Streptokinase Promotes Development of Dipeptidyl Peptidase IV (CD26) Autoantibodies after Fibrinolytic Therapy in Myocardial Infarction Patients

Dipeptidyl peptidase IV (DPP IV) (CD26) plays a critical rolein the modulation and expression of autoimmune and inflammatorydiseases. We recently reported that sera from patients withrheumatoid arthritis and systemic lupus erythematosus containedlow levels of DPP IV and high titers of anti-DPP IV autoantibodiesof the immunoglobulin A (IgA) and IgG classes and found a correlationbetween the low circulating levels of DPP IV and the high titersof anti-DPP IV autoantibodies of the IgA class. Since streptokinase(SK) is a potent immunogen and binds to DPP IV, we speculatedthat patients with autoimmune diseases showed higher DPP IVautoantibody levels than healthy controls as a consequence ofan abnormal immune stimulation triggered by SK released duringstreptococcal infections. We assessed this hypothesis in a groupof patients suffering from acute myocardial infarction, withouta chronic autoimmune disease, who received SK as part of therapeuticthrombolysis. Concomitant with the appearance of anti-SK antibodies,these patients developed anti-DPP IV autoantibodies. These autoantibodiesbind to DPP IV in the region which is also recognized by SK,suggesting that an SK-induced immune response is responsiblefor the appearance of DPP IV autoantibodies. Furthermore, wedetermined a correlation between high titers of DPP IV autoantibodiesand an augmented clearance of the enzyme from the circulation.Serum levels of the inflammatory cytokines tumor necrosis factoralpha (TNF- ${alpha}$ ) and interleukin 6 (IL-6) increased significantlyafter 30 days of SK administration, while the levels of solubleIL-2 receptor remained unchanged during the same period, suggestinga correlation between the lower levels of circulating DPP IVand higher levels of TNF- ${alpha}$ and IL-6 in serum in these patients.

INTRODUCTION

Top

Introduction

Dipeptidyl peptidase IV (DPP IV) (CD26) is a widely distributed,multifunctional, highly glycosylated membrane-bound ectoenzyme(10) that cleaves X-Pro dipeptides from the NH₂ terminus ofa number of proteins (31). Expression of DPP IV is highly associatedwith cell differentiation and activation, and it is involvedin T-lymphocyte activation and migration across the extracellularmatrix (1, 28). DPP IV is also found in human plasma (24), whereits enzymatic activity is correlated with the activity of theenzyme in normal T lymphocytes (37). Not only are plasma DPPIV isoforms analogous to isoforms found in T lymphocytes (23),but they bind adenosine deaminase with similar specificity andaffinity (21), suggesting that the plasma enzyme originatesfrom T lymphocytes (22, 38).

Due to the key role that the membrane-bound DPP IV plays inT-cell-mediated immune responses and lymphokine synthesis (8),the enzyme has been studied in several autoimmune disorderssuch as rheumatoid arthritis (RA) and systemic lupus erythematosus(SLE). In both RA and SLE there is a reduction in serum DPPIV activity (11, 29, 30, 41). In a recent report (6), we demonstratedthat reduction of serum DPP IV activity in RA patients was dueto hypersialylation of the enzyme, whereas in SLE patients asimilar reduction in activity was possibly the result of increasedclearance of DPP IV from circulation due to the high titersof circulating anti-DPP IV autoantibodies of the immunoglobulinA (IgA) class (6).

In previous studies we found that streptokinase (SK), a proteinsecreted by streptococci which facilitates the development offocal infection in association with plasminogen (Pg) of thehost, can induce anti-Pg autoantibodies (16). We also demonstratedthat SK binds to DPP IV expressed by rheumatoid synovial fibroblasts(17). Given these observations and since SK is a very potentimmunogenic protein, we hypothesized that high titers of DPPIV autoantibodies in plasma in patients with autoimmune diseasescould possibly result from immune stimulation by bacterial proteinssuch as SK. We assessed this hypothesis in a group of patientswithout chronic autoimmune disease who had suffered from acutemyocardial infarction and had received SK as part of therapeuticthrombolysis.

