Selective silencing of DNA-specific B lymphocytes delays lupus activity in MRL/lpr mice

Chemical coupling of peptides to the monoclonal antibodies

The comparison between the HPLC curves of the 2.4G2 antibody and of the antibody chimeras prepared using it showed that the latter was displaced as a result of the coupling of a certain number of peptides (Fig. 1A). Immunoglobulin aggregates were not formed during the conjugation procedure (data not shown). The final solutions of the chimeric antibody molecules contained 0.5 mg protein and only 0.24 EU/mL endotoxin.

There are many free carboxyl groups on the surface of the antibody molecules, and some of them are available for interaction with the reactive H₂N group of the peptide spacer. The peptides could theoretically bind to groups within the antigen-binding sites, and as a result the interaction of the parent antibody to mouse FcγRIIb could be affected. This possibility was excluded as we showed that the native 2.4G2 antibody and the DNA-like chimera recognized equally well the targeted receptor on mouse splenocytes (not shown). In a separate experiment, the DNA-like chimera inhibited, in a dose-dependent manner, the binding of a commercial FITC-labeled 2.4G2 antibody to the same cells (Fig. 1B). In an independent study, the pure heavy and light chains from the native 2.4G2 antibody and from the DNA-like antibody chimera were separated by HPLC. Comparative mass spectral analyses of both showed that 14–16 peptides were bound per one DNA-like chimeric IgG molecule (N. Mihaylova et al., manuscript in preparation).

The recognition of the DNA-like peptide by the dsDNA-specific monoclonal antibody 10F10 in a Western blot (Fig. 1C) confirmed that the peptides coupled to the rat IgG molecules retained their DNA-mimicking epitope. The latter is available for interaction with dsDNA-specific antigen receptors on lupus B lymphocytes.

As the abilities of the antibody and of the peptide part of the constructed DNA-like chimera to bind to both the FcγIIb receptors and the surface dsDNA-specific immunoglobulin receptors are preserved, it is expected that this artificial molecule would cross-link these receptors on the surface of the targeted disease-associated B cells.

The DNA-like chimera targets dsDNA-specific B lymphocytes only

MRL/lpr mice with lupus, immunized twice with diphtheria toxoid, have large numbers of both dsDNA- and toxoid-specific B and plasma cells. Their splenocytes were cultured for 5 days, and the numbers of anti-diphtheria toxoid and anti-dsDNA IgG antibody-producing cells were counted by the ELISpot technique (Fig. 2A). The addition of bacterial LPS did not increase further the numbers of antibody producers, showing that most B cells with the studied specificities were already activated. None of the constructed chimeras added to the medium during the period of cultivation influenced the numbers of diphtheria toxoid-specific IgG antibody-producing cells (Fig. 2A, lower panel). The DNA-like chimera caused a dose-dependent decrease of the numbers of in vitro differentiated anti-dsDNA antibody-producing cells (Fig. 2A, upper panel).

The DNA-like chimera was shown to bind only to B lineage cells at different stages of differentiation, but not to T cells (Fig. 2B). Culturing of splenocytes from a lupus mouse in the presence of the same chimera increased the percentage of apoptotic (Annexin V-positive) B lymphocytes (Fig. 3), but not T lymphocytes (not shown).

The DNA-like chimera decreases anti-dsDNA IgG antibody levels and disease activity

The in vivo effects of the DNA-like antibody chimera were studied in two groups of female lupus-prone MRL/lpr mice: in 7-wk- and in 18-wk-old ones. The younger animals were still disease free at that age but the appearance of the first signs of lupus were imminent. The older group had already high levels of IgG anti-dsDNA antibodies and of albuminuria at the start of the experiment. The period of treatment was limited to 6 wk, because of the appearance of anti-rat immunoglobulin antibodies after that point (data not shown).

The administration of the chimeric antibody to 7-wk-old mice resulted in prevention of the sharp increase of the IgG anti-DNA antibody levels during the next 6 wk (p <0.05, Student's t-test, compared to the PBS-treated group). Differences in proteinuria levels were also present (Fig. 4A). The DNA-like chimera-treated mice had low protein levels in the urine for the first 10 wk, which increased gradually in the next 10 wk. A rapid increase of the urine protein concentration was observed in the PBS-injected controls during the same period.

Treatment that started at the age of 18 wk resulted in the maintenance of a flat level of IgG anti-DNA antibodies in the next 6 wk. This result correlated with a flat level of proteinuria during the same period (Fig. 4B).

IgG antibodies against the SSA, SSB and Sm antigens were detected in the sera of all mice treated with the DNA-like chimera, with the control chimeras or with PBS. No significant differences were seen between the groups (not shown).

