To detect whether IFN-γ-producing CTLs could lyse target cells in vitro, an LDH assay was performed; the effector/target ratios were 10:1, 20:1 and 40:1. PBMCs from healthy donors, W02, W03, and C01, were stimulated with synthetic peptides (10 μg/ml) according to the previously mentioned method for CTLs induction. EC-9706 cells, p321-loaded T2A2 cells, KYSE-140 cells, and HT-29 cells were used

as target cells. As shown in Fig. 3, when EC-9706 cells used as target cells, the peptide-specific CTLs induced by p321-1Y9L showed more potent cytotoxic activity than that of p321 at the effector/target ratio of 20:1 and 40:1 in all ABT-888 price the tested donors, otherwise the peptide-specific CTLs induced by p321-9L showed more potent cytotoxic activity than that of p321 at the effector/target ratio of 20:1 and 40:1 in two donors (W02, W03). In addition, as shown in Fig. 4, in all the tested donors, the CTLs induced by the analogue p321-1Y9L showed more potent cytotoxic activities on p321-loaded T2 cells than that of p321 at the effector/target ratio of 40:1, but not on T2 cells without peptide-loaded at all the effector/target ratios. p321-9L showed the equal cytotoxic activity with p321-1Y9L in donor W03, but in other two donors p321-9L showed the equal cytotoxic activity with p321. These results showed that in all tested donors, the peptide-specific

CTLs induced by p321-1Y9L showed more potent cytotoxic activity than that of p321, and in donor W03, p321-9L showed more potent cytotoxic activity than that of p321. To further confirm the COX-2 specificity and HLA-A2 restriction of the CTLs, Z-IETD-FMK in vitro KYSE-140 (HLA-A2-positive, COX-2-negative) and HT-29 (HLA-A2-negative, COX-2-positive) were used as target cells. As shown in Fig. 5, the CTLs induced by p321 and its analogues p321-9L and p321-1Y9L could not lyse (a) KYSE-140 cells and (b) HT-29 cells, which Tenoxicam showed that the induced CTLs were peptide specific and HLA-A2 restricted. In addition, monoclonal antibody inhibition assay was carried out to further determine

whether the effectors recognized COX-2 positive target tumour cells in an HLA-A2-restricted manner. As shown in Fig. 6, our results showed that the specific killing effects of the CTLs could be significantly eliminated when the HLA-A2 molecules on the target cells were blocked by HLA-A2 monoclonal antibody, BB7.2. To investigate whether the peptides could induce specific CTLs in vivo, HLA-A2.1/Kb transgenic mice were immunized three times with p321 and p321-1Y9L emulsified in IFA in the presence of HBVcore128 T helper epitope. After immunization, spleen lymphocytes were pooled and re-stimulated in vitro with the related peptides, respectively. Then, LDH release assay (Fig. 6) and ELISPOT assay (Fig. 7) were carried out to test the cytotoxic activity of the CTLs induced by p321, p321-9L and p321-1Y9L.