A serious complication in the treatment X-linked bleeding disorder hemophilia A is the formation of inhibitory antibodies against factor VIII (FVIII), which compromise traditional replacement therapy. We previously developed an Oral immunotherapy (OTI) based on repeated uptake of a mixture of lettuce plant cells transgenic for heavy chain (HC) or C2 domain of human FVIII fused to cholera toxin B (CTB) subunit [Blood 124:1659; Plant Biotechnol J. 16:1148]. Fusion proteins were transgenically expressed in the chloroplasts. Repeated oral uptake of a mixture of freeze-dried powder of lettuce cells accomplished antigen delivery to the immune system of the small intestine by targeting of the GM1 receptor that is highly expressed on the surface of the gut epithelium, resulting in induction of regulatory T cells (Treg) that suppress inhibitor formation upon subsequent intravenous (iv) FVIII replacement therapy. An alternative to oral antigen delivery is the oral delivery of immune modulatory antibodies. Here, we compared the plant cell-based method with oral delivery of anti-CD3, which has been successful in murine models of autoimmune disease and is currently evaluated in clinical trials. Unlike in iv administration, oral anti-CD3 does not systemically deplete T cells. Hemophilia A BALB/c mice (F8 e16 gene deletion) received oral gavage of a mixture of CTB-FVIII-HC/-C2 (1.5 µg/antigen) expressing lettuce leaf cells 2x/week for 9 weeks. Starting at 4 weeks into the experiment, 1 IU/mouse of BDD-FVIII (Xyntha) was given iv, once per week for 5 weeks. Alternatively, following a published protocol that was successful in other models, 5 µg of hamster anti-murine CD3 was given by oral gavage daily for 5 straight days, followed by 5 weekly iv injections of BDD-FVIII. Control animals (no OTI) developed inhibitors with an average titer of 18 ± 3 BU/ml (n=16). Of these, 88% were high-titer (i.e >5 BU/ml, up to 43 BU/ml). Inhibitor formation was significantly reduced in plant cell-treated mice (10 ± 2.5 BU/ml, n=17), with 47% showing no or low-titer inhibitors (<5 BU/ml). Interestingly, inhibitor formation in anti-CD3 treated mice showed a unique pattern, with two distinct outcomes. One subset of animals developed inhibitors of 0-15 BU/ml (on average 2.5-fold lower than control group), while the second subset showed increased titers of 41-89 BU/ml. There were no animals with intermediate titers. In general, FVIII-specific IgG titers followed the same patterns as inhibitor responses. When the 2 approaches (plant cells and anti-CD3) were combined, 50% of mice developed <5 BU/ml (as compared to 36% for anti-CD3 only). However, this did not represent an improvement over giving lettuce cells only. Finally, we tested anti-CD3 Fab fragment in place of FL anti-CD3. While a daily dose of 5 µg anti-CD3 Fab/mouse was unsuccessful in suppressing inhibitor formation, a dose of 0.5 µg marginally decrease titers. In summary, oral anti-CD3 delivery was not as effective in reducing inhibitor formation as plant cells expressing CTB-FVIII antigen fusions. While FL anti-CD3 showed promise in a subset of animals, it also posed a risk of an increased response in others, which complicates development of this approach. Distinct from all other treatment groups, lettuce-fed mice showed elevated frequencies in Foxp3+ Treg, a reduction in Teff cells, and the highest increase in LAP+ Treg. Further studies on induced CD4+LAP+FoxP3- Treg cells from lettuce-fed mice showed high expression of ICOS, CD69 and Ki-67, but not CTLA-4. Among the different subsets of Treg, LAP+ T cell contained the highest proportion of IL-10 expressing cells, and oral antigen delivery induced IL-10+LAP+ Treg, which are likely key to the suppression of the immune response against FVIII.


Daniell:Takeda Pharmaceuticals: Patents & Royalties. Herzog:Takeda Pharmaceuticals: Patents & Royalties.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution