plaut. 1982. Relationship between the specificity of IgA proteases and serotypes in Haemophilus influenzae. J. Infect. Dis. 146:266-274.
22. Mulks, M. H., and A. G. Plaut. 1978. IgA protease production as a characteristic distinguishing pathogenic from harmless Neisseriaceae. N. Engl. J. Med. 299:973-976.
23. Mulks, M. H., A. G. Plaut, H. A. Feldman, and B. Frangione. 1980. IgA proteases of two distinct specificities are released by Neisseria meningitidis. J. Exp. Med. 152:1442-1447.
24. Plaut, A. G. 1983. The IgA1 proteases of pathogenic bacteria. Annu. Rev. Microbiol. 37:603-622.
25. Plaut, A. G., J. V. Gilbert, M. S. Artenstein, and J. D. Capra. 1975. Neisseria gonorrhoeae and Neisseria meningitidis: extracellular enzyme cleaves human immunoglobulin A. Science 193:1103-1105.
26. Plaut, A. G., R. Wistar, and J. D. Capra. 1974. Differential susceptibility of human IgA immunoglobulins to streptococcal IgA protease. J. Clin. Investig. 54:1295-1300.
27. Pohlner, J., R. Halter, K. Beyreuther, and T. F. Meyer. 1987. Gene structure and extracellular secretion of Neisseria gonorrhoeae IgA protease. Nature 325:458-462.
28. Pohlner, J., T. Klauser, E. Kuttler, and R. Halter. 1992. Sequence-specific cleavage of protein fusions using a recombinant Neisseria type 2 IgA protease. Biotechnology 10:799-804.
29. Polissi, A., A. Pontiggia, G. Feger, M. Altieri, H. Mottl, L. Ferrai, and D. Simon. 1998. Large-scale identification of virulence genes from Streptococcus pneumoniae. Infect. Immun. 66:5620-5629.
30. Poulsen, K., J. Brandt, J. P. Hjorth, H. C. Thogersen, and M. Kilian. 1989. Cloning and sequencing of the immunoglobulin A1 protease gene (iga) of Haemophilus influenzae serotype b. Infect. Immun. 57:3097-3105.
31. Poulsen, K., J. Reinholdt, C. Jespersgaard, K. Boye, T. A. Brown, M. Hauge, and M. Kilian. 1998. A comprehensive genetic study of streptococcal immunoglobulin A1 proteases: evidence for recombination within and between species. Infect. Immun. 66:181-190.
32. Poulsen, K., J. Reinholdt, and M. Kilian. 1996. Characterization of the Streptococcus pneumoniae immunoglobulin A1 protease gene (iga) and its translation product. Infect. Immun. 64:3957-3966.
33. Qiu, J., G. P. Brackee, and A. G. Plaut. 1996. Analysis of the specificity of bacterial immunoglobulin A (IgA) proteases by a comparative study of ape serum IgAs as substrates. Infect. Immun. 64:933-937.
34. Reinholdt, J., M. Tomana, S. B. Mortensen, and M. Kilian. 1990. Molecular aspects of immunoglobulin A1 degradation by oral streptococci. Infect. Immun. 58:1186-1194.
35. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
36. Senior, B. W., M. Albrechtsen, and M. A. Kerr. 1987. Proteus mirabilis strains of diverse type have IgA protease activity. J. Med. Microbiol. 24:175-180.
37. Senior, B. W., J. I. Dunlop, M. R. Batten, M. Kilian, and J. M. Woof. 2000. Cleavage of a recombinant human immunoglobulin A2 (IgA2)-IgA1 hybrid antibody by certain bacterial IgA1 proteases. Infect. Immun. 68:463-469.
38. Senior, B. W., L. M. Loomes, and M. A. Kerr. 1991. Microbial IgA proteases and virulence. Rev. Med. Microbiol. 2:200-207.
39. Senior, B. W., W. W. Stewart, C. Galloway, and M. A. Kerr. 2001. Cleavage of the hormone human chorionic gonadotropin, by the type 1 IgA1 protease of Neisseria gonorrhoeae, and its implications. J. Infect. Dis. 184:922-925.
40. Wood, S. G., and J. Burton. 1991. Synthetic peptide substrates for the immunoglobulin A1 protease from Neisseria gonorrhoeae (type 2). Infect. Immun. 59:1818-1822.(Bernard W. Senior and Jen)