We analyzed the expression and titers of anti-DPP IV antibodiesin serum for 90 days after administration of SK and found thatthese autoantibodies bind preferentially to an epitope in DPPIV which is also recognized by SK. We also analyzed serum levelsof the inflammatory cytokines tumor necrosis factor alpha (TNF- ${alpha}$ ),interleukin 6 (IL-6), and soluble interleukin 2 receptor (IL-2sR ${alpha}$ ), which are sensitive to DPP IV levels in the circulation(21). Since superantigens of bacterial origin have been postulatedas participants in the pathogenesis of autoimmune disease inhumans (32), our data suggest a potential role for SK in theseprocesses.

MATERIALS AND METHODS

Top

Materials and Methods

Patients. The protocol was approved by the Ethics Committees of both participatingcenters. After informed consent was obtained, a group of 10patients who were treated with SK due to myocardial infarctionduring 1994 and 1995, at the University of Chile Clinical Hospital,were studied to assess the induction of anti-SK and anti-DPPIV autoantibodies. All of them were men, with a mean age of47 years (age range, 39 to 72 years). Seven patients were chroniccigarette smokers. Three of the patients had arterial hypertension;two of them were being treated with atenolol and hydrochlorotiazide,and the third one was receiving enalapril. One of the patientshad diabetes mellitus and was being treated with tolbutamide.None of them had a history of streptococcal infection duringthe 6 months previous to the acute myocardial infarction. Theyreceived thrombolytic therapy as a single intravenous dose of1.5 x 10⁶ U of SK. All patients received aspirin, 250 mg/day,after administration of SK. Analyses of the levels of anti-DPPIV IgA, IgG, or IgM were performed with serum samples collectedimmediately before administration of SK. Additional serum sampleswere collected at 10, 30, and 90 days after administration ofthe thrombolytic agent. None of the patients had a prior historyof connective tissue disease or suffered from preexisting lymphoproliferativediseases. No one developed arthritis or any signs of connectivetissue disease during the 2 years following SK administration.

Peptides. The octapeptides LTSRPAHG and its randomized sequence APHLSTGRwere purchased from Multiple Peptide Systems, San Diego, Calif.

Proteins. DPP IV from a pool of human plasma (2 liters) was isolated bysequential purification using DEAE-Sepharose ion-exchange chromatography,Gly-Leu affinity chromatography, and gel filtration over SephacrylS-200 as described by Shibuya-Saruta et al. (38), followed byaffinity chromatography on immobilized adenosine deaminase asdescribed by De Meester et al. (7), with an overall yield of60% over that from serum. Calf intestinal mucosa adenosine deaminasewas obtained from Sigma Chemical Co., St. Louis, Mo.

Enzyme assays. DPP IV activity was measured in 96-well culture plates usingGly-Pro-p-nitroanilide (0.2 mM) as the substrate in reactionmixtures (200 µl) containing serum samples (40 µl)and 50 mM Tris-HCl, pH 8.0. The hydrolysis of the substratewas monitored at a wavelength of 405 nm using an Anthos Labteckinetic plate reader. Activity was expressed as ${Delta}$ A₄₀₅/min. Specificactivity was calculated using a molar extinction coefficientof 8,800 M^-1 cm^-1 for p-nitroanilide at 405 nm (9) and a molecularmass of 220,000 Da for DPP IV. All experiments were done intriplicate unless otherwise specified.

Antibodies. Antibodies to purified DPP IV from human serum were preparedin rabbits according to standard protocols (19). The polyclonalanti-DPP IV IgG was purified by affinity chromatography on proteinA-Sepharose (13) followed by immunoadsorption to DPP IV coupledto Sepharose 4B. The monoclonal anti-DPP IV IgG, IOA26-cloneBA5, was purchased from AMAC, Inc. (Westbrook, Maine). HumanIgG Fc fragment and goat affinity-purified F(ab')₂ fragmentsagainst human IgA, IgG, and IgM were purchased from ICN Pharmaceuticals,Inc. (Aurora, Ohio). Secondary antibodies to human IgA, IgG,and IgM were purchased from Sigma Chemical Co. Human anti-DPPIV IgA was purified from patient sera by chromatography on Jacalin-Sepharose(33) followed by immunoadsorption to DPP IV coupled to Sepharose4B. Human anti-DPP IV IgG was purified from patient sera bychromatography on protein A-Sepharose (13) followed by immunoadsorptionto DPP IV coupled to Sepharose 4B. Human anti-DPP IV IgM waspurified from patient sera by ion-exchange chromatography onmaltose binding protein-Sepharose (35), followed by immunoadsorptionto DPP IV coupled to Sepharose 4B.