By 5 months of age, all mice in the control groups had severe lymphadenopathy, while the DNA-like chimera-treated animals had smaller peripheral lymph nodes and spleens (Fig. 5A, B). We are presently investigating the mechanisms of this effect.

MRL/lpr mice develop a spontaneous progressive skin disease (Fig. 5C, upper panel), and by 5 months of age all have large plaque-like cutaneous lesions on their posterior neck. No lesions were detected in any of the DNA-like chimera-injected animals (Fig. 5C, lower panel).

The same treatment affected the deposition of IgG-containing immune complexes in the glomeruli. While massive mesangial deposits were observed in the animals treated with the control chimeras, the depositions detected in the chimera-treated mice were of the segmentary mesangial type (Fig. 6, upper panel). The 6-wk-long administration of the DNA-like chimera resulted also in a significant improvement in the kidney histology. The microscopical structure of the organs of chimera-treated mice was preserved although an interstitial infiltrate was present. In contrast, the kidney sections of irrelevant peptide chimera-treated animals were characterized by a mononuclear infiltrate and a massive mesangial proliferation, and those of the irrelevant antibody chimera-treated ones by a massive mononuclear infiltrate surrounding the small blood vessels. Finally, PBS-treated MRL/lpr controls presented mononuclear interstitial infiltrate and glomerulopathy with a global mesangial proliferation and epithelial crescents (Fig. 6, lower panel).

The administration of the DNA-chimera to 7-wk-old MRL/lpr mice did not prolong their lives (data not shown). However, the same treatment of 18-wk-old sick animals significantly prolonged their survival (Fig. 7; p <0.02, Log Rank test), proving that an approach mimicking the natural mechanism for avoiding unwanted immune responses could be used for the treatment of autoimmune disorders.

Antibodies

The rat 2.4G2 hybridoma producing a monoclonal IgG2b antibody specific to mouse FcγRII (CD32) (ATCC HB-197) was adapted to grow in the serum-free CHO medium (Gibco, Gaithersburg, MD). The antibodies from the supernatant were isolated by 50% ammonium sulfate precipitation and subsequent dialysis. Their purity was determined by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) under non-reducing conditions using a silver staining kit (PlusOne, Pharmacia Biotech AB, Uppsala, Sweden) and by blotting using an anti-rat κ chain alkaline phosphatase conjugate (Pharmingen BD, San Diego). Care was taken to diminish all possible sources of endotoxin during the purification and subsequent conjugation steps.

The rat hybridoma I/9 producing a monoclonal anti-idiotype [anti-B10-anti-(T,G-A-L)] IgG2b antibody 27 (kindly provided by Dr. Gloria Laszlo, Immunology Research Group, Hungarian Academy of Sciences, Budapest, Hungary) was cultured and the antibody purified as described above. The following antibodies from Pharmingen BD (San Diego, CA) were used in the flow cytometry experiments: 2.4G2-FITC, anti-mouse CD3-FITC, PE conjugates of anti-mouse CD24, CD27 and CD38, as well as a PE-Cy5 conjugate of anti-mouse CD19.

Mice

Groups of female MRL/lpr mice (7-wk–old ones with initial disease and 18-wk-old ones with severe disease) were obtained from the Taiwan National University Animal Service and from Harlan Farm, Blackthorn, UK). The mice were kept under specific pathogen-free conditions, and the manipulations were approved by the Animal Care Commission at the Institute of Microbiology in accordance with International Regulations.

The animals (five and ten per group in the first and second experiment, respectively) were injected i.v. twice weekly for 6 wk with 20 µg of the DNA-like chimera, with the same amount of the control chimeras or with PBS alone. Every 14 days, they were bled and the sera kept frozen at –20°C. Albuminuria was measured using Multistix strips (Beyer Diagnostics Manufacturing Ltd., Bridgend, UK) and graded semi-quantitatively (0, none; 1, 30–100 mg/dL; 2, 100–300 mg/dL; 3, 300–500 mg/dL; 4, >500 mg/dL).

