Human T Lymphocyte Response to Borrelia burgdorferi Infection: No Correlation between Human Leukocyte Function Antigen Type 1 Peptide Response and Clinical St

Departments of ¹Dermatology, ²Medicine, ³Neurology, and ⁴Preventive Medicine, State University of New York at Stony Brook, Stony Brook, and ⁵Plum Island Animal Diseases Center, Greenport, New York; ⁶Neurology Service, New Mexico Veterans Affairs Health Care System, Albuquerque

Received 12 July 2002; revised 25 September 2002; electronically published 13 December 2002.

     US Department of Health and Human Services and institutional guidelines for informed consent were followed. Protocols were approved by the State University of New York at Stony Brook Committee on Research Involving Human Subjects and the University of New Mexico School of Medicine Human Research Review Committee.
     Financial support: National Institutes of Health (program project grant 1PO1NS3409201).
     Reprints or correspondence: Dr. Richard S. Kalish, Dept. of Dermatology, Health Sciences Center T-16, 060, SUNY at Stony Brook, Stony Brook, NY 11794-8165 

     The mechanisms of chronic sequelae of Lyme disease in humans, particularly neurological disease and arthritis, have not been well elucidated. It is hypothesized that chronic sequelae result from an autoimmune process triggered by the infection. Alternatively, chronic sequelae may result from the inability to rid the host of the Borrelia burgdorferi spirochete. Other possibilities include bacterial antigens resistant to enzymatic degradation and irreversible damage to joints.

     The human immune response to B. burgdorferi infection depends at least in part on T cell responses. Evidence from passive transfer studies and active immunization with outer surface protein A (OspA) and OspC has indicated that protection from B. burgdorferi can be antibody mediated [14]. T cell help is essential for the generation of IgG, and a human T cell response to B. burgdorferi has been well documented. Furthermore, it is possible that the T cell response to B. burgdorferi contributes to chronic sequelae after infection. Peripheral blood T cells from patients with Lyme disease exhibit proliferation in vitro to B. burgdorferi antigens [57]. These T cell responses can be detected after proven infection in patients with negative antibody responses, as measured by ELISA [8]. T cells from synovium [9] and cerebrospinal fluid (CSF) [10, 11] of patients with chronic Lyme disease also proliferate in response to B. burgdorferi antigens.

     Multiple B. burgdorferi antigens are recognized by T cells, including OspA, OspB, OspC, flagellin, and Hsp70 proteins [1215]. T cells from affected patients respond to many of these antigens [16], and human T cell clones have been generated that proliferate specifically to these antigens. The response is greater among patients with long-standing disease. There is also a compartmentalized response, with synovial T cells demonstrating a greater response than peripheral blood T cells. During murine infection with B. burgdorferi, OspA expression is lost by day 30, whereas OspC expression is up regulated [17]. For this reason, it has been thought that the immune response is primarily directed toward OspC, in preference to OspA.

     OspA from B. burgdorferi sensu stricto contains HLA-DR4restricted epitopes that cross-react with human LFA-1 [1820]. It has been proposed that this molecular mimicry explains the rare occurrence of antibiotic-resistant chronic arthritis in patients with HLA-DR4. The purpose of the present project was to test the clinical relevance of T cell recognition of the cross-reactive epitope of human LFA-1 peptide.

SUBJECTS, MATERIALS, AND METHODS

     Human subjects.     The study was confined to subjects aged between 1870 years who were residents of New York, New Jersey, or Connecticut, and control subjects who were residents of New Mexico. This was part of a larger prospective study of the clinical outcome of neurological Lyme disease that is still ongoing and will be reported elsewhere.

     Early Lyme disease criteria.     Subjects presented with the erythema migrans rash of Lyme disease, as confirmed by an experienced physician. Work included documentation of the size and number of erythema migrans lesions, review of systems, and neurological examination. Serologic test results were obtained, but positive results were not necessary for inclusion in the presence of a documented rash of erythema migrans. Lumbar puncture was performed when indicated on patients with neurological findings. Subjects with multiple lesions of erythema migrans were included. Subjects with neurological findings (meningitis or cranial neuropathy) on presentation with erythema migrans were classified as having early neurological disease and were analyzed separately. Forty-eight patients with early infection presented with erythema migrans. Thirty-six had erythema migrans without neurological involvement, whereas 12 had concurrent neurological disease. Blood samples were obtained for studies on the day of presentation. Convalescent (4 weeks of follow-up) blood samples also were obtained from 16 of these patients.

