Skip to main content
NCBI home page
Infection and Immunity logoLink to Infection and Immunity
. 2003 May;71(5):2563–2570. doi: 10.1128/IAI.71.5.2563-2570.2003

Cleavage of the Human Immunoglobulin A1 (IgA1) Hinge Region by IgA1 Proteases Requires Structures in the Fc region of IgA

Koteswara R Chintalacharuvu 1,*, Philip D Chuang 1, Ashley Dragoman 1, Christine Z Fernandez 1, Jiazhou Qiu 2, Andrew G Plaut 2, K Ryan Trinh 1, Françoise A Gala 1, Sherie L Morrison 1
PMCID: PMC153282  PMID: 12704129

Abstract

Secretory immunoglobulin A (IgA) protects the mucosal surfaces against inhaled and ingested pathogens. Many pathogenic bacteria produce IgA1 proteases that cleave in the hinge of IgA1, thus separating the Fab region from the Fc region and making IgA ineffective. Here, we show that Haemophilus influenzae type 1 and Neisseria gonorrhoeae type 2 IgA1 proteases cleave the IgA1 hinge in the context of the constant region of IgA1 or IgA2m(1) but not in the context of IgG2. Both Cα2 and Cα3 but not Cα1 are required for the cleavage of the IgA1 hinge by H. influenzae and N. gonorrhoeae proteases. While there was no difference in the cleavage kinetics between wild-type IgA1 and IgA1 containing only the first GalNAc residue of the O-linked glycans, the absence of N-linked glycans in the Fc increased the ability of the N. gonorrhoeae protease to cleave the IgA1 hinge. Taken together, these results suggest that, in addition to the IgA1 hinge, structures in the Fc region of IgA are required for the recognition and cleavage of IgA1 by the H. influenzae and N. gonorrhoeae proteases.

Immunoglobulin A (IgA) is a primary line of immune defense against inhaled and ingested pathogens and their toxins at the mucosal surfaces of the respiratory, gastrointestinal, and genitourinary tracts. IgA functions to inhibit mucosal colonization by microorganisms in vivo and neutralize a variety of microbial toxins and enzymes, including those produced by Vibrio cholerae and oral streptococci (29, 30). While IgA binds to the antigen through the Fab region, the Fc regions interact with the Fc receptors present on NK cells, monocytes, macrophages, neutrophils, and B and T lymphocytes (17, 18). The carbohydrates in the Fc region of IgA interact with mucins to enable effective removal of the antigen from the host (22).

There are two subclasses of human IgA, IgA1 and IgA2. IgA1 has a 22-amino-acid hinge region containing up to five O-linked glycans at serine and threonine residues. In contrast, IgA2 has a 9-amino-acid hinge and lacks the O-linked glycans (8, 16, 37). While IgA1 contains two N-linked glycans in the Fc region, those at Asn263 and Asn459, IgA2m (1) contains four and IgA2m (2) and the novel IgA2 allotype IgA2(n) contain five N-linked glycans. Heterogeneity in glycan structures has been observed (2, 3, 9, 16).

The effectiveness of IgA is dependent on its structural integrity. However, some bacterial species such as Haemophilus influenzae, Neisseria meningitidis, Neisseria gonorrhoeae, Streptococcus pneumoniae, and Prevotella melaninogenica that cause diseases by colonizing or infecting human mucosal membranes secrete IgA1 proteases (10, 12, 13, 24, 25). The IgA1 proteases cleave proline-serine or proline-threonine peptide bonds in the hinge region of human IgA1 and secretory IgA1, thereby separating the Fab region from the Fc region and thus reducing the effectiveness of IgA (21, 35). IgA proteases from a few bacteria also cleave IgA2 that has a 13-amino-acid-shorter hinge lacking O-linked glycans (11, 12). Sequence analysis of the fragments obtained by cleaving serum IgA from gorillas, chimpanzees, and orangutans with H. influenzae, N. meningitidis, Clostridium ramosum, or S. pneumoniae indicated that both the composition and location of the scissile bond are critical in determining the specificity of IgA proteases (27). The IgA1 hinge contains two repeat sequences of 8 amino acids each. The proteases cleave at only one of the duplicated sequences and not at the other equivalent site (33). Carbohydrates in IgA1 may contribute to the recognition by IgA1 proteases (28).

In the present study, we used recombinant IgA1 with the IgA2 hinge, IgA2m (1) with the IgA1 hinge, and IgG2 with the IgA1 hinge to show that, while the IgA1 hinge is necessary for cleavage, in the context of IgG2, it is not sufficient for cleavage by H. influenzae or N. gonorrhoeae IgA1 proteases. Using IgA1/IgG2-domain-exchanged proteins, we showed that both Cα2 and Cα3 are required for H. influenzae and N. gonorrhoeae proteases to cleave IgA1. Using proteins lacking N-linked glycans and IgA1 containing O-linked carbohydrates only with GalNAc residues, we showed that, while there is no difference in the cleavage kinetics between wild-type IgA1 and IgA1 with truncated O-glycans, the presence of N-linked glycans in the Fc can have an impact. Taken together, these results suggest that structures in the Fc region of IgA1 are important for the recognition and cleavage of IgA1 by H. influenzae and N. gonorrhoeae proteases and thus provide insights for developing antibodies that are more resistant to protease cleavage and hence are more effective against bacterial infection.

MATERIALS AND METHODS

Reagents and cells.

Restriction endonucleases and molecular cloning enzymes were obtained from either New England Biolabs (Beverly, Mass.) or Stratagene (La Jolla, Calif.). [35S]Methionine was obtained from ICN Research Products (Costa Mesa, Calif.). The Sp2/0 myeloma cells and the transfectants were cultured in Iscove's modified Dulbecco's medium (IMDM) containing 5% bovine calf serum (BCS; HyClone Laboratories, Logan, Utah).

Generation of recombinant genes.

The α1 and α2 genes in Bluescript II (Stratagene) were used to generate IgA1-A2H and IgA2-A1H by exchanging SacII-XhoI fragments, including the hinges between the two plasmids.

To generate IgG2-A1H, a PCR product spanning Cγ1 and a second PCR product containing the α1 hinge and Cγ2 were generated and cloned at the 5′ end of Cγ3 in pBR322. The PCR fragment containing Cγ1 was generated by using IgG2 in Bluescript II as a template and the following primers: the T3 primer that anneals at the 5 ′ end of γ2 in Bluescript II and 5′-TGCAGAGAGAAGAATTCGGGAGTTACTCGGATCTGGGAGGAGAG-3′, designed to anneal between Cγ1 and the γ2 hinge and containing an EcoRI site (underlined).

The PCR fragment containing the α1 hinge and Cγ2 was produced by using pBR, containing a genomic fragment of a γ2 constant region and the following primers: 5′-CTCTGCAGTTCCCTCAACTCCACCTACCCCATCTCCCTCAACTCCACCTACCCCATCTCCCTCATGCTGCCACGGTAAGCCAGCCCAGGCCTCGC-3′, designed to anneal to the intronic sequence between the hinge, and Cγ2, containing the α1 hinge (underlined) and 5′-TACCCCGCGGGTCCCACCTTTGGTTTTGGAGATGGTTTTCTCGATGGGGGCTGGGAGCCCTTTGTTGGAGACCTTGCAC-3′, including a SacII site (underlined) designed to anneal to the reverse stand at the 3′ end of Cγ2.

The PCR products were cloned into TA vector and sequenced. A SalI-EcoRI fragment, containing Cγ1, and an EcoRI-SacII fragment, containing the α1 hinge and Cγ2, were ligated into the SalI-SacII sites present at the 5′ end of Cγ3 in pBR322. The complete constant region genes were cloned into the pSV2gpt mammalian expression vectors containing the dansyl variable regions (20, 34). The constant regions in the expression vectors were sequenced again to confirm the sequences of the hinge and constant regions.

The production of IgA1 lacking the N-linked carbohydrate sites by changing Asn263 and Asn459 to Gln was described previously (6), as was the production of IgA1/IgG2-domain-exchanged proteins (39).

Expression of recombinant antibodies in Sp2/0 myeloma cells.

The heavy (H)-chain expression vector was transfected into Sp2/0 cells expressing the dansyl-specific chimeric κ light (L)-chain gene by electroporation (34). Approximately 6 × 106 cells were washed in cold 0.02 M phosphate-buffered saline (PBS), pH 6.8, and incubated on ice for 10 min with 20 μg of DNA. Cells were pulsed with an electric field of 200 V and 960 μF in a Gene Pulser apparatus (Bio-Rad Laboratories, Richmond, Calif.); washed once; resuspended in 12 ml of IMDM containing 10% fetal calf serum (FCS), 100 μg gentamicin (Gibco-BRL, Grand Island, N.Y.)/ml, and 100 U nystatin (Gibco-BRL)/ml; and plated in 96-well tissue culture plates at 125 μl/well. After 2 days of growth, an equal volume of medium containing 15 μg of hypoxanthine/ml, 250 μg of xanthine/ml, and 6 μg of mycophenolic acid/ml was added to the wells to select for mycophenolic-acid-resistant colonies. After 2 weeks, the surviving colonies were screened for antibody production by enzyme-linked immunosorbent assay on microtiter plates coated with dansyl coupled to bovine serum albumin (BSA) with bound antibody detected by alkaline phosphatase-conjugated goat antiserum to human κ L chain (Sigma Chemical Co., St. Louis, Mo.). Color was developed by adding p-nitrophenyl phosphate (Sigma Chemical Co.), and the absorbance at 410 nM was determined in a microplate reader (MR 700; Dynatech, Chantilly, Va.). Clones producing the highest quantities of antibody were expanded in IMDM containing 10% (vol/vol) BCS.

Transient expression in 293T cells.

Transient transfection of human kidney epithelium 293T cells was performed by Ca2HPO4 precipitation. Reagents from 5 Prime→3 Prime, Inc. (Boulder, Colo.), or from Edge Biosystems (Gaithersburg, Md.) were used. Cells were grown to 60% confluency in 100 mM dishes overnight. Medium was removed, and 10 ml of fresh growth medium was added 2 h prior to transfection. A reagent mix containing 3 μg each of H- and L-chain expression vectors in 375 μl of distilled water, 125 μl of 1 M CaCl2, and 500 μl of phosphate buffer (50 mM HEPES [pH 7.05], 1.26 mM Na2HPO4, 140 mM NaCl) was added to the cell medium. After 4 h of incubation, the cells were rinsed and incubated in growth medium overnight before biosynthetic labeling.

Purification and characterization of recombinant proteins.

