Mycobacterial Hsp65-IgG–expressing tolerogenic B cells confer protection against adjuvant-induced arthritis in Lewis rats




Tolerization of T cells directed against a target autoantigen is a desired goal of experimental approaches for the treatment of autoimmune diseases, and novel and improved methods of tolerance induction are continuously being sought. Because most traditional methods of tolerance induction using soluble antigen are effective in the prevention of autoimmunity but fail to control established disease, this study was carried out to explore an innovative tolerogenic approach for the treatment of ongoing disease, using the rat adjuvant-induced arthritis (AIA) model of human rheumatoid arthritis.


Lewis (RT.1l) rats were injected subcutaneously with heat-killed Mycobacterium tuberculosis H37Ra to induce AIA. Before or after AIA induction, Lewis rats were treated intraperitoneally (IP) with tolerogenic B cells expressing a fusion construct of mycobacterial 65-kd heat-shock protein (Hsp65) and IgG heavy-chain. For comparison, control rats were treated IP with ovalbumin (OVA)–IgG–expressing B cells or soluble mycobacterial Hsp65, and the effects on AIA were observed. We also tested the immune response to mycobacterial Hsp65 in B cell–tolerized rats.


Administration of tolerogenic mycobacterial Hsp65–expressing B cells as well as soluble mycobacterial Hsp65, but not OVA-expressing B cells, resulted in a significant decrease in the severity of subsequent AIA. However, in rats with established disease, only the B cell regimen of mycobacterial Hsp65, but not the soluble antigen, suppressed ongoing AIA.


Mycobacterial Hsp65-IgG–expressing B cells can successfully attenuate the progression of AIA. This study introduces a promising approach for the treatment of arthritis that should be further explored.

Adjuvant-induced arthritis (AIA), an experimental model of human rheumatoid arthritis (RA), can be induced in the Lewis (RT.1l) rat by subcutaneous (SC) challenge with heat-killed Mycobacterium tuberculosis (Mtb) H37Ra (1). The 65-kd mycobacterial heat-shock protein (Hsp65) is one of the targets of the T cell response in arthritic Lewis rats (2, 3). Interestingly, patients with RA also show T cell reactivity to mycobacterial Hsp65 (4). Induction of antigen-specific T cell tolerance has been used extensively for the immunomodulation of autoimmune diseases in various animal models (5). However, most of these tolerogenic approaches are successful in the prevention of autoimmunity but generally fail to control ongoing, established disease. Gene therapy has emerged as a potent method for the modulation of autoimmune diseases (6–8).

In this study we tested an alternative and effective tolerogenic approach for modulation of AIA using B cells that express a defined antigen (9–11). We introduced a gene fusion construct of mycobacterial Hsp65 and IgG heavy-chain retrovirally ex vivo into B cells of Lewis rats, and then injected these cells intraperitoneally (IP) into syngeneic Lewis rat recipients, followed by induction of AIA by Mtb H37Ra injection. In parallel, we induced tolerance by injecting soluble mycobacterial Hsp65 IP in Lewis rats. We then tested the relative efficacy of these 2 approaches for both the prevention and the treatment of AIA. We also examined the immune response to mycobacterial Hsp65 in Lewis rats that were tolerized with mycobacterial Hsp65-IgG–expressing B cells.



Inbred male Lewis (RT.1l) rats (LEW/SsNHsd, 4–6 weeks old and weighing 120–200 grams) were obtained from Harlan Sprague-Dawley (Indianapolis, IN) and housed in the vivarium of the University of Maryland School of Medicine (UMB). All experimental procedures performed on these animals were in compliance with the guidelines of the UMB institutional animal care and use committee.


Recombinant mycobacterial Hsp65 was expressed by growing Escherichia coli cells (BL21pLysS) (Novagen, Madison, WI) transformed with pET23b-GroEL2 (NIH Tuberculosis Vaccine Testing and Research Materials Contract; Mycobacteria Research Laboratories, Colorado State University, Fort Collins, CO) followed by elution on a nickel column and purification by dialysis as described elsewhere (1). The purified mycobacterial Hsp65 protein was passed through an endotoxin-removal column (Sterogene, Carlsbad, CA) until the endotoxin level was below 0.25 EU/ml. Ovalbumin (OVA), hen eggwhite lysozyme, and concanavalin A (Con A) were obtained from Sigma-Aldrich (St. Louis, MO).