ELISA. Enzyme-linked immunosorbent assays (ELISA) were performed in96-well culture plates. Quantification of anti-SK antibodieswas performed in 96-well culture plates coated with SK as previouslydescribed (15). For quantification of DPP IV, plates were coatedfirst with 200 µl of a solution containing 5 µgof an anti-DPP IV monoclonal antibody (MAb) (IOA26-clone BA5)per ml in 0.1 M Na₂CO₃ and incubated overnight at 4°C. TheMAb anti-DPP IV (clone BA5) is well characterized and has beenused for immunoprecipitations and Western blot analyses of full-lengthDPP IV (20). After coating, plates were rinsed with 200 µlof 10 mM sodium phosphate-0.1 M NaCl (pH 7.4) containing 0.05Tween 80 (PBS-Tween) to remove unbound protein. Nonspecificsites were blocked by incubating with PBS-Tween containing 2%bovine serum albumin (PBS-Tween-2% BSA) at room temperaturefor 1 h. Plates were washed again with PBS-Tween, air dried,and stored at 4°C. For assays, increasing concentrationsof sera were added in triplicate in a 200-µl final volumeof PBS-Tween-2% BSA and incubated al 37°C for 2 h, followedby rinsing with PBS-Tween and incubation with an affinity-purifiedpolyclonal anti-DPP IV rabbit IgG (100 ng/well) at 37°Cfor 1 h. Plates were then washed with PBS-Tween and 200 µlof alkaline phosphatase substrate (1-mg/ml p-nitrophenylphosphate)in 0.1 M glycine-1 mM MgCl₂-1 mM ZnCl₂ (pH 10.4), was addedto the plate, and absorbance was monitored at a wavelength of405 _nm using an Anthos Labtec Kinetic Plate reader. Bound DPPIV was expressed as ${Delta}$ A₄₀₅/min. Concentrations of DPP IV werecalculated from a calibration curve constructed with affinity-purifiedDPP IV. The concentrations of anti-DPP IV IgA, IgG, and IgMin sera of patients were determined by ELISA using 96-well cultureplates coated with DPP IV. Provisions were made to avoid reactivityof specific DPP IV autoantibodies with rheumatoid factors. Plateswere incubated with serum samples (1:200 dilution) in the presenceof IgG Fc fragments (1 µg/ml) at 37°C for 2 h, rinsedextensively with PBS-Tween, and finally incubated with alkalinephosphatase-conjugated anti-human IgA, IgG, and IgM F(ab')₂fragments at 37°C for 1 h, and the concentrations of boundimmunoglobulins were calculated from calibration curves expressingthe rate of hydrolysis of the alkaline phosphatase substratep-nitrophenylphosphate ( ${Delta}$ A₄₀₅/min) versus the concentration ofpure immunoglobulins of the three classes. The total concentrationof these immunoglobulins in patient sera was determined by ELISAin 96-well culture plates coated with F(ab')₂ specific anti-humanIgA, IgG, and IgM. After incubation of sera in the presenceof Fc fragments (1 µg/ml) to avoid cross-reactivity ofimmunoglobulins, the plates were rinsed with PBS-Tween and incubatedwith Fc-specific alkaline phosphatase-conjugated anti-humanIgA, IgG, and IgM F(ab')₂ fragments. Bound antibodies were measuredas described above and expressed as ${Delta}$ A₄₀₅/min. Final concentrationswere calculated from calibration curves constructed with purehuman IgA, IgG, and IgM. For peptide competition experiments,plates coated with DPP IV were first incubated with increasingconcentrations of the octapeptides LTSRPAHG or its randomizedsequence APHLSTGR in PBS-Tween-2% BSA and incubated for 1 hat 37°C before addition of a single concentration (100 ng/well)of anti-DPP IV autoantibodies. Concentrations of bound antibodieswere determined as described above.

