For years, scientists had known that b-catenin acted at the cell membrane as an adhesion molecule. Then Hans Clevers' lab at Utrecht University in the Netherlands, which had spent years characterizing the Tcf set of transcription factors, reported that those transcription factors also interacted with b-catenin.4 Another group reported that a related transcription factor also interacted with the signaling protein.5 The sum of those findings showed that b-catenin, besides acting on the cell membrane, could act on the nucleus as a transcription factor. Those two papers also suggested that APC could mediate transcription. Then Kinzler's and Clevers' labs joined forces to illustrate APC/b-catenin activity in colon cancer, while Polakis' group examined that in melanoma. All three published different yet related papers in the same issue of Science. Polakis' paper emphasized the role of APC in regulating the stability—and there-fore levels and distribution—of b-catenin in the cell, while the Kinzler- and Clevers-led studies focused on the transcriptional activation mediated by b-catenin/Tcf and the ability of APC to control it. The three senior authors agree that the papers have received considerable attention because they examine the mechanisms behind one of the most frequently mutated tumor suppressor genes in humans. Although they have reached similar destinations, each group experienced a different trip, and each has since embarked on separate but related investigations. The three papers here focus on those scientific voyages, past and future.
With that discovery—and a clearer relationship established between APC and b-catenin—Clevers teamed up with Kenneth Kinzler's lab. The colleagues found that Tcf-4 was expressed in normal gut epithelium and in colorectal cancer cells. They then applied a reporter assay, which Clevers' team had developed earlier, to find which of the cloned Tcf factors were expressed in cancer samples that Kinzler's lab had collected. That assay ultimately showed that colorectal cancer cells without APC (APC-/-) contained catenin/Tcf-4 complexes in their nuclei. It also shows that reexpressing APC in those cells inhibited the reporter assay to normal levels. That assay led to a better understanding of how colorectal cancer emerges. "Colorectal cells express Tcf-4, a transcription factor [that] becomes active only upon binding tob-catenin," Clevers comments. "This occurs normally only during development upon signaling through the wnt pathway. In colorectal cancer, loss of APC (or mutation of b-catenin, reported in the Kinzler paper) inappropriately and constitutively activates Tcf-4, which then results in malignant transformation."
Since the 1997 paper, Clevers and colleagues have knocked out Tcf-4, showing that the protein maintains stem cells in the crypts of the intestine.2 "It is simple to see how the excess of Tcf-4 activity would lead to excessive expansion of stem cells and initiate polyp formation and cancer," Clevers notes. In the future, Clevers suspects the reporter assay may help find a drug that can inhibit Tcf-4 activation—and thus, colon cancer. "That could be the next generation of colon cancer therapeutics," Clevers remarks. However, specificity must first be addressed. "Is it possible to develop a compound that would inhibit b-catenin/Tcf-4 activity, and will such a corn-pound be a specific anticancer drug?"
"We next asked what part of the b-catenin sequence received the initial signal for its APC-mediated destruction and localized this to 89 N-terminal amino acids."2 When Polakis and colleagues presented that data to the National Cancer Institute (NCI), Polakis had a fortuitous encounter with an NCI scientist. "Paul Robbins told me he had found a mutation in the N-terminal sequence of b-catenin (Ser 37) in a melanoma cell while searching for potential vaccine antigens," Polakis recalls. "We checked the cell and found huge levels of b-catenin and then tested the mutation ourselves in recombinant b-catenin and found that it greatly stabilized the protein." Those observations led to their experiments in the melanoma cell line, which formed the basis for this paper. "The most appealing human interest aspect of our study was the rare event in which one scientist giving a seminar and another scientist listening can exchange thoughts in a manner that results in a highly relevant connection," Polakis comments.
Polakis stresses that his 1997 Science paper and his work prior to it emphasized the role of APC in regulating the stability—and therefore levels and distribution—of b-catenin in the cell. The Kenneth Kinzler and Hans Clevers work focused on transcriptional activation mediated by b-catenin/TCF and the ability of APC to control that process. The Polakis and Kinzler papers, however, found similar outcomes, but in different cell lines. Both reported tumor cells with mutated b-catenin that led to inappropriate activation of Tcf-4.
As a postscript, Polakis adds that the story circles back to its beginning, when both he and Kinzler reported b-catenin and APC bound together. At that point, their paths diverged for a time. Polakis had next reported regulation of b-catenin by APC, while Kinzler's group did not. "Quite to the contrary, they indicated they could not observe this effect in their paper on induction of apoptosis by APC."3 However, this diversion ended when their three Science papers were published.
"There are rare cancers that express full-length APC," Kinzler recalls thinking. "What's going on in these rare cancers?" To find out, Kinzler and colleagues examined two colorectal cancer cell lines that had intact copies of this tumor suppressor gene. "We measured the catenin-regulated transcription in these cell lines. Both of them expressed significant levels. So what's going on?"
The answer emerged when Hopkins researchers Patrice J. Morin and Andrew B. Sparks sequenced those rare tumor cell lines and found mutations in the regulatory domain located at the amino terminus of 3-catenin. "These regulatory domains resulted in a super-active form of b-catenin that was resistant to APC suppression," Kinzler comments. "The take-home message was, 'Here's a pathway that is elevated in colon cancer cells, that can be suppressed by wild-type APC; and in rare tumors that have APC, there [are] mutations in b-catenin that render it resistant to this suppression," Kinzler adds that the collaboration with Clevers' lab made both papers possible, since each shared factors and cooperated on creating assays.
Kinzler remarks that both papers help disentangle a complicated pathogenesis. "APC interacts with almost a dozen proteins. How do you know which of these interactions are important to its neoplastic process?" All three papers provide key genetic mutations that help point the way. Since these papers, many more such genetic guideposts have been found—all of which serve as potential points of therapeutic intervention. Kinzler's group has identified c-myc as a target gene of Tcf-4.1 Frank McCormick, founder of Onyx Pharmaceuticals in Richmond, Calif., and colleagues published another target of the pathway, the cyclin-D1 gene, which also regulates cell growth.2 Many more such targets likely exist. "c-myc and cyclin D1 are not going to be the only targets in this pathway. People are going to continue to identify what else b-catenin/Tcf-4XX1XX turns on and then the genes that those turn on. People will be defining this pathway for many years," Kinzler concludes.