Genetics

Edited by: Paul Smaglik S.C. Chandrasekharappa, S.C. Guru, P. Manickam, S. Olufemi, F.S. Collins, M.R. Emmert-Buck, L.V. Deblenko, Z. Zhuang, I.A. Lubensky, L.A. Liotta, J.S. Crabtree, Y. Wang, B.A. Roe, J. Weisemann, M. Boguski, S.K. Agarwal, M.B. Kester, Y.S. Kim, C. Heppner, Q. Dong, A.M. Spiegel, A.L. Burns, S.J. Marx, "Positional cloning of the gene for Multiple Endocrine Neoplasia-Type I," Science, 276:404-7, 1997. (Cited in more than 200 papers since publication) Francis S. Collins Comments by Francis S. Collins, director of the National Human Genetics Research Institute (NHGRI), Settara C. Chandrasekharappa, staff scientist at the genetics and molecular biology branch at the NHGRI, Allen M. Spiegel, director of intramural research at the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), and Stephen J. Marx, chief of the genetics and endocrinology section of the metabolic diseases branch at NIDDK A misleading map. Inaccurate data. Red herrings. Just plain bad luck. Those factors made the search for the tumor suppressor gene multiple endocrine neoplasia-type I (MEN1) daunting. But hard work and fruitful collaboration made it possible. The hunt began after a Swedish group suggested that MEN1 was a tumor suppressor gene and provided a general location for it.1 That first-generation chromosome location (near the centromere of chromosome 11, at band 11Q13) attracted the attention of an NIDDK group at the lab of the late Gerald D. Aurbach, to whom this paper was dedicated. "There was a flurry of work, including a confirmation by Steve Marx of their chromosomal localization," recalls Allen Spiegel.2 An NIDDK paper was the first of several to show loss of heterozygosity in tumors striking the endocrine system, suggesting that the loss of a gene in that location might also spur on sporadic tumors.3 Although the NIDDK group had collected much DNA from many samples of the tumor types and MEN1 families, it couldn't find the suspect gene without help. "We are not a positional cloning lab. We couldn't carry that work further." So Francis Collins' group joined forces to help with the labor-intensive positional cloning. A bioinformatics group, led by Mark Boguski, senior investigator at the computational biology branch of the National Center for Biotechnology Information, teamed up to manage the data. Collins and Settara C. Chandrasekharappa (Chandra) were initially optimistic that they could use the NIDDK tumor library to narrow down the 10-megabase interval, which was considered quite large. Still, they thought that with the aid of the existing map and its few markers they could be successful. Then they learned that the physical map contained errors. "The published order of markers was incorrect," Collins notes. "Having encountered similar difficulties in previous projects, we decided we'd better build our own map. We had to generate a lot of new markers in order to be sure the map was right." To generate those markers, the team reached into a grab bag of tricks: applying shotgun sequencing by collaborating with Bruce Roe at the University of Oklahoma; drawing from various databases generated by the Human Genome Project, "which in 1995 was good, but not fabulous"; acquiring yeast artificial chromosomes (YACs) from a Massachusetts Institute of Technology effort; deriving smaller clones from the YACs; pitting those smaller clones against radiation hybrids; and finally, confirming the map using fluorescence in situ hybridization (FISH)--"every trick in the book to try to really nail down the anatomy of this interval," Collins notes. "Once we had it nailed down, we then looked at all of these families to try to narrow the interval [further], simply by looking for critical recombinants, as one does in this business. That got us down to something in the neighborhood of two or three [megabases]." Chandra recalls the time-consuming process: "We started using all these physical mapping libraries, which allowed us to put together a contig, isolate the genetic markers, then order those markers. A paucity of material in terms of markers and YAC clones for this region delayed it." A fourth NIH team, pathologists from the National Cancer Institute (NCI) led by Lance Liotta, director of NCI's laboratory of pathology, used those markers as they were generated to screen familial tumor libraries for loss of heterozygosity. But even with innovative microdissection methods they were stymied. "In many cases, they weren't very helpful, because whole arms of chromosome 11 were lost," Spiegel notes. So they made what he describes as a "leap of faith"--turning to sporadic tumors instead of familial ones. "The leap of faith is, it's a little bit of circular reasoning. You are assuming that the loss is pathogenic, that even though this is not a bona fide familial MEN1, it's going to be driven by the same gene." But the assumption proved fruitful. Screening one sporadic gastronoma resulted in finding a small interstitial deletion, rather than the loss of an entire arm of the chromosome. "That got us down to a much more manageable interval, about half a megabase, which you could imagine exhaustively looking through to find the responsible gene," Collins states. He thought that the next step--finding the actual gene--would be easy. Previous gene hunts presented