A novel pancreatic β-cell targeting bispecific-antibody (BsAb) can prevent the development of type 1 diabetes in NOD mice

Abstract

To prepare a novel Bispecific Antibody (BsAb) as a potential targeted therapy for T1D, we produced a "functionally inert" monoclonal antibody (mAb) against Glucose transporter-2 (GLUT-2) expressed on β-cells to serve as an anchoring antibody. The therapeutic arm is an agonistic mAb against Cytotoxic T-Lymphocyte Antigen 4 (CTLA-4), a negative regulator of T-cell activation expressed on activated CD4+ T-cells. A BsAb was prepared by chemically coupling an anti-GLUT2 mAb to an agonistic anti-CTLA-4 mAb. This BsAb was able to bind to GLUT2 and CTLA-4 in vitro, and to pancreatic islets, both in vitro and in vivo. We tested the safety and efficacy of this BsAb by treating Non-Obese Diabetes (NOD) mice and found that it could delay the onset of diabetes with no apparent undesirable side effects. Thus, engagement of CTLA-4 on activated T cells from target tissue can be an effective way to treat type-1 diabetes.

Keywords: Anti-CTLA-4; Anti-Glut2; Dendritic cells; Diabetes; Regulatory T cells; Tolerance.

Figure 1. Expression and purification of recombinant GLUT2

A) Upper panel: Recombinant GLUT2 ectodomain (Ecto-GLUT2; position on gel shown by black triangle) was produced in E. coli, purified using Ni++ affinity chromatography and resolved on SDS-PAGE followed by Coomassie staining (left) and western blot (right). Lower panel: Recombinant full length GLUT2 (FL-GLUT2) was produced in HEK-293 cells, solubilized by re-suspending membrane fraction in 2% tween-20, purified using Ni++ affinity chromatography (left; position on gel shown by white triangle) and resolved on SDS-PAGE followed by Coomassie blue staining (left) and western blot (right). FT-flow through; W-wash; E-elution. B) Ecto-GLUT2 bulk purified by Ni-NTA agarose (left) was further purified by Reverse Phase HPLC (right) to >95% purity. Dilutions of BSA standard and Ecto-GLUT2 were loaded for comparison of protein quantities (M- protein marker).

Figure 2. Characterization of anti-Glut2 mAbs

A) Supernatants from hybridomas D2, E7 and H5, which contained anti-Ecto-GLUT2 antibodies were used to probe for their ability to bind Ecto-GLUT2 and FL-GLUT2 in lysates from recombinant E. coli and HEK/293 cells respectively. Antibodies H5, E7 and D2 which recognized both recombinant Ecto-GLUT2 and FL-GLUT2 (positions shown by white and black triangles respectively) were selected for further characterization (E-ectodomain; FL-full length). B) Anti-GLUT2 monoclonal antibodies recognize native GLUT2 expressed on Min-6 cells as revealed by Flow cytometry. 3B11 served as a non-specific control antibody. C) Glucose Stimulated Insulin Secretion in β-TC-6 cell line treated with 50 μg/ml anti-Glut2 mAbs (D2, E7 and H5) show no defect in insulin secretion. D) Anti-GLUT2 mAb H5 binds to islet cells as shown by immunofluorescence using pancreatic tissue sections. Slides were co-stained with an anti-insulin antibody.

Figure 3. Expression Characterization of anti-CTLA-4/anti-GLUT2 BsAb

A) T-BsAb can recognize CTLA-4 on CD4⁺ T-cells. Splenocytes were incubated with T-BsAb and the antibody binding was detected using FITC-conjugated secondary anti-hamster or anti-mouse Abs. Commercial fluorescence-conjugated anti-CD4 and anti-CTLA-4 (upper right panel) were used for co-staining. B) T-BsAb retains its capacity to bind to GLUT2 on Min6 cells. Min 6 cells were incubated with T-BsAb and anti-insulin antibodies and antibody binding was detected using either FITC-conjugated secondary anti-hamster or anti-mouse Abs (for BsAb) and TRITC-conjugated secondary anti-guineas pig antibody (for anti-insulin IgG). C) T-BsAb can bind to GLUT-2 on pancreatic β cells in vivo. CB.17-SCID mice were injected with T-BsAb and pancreatic tissue isolated. Fixed pancreatic tissue sections were incubated with anti-insulin (TRITC) and anti-hamster-IgG-FITC.

Figure 4. Lack of undesirable side effects of BsAb

A) NOD mice were either left untreated or treated with T-BsAb bi-weekly. Sera from treated mice were analyzed for evidence of perturbation of liver or kidney function by ALT and BUN tests respectively. B) Histology of liver and kidney tissues as revealed by H&E staining showed no apparent difference between untreated and T-BsAb treated mice. C) GSIS on mouse β-TC-6 cells (left) or isolated mouse islets (right) either in the presence of 50μg/ml of T-BsAb or the C-BsAb show no defect in insulin secretion in response to glucose stimulation.

Figure 5. Therapeutic Efficacy of BsAb in NOD mice

A and B show percentages of NOD mice that remained normoglycemic up to 23 weeks of age upon initiation of treatment beginning at the age of 8 weeks (N=10 for untreated group; N=9 for C-BsAb and T-BsAb groups) or 10 weeks (N=10 for all groups) respectively. Statistical significance was determined by comparing T-BsAb group with control group using log rank test C) (Upper panel) Representative picture of H&E stained sections showing scheme of insulitis score. (Lower panel) Bar diagrams show insulitis of islets (percentage) in T-BsAb treated group relative to either the C-BsAb treated or untreated controls. Statistical significance was determined using Fisher’s exact test.

Figure 6. Immunomodulation upon treatment with T-BsAb

A) CD4⁺ T-cells were isolated from spleens of NOD mice, labeled with CFSE and incubated with splenic APCs, immunodominant β-cell Ag peptides and total pancreatic cells in the presence or absence of C-BsAb or T-BsAbs. Foxp3 Tregs are increased in cultures supplemented with T-BsAbs (upper panel). CD4⁺Foxp3⁻ T-cells were gated for measuring the extent of CFSE dilution which shows reduced proliferation in the presence of T-BsAbs. (lower panel). B) Supernatants from co-cultures were analyzed for the presence of IFN-γ and IL-10 by ELISA. Data are representative of two independent experiments with cells pooled from 3–5 mice (* p<0.05; ** p<0.01).