     Chronic Lyme syndrome criteria.     Subjects had a history of symptoms of >3 months' duration. The history of Lyme disease was confirmed either by physician-documented erythema migrans rash or late clinical manifestations, as well as laboratory confirmation of infection. Late manifestations included recurrent briefly swollen joints, acute onset of secondary or tertiary atrioventricular conduction defect, meningitis, cranial neuropathy, radiculoneuritis, and encephalomyelitis with intrathecal antibody production. Chronic antibiotic-resistant arthritis was not essential for inclusion. Acceptable laboratory confirmation included culture of B. burgdorferi, positive antibody titer in CSF, or acute and convalescent serum documentation of increasing titers. Serologic test results were confirmed by Western blot according to the Centers for Disease Control and Prevention criteria. Twenty patients with chronic Lyme disease syndrome were studied. LFA-1 peptide data were available from 13 of these patients.

     Endemic-area control subjects from Suffolk County, New York.     Fifty Lyme endemicarea healthy control subjects, aged 1870 years, were recruited by random digit dialing and matched for area code with the patients with Lyme disease. Control subjects were excluded if they had a history of or syndromes consistent with Lyme disease. Evaluation included physical examination, including neurological examination, and serologic testing.

     Nonendemic-area control subjects from New Mexico.     Eighteen nonendemic-area (New Mexico) control subjects were recruited by Dr. Larry Davis, New Mexico Veterans Affairs Health Care system. New Mexico control subjects had always lived in New Mexico, never extensively vacationed in the northeastern United States, and never been diagnosed as having Lyme disease. Blood was obtained from 18 healthy subjects (2 women and 16 men), with a mean age of 50 years. Subjects had no history of immunosuppressive medication or acute illness.

     Recombinant B31 OspA and OspC.     Recombinant antigens from a B31 strain isolate of B. burgdorferi were prepared, as described elsewhere [21, 22].

     LFA-1 peptide.     The human LFA-1 peptide IYVIEGTSKQDLTSF recognized by OspA-reactive T cells [1820] was synthesized commercially by Chiron Technologies. The peptide was 92% pure by high-performance liquid chromatography analysis, with free (unblocked) C- and N-terminal amino acids.

     Proliferation assays.     PBMC were isolated from heparinized blood by centrifugation on Ficoll-hypaque gradients (Pharmacia). They were then washed and resuspended in AIM V media (Gibco) at 1 × 10⁶ cells/mL for proliferation assays. The PBMC (0.1 mL/well) were added to U-bottom 96-well trays (Linbro), with 0.1 mL of appropriate antigen. Antigens were added at the following final concentrations: OspA and OspB (0.1 g/mL) and LFA-1 peptide (10 g/mL). At the initiation of the study, a dose-response curve was tested for each donor. After considerable experience, the concentrations were narrowed to those shown below, because they gave the maximal results. The trays were incubated at 37°C in 7.5% CO₂ for 5 days, then pulsed with ³H-thymidine (1 Ci/well), and harvested 18 h later.

     Cytokine production assay.     Freshly isolated PBMC were resuspended at 2 × 10⁶ cells/mL in AIM V medium (Gibco) and were added (0.1 mL/well) to U-bottom 96-well trays, with 0.1 mL of either AIM V media or appropriate antigen. At 5 days, culture supernatant was harvested, centrifuged, aliquoted, and frozen at -70°C for cytokine assay.

     Cytokine ELISA assays.     Interleukin (IL)4, IL-5, and interferon (IFN) were assayed by ELISA. Monoclonal anti-human cytokine antibody pairs of purified antibody and biotinylated antibody were obtained from Pharmingen and were used according to the specified protocol. Capture and biotinylated antibodies were both used at 1 g/mL. Quadruplicate wells were used for each sample, and a standard curve was generated from 0 to 1.0 ng/mL. ELISA plates were read on a Molecular Devices microplate reader. Final cytokine concentrations represent the product of 2 × 10⁵ PBMC/0.2 mL of medium or 1 × 10⁶ PBMC/mL. Antigen-driven cytokine production was determined by subtracting cytokine production in the presence of media alone. For the purposes of scoring positive or negative production, 0.1 ng/mL was taken as a threshold.