Construction of chimeric antibody molecules

Two synthetic peptides – the DNA mimotope peptide (DWEYSVWLSN) and an irrelevant peptide containing the same amino acids but in a shuffled order (WSLDYWNEVS) – were used in the study. The synthesis of Ac-DWEYSVWLSN-NH-(CH₂)₆-NH₂ and Ac-WSLDYWNEVS-NH-(CH₂)₆-NH₂ was carried out using Fmoc-based manual solid-phase peptide synthesis protocols on 2-Cl-Trt resin. The peptides were purified (⩾95% purity) by reversed-phase sample displacement chromatography 28 on a series of 3 × 50 mm Nucleodur 100-5 C₁₈ columns (Macherey-Nagel, Germany). The DNA-mimicking DWEYSVWLSN peptide was coupled to the 2.4G2 antibody (DNA-like chimera) and to the irrelevant antibody I/9 (irrelevant chimera 2). The irrelevant peptide WSLDYWNEVS was coupled to the 2.4G2 antibody (irrelevant chimera 1).

The coupling of the antibodies to the peptides was carried out using the classical 1-ethyl-3-(3′-dimethylaminopropyl) carbodiimide.HCl (EDC; Fluka AG, Buchs, Switzerland) cross-linking technique 29, 30. During the synthesis of the peptides, a spacer [H₂N(CH₂)₆H₂N] was added to their C terminus with the aim to allow the peptide to form its own structure. To avoid cross-linking of the immunoglobulin molecules during the conjugation step, their concentration was kept low. The antibody (in concentration 0.1 mg/mL in sterile 0.1 M sodium phosphate buffer, pH 6.0) was mixed with a 20-fold molar excess of the peptide [dissolved in 10% v/v N,N-dimethylformamide (Sigma-Aldrich, Taufkirchen, Germany) in the same buffer to 0.02 mg/mL final concentration]. The reaction was started by the addition of carbodiimide at 60-fold molar excess over the antibody. The reaction mixture was stirred overnight at 4°C, dialyzed against PBS and concentrated by ultrafiltration. HPLC analyses using a model 1046A fluorescence detector (Hewlett-Packard), operating at excitation and emission wavelengths of 280 and 340 nm, respectively, were performed. The level of endotoxin in the final chimera solutions was determined by a Limulus amebocyte lysate assay (E-Toxate, Sigma).

Flow cytometry

The binding of the DNA-like chimera to mouse spleen cells was compared to that of the unconjugated 2.4G2 antibody itself. Cells were washed with PBS (containing 2.5% FCS and 0.05% sodium azide) and incubated with the 2.4G2 antibody or with the DNA-like chimera (1 µg/10⁶ cells) for 30 min at 4°C, followed by two washes. Next, the cells were incubated with goat anti-rat κ light chain biotin-conjugated antibody (Pharmingen BD), followed by incubation with FITC-conjugated avidin (Calbiochem, Darmstadt, Germany). Flow cytometric analysis was carried out on a FACSCalibur flow cytometer; 10 000 cells were collected for each sample.

The retained ability of the DNA-like chimera to interact with FcγRII was confirmed as it competed successfully with a FITC-conjugated antibody with the same specificity for binding to FcγRII on mouse spleen cells. The splenocytes were incubated with increasing concentrations (1.0, 0.5, 0.25 and 0.1 µg/10⁶ cells) of the DNA-like chimera (as described above), washed, and incubated with the same amount of a FITC-conjugated 2.4G2 antibody (Pharmingen BD).

The binding of the DNA-like chimera to mouse T cells or subsets of B cells was studied. Spleen cells were washed with PBS (containing 2.5% FCS and 0.05% sodium azide) and incubated with the DNA-like chimera (at 1 µg/10⁶ cells) for 30 min at 4°C, followed by two washes. Next, the cells were incubated with a goat anti-rat IgG FITC-conjugated antibody, a PE-Cy5 conjugate of an anti-mouse CD19 antibody, and with one of the following anti-mouse antibody-PE conjugates: anti-CD3, anti-CD24, anti-CD27 or anti-CD38. The analysis was performed on a FACSCalibur flow cytometer; 10 000 cells were collected for each sample.

Detection of apoptosis

Spleen cells from a sick 12-wk-old female MRL/lpr mouse were cultured for 4 days at 2 × 10⁶ cells/mL in complete RPMI 1640 medium or in the same medium containing 200 ng/mL of the DNA-like chimera. The apoptosis of gated B (CD19⁺) and T (CD3⁺) cells was measured using the Annexin V-FITC apoptosis detection Kit I (Pharmingen BD).

