Lettuce Plants Expressing High Levels of Factor VIII Antigen in the Chloroplast for Oral Tolerance in Hemophilia a

Inhibitor formation is a serious complication in factor VIII (FVIII) replacement therapy for the X-linked bleeding disorder hemophilia A and occurs in 20-30% of patients. No prophylactic tolerance protocols exist. In previous proof-of-principle work, we demonstrated oral tolerance induction to FVIII in hemophilia A mice using frozen tobacco leaf cells expressing the heavy chain (HC) and C2 domain of human B domain-deleted (BDD) FVIII (Blood 124:1659-68, 2014). This approach is based on expression of heterologous proteins in the chloroplasts of green leaves and natural bioencapsulation provided by plant cell wall, so that the antigen is protected until its release in the small intestine. The antigens are expressed as fusions with the transmucosal carrier CTB proteins for effective targeting of the gut epithelium and translocation to the immune system. However, significant hurdles toward clinical feasibility remained that we now successfully addressed. Until now, methods to express large human proteins in chloroplasts of edible crop plants did not exist. Building on our recently established gene transfer technology for lettuce chloroplasts and identification of powerful posttranscriptional regulatory elements, we focused on the development of chloroplast genomics tools (e.g. bioinformatics tools for codon optimization and assays that identify sites of ribosome stalling), which ultimately allowed us to express high levels of FVIII antigen in edible crop plants that are suitable for clinical application. Remarkably, codon-optimization improved expression of FVIII heavy chain (HC) and light chain (LC) by15- to 42-fold. Hence, the entire BDD-FVIII sequence can be orally delivered, thereby covering the entire patient population regardless of individual CD4⁺ T cell epitope usage. A clinical grade manufacturing process is in place, and lyophilization of leaf cells generates material that is stable at ambient temperature for at least 2 years. Testing in hemophilia A mice showed effectiveness of low FVIII antigen doses upon delivery of lyophilized lettuce cells. Repeated oral gavage (twice/week for 2 months) of a mixture of CTB-HC and CTB-C2 suppressed subsequent inhibitor formation against intravenous FVIII (1IU, once/ week for 4 weeks) in hemophilia A mice as well as the original tobacco material, using very low doses of 0.15-0.5 ug/antigen/mouse. When using CTB-HC/CTB-LC combination, a dose of 1.5 ug was most optimal and suppressed inhibitor titers by ~1 log (avg. 4.7 BU/ml vs 44 BU/ml in controls; 50% of tolerized mice had 0-3 BU/ml, while 86% of controls developed 6-126 BU/ml, n=12-14, P <0.05). Translational oral tolerance studies in humans would benefit from cellular biomarkers. Our prior studies showed induction of two subsets of Treg that suppress antibody formation against FVIII, namely CD4⁺CD25⁺FoxP3⁺ Treg and LAP⁺CD4⁺CD25^-FoxP3^- Treg. The latter stain positive on the cell surface for LAP (latency-associated peptide) because of overexpression of TGF-b. Although we found increases in the frequencies of LAP⁺CD4⁺CD25^-FoxP3^- Treg in peripheral blood, a statistical significance was not always observed when compared to the blood of control mice that received intravenous FVIII but not oral tolerance. However, we found significant and consistent increases in the frequencies of circulating LAP-expressing CD4⁺CD25⁺FoxP3⁺ Treg in tolerized mice, suggesting that LAP may be a suitable cellular biomarker for plant-based oral tolerance induction. In conclusion, clinical candidates for oral tolerance to FVIII are now available in form of lettuce plants expressing immunogenic domains or the entire antigen.