Bifunctional fusion proteins of the human engineered antibody domain m36 with human soluble CD4 are potent inhibitors of diverse HIV-1 isolates

Abstract

Currently used antiretroviral therapy is highly successful but there is still a need for new effective and safe prophylactics and therapeutics. We have previously identified and characterized a human engineered antibody domain (eAd), m36, which exhibits potent broadly neutralizing activity against HIV-1 by targeting a highly conserved CD4 binding-induced (CD4i) structure on the viral envelope glycoprotein (Env) gp120. m36 has very small size (∼15kDa) but is highly specific and is likely to be safe in long-term use thus representing a novel class of potentially promising HIV-1 inhibitors. Major problems with the development of m36 as a candidate therapeutic are possible short serum half life and lack of effector functions that could be important for effective protection in vivo. Fusion of m36 to human IgG1 Fc resulted in dramatically diminished neutralization potency most likely due to the sterically restricted nature of the m36 epitope that limits access of large molecules. To confer effector functions and simultaneously increase the potency, we first matured m36 by panning and screening a mutant library for mutants with increased binding to gp120. We next fused m36 and its mutants with the first two domains (soluble CD4, sCD4) of the human CD4 using a polypeptide linker. Our results showed that the selected m36 mutants and the sCD4 fusion proteins exhibited more potent antiviral activities than m36. The m36-sCD4 fusion proteins with human IgG1 Fc showed even higher potency likely due to their bivalency and increased avidity although with a greater increase in molecular size. Our data suggest that m36 derivatives are promising HIV-1 candidate therapeutics and tools to study highly conserved gp120 structures with implications for understanding mechanisms of entry and design of vaccine immunogens and small-molecule inhibitors.

Fig. 1

Amino acid sequence alignment of m36 mutants with m36. The sequences are numbered and antibody FRs and CDRs are indicated according to the ImMunoGeneTics (IMGT) numbering system (http://imgt.cines.fr/IMGT_vquest/vquest?livret=0&Option=humanIg). The residues in the mutants, which are identical to those in m36, are indicated by dots.

Fig. 2

ELISA binding of m36 and its mutants to gp120_Bal (A), gp140_JRFL (B), gp140_SC (C) and gp120_Bal-CD4 (D). Antibody specificity is determined by using an unrelated antigen, bovine serum albumin (BSA).

Fig. 3

Dose-dependent neutralization of Bal, JRFL and 89.6 by m36h1Fc and m36.4h1Fc. The pseudotyped viruses were generated from 293T cells and the assays were performed on HOS-CD4-CCR5 cells.

Fig. 4

Design of fusion proteins of m36 and its mutants. (A) Schematic representation of fusion protein architectures. (B) Sequences of polypeptide linkers connecting m36 or m36.4 with sCD4. (C) Reducing SDS-PAGE of m36, m36.4 and their fusion proteins.

Fig. 5

Comparative analysis of ELISA binding. (A) m36-sCD4 fusion proteins with linkers of different length are compared to m36 or sCD4 alone, and unlinked m36 plus sCD4, for binding to gp120_Bal. (B) Difference in binding of m36L2CD4 and m36L2CD4Fc to gp120_Bal. (C) Dramatic increase in binding of m36L2CD4Fc compared to that of m36h1Fc and sCD4Fc. (D) Comparison between the fusion proteins of m36 and m36.4 for binding to gp120_Bal.

Fig. 6

Dose-dependent inhibition of 92UG037.8, Bal, JRFL and 89.6 by m36h1Fc, sCD4Fc and m36L2CD4Fc, respectively. The pseudotyped viruses were generated from 293T cells and the assays were performed on HOS-CD4-CCR5 cells