Generation of replication-defective retroviral constructs.

Mycobacterial Hsp65 complementary DNA (cDNA) inserted in pET23b plasmid (obtained from Mycobacteria Research Laboratories) was amplified by polymerase chain reaction (PCR). The PCR product was ligated into the BSSK-IgG plasmid, a murine IgG1 heavy-chain cassette plasmid containing Not I and Xho I cloning sites at the N-terminus of the IgG1 heavy-chain (9). This mycobacterial Hsp65-IgG–containing plasmid was Sal I– and Pvu I–digested and subcloned into Sal I–digested Moloney murine leukemia virus with β-actin promoter and enhancer (MBAE) retroviral vector. Mutations were ruled out by gene sequencing conducted at the UMB Biopolymer Core Facility, and the desired orientation of the insert was validated using specific enzyme digests. A similar construct was prepared with OVA to obtain OVA-IgG–containing MBAE vector. The mycobacterial Hsp65-IgG and OVA-IgG constructs thus obtained were transfected into a packaging cell line, GP+E86, which was then selected under neomycin (G418, 0.6 mg/ml in active form). High-titer clones (∼106–107 neomycin-resistant NIH3T3 colony-forming units/ml) of virus-producing packaging cell line were stored in liquid nitrogen and then freshly thawed for each experiment as needed.

Retroviral transfer of the antigen-IgG fusion gene into lipopolysaccharide (LPS)–treated B cells.

Splenocytes of naive Lewis rats were initially plated to remove adherent cells (11), then stimulated overnight with 5 μg/ml of bacterial LPS (E coli O55:B5; Sigma-Aldrich), and finally recultured (4 × 106/ml) for an additional 24 hours with irradiated viral packaging cells transfected with mycobacterial Hsp65-IgG– or OVA-IgG–containing MBAE vector in the presence of 6 μg/ml polybrene and LPS. The live B cell blasts were isolated using Lympholyte-rat (Cedarlane, Hornby, Ontario, Canada) gradient centrifugation (11), and then injected IP into Lewis rats as needed.

Induction and evaluation of AIA.

Lewis rats were immunized SC at the base of the tail with 200 μl of Mtb H37Ra (10 mg/ml; Difco, Detroit, MI) suspended in mineral oil (Sigma), and AIA was scored as described earlier (1). Briefly, beginning on day 7 after immunization, the rats were scored daily for clinical signs of arthritis. The severity of arthritis in each paw was evaluated according to the extent of erythema and swelling, and graded on a scale of 0 to 4 as follows: 0 = no erythema or swelling, 1 = slight erythema or swelling of the ankle or wrist, 2 = moderate erythema and swelling of the ankle or wrist, 3 = moderate erythema and swelling of the wrist and metacarpal joints or ankle and metatarsal joints, and 4 = severe erythema and swelling of the entire fore paw or hind paw. The highest score for each paw was 4, and the total arthritis score per rat was 16.

Treatment of Lewis rats with transduced B cells or soluble antigen for the modulation of AIA.

Antigen-Ig–expressing B cells.

The retrovirus-infected B cells were washed and transferred IP into syngeneic Lewis rats as a single injection (15–30 × 106 cells/rat). Retroviral treatment was administered either before or after the induction of AIA by injection of Mtb H37Ra.

Soluble antigen.

Lewis rats were injected IP on alternate days with soluble mycobacterial Hsp65 or OVA (200 μg/injection in each group, for a total of 3 injections). Soluble treatment was also administered either before or after the induction of AIA by injection of Mtb H37Ra.

Total RNA extraction from transduced cells and synthesis of cDNA.

Retrovirally transduced cells were isolated from the coculture by Lympholyte-rat gradient centrifugation, as described earlier. Two million cells were lysed, and then total RNA was extracted using TRIzol reagent (Invitrogen-Gibco, Grand Island, NY), precipitated by isopropanol, and washed with 75% ethanol. The RNA pellet was dried and resuspended in diethylpyrocarbonate-treated water. RNA was quantitated, and 1 μg total RNA was used to synthesize cDNA using the iScript cDNA synthesis kit (Bio-Rad, Hercules, CA).

Real-time PCR for quantification of mycobacterial Hsp65 messenger RNA (mRNA) in transduced cells.

Quantitative real-time PCR primers were designed using the Accelrys Gene Program (version 2.0; Accelrys Software, San Diego, CA) and then synthesized in the UMB Biopolymer Core Facility. Primers were as follows: for mycobacterial Hsp65, sense 5′-GAGGGCGTCATCACCGTC-3′ and antisense 5′-ACGAAGTACCCCGAGATGTAGC-3′, and for the HPRT housekeeping gene, sense 5′-ACCAGTCAACAGGGGACATAAAAG-3′ and antisense 5′-GTCTGCATTGTTTTGCCAGTGTC-3′. The cDNA was amplified from naive splenocytes, mycobacterial Hsp65-IgG–specific cells, and OVA-IgG–specific cells using an Applied Biosystems Prism 7900HT Sequence Detection System (Foster City, CA). The reaction mixture contained 20 ng cDNA, 0.3 μM of each primer pair, and 12.5 μl of 2× SYBR Green Master Mix (Applied Biosystems), under the following thermal conditions: 50°C for 2 minutes, 95°C for 10 minutes, and 40 cycles of 15 seconds at 95°C and 1 minute at 60°C, followed by a dissociation step. The mRNA levels of the genes of interest were normalized to the levels of the HPRT gene, and the relative gene expression levels were determined using the formula 2−ΔΔCt, Ct being the cycle threshold.

Lymph node cell (LNC) proliferation assay.

The draining LNCs (inguinal, paraaortic, and popliteal) of Lewis rats were harvested 9 days after immunization SC with mycobacterial Hsp65 emulsified in Freund's incomplete adjuvant. The LNCs were cultured for 96 hours at 1 × 105 cells/well in HL-1 serum-free medium (BioWhittaker, Walkersville, MD) in the presence or absence of the test (mycobacterial Hsp65) or control (OVA) antigen used at an optimal pretitrated concentration. Con A served as the positive control. The cells were pulsed with 1 μCi/well of 3H-thymidine for an additional 16 hours. The results were expressed as the stimulation index, calculated as the ratio of stimulation, in counts per minute, of cells in the presence of antigen to the cpm of cells in medium alone.

Determination of serum antibody levels by enzyme-linked immunosorbent assay (ELISA).

Blood samples were obtained at various time points before and after Mtb H37Ra immunization from 3 Lewis rats in each test group (those treated with mycobacterial Hsp65-IgG–expressing B cells and control rats treated with OVA-IgG–expressing B cells). Sera were tested in ELISA as described in detail elsewhere (12). Briefly, an ELISA plate (Greiner Bio-One, Longwood, FL) was first coated with 100 ng/well of mycobacterial Hsp65 overnight at 4°C, and the wells were then saturated with 10% bovine serum albumin (Sigma-Aldrich). Sera were added to the wells at 1:100 dilution and then incubated for 1 hour at room temperature. Following thorough washings of the wells, the plate-bound total Ig was detected by horseradish peroxidase–conjugated goat anti-rat Ig (BD PharMingen, San Diego, CA). The absorbance was read at 450 nm using a Vmax ELISA autoreader (Molecular Devices, Sunnyvale, CA).

Statistical analysis.

Student's 2-tailed t-test was performed with the assumption of equal or unequal variance of the data, as appropriate. Initially, the F test was carried out to determine whether the variance was equal or unequal, and then an appropriate Student's t-test was used. P values less than 0.05 were considered significant.


Expression of mycobacterial Hsp65 mRNA in retrovirally transduced cells.