Transfectomas producing IgA were cultured in roller bottles (Becton Dickinson Labware, Lincoln Park, N.J.) in IMDM supplemented with 1% BCS (HyClone) and 6 mM Glutamax (Gibco-BRL). Supernatants were filtered to remove cells and cell debris and supplemented with 10 mM phosphate buffer (pH 6.8), 0.45 NaCl, 0.02 M EDTA, and 0.02% NaN3. Proteins were purified by affinity chromatography with AH-Sepharose beads coupled with the dansyl isomer 2-dimethyl-aminonapthalene-5-sulfonyl chloride (Molecular Probes Inc., Eugene, Ore.). Bound antibodies were eluted with N-(5-carboxy-pentyl)-2-dimethyl-aminonaphthyl-5-sulfonamide and concentrated, and the hapten was removed by extensive dialysis against Tris buffer, pH 7.8, containing 0.45 M NaCl and 0.02% NaN3. The concentration of proteins was determined by a bicinchoninic acid assay (Pierce, Rockford, Ill.) and was confirmed by intensity comparison with an Ig standard of known concentration following sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and staining with Coomassie blue.

To determine the aggregation status of IgG2-A1H, purified IgG2 and IgG2-A1H were analyzed by gel filtration on a 30- by 1.5-cm Superose 6 column (Pharmacia, Piscataway, N.J.) in PBS plus 0.02% NaN3. The fractions containing the peaks were analyzed by SDS-PAGE under nonreducing conditions (data not shown).

Preparation of bacterial IgA proteases.

Type 2 IgA1 protease of N. gonorrhoeae was obtained from the spent cell medium after culture in brain heart infusion broth containing IsoVitaleX (BBL Microbiology Systems; Becton Dickinson and Co., Cockeysville, Md.) for 16 h at 37°C. The type 1 IgA protease of H. influenzae strain Rd was obtained from the medium of cells cultured in brain heart infusion broth supplemented with 10 mg of beta-NAD/ml and hemin (both from Sigma Chemical Co.) for 16 h at 37°C. The enzyme in the medium was precipitated by ammonium sulfate, dialyzed, and resuspended in 50 mM Tris-HCl buffer, pH 7.5, to obtain a 25× concentration of the protease.

Biosynthetic labeling and immunoprecipitation.

To biosynthetically label transfectants with [35S]methionine, 6 × 106 cells were washed twice, incubated at 37°C for 30 min in methionine-free medium (Mediatech, Washington, D.C.), and then incubated for 16 h with 3 ml of methionine-free medium containing 1% (vol/vol) FCS and 37.5 μCi of [35S]methionine. To label transiently transfected 293T cells, cells grown on 5-ml petri dishes were washed twice with cold PBS and incubated overnight with 3 ml of methionine-free medium containing 1% (vol/vol) FCS and 37.5 μCi of [35S]methionine.

After labeling, supernatants were incubated for 1 h at 4°C with 30 μl of 50% (vol/vol) dansylated BSA coupled to Sepharose beads (DNS-BSA Sepharose). Antibodies bound to Sepharose beads were pelleted by centrifugation at 13,000 × g for 2 min and washed twice with 1 ml of phosphate buffer, pH 7.8, containing 0.45 M NaCl and twice with 1 ml of PBS. Antibodies were eluted by incubating the beads for 10 min on ice in 40 μl of 3 mM Nɛ-dansyl-l-lysine (Sigma Chemical Co.) in phosphate buffer, pH 7.8, containing 0.45 M NaCl.

Cleavage of antibodies by bacterial IgA proteases.

[35S]Methionine-labeled proteins were mixed with 2 μg of unlabeled IgA1-2 μl of concentrated protease supernatant and brought to a final volume of 20 or 25 μl in 50 mM Tris buffer, pH 7.5, containing 0.05% NaN3 and 10 mM EDTA; the reactions were set up on ice. Note that the incubations were performed in the presence of 2 μg of unlabeled IgA1 to ensure that equivalent amounts of substrate were always present. As a negative control, the reactions received 2 μl of PBS instead of protease supernatant. As a complete-digestion positive control, the reactions were performed without the 2 μg of IgA1 competitor. A mixture for all the time points was incubated in a 37°C water bath. At various time points, aliquots were removed and the proteolysis was stopped by the addition of sample buffer (2% SDS, 10% glycerol, 0.008% bromophenol blue, 25 mM Tris [pH 6.7]) containing 0.15 M β-mercaptoethanol and by boiling for 5 min.

SDS-PAGE analysis and quantification of cleaved products obtained after digestion with IgA1 proteases.

Samples were analyzed by SDS-PAGE in 12.5% Tris-glycine gels, the gels were dried, and the bands were detected by exposing the gels to Kodak XAR-5 film or by phosphorimager analysis. Films were photographed by using the ChemiImager model 4400, and bands were quantified by spot densitometry with the associated software (Alpha Innotech Corp., San Leandro, Calif.). Since the H. influenzae and N. gonorrhoeae IgA1 proteases do not cleave L chains, intact H-chain values were normalized against L-chain values in order to control for loading differences from lane to lane.

RESULTS

Production and characterization of chimeric IgA1-A2H, IgA2-A1H, and IgG2-A1H.

The hinge region of IgA1 is uniquely susceptible to cleavage by a number of bacterial proteases. A question of continuing interest is what the structural requirements for recognition and cleavage are. To further investigate the ability of IgA1 proteases from H. influenzae and N. gonorrhoeae to cleave the IgA1 hinge, we placed the hinge in the context of the IgA2 (IgA2-A1H) and IgG (IgG2-A1H) constant regions. We also replaced the hinge of IgA1 with that of IgA2 (IgA1-A2H). Transfectants producing IgA1-A2H, IgA2-A1H, and IgG2-A1H were generated as described in Materials and Methods. Biosynthetic labeling by growth in medium containing 35[S]methionine and analysis by SDS-PAGE showed that all of the transfectants secreted Igs (Fig. 1). Under nonreducing conditions, IgA1-A2H migrated predominantly as monomers and dimers with covalently associated L chain. IgA2-A1H migrated as monomers, dimers, and larger polymers and resembled IgA2m (1) in that the L chains were noncovalently associated dimers (Fig. 1A). Thus, the presence of the hinge from IgA1 did not alter the H-L disulfide bonding pattern of IgA2m (1). Interestingly, IgG2-A1H migrated predominantly as H-L with only a small amount of completely covalently assembled H2L2. Under reducing conditions, all H chains were of the expected molecular weight, with the IgA1-A2H H chain migrating with an Mr smaller than that of the IgA1 H chain, consistent with the absence of the longer IgA1 hinge region containing the O-linked glycans, and with the IgA2-A1 H chain migrating with an Mr larger than that of the IgA1 H chain due to the presence of four N-linked glycans and O-linked glycans in the IgA1 hinge. IgG2-A1H H chain migrated with Mr corresponding to the wild-type IgG2 H chain.

