Re-engineering and evaluation of anti-DNA autoantibody 3E10 for therapeutic applications.

A key challenge in the development of novel chemotherapeutics is the design of molecules capable of selective toxicity to cancer cells. Antibodies have greater target specificity compared to small molecule drugs, but most are unable to penetrate cells, and predominantly target extracellular antigens. A nuclear-penetrating anti-DNA autoantibody isolated from the MRL/lpr lupus mouse model, 3E10, preferentially localizes to tumors, inhibits DNA repair, and selectively kills cancer cells with defects in DNA repair. A murine divalent single chain variable fragment of 3E10 with mutations for improved DNA binding affinity, 3E10 (D31N) di-scFv, has previously been produced in P. pastoris and yielded promising pre-clinical findings, but is unsuitable for clinical testing. The present study reports the design, expression and testing of a panel of humanized 3E10 (D31N) di-scFvs, some of which contain CDR substitution. These variants were expressed in a modified CHO system and evaluated for their physicochemical attributes and ability to penetrate nuclei to selectively cause DNA damage accumulation in and kill cancer cells with DNA repair defects. Secondary structure was conserved and most variants retained the key characteristics of the murine 3E10 (D31N) di-scFv produced in P. pastoris. Moreover, several variants with CDR substitutions outperformed the murine prototype. In conclusion, we have designed several humanized variants of 3E10 (D31N) di-scFv that have potential for application as monotherapy or conjugates for targeted nuclear drug delivery.