Serum cytokinase assays. Concentrations of TNF- ${alpha}$ , IL-6, and IL-2 sR ${alpha}$ in serum were determinedby quantitative sandwich ELISA (R&D Systems, Minneapolis,Minn.) according to manufacturer's instructions. Analyses ofassay reproducibility showed intra-assay coefficients of variationfor TNF- ${alpha}$ , IL-6, and IL-2 sR ${alpha}$ equal to or less than 4.2, 2.0,and 4.6%, respectively, while interassay coefficients of variationwere equal to or less than 3.8, 2.6, and 1.8%, respectively.

Statistics. The statistical significance of differences between the variablesserum DPP IV levels, cytokine levels, and basal levels of IgG,IgA, and IgM antibodies against DPP IV was evaluated by meansof Kruskal-Wallis analysis of variance. The specific post hoccomparisons within each group of variables was performed bymeans of the Mann-Whitney U test. The changes between the baselineIgG, IgA, and IgM antibody levels among the SK-treated patientsand levels at 30, 60, and 90 days' time was evaluated by meansof the Friedman analysis of variance test.

RESULTS

Top

Results

Anti-SK antibody levels in patients who received SK therapeutically. Analysis of anti-SK IgA, IgG, and IgM levels (Fig. 1) showsthat IgA and IgG titers peaked 30 days after administrationof SK, followed by a slow decrease up to 90 days after the infusionof the thrombolytic agent. There were statistically significantdifferences between the IgA baseline levels and those at 10,30, and 90 days (medians, 34.99, 72.02, 74.41, and 60.98 µgof anti-SK IgA/mg of IgA, respectively; P < 0.0003 overalland P < 0.005 for each comparison). The differences betweenthe baseline IgG anti-SK antibodies and those at 10, 30 and90 days were also statistically significant (medians, 20.86,64.83, 78.04, and 74.38 µg of anti-SK IgG/mg of IgG, respectively;P < 0.0002 overall, and P < 0.008 for each comparison).No statistically significant differences between baseline and10, 30, and 90 days were observed for IgM anti-SK autoantibodies(medians, 28.88, 29.37, 26.99, and 27.32 µg of IgM/mg,respectively).

View larger version (29K):
[in this window]
[in a new window]
FIG. 1. Levels of anti-SK antibodies in serum from patients (n = 10) who received SK therapeutically. Serum samples (1:500 dilution) in 0.2 ml of PBS-Tween were added separately to 96-well microtiter plates coated with SK and analyzed as described in Materials and Methods. (A) Anti-SK IgA titers; (B) anti-SK IgG titers; (C) anti-SK IgM titers. n.s., not significant.

Anti-DPP IV autoantibodies levels in patients who received SK therapeutically. Concomitant with the increase of anti-SK IgA and IgG in serum,we observed a statistically significant increase of anti-DPPIV IgA from the baseline to 10, 30, and 90 days after SK administration(median concentrations in serum, 14.76, 24.34, 27.97, and 27.13µg of anti-DPPV IV IgA/mg of IgA respectively; P <0.0002 overall and P < 0015 for all comparisons) (Fig. 2A).Anti-DPP IV autoantibodies of the IgG class were also increased10 days after SK administration, reaching statistically significantlevels at 30 and 90 days after SK administration (median concentrationsin serum, 0.72, 0.78, 0.90, and 0.93 µg of anti-DPP IVIgG/mg of IgG, respectively; P < 004 overall; P < 0.057at 10 days; P < 0.012 at 30 and 90 days) (Fig. 2B). No statisticallysignificant increase in DPP IV autoantibodies of the IgM classwas found over time (median concentration in serum, 28.87, 25.45,24.03, and 25.96 µg of anti-DPP IV IgM/mg of IgM, respectively)(Fig. 2C).

View larger version (27K):
[in this window]
[in a new window]
FIG. 2. Analysis of anti-DPP IV autoantibodies from patients (n = 10) who received SK therapeutically. Serum samples (1:500 dilution) in 0.2 ml of PBS-Tween were added separately to 96-well plates coated with DPP IV and analyzed as described in Materials and Methods. (A) Anti-DPP IV IgA titers; (B) anti-DPP IV IgG titers; (C) anti-DPP IV IgM titers. n.s., not significant.

