By David N. Leff

Second only to lung cancer, colorectal carcinoma is the most common malignancy in the U.S. It strikes 140,000 people a year, and kills 60,000 of them. Yet this medical terror is potentially curable, if diagnosed early and treated by surgery and chemotherapy.

If not, the tumors growing in the large intestine strike out on their own, and land most often in the liver. To get there, they may break out through the bowel wall or via the blood or lymphatic route. Once this hepatic metastasis takes hold, it becomes the main cause of death.

¿Why the liver is a preferred target for metastasis,¿ suggested molecular oncologist Kenneth Kinzler, at the Johns Hopkins Medical Institutions in Baltimore, ¿may simply be an anatomical issue. A lot of the blood supply that is removed from the gut goes to the liver.¿

Kinzler is co-senior author of a paper in Science, dated Oct. 12, 2001. Its title: ¿A phosphatase associated with metastasis of colorectal cancer.¿

¿I think the take-home, cut-to-the-chase point of our Science article,¿ Kinzler told BioWorld Today, ¿is distinct molecular differences between the metastatic regions and primary cancers. And it appears that some of these may be due to genetic changes.

¿In particular,¿ he continued, ¿this finding is based mainly on one of these genes that we identified, called PRL-3, which is consistently elevated in liver metastases from colorectal cancer. They went up in all 18 liver metastases that we looked at. Also, in about 25 percent of the cases where we could look at the gene, we saw an increase in its expressed copy numbers. It¿s one of the first times that a gene has been shown to be differentially expressed in a metastatic lesion due to a potential genetic alteration.

¿There are many differences between cancer cells and metastatic cells, as well as in normal tissue,¿ Kinzler pointed out, ¿and it¿s always difficult to sort out which ones are passengers, and which are critical to the growth of the tumors. The fact that we saw in some cases an increase in PRL-3 copy number suggests that there¿s a genetic basis for this increase in gene expression.¿

PRL-3 Gene Stars In SAGE Search

¿PRL-3 encodes a protein phosphatase,¿ Kinzler observed.¿ The mouse gene was first cloned in 1998, the human one in 2001. It resides on the long arm of human chromosome 8.

To begin with, he and his co-authors comparatively quantified the gene¿s expression by a screening technique called SAGE ¿ serial analysis of gene expression, which was invented by Kinzler and other co-authors at Johns Hopkins and licensed to Genzyme Molecular Oncology.

¿We isolated messenger RNA from purified tumor cells in Stage 4 liver metastases,¿ Kinzler recounted, ¿and compared the pattern of gene expression in those lesions to those we observed in normal colon tissue and primary colorectal cancer. We compared over 17,000 different gene transcripts from tissue samples, and identified several that were differentially expressed. They included 79 that were significantly lower, and 38 significantly higher, than in primary cancers. That¿s where the SAGE technology came in.

¿We started out just looking at a single metastasis, comparing it to two primary tumors and two normal colon samples. We went on then, after having identifying this list of candidate genes, and focused on those that were increased, because we thought they had the most interesting therapeutic and diagnostic potential. So we focused on 38 of the transcripts that were higher in the metastatic lesion. But only one of them, PRL-3, was found to be elevated in all the lesions we looked at. Eventually we wound up looking at 18 different liver metastases, and PRL-3 was increased in each of those 18 cases.

¿We confirmed these results,¿ Kinzler continued, ¿by looking at a series of six patients. The colons in all six had normal tissue, a primary cancer and liver metastases. PRL-3 was preferentially expressed in all six metastatic lesions. We also were prompted to figure out why this gene was expressed at such high levels. One way is genetic amplification ¿ extra copies of the gene cloned. We found that in these lesions, in some cases, the gene was increased more than 20-fold ¿ more than 20 copies, compared to the normal two.

¿That PRL-3 increased expression we saw in liver metastases is very exciting,¿ Kinzler said, ¿because it suggests a possible therapeutic target in the future, and insights into the mechanisms for metastases to occur. We also want to know whether metastases from other types of primary tumors, such as breast or lung cancer, overexpress PRL-3. We want to do functional studies of the gene to try to figure out what role it may be playing in the metastatic process. Also, what gene product is the protein phosphatase enzyme targeting.¿

Kinzler proposes that PRL-3 has the theoretical potential for colorectal cancer diagnosis ¿ including imaging ¿ and staging of patients. ¿There¿s also the possibility of prognosis,¿ he pointed out. ¿We really don¿t have enough data yet, but it¿s possible that slight increases of PRL-3 expression in primary cancers may predict the propensity for metastasizing.¿

Drug Discovery To Block Enzyme Simple¿

¿A third, more promising, possibility,¿ he added, ¿is chemotherapy. The fact that we see PRL-3 consistently elevated, the fact that it¿s an enzyme, and the fact that at least in some cases it seems to be driven by genetic alteration ¿ amplification ¿ suggests that it may be critical to the growth of metastases. Therefore, it would seem to be a good target for development of drugs, because as an enzyme, it could be inhibited. It¿s very simple. There are substrates for phosphatases; this one is no exception. One could look for a small macromolecule that inhibits this protein¿s activity. And one might want to further look for ones that are specific for PRL-3, relative to other members of its gene family, such as PRL-1 or 2.

¿It could be done very soon,¿ he observed, ¿but I think from the point of view of the research, people would first like to see some more data about its functional role in developing these tumors, and other metastases. We¿d probably want to do these functional and epidemiological studies before exploring the inhibitor compounds.

¿We have filed a patent application very recently,¿ Kinzler allowed, ¿covering the use of PRL-3 in the diagnosis, prognosis and treatment of metastatic disease.¿