     Statistical analysis.     Positive response rates in  and  were compared by use of Fisher's exact test. The results for which the P values were .05 for groups in both tables and for which the P values are listed in the footnotes of these tables remained significant at = .05, after Simes' improved Bonferroni procedure was used to adjust for multiple comparisons [23]. Correlations among continuous variables were tested using Spearman's correlation coefficient. Categorical measures of agreements were tested using the statistic.

fig.ommitted

Table 1.          Stimulation indices in response to Borrelia antigens.

fig.ommitted

Table 2.          Interferon (IFN) production to Borrelia antigens.

RESULTS

     Acute versus convalescent T cell responses to recombinant OspA and OspC.     PBMC of patients with Lyme disease at both the acute stage and at 4 weeks of follow-up were tested for proliferation to recombinant OspA and OspC . Of 39 patients presenting with early erythema migrans but without neurological disease, 11 (28%) had a proliferation response  and 3 (8%) had a IFN- response to OspA . At 4 weeks of follow-up, 16 of 32 had a proliferation response, with an IFN- response (>0.100 ng/mL) in 3 (19%) of 16 patients ( and ). In contrast, IL-4 and IL-5 were produced at >0.100 ng/mL by <10% of subjects, and no boost was seen at 4 weeks of follow-up.

     T cell proliferation to recombinant OspC was not as frequent as the response to OspA . At presentation, 5 (13%) of 40 patients with erythema migrans exhibited proliferation in response to OspC. This increased to 6 (38%) of 16 patients at 4 weeks' follow-up. Patients tended to respond to both OspA and OspC. However, there was no correlation of stimulation indices.

     An IFN- response to recombinant OspC was rare, with the exception of those patients with chronic Lyme disease syndrome who showed a 35% response rate (7/20). This was significantly different (P = .005) from the response of other Lyme disease groups, as well as endemic-area control subjects.

     T cell response to human LFA-1 peptide.     It has been proposed that chronic arthritis can be induced by B. burgdorferi infection as a result of cross-reactivity between OspA and a human LFA-1 peptide, as described elsewhere [1820]. T cell proliferation to this LFA-1 peptide was tested in a subset of 85 subjects . There was a significant correlation between the stimulation index of the response to OspA and LFA-1 peptide (P < .001; Spearman's correlation coefficient). Of 23 subjects responding to recombinant OspA, 23 (78%) also responded to LFA-1 peptide. Conversely, of 20 subjects responding to LFA-1 peptide, 18 (90%) also responded to OspA. Response to both antigens was seen in 18 (72%) of 25 subjects responding to either antigen. The subjects responding to LFA-1 but not OspA were 1 control subject from New Mexico and 1 patient with chronic Lyme disease syndrome. The sensitivity of LFA-1 response using OspA response as stardard was 74%, with a specificity of 97%. Nonparametric analysis showed a significant (P < .001, statistic) association between the response to LFA-1 peptide and OspA.

fig.ommitted

Table 3.          T cell response to human leukocyte function antigen (LFA)1 peptide.

fig.ommitted

Figure 1.        Correlation of stimulation indices for outer surface protein A (OspA) and leukocyte function antigen (LFA) type 1 peptide. Spearman's correlation coefficient, 0.737 (P < .001).

     Response to LFA-1 peptide did not correlate with clinical category. Eight (30%) of 27 of the tested endemic control subjects gave a response to human LFA-1 peptide, as well as 3 (30%) of 10 4-week follow-up patients and 3 (23%) of 13 patients with chronic Lyme disease syndrome. The antigen specificity of the LFA-1 peptide response in endemic control subjects was supported by its association with response to recombinant OspA in all 8 subjects and the low response rate of control subjects from New Mexico.

