Toxin-Based Therapeutic Approaches

Shapira, Assaf; Benhar, Itai

Alexander Varshavsky suggested previously the construction of a new kind of toxins where a signal that inactivates the toxin, e.g. a degradation signal, can be cleaved off by a viral protease, resulting in selective intoxication of virally infected cells. He denoted such toxins "sitoxins" (signal-regulated, cleavage mediated toxins) [506] . For the construction of such viral-protease activated sitoxins, Falnes et al designed fusion proteins composed of a FLAG peptide containing an N-terminal phenylalanine (a destabilizing amino acid according to the N-end rule), followed by an HIV-1 protease (HIV-1 PR) cleavage sites that is positioned upstream to a chimeric sequence between DT-A and the first PA-binding 255 amino acids of the anthrax LF, the LF N fragment, which was also shown to be destabilized by the addition of a degradation signal for N-end-rule-mediated degradation [513] . The rationale behind this construct was that following anthrax PA-mediated translocation into the cytoplasm of uninfected cells, the modified toxin would be rapidly degraded by virtue of its destabilizing N-degron, resulting in attenuated cytotoxic activity against these uninfected cells. In contrast, cleavage of the construct by the HIV-1 protease whose activity has been documented in the cytosol of acutely infected cells [514] [515] [516] [517] [518] would result in the removal of the destabilizing N degron from the chimeric toxin, exposing a new N terminal amino acid. As HIV-PR cleavage in each of its two recognition sequences that were chosen (TATIM*MQRG or VSQNY*VIVQ) generates a stabilizing N terminal residue (methionine or valine, respectively) according to the N-end rule, a longlife toxin with a potent cytotoxic activity is generated in HIV infected cells which leads to their destruction. Indeed, in vitro and in vivo simulating experiments showed that when the constructs were pre-digested with HIV-PR, both stability and cytotoxicity of the chimeric toxins was considerably augmented, proving that the concept of stabilizing a protein through specific proteolytic removal of a degradation signal works in practice. However, no selective eradication of HIV infected cells was observed following treatment with non pre-treated constructs, probably because of low cytosolic HIV-1 PR activity (no proteolytic processing of the constructs were detected in HIV-infected cells) [433] .