DPP IV levels and activity over time in sera of patients who received SK therapeutically. The median serum DPP IV levels from a group of patients (n =10) who received SK therapeutically (Fig. 3) show a continuousdecrease in the enzyme concentration between 0 and 90 days afterSK administration (1.9, 1.12, 1.12, and 0.80 µg/ml at0, 10, 30, and 90 days, respectively). The difference between0 and 90 days is statistically significant (P < 0.0042).However, measurements of DPP IV enzymatic activity during thesame period (data not shown) show no statistically significantchanges.

View larger version (14K):
[in this window]
[in a new window]
FIG. 3. Levels of DPP IV in serum from patients (n = 10) who received SK therapeutically. Serum samples (40 µl) in 0.2 ml of PBS-Tween were added to 96-well plates coated with an anti-DPP IV MAb and incubated for 2 h at 37°C as described in Materials and Methods. The plates were washed with PBS-Tween and incubated with a polyclonal anti-DPP IV IgG for 1 h at 37°C, followed by incubation with an alkaline phosphatase-conjugated protein A for 30 min. Bound DPP IV was detected by monitoring the hydrolysis of the alkaline phosphatase substrate p-nitrophenylphosphate at a wavelength of 405 nm.

Binding specificity of anti-DPP IV autoantibodies. SK binds to DPP IV via the LTSRPA amino acid sequence (17).Therefore, if interactions with SK are involved in the mechanismsresponsible for the production of anti-DPP IV autoantibodies,either SK or the LTSRPA peptide should be able to interferewith autoantibody binding to DPP IV. A series of ELISA experimentswas carried out to establish the binding specificity of theanti-DPP IV autoantibodies. Plates coated with DPP IV were incubatedwith a single aliquot (5 µl), in triplicate, from eachSK-treated patient in the presence of increasing concentrationsof the octapeptide LTSRPAHG or its randomized sequence APHLSTGR.The concentrations of anti-DPP IV autoantibodies of the IgAand IgG classes bound under these conditions were determinedas described in Materials and Methods. Since the inhibitioncurves were similar for all the sera tested (data not shown),we purified anti-DPP IV IgA and IgG from one serum with thehighest titers. Both purified anti-DPP IV antibodies react withthe immobilized antigen in a dose-dependent manner (Fig. 4).Incubation of DPP IV plates with a single concentration of eitherIgA (Fig. 5A) or IgG (Fig. 5B) anti-DPP IV autoantibody (500ng/ml) in the presence of increasing concentrations of the peptideLTSRPAHG, inhibited binding of both classes of autoantibodies,whereas its randomized sequence APHLSTGR produced no effect.

View larger version (21K):
[in this window]
[in a new window]
FIG. 4. Binding of anti-DPP IV autoantibodies to immobilized DPP IV. Increasing concentrations of purified anti-DPP IV IgA ( ${triangleup}$ ) or IgG ( ${circ}$ ) autoantibodies in PBS-Tween (200 µl/well) were added to 96-well plates coated with DPP IV and incubated for 2 h at 37°C. The plates were rinsed twice with PBS-Tween followed by a 1-h incubation at 37°C with alkaline phosphatase-conjugated anti-IgA or anti-IgG secondary antibodies. Binding was detected by monitoring the hydrolysis of the alkaline phosphatase substrate p-nitrophenylphosphate at 405 nm. Error bars, standard deviations.

View larger version (21K):
[in this window]
[in a new window]
FIG. 5. Inhibition of autoantibody binding to DPP IV by a peptide structurally homologous to the SK binding region to DPP IV. Ninety-six-well plates coated with DPP IV were incubated with increasing concentrations of the octapeptide LTSRPAHG or APHLSTGR in PBS-Tween for 1 h at 37°C. At this time, a single concentration (500 ng/ml) of IgA or IgG was added to the wells and incubated for 1 h at 37°C. Excess reagents were removed by rinsing the plates with PBS-Tween, and bound autoantibodies were detected with specific alkaline phosphatase-conjugated anti-IgA or anti-IgG secondary antibodies as described in Materials and Methods. (A) Inhibition of IgA autoantibody binding to DPP IV in the presence of increasing concentrations of LTSRPAHG ( ${triangleup}$ ) or its randomized sequence APHLSTGR ( ${blacktriangleup}$ ). (B) Inhibition of IgG autoantibody binding to DPP IV in the presence of increasing concentrations of LTSRPAHG ( ${circ}$ ) or its randomized sequence (•). Error bars, standard deviations.