     Control populations.     Unexpectedly, 22 (44%) of 50 endemic-area control subjects (Suffolk County, Long Island, New York) showed a proliferation response to recombinant OspA , and 10 (21%) of 48 produced IFN- . In contrast, only 3 (17%) of 18 control subjects from New Mexico had a proliferative response to OspA (P = .049), and 2 (13%) of 16 produced IFN- (P = .013). The proportion of control subjects with a T cell response to Borrelia species exceeded that with a serological response. Only 20% of these endemic control subjects were seropositive, as confirmed by Western blot. Lack of correlation between serologic test results and proliferation response to Borrelia species has been reported elsewhere [8]. One of the control subjects from New Mexico was seropositive for Borrelia species by Western blot. This donor produced IFN- in response to whole Borrelia species, recombinant OspA, and recombinant OspC.

DISCUSSION

     It had been proposed that chronic Lyme arthritis may be mediated by a cross-reaction of OspA with human LFA-1 [1820]. We found that reactivity to the cross-reactive epitope of human LFA-1 was present in 18 (72%) of 25 subjects responding to OspA. There was a significant (P < .01) association between reactivity to OspA and reactivity to human LFA-1 peptide. However, there was no correlation between clinical status and response to the human LFA-1 peptide. The same strong correlation between LFA-1 and OspA response was present for all groups of subjects, including those without systemic symptoms and endemic control subjects with no history of disease.

     It should be noted that patients with chronic Lyme syndrome were not selected for antibiotic-resistant arthritis, and no conclusion can be made about the level of LFA-1 reactivity in the arthritis subset. However, among those patients reacting to OspA, almost all of them also reacted to human LFA-1 peptide, and LFA-1 peptide reactivity did not correlate with disease status.

     The human T cell response to infection with B. burgdorferi was marked by an early proliferative response to recombinant OspA and OspC, which approximately doubled in frequency at 4 weeks after presentation. Patients with neurological disease at presentation exhibited higher levels of T cell reactivity, close to that of the 4-week follow-up patients. This may simply reflect a time delay in diagnosis. The cytokine response to OspA and OspC was characterized by IFN- production, with minimal IL-4 or IL-5, and exhibited a similar boost at 4 weeks. There was no difference in the IFN- bias of the cytokine response between patient groups. The only significant difference between patients with chronic Lyme disease syndrome and patients with early disease was the production of IFN- in response to OspC by patients with chronic Lyme syndrome.

     The high rates of T cell response of endemic-area control subjects to recombinant OspA and OspC raises the question of cross-reactivity of the T cell response with other Borrelia species. Evidence that the observed T cell response to OspA was related to Borrelia species exposure includes the low level of reactivity found in nonendemic New Mexico control subjects, as well as the significant boost in reactivity detected between presentation and at 4 weeks' follow-up (acute vs. convalescent).

     We have reported elsewhere that, of >20 genotypes of OspC expressed in ticks, only a minority are associated with human skin infection, and only 4 genotypes are associated with invasive disease [24]. This may explain both the low rate of T cell reactivity to the B31 strain OspC used for the present study and the high incidence of T cell reactivity to Borrelia antigens in endemic-area control subjects with no history of Lyme disease. Cross-reactivity between the OspC genotype proteins is minimal [25], whereas Lyme serologies in North America recognize the B31 strain of OspA [26]. The high level of T cell reactivity to Borrelia species in endemic-area subjects with no history of Lyme disease may result from infection with Borrelia OspC genotypes not associated with clinical disease.

     OspA was demonstrated to be an important target of the early T cell immune response to B. burgdorferi, and OspA T cell responses increased over the first 4 weeks after infection. Human T cell responses to OspA have been reported elsewhere and do not correlate with serum antibody responses to OspA [12, 14]. OspA expression is down regulated by Borrelia in the mammalian host [17]. For this reason, the maintenance and expansion of T cell responses to OspA in humans is not well understood. Many patients with a T cell response exhibited proliferation to both OspA and OspC. However, there was no correlation in the stimulation index to these 2 antigens.

