Establishment of liver metastasis model of human gallbladder cancer and isolation of the subpopulation with high metastatic potential

Abstract

To establish an experimental liver metastatic model of gallbladder cancer and isolate the subpopulation with high metastatic potential from the model, which may serve as a reliable tool in research on liver metastasis of gallbladder cancer in vivo and in vitro. Human gallbladder cancer cells of the line GBC-SD were cultured. Ten nude athymic BALB/c mice, aged 4 approximately 5 weeks, underwent inoculation of suspension of GBC-SD cells into the spleens and then their spleens were resected. Three weeks later laparotomy was performed to observe if liver metastasis occurred. Once liver metastasis was discovered, the mice were killed and the livers were taken out to undergo microscopy, tumor cells were isolated from the metastatic foci and cultured in vitro, and then inoculated into other 10 mice in the same way as their parents cells for the second round of selection. The similar steps were repeated, altogether for three rounds of selection and a total 30 mice were inoculated in 3 rounds so as to find subpopulation with high metastatic potential. Another 10 mice underwent subcutaneous inoculation, 90 days later the mice were killed to observe if pulmonary metastasis occurred. PCR was used to detect the 3 microsatellite sequences D14S68, D18S69, and D20S199 in the DNA samples of the gallbladder cancer cells of the parent cell line GBC-SD, the isolated cells of the subpopulation with high metastatic potential, and the liver of experimental animal. 90% of the mice inoculated subcutaneously with the GBC-SD cells developed subcutaneous tumors, however, no mouse in this group died and no pulmonary metastasis was found. The liver metastatic rate was 65% in the 10 mice undergoing intrasplenic inoculation. Thus a metastatic model of human gallbladder cancer in nude mice was established. The liver metastatic tumors were uniformly distributed throughout the liver parenchyma with predominance to the periphery. Despite multiple sites of involvement, the left lobes were most commonly affected in all experimental animals. Histological examination of the metastatic lesion demonstrated adenocarcinoma. Gross hepatic metastasis was detected 10, 7, and 5 weeks after the inoculation respectively in the first, second, and third round selection respectively with an incidence rate of metastases of 60%, 70%, and 90% respectively. From the third round metastatic model a subpopulation with high metastatic potential was isolated and designated as GBC-SD/M, which exhibited the similar histological characteristics as its parent cell line GBC-SD under inverted light microscopy. The three amplified bands at the sites 14S68, D18S69, and D20S199, amplified with the three pairs of specific primers for the three homo-specific microsatellites, were detected in the GBC-SD cells and GBC-SD/M cells, but not in the liver of tumor-bearing animal. A liver metastasis model of human gallbladder cancer has been established, a reliable and efficient tool for study on the metastasis of gallbladder cancer to liver in vivo. Isolated from hepatic metastasis, the line of GBC-SD/M is a subpopulation with high metastatic potential, retaining the histological properties and identification of genetic background of its parent cell line GBC-SD.