Serum cytokine levels. Analyses of levels of TNF- ${alpha}$ , IL-6, and IL-2 sR ${alpha}$ in serum afterSK administration (Fig. 6A) show that TNF- ${alpha}$ increased from amedian of 1.55 pg/ml at baseline to 1.86 pg/ml at 10 days and7.04 pg/ml at 30 days, returning to near baseline (1.50 pg/ml)at 90 days after SK administration. With regard to serum IL-6levels (Fig. 6B), their medians also increased from 36.91 pg/mlat baseline to 93.57 and 117.80 pg/ml at 10 and 30 days afterSK administration, respectively. The median serum IL-6 levelsreturned close to baseline (38.86 pg/ml) at 90 days. For boththese cytokines, the differences between their levels at baselineand 10 and 30 days post-SK administration were statisticallysignificant, with a P of <0.02 and <0.006 for TNF- ${alpha}$ at10 and 30 days, respectively, and a P of <0.006 and <0.006for TNF- ${alpha}$ at10 and 30 days, respectively. The median IL-2 sR ${alpha}$ levels in serum changed from 610.68 pg/ml at baseline to 694.84,676.41, and 655.03 pg/ml at 10, 30, and 90 days after SK administration,respectively (Fig. 6C). None of these differences were statisticallysignificant.

View larger version (22K):
[in this window]
[in a new window]
FIG. 6. Analyses of levels of TNF- ${alpha}$ , IL-6, and IL-2 sR ${alpha}$ in serum. The concentrations of the three cytokines in serum samples (n = 10) from patients who received SK therapeutically were determined by an ELISA method as described in Materials and Methods. (A) TNF- ${alpha}$ titers; (B) IL-6 titers; (C) IL-2 sR ${alpha}$ titers. n.s., not significant.

DISCUSSION

Top

Discussion

In addition to its multiple roles in the expression or modulationof autoimmune and inflammatory responses (29), DPP IV functionsas a receptor for both Pg (14) and SK (17). In a previous study(16), we found evidence suggesting that high titers of anti-Pgautoantibodies in sera from patients with RA and SLE may bethe result of associations between SK (liberated during streptococcalinfections) and Pg of the host (16). Recently, we reported thepresence of anti-DPP IV autoantibodies in sera of patients withseveral autoimmune disorders (6). Since we found a correlationbetween appearance of Pg autoantibodies and anti-SK antibodiesin SK-treated patients (16) and taking into consideration thatSK associates with DPP IV (17), we investigated whether similarcorrelations existed in the appearance of anti-DPP IV autoantibodiesin patients who received SK as a thrombolytic agent. This studyshows a correlation between the appearance of anti-SK antibodiesand anti-DPP IV autoantibodies of the IgA class during the sameperiod. Inhibition of binding of these antibodies to DPP IVby the peptide LTSRPAH, an octapeptide containing the SK regionresponsible for binding to DPP IV (17), suggests SK as a promoterin the production of these autoantibodies. Although the specificactivity of DPP IV in these patients was not changed over time,there is a continuous reduction in the concentration of circulatingDPP IV, suggesting that the increase in levels of IgA autoantibodiesmay play a role in the augmented clearance of the enzyme fromcirculation. Levels of the inflammatory cytokines TNF- ${alpha}$ and IL-6in serum were significantly elevated after 30 days of SK administration,while the levels of IL-2 sR ${alpha}$ remained unchanged during the sameperiod, suggesting a correlation between the lower levels ofcirculating DPP IV and high levels of TNF- ${alpha}$ and IL-6 in thesepatients.

Newly synthesized DPP IV is cleared from circulation by theliver basolateral endosomes in association with polymeric IgAreceptors, responsible for the transportation of IgA to bile,via a joint transcytotic process (2). This process has beenalso suggested as a major mechanism to clear harmful antigensfrom the circulation in the form of IgA-antigen complexes viabile secretion (4) and clears both the polymeric IgA receptorsand the DPP IV together with the same kinetics (2). In thiscontext, we should mention that SK is also rapidly removed fromthe circulation via secretion into the bile (5).