The mycobacterial Hsp65-IgG heavy-chain fusion gene construct and the control construct containing OVA instead of mycobacterial Hsp65 were inserted into the retroviral vector MBAE (Figure 1A). Rat splenocytes blasted with LPS and transduced with the mycobacterial Hsp65-IgG–containing MBAE virus were then tested for the expression of mycobacterial Hsp65 transcripts by real-time PCR. The results showed the specificity of antigen expression; the relative expression of mycobacterial Hsp65 mRNA from the mycobacterial Hsp65-IgG construct was ∼2,000 times higher than that from the OVA-IgG construct (Figure 1B).

Figure 1.

Retroviral vector constructs and the expression of antigen in retrovirally transduced splenocytes. A, A schematic diagram of the Moloney murine leukemia virus with β-actin promoter and enhancer (MBAE) vector DNA carrying mycobacterial Hsp65 (Bhsp65)–IgG and ovalbumin (OVA)–IgG constructs. LTR = long terminal repeats; NeoR = gene for neomycin resistance; P&E = promoter and enhancer. B, Mycobacterial Hsp65 mRNA expression, as measured in splenocytes from naive Lewis rats and in those transduced with mycobacterial Hsp65-IgG– or OVA-IgG–specific MBAE virus. The results are shown as the mean and SEM fold increase in mRNA compared with the OVA-IgG MBAE–treated and naive spleen cells. ∗ = P < 0.05 versus control groups.

Prevention of subsequent arthritis by mycobacterial Hsp65-IgG–expressing B cells or soluble mycobacterial Hsp65.

LPS-treated splenocytes from naive Lewis rats were infected with mycobacterial Hsp65-IgG– or OVA-IgG–containing MBAE virus as described in Materials and Methods. These cells were then injected IP into syngeneic Lewis rats 7 days prior to immunization with Mtb H37Ra. Rats treated with mycobacterial Hsp65-IgG–expressing cells had significantly reduced arthritis scores (P < 0.05) compared with rats that received either OVA-IgG–expressing cells (Figure 2A) or phosphate buffered saline (results not shown). Similarly, rats pretreated with soluble mycobacterial Hsp65 (200 μg/injection) exhibited significantly reduced severity of arthritis (P < 0.05) compared with OVA-pretreated control rats (Figure 2B). Thus, both of the tested regimens led to decreased severity of subsequent AIA, thereby representing a strategy of disease prevention.

Figure 2.

Protection against subsequently induced adjuvant-induced arthritis (AIA) in Lewis rats by pretreatment with either mycobacterial Hsp65-IgG–expressing B cells or soluble mycobacterial Hsp65. A, Naive Lewis rats (n = 7 per group) received either mycobacterial Hsp65-IgG–expressing cells (●) or ovalbumin (OVA)–IgG–expressing cells (▪) intraperitoneally (IP) (indicated by ↑) 7 days prior to immunization with heat-killed Mycobacterium tuberculosis (Mtb) H37Ra. Thereafter, all rats were observed regularly and scored for signs of AIA. B, Lewis rats (n = 3 per group) were injected IP at 3 time points (indicated by ↑) with either mycobacterial Hsp65 (○) or OVA (□) in phosphate buffered saline. On day 9 after the first injection, all rats were immunized with Mtb H37Ra and then observed and graded regularly for signs of AIA. ∗ = P < 0.05 versus control groups.

Amelioration of ongoing AIA by mycobacterial Hsp65-IgG–expressing B cells, but not soluble mycobacterial Hsp65.

We tested the effectiveness of the above-mentioned AIA-preventive regimens in controlling the ongoing, established disease in Lewis rats. Injection of mycobacterial Hsp65-IgG–expressing cells into Lewis rats after the appearance of signs of AIA resulted in a significantly reduced severity of the disease (P < 0.05) compared with that in rats that received OVA-IgG–expressing cells (Figure 3A). In contrast, administration of soluble mycobacterial Hsp65 IP after the appearance of signs of AIA failed to down-modulate the severity of AIA (Figure 3B).

Figure 3.