FIG. 1.

FIG. 1.

Open in a new tab

SDS-PAGE analysis of secreted wild-type and hinge exchanged antibodies. Transfectants were biosynthetically labeled with [35S]methionine for 16 h, and the labeled Igs were precipitated from the culture supernatants with dansylated BSA coupled to Sepharose. Labeled proteins were analyzed by SDS-PAGE in 5% PO4 gels under nonreducing conditions (A) and on 12.5% Tris-glycine gels after reducing with β-mercaptoethanol (B). pIg, Igs larger than dimers; dIg, dimeric Igs; mIg, monomeric Igs; Mr, molecular weight marker.

To address the issue of whether there was noncovalent association of the H-L molecules produced by IgG2-A1H, we compared the molecular weight of IgG2-A1H to that of wild-type IgG2 in solution under nondenaturing conditions by gel filtration on a Superose 6 column (Pharmacia). IgG2-A1H showed a major peak eluting at the same position as monomeric IgG2 (Fig. 2), indicating that, under nondenaturing conditions, noncovalent forces result in the formation H2L2. A similar noncovalent association has been observed for the H-L half-molecules produced by human IgG4 (1, 36).

FIG. 2.

FIG. 2.

Open in a new tab

Fast protein liquid chromatography analysis of wild-type IgG2 (A) and IgG2-A1H (B). A total of 100 μg of purified protein was analyzed by gel filtration on a 30- by 1.5-cm Superose 6 column in PBS. Note that both proteins migrated as a single peak with a retention volume of 17.5 ml.

IgA1 hinge is necessary but not sufficient for cleavage by H. influenzae and N. gonorrhoeae proteases.

To determine if the IgA1 hinge was necessary and sufficient for cleavage by H. influenzae and N. gonorrhoeae IgA1 proteases, the susceptibility of IgA1-A2H, IgA2-A1H, and IgG1-A1H to these enzymes was determined. [35S]Methionine-labeled proteins were incubated at 37°C with each enzyme for 2 h. The reaction was stopped by boiling the samples with β-mercaptoethanol. All incubations were performed in the presence of 2 μg of unlabeled IgA1 to ensure that equivalent amounts of substrate were always present. In addition, the degradation of unlabeled IgA1 served as a control for enzyme activity in each sample. Samples with no enzyme were used as negative controls. Both enzymes cleaved wild-type IgA1 and IgA2-A1H to produce bands corresponding to the Fab and Fc fragments (Fig. 3). The enzymes were unable to cleave IgA2 (data not shown) and IgA1-A2H. Therefore, in the context of the IgA1 and IgA2 constant regions, the IgA1 hinge was both necessary and sufficient for cleavage. In contrast, IgG2-A1H was resistant to cleavage by the enzymes, indicating that, in the context of the IgG2 constant region, the hinge of IgA1 was not sufficient for cleavage by H. influenzae and N. gonorrhoeae proteases. Therefore, in addition to the hinge region, structures present in both IgA1 and IgA2 and absent in IgG2 are required for the H. influenzae and N. gonorrhoeae enzymes to recognize and cleave in the IgA1 hinge.

FIG. 3.

FIG. 3.

Open in a new tab

SDS-PAGE analysis of wild-type IgG2, IgA1, and IgA1-A2H and the IgA2-A1H and IgG2-A1H proteins digested with H. influenzae (HI) and N. gonorrhoeae (NG) IgA1 proteases. 35[S]Methionine-labeled proteins were mixed with 2 μg of unlabeled IgA1-2 μl of concentrated protease supernatant in 50 mM Tris buffer, pH 7.5, containing 0.05% NaN3 and 10 mM EDTA in a total volume of 20 μl and incubated at 37°C for 2 h. The samples were boiled in sample buffer containing β-mercaptoethanol and analyzed by SDS-PAGE in 12.5% Tris-glycine gels. Mr, molecular weight marker.

Structures in Cα2 and Cα3 are important for cleavage by H. influenzae and N. gonorrhoeae proteases.

To further investigate the requirement of different domains in IgA for substrate recognition and cleavage by proteases, we determined the ability of H. influenzae and N. gonorrhoeae proteases to cleave the IgA1/IgG2-domain-exchanged proteins produced as described previously (39). Biosynthetically labeled proteins were incubated with each enzyme in the presence of 2 μg of unlabeled IgA1 to ensure that equivalent amounts of substrate were always present. Control samples were also prepared; a negative control consisted of antibody incubated without enzyme. Like wild-type IgA1, Cγ1[αH]Cα2Cα3 containing the α hinge, Cα2, and Cα3 was cleaved by H. influenzae and N. gonorrhoeae proteases (Fig. 4A). Not surprisingly Cγ1[γH]Cγ2Cα3, Cα1[γH]Cγ2Cγ3, and Cα1[γH]Cγ2Cα3 lacking the hinge of IgA1 were not cleaved (Fig. 4B and C). Interestingly, Cα1[αH]Cα2Cγ3 and Cα1[αH]Cγ2Cα3 containing the hinge of IgA1 and either Cγ2 or Cγ3 were not cleaved (Fig. 4C and D). These results suggest that structures from both Cα2 and Cα3 but not Cα1 are required for H. influenzae and N. gonorrhoeae proteases to recognize and cleave IgA1 in the hinge region.

