Engineered Adenovirus Fiber-TRAIL Fusion Proteins With Enhanced Antitumor Activity

Successful cancer therapies aim to induce selective apoptosis in neoplastic cells. TNF-related apoptosis-inducing ligand(TRAIL) is considered an attractive anticancer agent due to its tumor cell-specific cytotoxicity. Therefore, TRAIL is a promising agent for cancer therapy, as TRAIL-based therapeutics have been shown to be more effective inducers of apoptosis in cancer cells than normal cells. The soluble TRAIL(s TR) protein exists as a homotrimer, and a single zinc atom is chelated by the Cys230 of each monomer. The trimeric structure is a critical requirement for its biological function. Importantly, TRAIL induces apoptosis in malignant cells via both extrinsic and intrinsic pathways, and its effect is independent of the functional status of p53, thus increasing the probability of the apoptotic outcome. However, earlier studies with recombinant TRAIL revealed many shortcomings, including a short half-life, off-target toxicity and these deficiencies are further aggravated by a short half-life in the blood as well as the original and acquired resistance of certain cancers to TRAILIn this study, we developed a novel engineering strategy for recombinant soluble TRAIL by redesigning its structure with the adenovirus knobless fiber motif to form a stable homotrimer with improved antitumor activity. The human adenovirus type 5(Ad5) fiber protein is encoded by a single gene that expresses a polypeptide of 581 amino acid(aa) residues and exists as a homotrimer. The monomeric fiber protein is composed of an N-terminal tail of approximately 47 aa residues that interact with the penton base protein of the capsid, a shaft comprising 21 pseudo repeats of 15 aa and an ~180 aa C-terminal knob for cellular attachment. The formation of the trimeric fiber is less essential for its function and its assembly into virus particles. The fiber shaft at the C terminus was found to be required for fiber knob trimer formation, since knob proteins containing two or more shaft repeats can form functional trimmers.Here, to obtain a highly stable and active trimeric TRAIL protein while avoiding hepatocyte transduction in vivo by the fiber shaft, several trimerization elements from human Ad5 and fowl Ad1 fiber were first used to engineer the s TR protein. A human Ad5 fiber s TR was reengineered to contain the N-terminal tail and the shortened shaft repeats(HA5FT) or to contain the C-terminal shaft repeat(HA5ST). A fowl Ad1 s TR(FA1FT) containing the full-length knobless fiber was also constructed. Due to the reported toxicity of the polyhistidine tag in normal hepatocytes and keratinocytes and lack of the optimum amount of zinc, all purified proteins were digested by thrombin, and zinc was added in the s TR samples.In vitro, the recombinant TRAIL variant FA1 FT and HA5 ST showed the greatest apoptotic activity in human tumor cell lines and a negligible toxic effect in normal cells. In addition, FA1 FT and HA5 ST were found to be more stable than s TR at 37°C or after freeze/thaw cycles in vitro. Furthermore, HA5 ST displayed a better pharmacokinetic profile than s TR with minimal activity loss. These enhanced physiological stability and better pharmacokinetic profiles than s TR resulted in improved therapeutic effects in an animal breast tumor model without detectable side effects. We demonstrated potent apoptosis-inducing activities of FA1 FT and HA5 ST, similar to that of s TR but with an enhanced stability profile in vitro compared to the latter. In fact, an improved half-life of HA5 ST was observed, now exceeding 6 h in mice compared with a half-life of only half an hour reported earlier for recombinant TRAIL. Furthermore, HA5 ST did not induce cell death in human primary hepatocytes. Our results clearly indicate that HA5 ST has therapeutic potential for the treatment of cancers. Another important finding of our current work is the fact that the last repeat of the shaft in human Ad5 fiber could be further genetically modified as a trimerization element.In vivo, FA1 FT was found to have substantial loss of activity in vivo after i.p. injection. In living animals, circulation plays an important role in the metabolism of proteins. Unlike that found with HA5 ST, FA1 FT did not induce tumor cell apoptosis even at high protein concentrations. The local treatment results suggested that FA1 FT could sufficiently induce tumor apoptosis in vivo. The biodistribution analysis revealed that FA1 FT was concentrated in the liver, which we believe was responsible for the remarkably reduced antitumor activity and the weak pharmacokinetic characteristice. These unexpected results presumably was due to the shaft sequence, which is a repetitive heparin-binding motif. The human Ad5 shaft domain has been demonstrated to target to hepatic cells and cytokine release when administered through intravenous injection in mice. Although most of the FA1 FT proteins were detected in the liver, the fowl Ad1 shaft did not induce cell death in primary hepatocytes.The result is a highly stable fiber-TRAIL fusion protein that could formhomotrimers similar to natural TRAIL. The recombinant fusion TRAIL developed here displayed high specific activity in both cell-based assays in vitro and animal tests in vivo. This construct will serve as a foundation for a new generation of recombinant proteins suitable for use in preclinical and clinical studies and for effective combination therapies to overcome tumor resistance to TRAIL.