     The association of OspA T cell response and chronic sequelae has been theorized to result from cross-reaction between a DRB1*0401-restricted OspA T cell epitope and human LFA-1 [18, 19]. This cross-reactivity to LFA-1 has been demonstrated for T cells from patients with Lyme disease who have chronic arthritis. Inflammatory cells within the synovium express LFA-1 [27]. Only a subset of human OspAreactive T cell clones recognize this human LFA-1 peptide, and the response to LFA-1 peptide tends to be suboptimal, which suggests that it is a partial agonist [20]. T cell clones proliferate less well to LFA-1 peptide than OspA, and the cytokine response favors IL-13 rather than IFN-. It has been proposed that human LFA-1 peptide is a partial agonist [20]. It is likely that the human LFA-1 peptide is not processed and presented in vivo. When autologous Epstein-Barr virus (EBV)transformed B cells were used as antigen-presenting cells, it was necessary to add exogenous LFA-1 peptide to induce a T cell response, despite the expression of LFA-1 on EBV lymphoblasts [20].

     Our data demonstrate that T cell recognition of human LFA-1 peptide correlated significantly with T cell response to OspA. T cell response to both LFA-1 peptide and OspA was detected in all groups, including those without neurological disease or arthritis and healthy endemic control subjects. The antigen specificity of the LFA-1 response of endemic control subjects was supported by a response to recombinant OspA in all 18 of these subjects, as well as the low level of LFA-1 response among nonendemic control subjects. The principle limitation of our study was that the subjects were not typed for HLA-DR, precluding conclusions specific for HLA-DR4 patients. The strength of our study was that it was a prospective epidemiological study with matched endemic-area control subjects. None of the previous studies on LFA-1 reactivity have had similar numbers of patients or control subjects. There was no correlation between LFA-1 response and clinical status, arguing that the T cell response to human LFA-1 peptide is not clinically relevant. This is supported by the evidence, discussed above, that the human LFA-1 peptide epitope is not processed by LFA-1positive human antigen-presenting cells and must be added as exogenous peptide [20].

Acknowledgment

     We thank Julia Coleman for her assistance in collecting blood from the New Mexico subjects.

References

1. 

Schaible UE, Wallich R, Kramer MD, et al. Protection against Borrelia burgdorferi infection in SCID mice is conferred by presensitized spleen cells and partially by B but not T cells alone. Int Immunol 1994; 6:67181.  

2. 

Schaible UE, Wallich R, Kramer MD, Museteanu C, Simon MM. A mouse model for Borrelia burgdorferi infection: pathogenesis, immune response and protection. Behring Inst Mitt 1991; 88:5967.

3. 

Sigal LH, Zahradnik JM, Lavin P, et al. A vaccine consisting of recombinant Borrelia burgdorferi outer-surface protein A to prevent Lyme disease. N Engl J Med 1998; 339:21622.

4. 

Steere AC, Sikand VK, Meurice F, et al. Vaccination against Lyme disease with recombinant Borrelia burgdorferi outer-surface lipoprotein A with adjuvant. N Engl J Med 1998; 339:20915.

5. 

Schempp C, Owsianowski M, Lange R, Gollnick H. Comparison of Borrelia burgdorferi ultrasonicate and whole B. burgdorferi cells as a stimulus for T-cell proliferation and GM-CSF secretion in vitro. Zentralbl Bakteriol 1993; 279:41725.  

6. 

Dressler F, Yoshinari NH, Steere AC. The T-cell proliferative assay in the diagnosis of Lyme disease. Ann Intern Med 1991; 115:5339.

7. 

Krause A, Brade V, Schoerner C, Solbach W, Kalden JR, Burmester GR. T cell proliferation induced by Borrelia burgdorferi in patients with Lyme borreliosis: autologous serum required for optimum stimulation. Arthritis Rheum 1991; 34:393402.  

8. 

Dattwyler RJ, Volkman DJ, Luft BJ, Halperin JJ, Thomas J, Golightly MG. Seronegative Lyme disease. Dissociation of specific T- and B-lymphocyte responses to Borrelia burgdorferi. N Engl J Med 1988; 319:14416.

9. 

Yssel H, Nakamoto T, Schneider P, et al. Analysis of T lymphocytes cloned from the synovial fluid and blood of a patient with Lyme arthritis. Int Immunol 1990; 2:10819. 

10. 