FIG. 4.

FIG. 4.

Open in a new tab

SDS-PAGE analysis of domain-exchanged proteins digested with H. influenzae (HI) and N. gonorrhoeae (NG) IgA1 proteases. (A through C) a mixture containing 35[S]methionine-labeled proteins, 2 μg of unlabeled IgA1, and 2 μl of protease for each time point was incubated in a 37°C water bath. At various time points, aliquots were removed and the proteolysis was stopped and analyzed by SDS-PAGE as described in the legend to Fig. 3. (D) The H-chain and L-chain expression vectors were expressed in 293T cells and biosynthetically labeled with [35S]methionine as described in Materials and Methods. Labeled protein was digested and analyzed as described in the legend to Fig. 3. To eliminate the possibility of differences due to expression in different cell types, the wild-type IgA1 H chain and L chain were expressed in 293T cells and biosynthetically labeled and the labeled protein was digested with proteases and analyzed as described in the legend to Fig. 3. (data not shown).

Role of N-linked glycans in the cleavage of IgA1 by H. influenzae and N. gonorrhoeae proteases.

Carbohydrates in IgA have been proposed to contribute to the recognition by streptococcal IgA1 proteases (28). The Fc region of IgA1 contains two N-linked glycans at Asn263 and Asn459. To determine the role of N-linked glycans in the cleavage of IgA1 by H. influenzae and N. gonorrhoeae proteases, the kinetics of the cleavage of wild-type IgA1 and IgA1 lacking both N-linked glycans was determined. Biosynthetically labeled proteins were incubated with each enzyme, and aliquots were removed at various time points in order to assess degradation over time. Incubations were performed in the presence of unlabeled IgA1 to ensure that equivalent amounts of substrate were always present. A negative control consisted of antibody incubated without enzyme. Interestingly, under the experimental conditions used, we were never able to achieve 100% cleavage of the proteins, even in the absence of the competitor (data not shown). Cleavage of labeled IgA1 was similar in the presence and absence of unlabeled IgA1 (Fig. 5). Cleavage by H. influenzae protease was unaffected by the lack of N-linked glycans. In contrast, antibodies lacking N-glycans were more rapidly cleaved by the N. gonorrhoeae protease (Fig. 5 and 6), suggesting that the N-linked glycans play an important role in the cleavage by this protease.

FIG. 5.

FIG. 5.

Open in a new tab

Kinetics of digestion of wild-type IgA1 and IgA1 lacking N-linked glycans digested with H. influenzae (HI) (A) and N. gonorrhoeae (NG) (B) proteases. 35[S]Methionine-labeled IgA1 lacking N-linked glycans was digested and analyzed by SDS-PAGE as described in the legend to Fig. 3. Molecular weights are indicated to the left of the gels.

FIG. 6.

FIG. 6.

Open in a new tab

Degradation of wild-type IgA1, IgA1 lacking N-linked glycans, and IgA1 lacking N-linked glycans and containing truncated O-linked glycans by H. influenzae (A) and N. gonorrhoeae (B) proteases. Bands in Fig. 5 were quantified by spot densitometry. Since the H. influenzae and N. gonorrhoeae IgA1 proteases do not cleave L chains, intact H-chain values were normalized against the L-chain values obtained in order to control for loading differences from lane to lane. wt, wild type.

Role of O-linked glycans in the cleavage of IgA1 by H. influenzae and N. gonorrhoeae proteases.

IgA1, unlike IgA2 and IgG, contains O-linked glycans in its hinge region, and many proteases cleave directly adjacent to a serine or threonine bearing the O-linked glycan. To determine the role of O-linked glycans in the cleavage of IgA1 by N. gonorrhoeae and N. gonorrhoeae proteases, wild-type IgA1 and IgA1 lacking N-linked glycans and with truncated O-linked glycans produced by culturing cells in the presence of benzyl 2-acetamido-2-deoxy-α-d-galactopyranoside (BADG) were analyzed. BADG, a derivative of N-acetylgalactosamine, serves as a competitive inhibitor of UDP-GlcNAc:GalNAc-β1,3-N-acetylglucosaminyl-transferase, yielding O-glycans with only GalNAc residues (15). Our results indicate that the truncation of the O-linked glycans had no effect on the kinetics of cleavage (Fig. 6).

DISCUSSION

Certain bacteria constitutively secrete IgA1 proteases that may be virulence factors (23, 33). The IgA1 proteases cleave in the hinge of IgA and secretory IgA H chains, generating Fab and Fc fragments and thus enabling the bacteria to circumvent the IgA-mediated host defense. As IgA-based therapeutics gain importance, it is essential to determine the structures in IgA that are recognized by the proteases. Here, we have identified structures in IgA essential for IgA1 proteases to bind and cleave IgA1.

