Biologics consisting of CTLA4 fused to the Fc domain of IgG (CTLA4-Ig), marketed as abatacept and belatacept, are immunosuppressive therapies approved for treatment of rheumatoid arthritis (RA) and kidney transplant, respectively. These biologics function by selectively modulating the CD80/CD86:CD28 costimulatory signal needed for full T-cell activation. CTLA4-Ig is significantly less potent at inhibiting CD86-dependent costimulation as opposed to CD80-dependent costimulation, due to CTLA4 binding with much higher avidity to CD80 than CD86. Abatacept is administered by a monthly IV infusion, which is less convenient than the subcutaneous self-injections of competing anti-TNF biologics. We have therefore developed a new optimized CTLA4-Ig with an engineered CTLA4 domain that binds preferentially and with higher affinity to CD86. In addition, by increasing affinity of the CTLA4-Ig Fc domain for the antibody salvage receptor FcRn, we generated a biologic with significantly longer half-life (Zalevsky et al. Nat. Biotech. 2010; 28:157). This biologic has the potential to increase efficacy, reduce dosing frequency, and enable more convenient subcutaneous dosing.
XENP9523 is a CTLA4-Ig with optimized affinity for CD86 and containing an extended half-life Ig domain. Using rational structure-based engineering, 21 positions in the extracellular domain of CTLA4 were identified as targets for mutagenesis. A total of 149 variants were constructed at these positions during two rounds of affinity optimization. After each round, CD80 and CD86 affinity was measured using SPR, and variants with higher CD86 affinity were identified. For mixed lymphocyte reaction (MLR) assays, PBMC at 1.2x106 per well from two different donors were mixed with CTLA4-Igs and incubated for 6 days, followed by measurement of IL2 release by ELISA. For PK studies, same-sex mice transgenic for human FcRn (mFcRn-/- hFcRn Tg 276 heterozygote on a B6 background) were obtained from The Jackson Laboratory and a single IV dose of CTLA4-Ig at 10 mg/kg given, followed by regular blood collection. PK parameters were determined for individual mice with a non-compartmental model using WinNonlin.
CTLA4-Ig variants with 20-fold increased binding to CD86 were identified and variant XENP9523 was chosen based on favorable CD80/CD86 binding (improvement in CD86 >> CD80). This variant showed a 4-fold improvement in IC50 for the ability to inhibit IL2 production in MLR assays compared to abatacept. Variant XENP9523 when fused to our extended half-life Fc domain showed a 2-fold improvement in half-life and similar exposure compared to abatacept in a PK study using huFcRn transgenic mice.
We show that CTLA4-Ig engineered for enhanced binding to CD86 via its CTLA4 domain, and for extended half-life via its Ig domain, is a potent suppressor of T-cell activation in vitro. Our results demonstrate the importance of high affinity binding to CD86 for modulating the CD80/CD86:CD28 costimulatory signal. The combination of increased potency and extended in vivo half-life suggests that XENP9523 has potential clinical advantages compared to abatacept and belatacept, while simultaneously enabling the convenience of subcutaneous dosing.
Disclosure: M. J. Bernett, Xencor, Inc., 3, Xencor, Inc., 1 ; S. Y. Chu, Xencor, Inc., 3, Xencor, Inc., 1 ; H. M. Horton, Xencor, Inc., 3, Xencor, Inc., 1 ; E. Pong, Xencor, Inc., 3, Xencor, Inc., 1 ; I. Leung, Xencor, Inc., 3, Xencor, Inc., 1 ; G. L. Moore, Xencor, Inc., 3, Xencor, Inc., 1 ; U. S. Muchhal, Xencor, Inc., 3, Xencor, Inc., 1 ; G. A. Lazar, Xencor, Inc., 3, Xencor, Inc., 1 ; D. E. Szymkowski, Xencor, Inc., 3, Xencor, Inc., 1 ; J. R. Desjarlais, Xencor, Inc., 3, Xencor, Inc., 1 .