Suppression of ongoing adjuvant-induced arthritis (AIA) after treatment of arthritic Lewis rats with mycobacterial Hsp65-IgG–expressing cells but not with soluble mycobacterial Hsp65. A, Mycobacterial Hsp65-IgG–expressing cells (●) and ovalbumin (OVA)–IgG–expressing cells (▪) were injected intraperitoneally (IP) (indicated by ↑) into Lewis rats (n = 6 per group) at the onset of AIA. All rats were observed regularly and scored for signs of AIA. ∗ = P < 0.05 versus control group. B, Mycobacterial Hsp65 (○) or OVA (□) in phosphate buffered saline was injected IP at 3 time points (indicated by ↑) after immunization with heat-killed Mycobacterium tuberculosis H37Ra. P > 0.05 between groups.

Decreased mycobacterial Hsp65–specific T cell responses in parallel with increased anti–mycobacterial Hsp65 antibody responses after pretreatment of Lewis rats with mycobacterial Hsp65-IgG–expressing B cells.

Lewis rats pretreated with mycobacterial Hsp65-IgG–expressing cells prior to injection of Mtb H37Ra developed antigen-specific T cell tolerance, as evident from the decreased T cell response to mycobacterial Hsp65 exhibited by these rats when compared with the response of syngeneic control rats pretreated with OVA-IgG–expressing cells (Figure 4A). In contrast to the T cell response, this pretreatment strategy using mycobacterial Hsp65-IgG–expressing B cells facilitated the generation of increased anti–mycobacterial Hsp65 antibody responses, which occurred much earlier during the course of AIA in these rats as compared with control rats (Figure 4B). We (12) and others (13) have previously shown that antibodies produced during the course of AIA are protective against AIA rather than being arthritogenic. In this context, a decreased T cell response coupled with an increased antibody response to mycobacterial Hsp65 might contribute to the protection against AIA in Lewis rats that are tolerized with mycobacterial Hsp65-IgG–expressing B cells.

Figure 4.

Effects of pretreatment of Lewis rats with mycobacterial Hsp65 (Bhsp65)–IgG–expressing cells on the T cell response and antibody response. Mycobacterial Hsp65-IgG–expressing or ovalbumin (OVA)–IgG–expressing cells were injected intraperitoneally into naive Lewis rats, followed 7 days later by immunization subcutaneously with Mycobacterium tuberculosis (Mtb) H37Ra in Freund's incomplete adjuvant. A, The draining lymph node cells were harvested 9 days after the last injection and tested in a proliferation assay. The medium background ranged 1,000–2,000 counts per minute. B, Sera collected from the test and the control Lewis rats at various time points after heat-killed Mtb H37Ra challenge were tested in enzyme-linked immunosorbent assay for total anti–mycobacterial Hsp65 antibodies, with results expressed as the optical density (OD) at 450 nm. The pre-Mtb time point was day 0, just before injection of Mtb H37Ra. ∗ = P < 0.05 versus controls.


Systemic or mucosal administration of an antigen can readily inactivate potentially pathogenic antigen-specific T cells and prevent the development of autoimmunity in animal models (5, 14). However, some of these tolerogenic approaches may cause untoward effects such as anaphylaxis (15) or exacerbation of disease (16, 17). Moreover, the effectiveness of these tolerogenic regimens in modulating ongoing disease in the clinical setting has not yet been fully established (14).

In this study, we observed that the injection of soluble mycobacterial Hsp65 in Lewis rats significantly reduced the severity of the subsequently induced AIA but failed to ameliorate established AIA. In contrast, the administration of mycobacterial Hsp65-IgG–expressing tolerogenic B cells not only was effective in reducing the severity of subsequent AIA, but also ameliorated established AIA in Lewis rats. Our findings corroborate the results from earlier reports of the success of antigen/epitope-IgG–expressing B cell–based therapy for the prevention and treatment of autoimmunity in animal models of experimental autoimmune encephalomyelitis, experimental autoimmune uveitis (EAU), and type I diabetes (9–11). In the EAU model, the B cell–mediated antigen-specific tolerance has been reported to last for more than 6 months (9).

Attempts have been made in the past to build protective immunity against Hsp65 by administration of human Hsp60–containing vaccinia virus (18) or injection of naked DNA encoding mycobacterial Hsp65 (19) or human Hsp60 (20). These attempts were successful in preventing subsequent induction of AIA. However, the effectiveness of such approaches in established disease was not demonstrated. In this regard, our results show that the B cell–mediated treatment was effective in suppressing ongoing AIA.

At present, the precise mechanisms involved in the modulation of AIA by mycobacterial Hsp65-IgG–expressing B cells are not fully understood. However, the results of the above-mentioned studies in other models of autoimmunity have shown that the presence of B cells is essential, and that presentation of class II major histocompatibility complex, engagement of CTLA-4, and B7 signaling are involved in B cell–mediated tolerization of the disease-related T cells (21). It has also been reported that the number of CD4+,CD25+ regulatory T cells is increased in NOD mice after application of this regimen (22). However, no evidence of a deviation in the cytokine response has been observed (10).

Mauri and coworkers demonstrated that B cells producing interleukin-10 (IL-10) can ameliorate collagen-induced arthritis (23). These B cells were nonspecifically activated in vivo with anti-CD40. Although production of IL-10 by LPS-activated B cells, as used in the present study and in our previous investigations (Scott DW, et al: unpublished observations), may be involved in the mechanisms of suppression, it should be pointed out that our approach involves induction of antigen-specific tolerance by the administration of genetically modified B cells, and therefore it is not a generalized immunosuppressive regimen. With regard to the T cell tolerance induced by soluble antigen administration, other investigators have shown that it involves the induction of anergy that is attributed to a lack of adjuvant-mediated costimulation, immune deviation (Th1 to Th2 type), and CD4+,CD25+ regulatory T cells (14).

We observed that the pretreatment of Lewis rats with mycobacterial Hsp65-IgG–expressing B cells induced an anti–mycobacterial Hsp65 antibody response but a decrease in the mycobacterial Hsp65–directed T cell proliferative response during the early phase of AIA. Furthermore, this modulation of the humoral and cellular immune responses to mycobacterial Hsp65 corresponded to the protection against AIA observed with this pretreatment. We (12) and others (13) have previously shown that antibodies produced during the course of AIA are protective rather than being pathogenic. Therefore, our results suggest that tolerogenic mycobacterial Hsp65-IgG–expressing B cell treatment offers protection from AIA, in part through suppression of the T cell response along with an enhancement of the protective antibody response to mycobacterial Hsp65.

At this time, we do not know the precise mechanistic differences that might account for the differing effects on established AIA of B cell–mediated compared with soluble antigen–induced tolerogenic approaches. However, a major difference in immune activation exists in this context, in that there is a continuing in vivo antigen presentation by mycobacterial Hsp65–expressing B cells, as compared with a single antigen exposure in the case of soluble antigen administration. The former regimen might facilitate the induction of disease-regulating T cells at a regular pace, which in turn might successfully down-regulate the activity of continuously emerging arthritogenic T cells following Mtb H37Ra challenge. In contrast, the latter regimen, involving the inactivation and/or deletion of pathogenic T cells by soluble antigen–induced tolerance, might be ineffective in suppressing the activity of the emerging pathogenic T cells. Mtb is a lipid-rich particulate antigen that tends to provide an antigen depot for sustained release of arthritogenic stimulus, and therefore a disease-modulating regimen that can provide a similarly sustained protective stimulus (e.g., mycobacterial Hsp65-Ig–expressing B cells) might have a better chance to succeed, compared with a one-time protective challenge (e.g., soluble antigen injection), in suppressing ongoing AIA.

Detailed mechanistic studies comparing the antiarthritis activities primed by these 2 tolerogenic regimens are currently in progress in our laboratory. Since T cell responses to mycobacterial Hsp65 have also been reported in patients with RA as well as in patients with juvenile RA (4), the use of mycobacterial Hsp65-IgG–expressing tolerogenic B cells might be an effective approach for the treatment of these debilitating autoimmune disorders.


Dr. Moudgil had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Satpute, Scott, Moudgil.

Acquisition of data. Satpute, Soukhareva.

Analysis and interpretation of data. Satpute, Scott, Moudgil.

Manuscript preparation. Satpute, Scott, Moudgil.

Statistical analysis. Satpute.

Principal investigator. Moudgil.


We thank Drs. Hong Ro Kim, Agnes Azimzadeh, Jun Hayashi, Swamy Polumuri, and Ricardo Feldman for their helpful discussions.