PAGE and Western blotting

All constructed chimeras were subjected to SDS-PAGE and Western blot analysis. SDS-PAGE was performed using 10% gels and a MiniProtean II system (Bio-Rad, Richmond, CA) in the presence of 0.1% SDS. After the electrophoresis, the proteins were transferred to a nitrocellulose membrane (0.45 µm; Sartorius, Germany) using a MiniTrans Blot device (Bio-Rad) in a buffer containing 48 mM Tris and 110 mM glycine in the presence of 20% v/v methanol. The membranes were blocked overnight in a TBS buffer containing 0.4% Tween-20, and were incubated for 1 h with the dsDNA-specific mouse IgG2a monoclonal antibody 10F10 (kindly provided by Dr. Bor-Luen Chiang, College of Medicine, National Taiwan University, Taipei, Taiwan). After washing, the membranes were incubated in an optimal dilution of a goat anti-mouse IgG antibody, conjugated to peroxidase (from Sigma). The reaction was developed using sodium nitroprusside and o-dianisidin-dihydrochloride-(3,3′-dimethoxybenzidine) (Sigma).

ELISA for anti-DNA antibodies

Methylated bovine serum albumin (Calbiochem; 10 µg/mL in PBS) was coated on 96-well Maxisorp immunoplates (Nunc, Roskilde, Denmark). The plates were washed and incubated with S1 nuclease-treated protein-free calf thymus DNA (Sigma) at a concentration of 2.5 µg/mL in PBS overnight at 4°C. The plates were then washed and blocked with 0.1% gelatin for 1 h at room temperature, followed by two washes. Next, the plates were incubated with sera dilutions for 1 h at 37°C. After five washes, the plates were incubated with peroxidase-conjugated anti-mouse IgG (γ chain-specific) antibody (Pharmingen BD) for 1 h at room temperature. The plates were washed and developed using a 2,2′-azido-bis (3-ethyl benz-thiazoline-6-sulfonic acid) solution (ABTS; Sigma) and read at 405 nm. All serum dilutions were tested in triplicates. Significance between IgG anti-dsDNA antibody levels in the DNA-like chimera and in the PBS-treated groups was determined by Student's t-test 31.

ELISA for other lupus antibodies

IgG anti-SSB/La, anti-SSA/Ro and anti-Sm IgG antibodies in sera of immunized animals were detected using commercial ELISA kits (from Alpha Diagnostic Int., San Antonio, USA) according to the manufacturer's instructions.

ELISpot assays

Nitrocellulose membranes (0.45 µm; Sartorius, Germany) with a size covering the working area of a blotting manifold apparatus (Bio-Dot, Bio-Rad) were coated with either 50 µg/mL calf thymus dsDNA or with 30 µg/mL diphteria toxoid (from BulBio, Sofia, Bulgaria) for 30 min at room temperature. Spleen cells were obtained from sick MRL/lpr mice 10 days after the second subcutaneous immunization with 2.5 µg alum-adsorbed diphtheria toxoid. These cells were cultured for 5 days at 2 × 10⁶ cells/mL in RPMI 1640 (Gibco) containing 10% FCS, 4 mM L-glutamine, 50 µM 2-mercaptoethanol and antibiotics in the presence of the DNA-like chimera, or of the control chimeras, or of 10 µg/mL lipopolysaccharide (LPS) from E. coli (Sigma, L-2630 – used as a positive control), or of medium alone. The same number of cells from each well was then transferred to a well of the manifold apparatus on the pre-coated membranes. Next, after washing, the cells were incubated for five additional hours in a humidified 5% CO₂ atmosphere at 37°C. After washing, the membranes were incubated with Fc-specific anti-mouse IgG conjugated with alkaline phosphatase (Sigma) for 2 h at room temperature and developed using the nitroblue tetrazolium/bromo-chloro-indolyl-phosphate chromogenic substrate (Sigma). The spots were counted blindly under a microscope by an independent observer.

Separately, samples from the same spleen cell suspension were added without additional in vitro culturing to the manifold apparatus containing DNA- and diphtheria toxoid-pre-coated membranes, and the IgG anti-dsDNA and anti-diphtheria toxoid antibody-producing cells were determined as described above. This number (showing the in vivo differentiated plasma cells) was subtracted from the number obtained after the 5-day-long in vitro cultivation.

Assays for glomerular IgG deposition

Kidneys from treated mice were frozen; cryostat sections were stained with a FITC-conjugated anti-mouse IgG antibody (Vector Laboratories Inc., Burlingame, CA) and viewed under a fluorescence microscope (Carl Zeiss, Jena, Germany). The second kidney of the animals was fixed, imbedded in a paraffin block and stained with hematoxylin/eosin using standard methods.

Statistical analysis

Animal experiments were repeated twice with five mice per group in the first and with ten mice per group in the second experiment. Differences in survival were compared using the Log Rank test. All in vitro experiments were repeated at least three times. Values in the figures were expressed as mean ± SD. Student's t-test was used to compare means between groups. A value of p <0.05 was considered to be statistically significant.