Our study also demonstrates that the IgA class antibody responseagainst SK and DPP IV persists for at least 90 days after thrombolytictherapy with SK. Taking into consideration the short half-lifeof IgA in circulation, these findings suggest that SK persistsfor long periods in the host tissues, although it is not clearwhether it remains intact or degraded. It has been postulatedthat a significant elevation of IgA antibody titers may resultfrom enterobacterial antigenic stimulation via the gastrointestinaltract (39). In our study, intravenous administration of SK,without any evidence of mucosal participation, produced a predominantIgA anti-DPP IV response, thereby suggesting that high IgA circulatinglevels do not necessarily involve the gut as the site of infection.In this regard, it may be noteworthy that a large proportionof autoantibodies developed against plasma proteins after SKadministration appear to be of IgA class. Proteins like Pg (16)and lactate dehydrogenase M (33) have in common their reactivitywith SK. Most streptococcal infections of humans occur at skinor mucosal surfaces without causing bacteremia. Therefore, ifthe original immunogenic stimulus by streptococci is at a mucosalsurface, the humoral immune response is primarily IgA.

The role of DPP IV in immune function is very complex. In someexperimental models, such as adjuvant-induced arthritis (41)specific inhibition of DPP IV activity suppresses disease expression,pointing to a role for DPP IV activity in the pathogenesis ofexperimentally induced arthritis. However, in rheumatoid synovialfibroblasts (17), SK binds to DPP IV and induces an increasein intracellular calcium independently of DPP IV enzymatic activity,thereby suggesting that its role in the pathogenesis of arthritisis not confined to only its enzymatic activity. DPP IV alsohas a key regulatory role in the metabolism of peptide hormonesinvolved in psychoneuroendocrine, nutrition, and immune functions(21). Therefore, an increase in the clearance rate of circulatingDPP IV may have either negative or positive pathogenic implications.For example, since DPP IV inactivates TNF- ${alpha}$ (3), lower levelsof DPP IV may increase bioactive levels of this cytokine inthe circulation, thereby promoting TNF- ${alpha}$ -induced tissue damage(21). Conversely, in patients with inflammatory bowel disease,reduced levels of circulating DPP IV lead to increased levelsof bioactive glucagon-like peptide-2, facilitating enhancedrepair of the intestinal mucosal epithelium in vivo (46).

Although SK administration induced anti-DPP IV autoantibodiesand higher circulating levels of TNF- ${alpha}$ and IL-6, none of thesepatients developed an autoimmune disease at the beginning orduring 2 years after the thrombolytic therapy with SK. Workby other investigators (40) shows that myocardial infarctionpatients have high titers of anti-SK antibody up to 90 monthsafter SK administration, but their immune functions were similarto those of the general population. Based on the experimentaland clinical data available, the apparently contradictory effectson the immune system may be explained by a bimodal action ofDPP IV in immune function; immune reactions that have been initiatedby other mechanisms are supported by DPP IV enzymatic activity,while other immune reactions are rather reduced, thus focusingthe immunosurveillance on processes that are already under way(21).

A direct participation of bacterial components in the onsetof arthritis has been reported, but the mechanisms involvedare unclear (18, 25, 26, 36, 43, 45). It appears that exposureto a single isolated bacterial peptide is insufficient to triggeran autoimmune disease. Other factors such as up-regulation ofthe immunostimulatory cytokines TNF- ${alpha}$ , IL-6, or IL-1 by microbialentry into phagocytic cells (44); a distinct genetic profilebetween streptococcal strains (12); or autoimmune genetic susceptibilityof the host (25) may also be necessary for this process.

ACKNOWLEDGMENTS

This work was supported by National Heart, Lung, and Blood InstitutesGrant HL-24066.

FOOTNOTES

* Corresponding author. Mailing address: Department of Pathology, Box 3712, Duke University Medical Center, Durham, NC 27710. Phone: (919) 684-8879. Fax: (919) 684-8689. E-mail: gonza002{at}mc.duke.edu.

REFERENCES

Top