Proteases secreted by H. influenzae specifically cleave the Pro231-Ser232 bond (type 1) and by N. gonorrhoeae specifically cleave the Pro235-Thr236 bond (type II) in the hinge of IgA1 (13, 24, 25) but not in the hinge of IgA2. However, the N. gonorrhoeae but not H. influenzae protease will cleave IgA2 containing a 7-amino-acid sequence corresponding to the amino half of the IgA1 hinge incorporated into the equivalent site in IgA2 (33). The hinge of IgA1 contains a repeat of two identical and contiguous STPPTPS sequences, and a recombinant hybrid, IgA2, containing STPPTPS between the CH1 and hinge of IgA2 was also sensitive to type 2 N. gonorrhoeae protease but resistant to type I H. influenzae protease (33). Here, we have shown that the complete IgA1 hinge is cleaved by H. influenzae and N. gonorrhoeae proteases in both the IgA1 and IgA2 frameworks but was not in the context of IgG2. Therefore, additional structures present in IgA1 and IgA2 and absent in IgG2 appear to be required for the proteases to cleave. Analysis of the fragments obtained by cleaving serum IgA from gorillas, chimpanzees, and orangutans by a number of proteases also suggests that the enzyme cleavage of specific hinge bonds depends on the location of the bonds within a much larger segment of the α chain (27).

The catalytic efficiency of IgA1 proteases for the hinge region depends on its context. When the hinge sequence of IgA1 was substituted for the linker connecting the two domains of endoglucanase A, the fusion protein was cleaved very slowly. Additionally, oligopeptides similar to the IgA1 hinge are cleaved much less efficiently than intact IgA1 (14, 19, 26, 38). Here, using IgA1/IgG2-domain-exchanged proteins, we found that the H. influenzae and N. gonorrhoeae proteases cleave antibodies containing both Cα2 and Cα3, but not antibodies containing Cα1, and lacking either Cα2 or Cα3. These results suggest that both Cα2 and Cα3 are essential for H. influenzae and N. gonorrhoeae proteases to bind and cleave in the hinge region. Because the proteases cleave the IgA1 hinge in the context of IgA2, the structures required by the H. influenzae and N. gonorrhoeae proteases are also present in Cα2 and Cα3 of IgA2. Perhaps this is not surprising, given the extensive conservation in the sequence between IgA1 and the different allotypes of IgA2 (4). Since antibodies with the constant region composition Cα1[αH]Cα2Cγ3 and Cα1[αH]Cγ2Cα3 are not cleaved by the proteases, structures present in both Cα2 and Cα3 are either required for binding by the enzymes or, when Cγ2 or Cγ3 is present, the conformation of the hinge is altered so that the enzyme cannot cleave. Extensive deglycosylation of the IgA1 hinge also decreased its susceptibility to oral streptococci (28). Because type I and type II IgA1 proteases are enzymes, it is reasonable to postulate that the IgA1 proteases may contain a catalytic site and a recognition site and that binding to the structures in the Fc region of IgA is the first step in cleavage.

It is interesting that the presence of N-linked glycans outside of the cleavage site has an impact on cleavage by the N. gonorrhoeae protease but not the H. influenzae protease. IgA1 has two N-linked glycans in its Fc region, one in CH2 and one at the carboxy terminus of CH3. Analysis of the glycan structure shows that the Cα2 N-glycosylation site contains mostly biantennary glycans, while the tailpiece site contained mostly triantennary structures (16). The glycans present in IgA1 are highly (94%) sialylated and are exposed to solvent. Since the IgA1 glycans are surface exposed, they could potentially form part of the recognition site for the proteases, but this does not seem to be the case since cleavage either remains unchanged or improves in their absence. Instead, it may be that, for N. gonorrhoeae, the surface-exposed carbohydrates impede access and cleavage. Alternatively, distal conformation differences resulting from the absence of N-linked glycans may render the antibody a better substrate for N. gonorrhoeae binding and cleavage. However, the same changes do not alter recognition and cleavage by H. influenzae.

Reducing the carbohydrate content of the hinge of IgA1 to only GalNAc attached to the Ser and Thr residues did not influence the recognition and cleavage by the N. gonorrhoeae and H. influenzae proteases. These results and the observation that type 2 proteases from N. meningitidis and H. influenzae were able to cleave IgA1 with the specific Thr mutated to a Val (32) suggest that O-linked glycans may not play an important role in rendering the hinge susceptible to cleavage by H. influenzae and N. gonorrhoeae proteases. In contrast, enzymatic removal of the entire O-linked carbohydrate from the hinge made IgA1 less susceptible to cleavage by the proteins from Streptococcus sanguinis and Streptococcus oralis (28).

The IgA1 hinge segment failed to lead to the formation of H-H disulfide bonds when placed in the context of IgG2, although the transferred sequence contained two cysteine residues, Cys241 and Cys242, in the IgA1 H chain. Structural analysis has shown that, for IgA1, Cys241 forms an inter-H-chain disulfide (5) while Cys242 forms an intradomain disulfide bond with Cys299 in CH2. The hinge of IgG2 contains four Cys residues. Within the context of IgG2, it is possible that the two Cys residues of the IgA1 hinge are not oriented in a way that makes it possible to effectively form inter-H-chain bonds. IgG3 lacking a genetic hinge but with a Cys between CH1 and CH2 also failed to assemble and was secreted primarily as H-L half-molecules (7). Human IgG4 also is secreted with approximately 30% of the protein as H-L half-molecules (36). This is in part because, within the hinge, the Cys in the sequence Cys-Pro-Ser-Cys forms intraH-chain bonds instead of covalently linking the H chains (31). However, as is observed for IgG4 (36), strong noncovalent interactions assure that, in solution, most of the protein is in the H2L2 form. It is possible that, within the context of a noncovalently assembled molecule, the conformation of the IgA1 hinge may not be recognized by the proteases. However, it should be noted that, when only the CH3 domain is from IgG2, covalent assembly occurs but the proteases still failed to cleave the IgA1 hinge.