a pattern--the larger genes in a candidate interval were usually the culprits. "In this case, the pattern was dead wrong," he notes. "We ended up characterizing about 25 candidate genes before we found the right one. Each one of those was a fairly painful experience." Spiegel, too, recalls the early misfires. "As the candidates came out, we were not lucky." After rejecting those 25 candidates, they finally found one with the long-sought mutations in MEN1 families. The researchers agree that the paper has received many citations because MEN1 was one of few major unidentified tumor suppressor genes. "This is a particularly intriguing human familial cancer syndrome where the tumors arise in endocrine glands," Collins notes. "So it's always held a certain fascination from a clinical point of view." He adds that tumor suppressor papers generally tend to attract references. "MEN1 is a classic tumor suppressor and the assumption is that uncovering its biology will explain not only the familial form of the disease, but maybe cancers that arise in these same tissues on a sporadic basis." Marx agrees that hereditary tumor suppressor genes usually turn out to be more important than their rare syndromes would suggest. Examples include the common tumor suppressors p53 and Rb. "Within endocrinology, MEN1 was the first and perhaps most common familial syndrome of endocrine excess." Since isolating MEN1, Marx's lab applied similar methodology to conduct mutational analyses in 50 families archived by NIDDK. They detected a variety of mutations to MEN1, including nonsense codons, frameshift truncations, and missense mutations in many tumor types. In all, they reported MEN1 mutations in 47 of 50 in the NIDDK archives.4 In another analysis, they looked for the MEN1 mutations that they had predicted a decade earlier in sporadic parathyroid adenoma, finding such mutations in about 20 percent of those tumors.5 "Our NIH collaboration and others have extended these observations to establish that MEN1 is the gene most commonly mutated in common-variety parathyroid adenomas, gastronomas, insulinomas, and bronchial carcinoids." They also learned that menin, the protein encoded by the MEN1 gene, is a nuclear protein.6 Most recently, they reported that menin binds to and inhibits JunD, part of the jun/fos transcription factor family.7 "This establishes an interesting nuclear transcriptional role for menin, opening up a pathway that has not previously been implicated in hereditary human neoplasia," Marx suggests. Spiegel agrees that the protein's transcriptional role appears especially intriguing; while c-Jun and JunB promoted tumor growth, JunD does not. "It's actually antiproliferative. It's found when the cells are not dividing. It's found when the cells are static." He adds that menin interacts only with JunD, not with other members of the jun/fos family. "There may be something intrinsic to the fact that these two guys tie up that has the opposite effect of c-Jun," he speculates. Indeed, despite the flurry of research since the paper went to press, much about the gene and its protein remains mysterious, notes Chandra. "We don't know much about the function or the structure, because the sequence doesn't resemble anything known at the moment." Spiegel agrees that the 610-amino acid protein had no obvious signature like, say, that of a phosphatase. Indeed, the gene lacks structural clues that would reveal the protein's biochemical activity. That may change soon. Knockouts could help place the gene in a pathway occupied by other known genes. They could also identify other steps in the pathway by which a normal cell becomes neoplastic, and perhaps also catalog other modifier genes that might help explain the variability of the disease in humans. At some point, the activation or inactivation of the gene could be used, in concert with other genetic analyses, as a diagnostic and prognostic tool in the management of cancer. "A compelling new paradigm in cancer diagnosis and treatment is to define the precise mutations present in a particular tumor so that you can individualize therapy," Collins notes. Still, the therapeutic consequences of such stratification of tumors likely remain years away. C. Larsson et al., "Multiple endocrine neoplasia type-1 gene maps to chromosome-11 and is lost in insulinoma," Nature, 332:85-7, 1988. S.J. Bale et al., "Linkage analysis of multiple endocrine neoplasia type 1 with INT2 and other markers on chromosome 11," Genomics, 4:320-2, 1989. E. Friedman et al., "Clonality of parathyroid tumors in familial multiple endocrine neoplasia type 1," New England Journal of Medicine," 321:213-8, 1989. S.K. Agarwal et al., "Germline mutations of the MEN1 gene in familial multiple endocrine neoplasia type 1 and related states," Human Molecular Genetics, 6:1169-75, July 1997. C. Heppner et al., "Somatic mutation of the MEN1 gene in parathyroid tumours," Nature Genetics, 16:375-8, 1997. S.C. Guru et al., "Menin, the product of the MEN1 gene, is a nuclear protein," Proceedings of the National Academy of Sciences, 95:1630-4, 1998. S.K. Agarwal et al., "Menin interacts with the AP1 transcription factor JunD and represses JunD-activated transcription," Cell, 96:143-52, 1999.

Genetics

The Scientist Staff

Interested in reading more?

Become a Member of