Gene Construction, Expression And Targeted Killing Of Immunoproapoptotic Molecule ScFv15-Fdt-HA2-tBid Against HBsAg Positive Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is one of the most common malignanttumors occurring worldwide. High recurrence rates of surgical treatment andnon-specific side-effects of radiotherapy and chemotherapy on normal tissueslead to limited clinical benefits of HCC therapy. We have established a novelantitumor strategy by generating a series of immunoproapoptotic molecules(IPMs) and confirmed its antitumor activity against HER2-positive tumors invitro and in vivo. Given its tumor-specificity and minimal side effect, the IPMsstrategy is really promising in HCC tumor therapy. The key issue for theHCC-targeting IPMs strategy is to identify a single-chain antibody specificallyrecognizing HCC markers and serving as a guide to kill HCC tumors.Fortunately, we previously identified such a humanized single-chainantibody against hepatitis B surface antigen (HBsAg). By screening a phage-display antibody library, scFv15was shown to bind specifically toHBsAg with high affinity, followed by internalization to the endosomes ofHBsAg positive tumor cells. Our previous work showed that thescFv15/protamine fusion protein effectively delivered small nucleic acidtherapeutants specifically to HBsAg positive tumor cells, resulting in potentantiviral effects. This makes us to believe that the scFv15is feasible to directIPMs to kill HBsAg positive HCC tumor cells.The HBsAg-targeting IPM comprises the scFv15, a translocation domainand a proapoptotic molecule. Although the IPM undergoes scFv15-mediatedinternalization to HBsAg positive HCC tumor cells, there are still two criticalissues to limit their antitumor activity.The first limiting issue is for the proapoptotic molecule to exert as strongapoptotic effect as possible. Our earlier study compared tumor cell apoptosisinduced by caspase-3, caspase-6, granzyme B, AIF, and truncated Bid (tBid) invitro and in vivo. It was shown that tBid induced apoptosis more quickly andefficiently than others. Besides, tBid has small size, easy to penetrate into solidtumor tissues. Therefore, tBid was chosen as a therapeutic domain for theHBsAg-targeting IPM.The second limiting issue is for the translocation domain to exert as strongtranslocative effect as possible. Apoptosis occurs only after the HBsAg-targetingIPM is translocated from the endosomes to the cytosol, and the endosomalescape involves two sequential steps. First, in the endosomes, furin proteasesrecognize and cleave the HBsAg-targeting IPM, leding to dissociation betweenthe scFv15moiety and the proapoptotic molecule. Second, endosomalmembranes are disrupted, causing the release of the proapoptotic molecule tothe cytosol. Based on the two sequential translocation steps, we optimized severaltranslocation domains and peptides derived from bacterial exotoxins and viralinfection-related proteins. For the furin cleavage step, Fdt motif, thefurin-cleavable sequence from diphtheria toxin, was shown to be cleaved mostefficiently. For the endosomal disruption step, HA2sequence of influenzaHemagglutinin was shown to destroy endosomes most potently. However, itremains unknown whether the location of HA2in the IPM affects itstranslocation activity.To address this issue, HA2was fused to the C-terminus or the N-terminusof the scFv15to generate scFv15-HA2and HA2-scFv15respectively the fusionproteins were purified and compared for their endosomal escape capability. Itwas shown that the HA2-scFv15, but not scFv15-HA2, caused ensodomeleakage and release of fluorescence-labeled indicator into the cytosol. This dataled us to localizing HA2to the N-terminus of the proapoptotic molecule in thisIMP strategy.In present study, the HBsAg-targeting IPM gene was generated by fusingthe gene fragments of the scFv15, Fdt, HA2and tBid sequentially. The resultingfusion protein gene was designated as scFv15-Fdt-HA2-tBid. In theory, afterentry into the endosomes, scFv15-Fdt-HA2-tBid would be cleaved by furinproteases at the Fdt motif, making the scFv15moiety and HA2-tBid dissociated.The exposure of the HA2-tBid would cause endosomal disruption, releasingtBid to the cytosol to induce apoptosis. scFv15-Fdt-tBid, an IMP without HA2sequence, was used as a control.In vitro study showed that both scFv15-Fdt-HA2-tBid and scFv15-Fdt-tBidspecifically recognized and bound to HBsAg positive tumor cells but not thoseHBsAg negative cells. scFv15-Fdt-HA2-tBid escaped from the endosomes more quickly and efficiently than scFv15-Fdt-tBid. Both of them killed HBsAgpositive tumor cells by inducing apoptosis, with stronger killing ability ofscFv15-Fdt-HA2-tBid. The IPM-induced apoptosis was shown to involvemitochondrial apoptotic pathway, given loss of mitochondrial transmembranepotential and activation of caspases.The in vivo antitumor activity of the IPMs was evaluated in orthotopichepatic tumor bearing nude mice. Intravenous injection of the IPMs led to rapidaccumulation of the fusion proteins within tumors instead of normal tissues. TheIPMs significantly inhibited tumor growth, and induced tumor cell apoptosis asassayed by TUNEL. scFv15-Fdt-HA2-tBid showed stronger antitumor activitythan scFv15-Fdt-tBid.Balb/c mice receiving10-fold therapeutic dose of scFv15-Fdt-HA2-tBidwere used as a biosafety model. There were neither acute toxic effects on serumbiochemistry and hematological parameters, nor pathological morphology inmajor organs. The data indicate the in vivo safety of the scFv15-Fdt-HA2-tBid.In summary, we generated the HBsAg-targeting IPM,scFv15-Fdt-HA2-tBid. It specifically recognized HBsAg positive HCC cells invitro and in vivo. The presence of HA2enhanced tumor cell apoptosis bymaximizing endosome escape of the tBid. scFv15-Fdt-HA2-tBid was relativelysafe and less immunogenic. Our data suggest the therapeutic potential ofscFv15-Fdt-HA2-tBid against HBV related HCC.