Potency and stability of LAP+FOXP3+ Tregs in control GVHD development (VAC3P.954)

Intro: We have developed previously a method to isolated bona fide Tregs from expansion cultures based on their selective surface expression of latency-associated peptide (LAP). In this study, we investigated their in vivo potency and stability using a humanized GVHD murine model. Methods: CD25+ cells isolated from PBMC via AutoMACS were stimulated with OKT3 loaded K562 aAPC expressing CD64 and CD86 in media +IL2 and rapamycin. 36h post-restimulation at wk2, the Tregs were re-isolated with antiLAP via AutoMACS. The LAP+ Tregs and LAP- nonTregs were expanded for 2wks. In vitro suppression, phenotype and TSDR demethylation were performed after 4wk expansion. CD25-depleted human PBMC were injected into NSG mice to induce xGVHD and compared with coinjection of LAP+ Tregs or LAP- nonTregs. Results: LAP+ repurification results in >85% LAP+FOXP3+ Tregs, leaving behind FOXP3- and FOXP3+ nonTregs within the LAP- population. After a total of 4wk expansion, LAP+ Tregs were >1 billion, >85% FOXP3, highly suppressive and anergic in vitro, and >95% demethylated TSDR (p<0.05). In GVHD model, PBMC alone resulted in a median survival of 29 days. The LAP+ Tregs increased median survival to 43 days (p<0.01), while the LAP- non-Tregs had a median survival of 28.5 days. Conclusions: LAP+ Tregs are highly purified, stable, and potent based on these preclinical studies. LAP+ Tregs are an ideal immunotherapeutic population for clinical trials to prevent or treat GVHD and other autoimmune diseases.