In summary, we have found that the context of the hinge of IgA1 is important for its cleavage by the N. gonorrhoeae and H. influenzae proteases. In particular, IgA1 with either a CH2 or CH3 domain from IgG2 is not cleaved, suggesting it will be possible to alter either the CH2 or CH3 of IgA so as to render the molecules resistant to protease cleavage. Whereas truncation of the O-linked carbohydrate does not impact cleavage by the proteases, lack of the N-linked carbohydrate makes the IgA1 more susceptible to cleavage by the N. gonorrhoeae protease. Increased understanding of the requirements for protease cleavage should make it possible to produce IgA that is a more effective therapeutic agent.

Acknowledgments

This work was supported by National Institutes of Health (NIH) Fellowship AI08994 to K.R.C. and by NIH grants DE09677 and NIDDK P30 DK34928 to A.G.P. and AI29470 and AI39187 to S.L.M.

Editor: J. N. Weiser

REFERENCES

  • 1.Aalberse, R. C., and J. Schuurman. 2002. IgG4 breaks the rules. Immunology 105:9-19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Baenziger, J., and S. Kornfeld. 1974. Structure of the carbohydrate units of IgA1 immunoglobulin. I. Composition, glycopeptide isolation, and structure of the asparagine-linked oligosaccharide units. J. Biol. Chem. 249:7260-7269. [PubMed] [Google Scholar]
  • 3.Baenziger, J., and S. Kornfeld. 1974. Structure of the carbohydrate units of IgA1 immunoglobulin. II. Structure of the O-glycosidically linked oligosaccharide units. J. Biol. Chem. 249:7270-7281. [PubMed] [Google Scholar]
  • 4.Chintalacharuvu, K. R., M. Raines, and S. L. Morrison. 1994. Divergence of human alpha-chain constant region gene sequences. A novel recombinant alpha 2 gene. J. Immunol. 152:5299-5304. [PubMed] [Google Scholar]
  • 5.Chintalacharuvu, K. R., L. J. Yu, N. Bhola, K. Kobayashi, C. Z. Fernandez, and S. L. Morrison. 2002. Cysteine residues required for the attachment of the light chain in human IgA2. J. Immunol. 169:5072-5077. [DOI] [PubMed] [Google Scholar]
  • 6.Chuang, P. D., and S. L. Morrison. 1997. Elimination of N-linked glycosylation sites from the human IgA1 constant region: effects on structure and function. J. Immunol. 158:724-732. [PubMed] [Google Scholar]
  • 7.Coloma, M. J., K. R. Trinh, L. A. Wims, and S. L. Morrison. 1997. The hinge as a spacer contributes to covalent assembly and is required for function of IgG. J. Immunol. 158:733-740. [PubMed] [Google Scholar]
  • 8.Endo, T., J. Mestecky, R. Kulhavy, and A. Kobata. 1994. Carbohydrate heterogeneity of human myeloma proteins of the IgA1 and IgA2 subclasses. Mol. Immunol. 31:1415-1422. [DOI] [PubMed] [Google Scholar]
  • 9.Field, M. C., R. A. Dwek, C. J. Edge, and T. W. Rademacher. 1989. O-linked oligosaccharides from human serum immunoglobulins A1. Biochem. Soc. Trans. 17:1034-1035. [DOI] [PubMed] [Google Scholar]
  • 10.Frandsen, E. V. G., E. Theilade, B. Ellegard, and M. Kilian. 1986. Proportions and identity of IgA1-degrading bacteria in periodontal pockets from patients with juvenile and rapidly progressive periodontitis. J. Periodontal Res. 21:613-623. [DOI] [PubMed] [Google Scholar]
  • 11.Fujiyama, Y., M. Iwaki, K. Hodohara, S. Hosoda, and K. Kobayashi. 1986. The site of cleavage in human alpha chains of IgA1 and IgA2: A2m(1) allotype paraprotein by the clostridial IgA protease. Mol. Immunol. 23:147-150. [DOI] [PubMed] [Google Scholar]
  • 12.Fujiyama, Y., K. Kobayashi, S. Senda, Y. Benno, T. Bamba, and S. Hosoda. 1985. A novel IgA protease from Clostridium sp. capable of cleaving IgA1 and IgA2m(1) but not IgA2m(2) allotype paraproteins. J. Immunol. 134:573-576. [PubMed] [Google Scholar]
  • 13.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. Bienenstock, and J. R. McGhee (ed.), Mucosal immunology. Academic Press, New York, N.Y.
  • 14.Klauser, T., J. Kramer, K. Otzelberger, J. Pohlner, and T. F. Meyer. 1993. Characterization of the Neisseria Igaβ-core. The essential unit for outer membrane targeting and extracellular protein secretion. J. Mol. Biol. 234:579-593. [DOI] [PubMed] [Google Scholar]
  • 15.Kuan, S. F., J. C. Byrd, C. Basbaum, and Y. S. Kim. 1989. Inhibition of mucin glycosylation by aryl-N-acetyl-alpha-galactosaminides in human colon cancer cells. J. Biol. Chem. 264:19271-19277. [PubMed] [Google Scholar]
  • 16.Mattu, T. S., R. J. Pleass, A. C. Willis, M. Kilian, M. R. Wormald, A. C. Lellouch, P. M. Rudd, J. M. Woof, and R. A. Dwek. 1998. The glycosylation and structure of human serum IgA1, Fab, and Fc regions and the role of N-glycosylation on Fc alpha receptor interactions. J. Biol. Chem. 273:2260-2272. [DOI] [PubMed] [Google Scholar]
  • 17.McGhee, J. R., M. E. Lamm, and W. Strober. 1999. Mucosal immune responses: an overview, p. 485-506. In P. L. Ogra, J. Mestecky, M. E. Lamm, W. Strober, J. Bienenstock, and J. R. McGhee (ed.), Mucosal immunology. Academic Press, New York, N.Y.
  • 18.Mestecky, J., and J. R. McGhee. 1987. Immunoglobulin A (IgA): molecular and cellular interactions involved in IgA biosynthesis and immune response. Adv. Immunol. 40:153-245. [DOI] [PubMed] [Google Scholar]
  • 19.Miller, P. B., H. Shen, N. R. Gilkes, D. G. Kilburn, R. C. Miller, Jr., 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. [DOI] [PubMed] [Google Scholar]
  • 20.Morrison, S. L. 1994. Cloning, expression, and modification of antibody V regions. Curr. Protoc. Immunol. 1:2.12.1.. [DOI] [PubMed] [Google Scholar]
  • 21.Mortensen, S. B., and M. Kilian. 1984. Purification and characterization of an immunoglobulin A1 protease from Bacteroides melaninogenicus. Infect. Immun. 45:550-557. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Phalipon, A., A. Cardona, J. P. Kraehenbuhl, L. Edelman, P. J. Sansonetti, and B. Corthesy. 2002. Secretory component: a new role in secretory IgA-mediated immune exclusion in vivo. Immunity 17:107-115. [DOI] [PubMed] [Google Scholar]
  • 23.Plaut, A. 1983. The IgA1 proteases of pathogenic bacteria. Annu. Rev. Microbiol. 37:603-622. [DOI] [PubMed] [Google Scholar]
  • 24.Plaut, A. G., R. J. Genco, and T. B. Tomasi, Jr. 1974. Isolation of an enzyme from Streptococcus sanguis which specifically cleaves IgA. J. Immunol. 113:589-591. [PubMed] [Google Scholar]
  • 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. [DOI] [PubMed] [Google Scholar]
  • 26.Pohlner, J., J. Kramer, and T. F. Meyer. 1993. A plasmid system for high-level expression and in vitro processing of recombinant proteins. Gene 130:121-126. [DOI] [PubMed] [Google Scholar]
  • 27.Qiu, J., G. P. Brackee, and A. Plaut. 1996. Analysis of the specificity of bacterial immunoglobulin A (IgA) proteases by comparative study of ape serum IgAs as substrates. Infect. Immun. 64:933-937. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Reinholdt, J., M. Tomana, S. B. Mortensen, and M. Kilian. 1990. Molecular aspects of immunoglobuliin A1 degradation by oral streptococci. Infect. Immun. 58:1186-1194. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Renegar, K. B., and P. A. Small, Jr. 1999. Passive immunization: systemic and mucosal, p. 729-738. In P. L. Ogra, J. Mestecky, M. E. Lamm, W. Strober, J. Bienenstock, and J. R. McGhee (ed.), Mucosal immunology. Academic Press, New York, N.Y.
  • 30.Russell, M. W., M. Kilian, and M. E. Lamm. 1999. Biological activities of IgA, p. 225-240. In P. L. Ogra, J. Mestecky, M. E. Lamm, W. Strober, J. Bienenstock, and J. R. McGhee (ed.), Mucosal immunology, 2nd ed. Academic Press, New York, N.Y.
  • 31.Schuurman, J., G. J. Perdok, A. D. Gorter, and R. C. Aalberse. 2001. The inter-heavy chain disulfide bonds of IgG4 are in equilibrium with intra-chain disulfide bonds. Mol. Immunol. 38:1-8. [DOI] [PubMed] [Google Scholar]
  • 32.Senior, B. W., M. R. Batten, M. Kilian, and J. M. Woof. 2002. Amino acid sequence requirements in the human IgA1 hinge for cleavage by streptococcal IgA1 proteases. Biochem. Soc. Trans. 30:516-518. [DOI] [PubMed] [Google Scholar]
  • 33.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. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Shin, S. U. a. M., S. L. 1989. Production and properties of chimeric antibody moleucules. Methods Enzymol. 178:459-476. [DOI] [PubMed] [Google Scholar]
  • 35.Spooner, R. K., W. C. Russell, and D. Thirkell. 1992. Characterization of the immunoglobulin A protease of Ureaplasma urealyticum. Infect. Immun. 60:2544-2546. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Tan, L. K., R. J. Shopes, V. T. Oi, and S. L. Morrison. 1990. Influence of the hinge region on complement activation, C1q binding, and segmental flexibility in chimeric human immunoglobulins. Proc. Natl. Acad. Sci. USA 87:162-166. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Toraño, A., Y. Tsuzukida, Y. S. Liu, and F. W. Putnam. 1977. Location and structural significance of the oligosaccharides in human Ig-A1 and IgA2 immunoglobulins. Proc. Natl. Acad. Sci. USA 74:2301-2305. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Wood, S. G., and D. R. Burton. 1991. Synthetic peptide substrates for the immunoglobulin A1 protease from Neisseria gonorrhoeae (type 2). Infect. Immun. 59:1818-1822. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Yoo, E. M., M. J. Coloma, K. R. Trinh, T. Q. Nguyen, L.-U. Vuong, S. L. Morrison, and K. R. Chintalacharuvu. 1999. Structural requirements for polymeric immunoglobulin assembly and association with J chain. J. Biol. Chem. 274:33771-33777. [DOI] [PubMed] [Google Scholar]

Articles from Infection and Immunity are provided here courtesy of American Society for Microbiology (ASM)

RESOURCES