To improve therapeutic indices and decrease systemic toxicity, more specific and selective anticancer agents that can discriminate between tumor and nonmalignant cells are urgently needed, along with the development of antitumor radiochemotherapy [16
]. The discovery that angiogenesis plays a crucial role in tumor formation, and that vascular targeting approaches exhibit the advantages of easy accessibility and lower incidence of drug resistance, provides a possible path to creating these new anticancer agents [19
]. Several studies have pursued this strategy, including the application of phage display technology to pan for peptides that bind specifically to defined tissue targets [8
]. Using this technique, several homing peptides have been identified, including RGD, NGR and F3, and many have showed promising results for imaging diagnosis and treatment of various tumors in preclinical or clinical investigations [22
]. Furthermore, some of these peptides have been conjugated to bioactive agents, including drugs, cytokines, procoagulant factors, photosensitizers and radionuclides, and have been included in antineoplastic therapies. Initial results of these studies showed more selective and targeted drug delivery and fewer side effects [7
]. However, to date, no such peptide has been identified that targets human gastric cancer.
Previously, we used in vivo
screening of a phage displayed peptide library to identify GX1, a cyclic 7-mer peptide CGNSNPKSC that binds specifically to the human gastric cancer vasculature [9
]. Autoradiography on different cell lines confirmed the targeting activity of GX1 toward the gastric cancer vascular endothelium, by showing that the binding affinity of GX1 was significantly higher in HUVEC cultured in tumor-conditioned medium than in HUVEC cultured in non-conditioned medium. No specific binding was observed in the human gastric carcinoma cell line SGC7901 or in the immortalized gastric epithelial cell line GES cells [28
]. Furthermore, immunohistochemical staining and immunofluorescence showed positive staining for GX1 in the vascular endothelium of human gastric adenocarcinoma, but not in heart, liver, muscle, spleen or normal gastric tissues [10
]. In another study, using single photon emission computed tomography (SPECT), GX1 labelled with 99
-GX1) was observed to concentrate in tumor xenografts in nude mice [28
]. Collectively, these results indicate that GX1 is a novel vascular marker of human gastric cancer, and may lead to a new way of imaging diagnosis and targeted gastric cancer therapy.
Since GX1 selectively targeted the vascular endothelium of gastric cancer, we investigated whether it had specific effects on tumor angiogenesis and growth. In this study, in addition to targeting, GX1 showed bioactivity by both MTT and CAM assay, inhibiting vascular endothelial cell proliferation and hampering neovascularization. To probe into the possible mechanisms of these effects, the cell cycle distribution, cell apoptosis and the expression level of apoptosis related molecule caspase3 were detected by FCM and western blot assays. Inhibition of vascular endothelial cell proliferation by GX1 was observed, at least in part, to be through the up-regulation of caspase 3 expression and the induction of apoptosis. Further tests including RT-PCR and gene microarray are underway to investigate the precise mechanisms.
analysis showed that HUVEC cultured in tumor-conditioned media partially acquire the characteristics of tumor vascular endothelial cells, such as enhanced tubule formation, cell proliferation, and migration [13
]. Furthermore, some proteins like vascular endothelial growth factor receptor and the integrin αvβ3 may be up-regulated in co-HUVEC, as is the case for other cancer endothelia [6
]. These findings lead us to the hypothesis that GX1 receptors are up-regulated in co-HUVEC, reflecting the case in tumor vessels, and that more receptors may lead to greater selective affinity and stronger anticancer effects. This hypothesis is consistent with the MTT assay results, in which GX1 showed more significant inhibitory effects on co-HUVEC than on the parental HUVEC culture that was not exposed to tumor-conditioned medium.
To assess the possibility of using GX1 as a targeted delivery vector in combination with another antitumor molecule for treatment of gastric cancer, GX1 was conjugated to rmhTNFα. TNFα is a well-known, antitumor cytokine whose clinical application is hampered by severe systemic toxicity [12
]. The novel mutant cytokine rmhTNFα shows higher antitumor efficacy and has been approved for clinical use in China [12
]. Our data showed that after fusion to GX1, rmhTNFα was selectively delivered to target tumor vasculature sites. Most important, GX1-rmhTNFα delayed tumor growth in vivo
, with less loss of body weight compared to rmhTNFα alone (Figure ). These results indicated that more targeted and efficient antitumor activity might be achieved by combining GX1 with other anti-tumor agents (e.g. rmhTNFα), for a significant reduction in systemic toxicity.
Despite the encouraging results, some questions are still open, such as what the receptor is for GX1 on vascular endothelial cells, and how ligand-receptor interaction interferes with tumor angiogenesis. Further studies are underway to answer these questions, and several candidate receptor molecules have been obtained. Identification of the GX1 receptor will be a great help in understanding the mechanism of GX1 and will accelerate the development of clinical applications for GX1 in diagnosis and targeted treatment of gastric cancer.