Pachner AR, Steere AC, Sigal LH, Johnson CJ. Antigen-specific proliferation of CSF lymphocytes in Lyme disease. Neurology 1985; 35:16424.  

11. 

Martin R, Ortlauf J, Sticht-Groh V, Bogdahn U, Goldmann SF, Mertens HG. Borrelia burgdorferispecific and autoreactive T-cell lines from cerebrospinal fluid in Lyme radiculomyelitis. Ann Neurol 1988; 24:50916.  

12. 

Krause A, Burmester GR, Rensing A, et al. Cellular immune reactivity to recombinant OspA and flagellin from Borrelia burgdorferi in patients with Lyme borreliosis. Complexity of humoral and cellular immune responses. J Clin Invest 1992; 90:107784.

13. 

Shanafelt MC, Hindersson P, Soderberg C, et al. T cell and antibody reactivity with the Borrelia burgdorferi 60-kDa heat shock protein in Lyme arthritis. J Immunol 1991; 146:398592.

14. 

Shanafelt MC, Anzola J, Soderberg C, Yssel H, Turck CW, Peltz G. Epitopes on the outer surface protein A of Borrelia burgdorferi recognized by antibodies and T cells of patients with Lyme disease. J Immunol 1992; 148:21824.  

15. 

Anzola J, Luft BJ, Gorgone G, et al. Borrelia burgdorferi HSP70 homolog: characterization of an immunoreactive stress protein. Infect Immun 1992; 60:370413. 

16. 

Yoshinari NH, Reinhardt BN, Steere AC. T cell responses to polypeptide fractions of Borrelia burgdorferi in patients with Lyme arthritis. Arthritis Rheum 1991; 34:70713.

17. 

Montgomery RR, Malazista SE, Feen KJM, Bockenstedt LK. Direct demonstration of antigenic substitution of Borrelia burgdorferi ex vivo: exploration of the paradox of the early immune response to outer surface proteins A and C in Lyme disease. J Exp Med 1996; 183:2619.

18. 

Gross DM, Forsthuber T, Tary-Lehmann M, et al. Identification of LFA-1 as a candidate autoantigen in treatment-resistant Lyme arthritis. Science 1998; 281:7036.  

19. 

Steere AC, Gross D, Meyer AL, Huber BT. Autoimmune mechanisms in antibiotic treatment-resistant Lyme arthritis. J Autoimmun 2001; 16:2638. 

20. 

Trollmo C, Meyer AL, Steere AC, Hafler DA, Huber BT. Molecular mimicry in Lyme arthritis demonstrate at the single cell level: LFA-1_L is a partial agonist for outer surface protein Areactive T cells. J Immunol 2001; 166:528691.

21. 

Schutzer SE, Coyle PK, Krupp LB, et al. Simultaneous expression of Borrelia OspA and OspC and IgM response in cerebrospinal fluid in early neurologic Lyme disease. J Clin Invest 1997; 100:7637.

22. 

Dunn JJ, Lade BN, Barbour AG. Outer surface protein A (OspA) from the Lyme disease spirochete, Borrelia burgdorferi: high level expression and purification of a soluble recombinant form of OspA. Protein Expr Purif 1990; 1:15968.  

23. 

Simes RJ. An improved Bonferroni procedure for multiple tests of significance. Biometrika 1986; 73:7514.

24. 

Seinost G, Dykhuizen DE, Dattwyler RJ, et al. Four clones of Borrelia burgdorferi sensu stricto cause invasive infection in humans. Infect Immun 1999; 67:351824.  

25. 

Wang IN, Dykhuizen DE, Qui W, Dunn JJ, Bosler EM, Luft BJ. Genetic diversity of ospC in a local population of Borrelia burgdorferi sensu stricto. Genetics 1999; 151:1530.  

26. 

Gomes-Solecki MJC, Dunn JJ, Luft BJ, et al. Recombinant chimeric Borrelia proteins for diagnosis of Lyme disease. J Clin Microbiol 2000; 38:253035.  

27. 

Akin E, Aversa J, Steere AC. Expression of adhesion molecules in synovia of patients with treatment-resistant lyme arthritis. Infect Immun 2001; 69:177480.