Functional characterization of N297A, a murine surrogate for low-Fc binding anti-human CD3 antibodies

Abstract

Several low- or non-FcR binding anti-human CD3 monoclonal antibodies have been under investigation for the treatment of autoimmune diseases. To model the mechanism of action of these anti-human CD3 mAbs in the murine system, an Fc-modified anti-mouse CD3 antibody (N297A) was generated. N297A exhibited similar biological effects as Fc-modified anti-human CD3 antibodies including rapid, reversible reduction in peripheral leukocyte numbers, differential modulation of activated versus resting T cells, and reduced levels of induced cytokine release compared to the non-Fc-modified parent antibody. In an in vivo model of colitis induced by adoptive transfer of IL-10-deficient cells, administration of N297A significantly reduced body weight loss. As N297A shared many functional characteristics of non-FcR binding anti-human CD3 mAbs both in vitro and in vivo, it provides a means to model the mechanisms of action of Fc-modified anti-human CD3 antibodies in mouse.

Figure 1

N297A did not bind to Fc gamma receptors. (A) RAW264.7, a mouse macrophage cell line which highly expresses Fcγ receptors, and (B) 3T12, a non-Fcγ receptor-expressing mouse cell line, were used to analyze binding of antibodies. Serial dilution of mouse antibodies IgG2a (□), IgG2b (▴), IgG1 (▾), WT (•) and N297A (○) were incubated with the cells for 20 minutes on ice. After washing, samples were incubated with secondary APC-conjugated goat anti-mouse IgG, F(ab′)₂ fragment specific antibody (2.5 μg/ml) for detection by flow cytometry.

Figure 2

Soluble N297A did not induce proliferation but induced apoptosis in activated T cells. Spleen cells from BALB/c mice were incubated with either plate-bound (A) or soluble (B) mAbs for 48 hours. Proliferation was determined following incubation with alamarBlue. WT (□), N297A (▴), or mIgG1 isotype control (). (C+D) Spleen cells from BALB/c mice were treated with WT (□), N297A (▴), or NCNS-1 negative control () mAb for 24 hours with either no prior activation (C) or prior stimulation with Con A for 96 hours (D). Apoptotic cells were then stained using TACS Annexin V/PI kit. Cells that did not stain for Annexin V or PI were calculated and shown as % of viable cells. Triplicate samples for each condition were analyzed and average values are shown in the graphs.

Figure 3

N297A caused a rapid and transient disappearance of blood lymphocytes. (A) C57/B6 mice were injected i.v. with 10 μg of N297A (▴), WT (□) or NCNS1 () mAb. Blood samples from 5 mice per group were collected in EDTA coated tubes at indicated time points for total lymphocyte counts. (B) Different dose levels of N297A (10 μg, black bars, or 1 μg, white bars) or NCNS-1 (dotted bars) were injected i.v. into naïve C57/B6 mice at day 0 (n=5). The mice were bled at the indicated time point, and CD4⁺ T cells were quantified by flow cytometry using TruCOUNT reagents; counts per μl are shown. An unpaired two-tailed t-test was used for statistical analysis. *p=0.0002 for N297A (10 μg) vs. NCNS-1; # p=0.019 for N297A (1 μg) vs. NCNS-1 at day 1. The differences at other time points were not statistically significant.

Figure 4

Anti-CD3 mAbs preferentially modulate activated T cells in vivo. Spleen cells from DO11.10 TCR transgenic mice were activated in vitro for 72 hours with OVA peptide prior to transfer to BALB/c mice. 24 hours after cell transfer, mice were dosed with 10 μg antibody i.v. (n=5 per group), and then sacrificed 18 hours after antibody treatment. Spleen and peripheral lymph nodes were harvested for analysis. Levels of pre-activated CD4+ donor cells (KJ1–26⁺) from spleen (A) or lymph node (C) were greatly reduced by treatment with both anti-CD3 mAbs. Levels of host CD4+ cells (KJ1–26-) showed little change in the spleen (B) or a moderate decrease in lymph node (D) following treatment of N297A (▴), NCNS-1 () and WT (□) mAb. Mice that did not receive any OVA-activated cells (No cells •) were used as control.

Figure 5

N297A treatment slowed bodyweight loss in the IL-10-deficient adoptive transfer IBD model. CD4⁺ T cells from IL-10 deficient mice were purified and transferred into NOD/scid mice to induce IBD. The antibody treatments were started at disease onset when the mice started losing body weight. (A) Mice were treated i.v. with 10 μg of N297A (n=9, ▴) or NCNS-1 (n=9, ) 2 to 3 times a week for a total of six doses starting at day 20. No cell control group (❍) did not receive any T cells and thus continued to grow and gain body weight (n=6). Percent change in body weight (BW) was calculated by subtracting body weight on day of measurement from ody weight on day 0 and divided by the initial body weight. Statistical significance was assessed at each time point using an unpaired, two-tailed t-test to derive the p values between N297A and NCNS-1 groups (*p < 0.05). (B-D) Representative histological images of the IBD study. (B) Normal colon of control mice that did not receive T cells. Colons of IBD mice that were treated with 10 μg of NCNS-1 (C) or N297A (D). The colons were harvested at the end of study and stained with H/E for morphology. The slides were reviewed by a certified pathologist. Severe inflammation and crypts filled with neutrophils were observed (arrows) in the colons of most NCNS-1 treated mice.