Abstract 1515: Disruption of circadian rhythmicity in liver metastasis of colorectal cancer

Background: Liver metastases of colorectal cancer are a major cause of cancer related mortality. Resection of hepatic metastasis prolongs life, but 25% of the functional reserve of the liver should remain after resection to allow adequate liver regeneration. An effective means to downsize the tumor and enable resection is the use of chemotherapeutics. Efficacy and adverse side effects of chemotherapy may vary considerably over a 24 hour period. Chronotherapy (circadian rhythm adjusted chemotherapy) could take advantage of the fact that there may be asynchronies in circadian rhythm between normal and cancer tissue. In addition, genetic or functional disruption of the circadian clock may result in genomic instability and accelerated carcinogenisis and tumor growth. A more thorough understanding of circadian rhytmicity may help in elucidating the role of the clock in tumorigenesis and tumor growth and develop more effective chronotherapeutic chemotherapy treatment schedules. Therefore, in this study we investigated the 24-hour expression levels of key circadian clock genes in colorectal liver metastasis and adjacent healthy liver tissue in the mouse. Materials and methods: Male BALB/c mice were injected with C26 colorectal carcinoma cells into the spleen. Three weeks after tumor inoculation, a hepatectomy was performed at 6 different timepoints over a 24-hour period (n=4 per timepoint) to compare circadian rhythmicity in tumor and adjacent healthy tissue. RNA was isolated and RT-PCR was performed on 5 clock genes: Rev-erbα, Per2, Per1, DBP, and Bmal1 and a clock independent gene induced by feeding (Acss2). Results: Healthy liver tissue showed normal 24-hour oscillations of all 5 clock genes, consistent with normal circadian rhytmicity. In colorectal liver metastasis however, 24-hour oscillations were completely absent in all clock genes except Per2. Per2 oscillations were similar in tumor and adjacent liver tissue, although expression levels in tumor tissue were lower. Since Per2 may also be induced by feeding, we determined the expression of the food induced gene Acss2, which is independent of circadian rhythm. Acss2 levels rose in response to feeding, followed by a rise in Per2, strongly suggesting that the increased expression of Per2 was induced by feeding. Conclusion: Our data demonstrate the absence of circadian rhythm in hepatic metastasis of C26 coloncarcinoma in vivo. The difference between tumor and healthy tissue in this model may provide new clues