Fragile histidine triad gene and skin cancer

ARTICLE

Several lines of evidence indicate that tumorigenesis is a multistep process and that these steps reflect molecular alterations that lead to the transformation of normal cells into their malignant counterparts. Initiation often occurs as an irreversible event due to the interaction of a tissue with carcinogens, supporting the idea that somatic mutations are the basis for cancer pathogenesis. Promotion, initially focused on the role of phorbol esters in skin carcinogenesis, was considered as a reversible process facilitating the expression of the initiated phenotype, and tumor progression was thought to represent further phenotypic alterations in initiated cells. Eventually it was thought that conversion from a pre-malignant phenotype to a malignant cell type is the major time-dependent stage of carcinogenic process whereas the acquisition of tumor heterogeneity and metastatic ability are relatively rapid events [1].

The basic understanding of the molecular pathology of cancer has been determined by the discovery of mutations that produce oncogenes with dominant gain of function and tumor suppressor genes (TSGs) with recessive loss of function. It was rapidly determined that these genes could be the targets for carcinogen-induced mutations in experimental models, and subsequently it was demonstrated that signature mutations occur in human cancers [1]. Experimental analysis revealed that oncogenes and TSGs involve critical pathways controlling cell growth, programmed cell death and differentiated functions. The importance of TSGs was first recognized from observations of hereditary and spontaneous human cancers. The high frequency of inactivating mutations in these loci resulted in the creation of experimental models to study their mechanism of action.

The discovery of human "cancer genes", oncogenes and tumor suppressor genes, has allowed the genetic manipulation of the mouse as a profound test that the predicted genetic alterations, when placed in the mouse would indeed produce cancer. Perhaps the least artificial models for the development of human cancer using genetically manipulated mice employ the reactivation of oncogenes and the inactivation of tumor suppressor genes. These models are least artificial, particularly for tumor suppressor genes because there are compelling human genetic data on the role of the loss of function of these genes in human cancer development and because of the fact that the heterozygous mouse has been created to model the human who carries a heterozygous germline inactivating mutation and who is thereby predisposed to develop cancer [2].

The FHIT tumor suppressor gene

In 1979, an Italian-American family was observed to be transmitting a constitutional reciprocal t(3;8)(p14.2;q24) chromosome translocation [3, 4], which segregated in the family with early onset, bilateral and multifocal renal cell carcinoma (RCC). Interestingly, the breakpoint at 3p14.2, interrupted the third intron of the FHIT gene, inactivating one of the two FHIT alleles [5].

The large (2.5 Mb, [5, 6]) FHIT gene is composed of 10 exons, of which five are protein-coding (exons 5 through 9); it encodes a small mRNA (1.1 kb) and a small protein (16.8 kd) of the histidine triad family of nucleotide-binding proteins [7], and at this time it is the only example of a TSG located at a chromosomal fragile region [5].The name FHIT comes from Fragile site of the HIstidine Triad family. Fragile sites are chromosome regions that reveal cytogenetically detectable gaps after exposure of cells to specific reagents. The distribution of common fragile sites parallels the positions of neoplasia-associated chromosomal rearrangements, prompting the proposal that fragility disposes to chromosomal rearrangements [8]. Implicit to this hypothesis is that oncogenes or TSGs at fragile sites are altered by chromosome rearrangements and thus contribute to clonal expansion of the neoplastic cells [9]. Thus the observation that the FHIT locus contains the most inducible common fragile site of the human genome, FRA3B, represents the first support for such speculation.

Expression of FHIT mRNA is detectable in most tissues, and the highest levels of expression of FHIT mRNA and protein are detectable in epithelial cells and tissues [10]. The FHIT gene is altered by deletion or translocation in a large fraction of many types of cancer, including lung, cervical, gastric, and pancreatic [5, 11-16] and less frequently by methylation [17]. Fhit protein is lost or reduced in the majority of these cancers, and in a large fraction of other cancer types [18-20]. Since both FHIT alleles are frequently altered in human cancers and since a family with hereditary cancer (RCC) associated with a translocation disrupting one FHIT allele has been described, it is reasonable to consider FHIT a bona fide TSG [5]. To demonstrate suppressor activity, the human FHIT cDNA has been transfected into four different tumor cell lines with homozygous deletions of the FHIT gene and then the Fhit-expressing transfectants were injected into nude mice, showing that Fhit expression results in the loss of the ability to form tumors [21].

The FHIT gene and the skin

FHIT gene abnormalities have so far been investigated in two human skin tumor types. Merkel cell carcinoma (MCC) is a rare neuroendocrine carcinoma of the skin which shares several features with small cell lung carcinoma (SCLC). Previously a high frequency of abnormalities of the FHIT gene in SCLCs was reported [11]. Fifty-seven per cent of MCCs displayed abnormal FHIT products that lacked three or more exons of the FHIT gene and the pattern of abnormal transcripts was similar to that observed in SCLCs [22]. MCC tumors frequently coexist with basal or squamous cell carcinoma of the skin. Other investigators [23] evaluated the role of the FHIT gene in non-melanoma skin cancer by screening for deletions in 16 tumors including basal cell carcinomas, squamous cell carcinomas and actinic keratoses. A normal transcript was found to be expressed in 14 of 16 tumors suggesting that the FHIT gene is not a very common target in human non-melanoma skin cancer, although no immunohistochemistry studies have been carried out, and the presence of normal FHIT transcripts in human skin cancers could be due to contamination of normal human cells.

When the DNA sequence of a human TSG is known, a common approach to study its role in cancer development is to isolate the mouse counterpart and eliminate its function in a mouse model. The murine Fhit locus is similar to its human homolog, encompasses a common fragile site, and is altered in murine cancer cell lines [24, 25]. Recently, we knocked out the murine Fhit gene and have established a strain of Fhit +/- mice. A full 100% of these mice developed tumors when given N-nitrosomethylbenzylamine (NMBA) intragastrically, compared with 25% of the treated Fhit +/+ mice. Because the only genetic difference between the Fhit +/+ and +/- mice is the targeted Fhit allele in the +/- mice, we believe that the second Fhit allele is the gatekeeper in tumor development. By ten weeks after NMBA exposure, the spectrum of visceral and skin tumors developed by all the Fhit +/- mice was similar to those observed in a rare human familial skin cancer syndrome [26].

In 1967, Muir [27] and Torre [28] described independently a patient with both sebaceous gland tumors and intestinal malignancies. Since then, more than 150 cases have been reported with Muir-Torre syndrome (MTS). MTS is defined as the presence of: 1) a sebaceous gland benign or malignant tumor; and 2) an internal malignancy. Internal malignancies that are most often seen in MTS patients are colorectal carcinomas and genitourinary tract neoplasms predominantly originating from the bladder, uterus, renal pelvis or ovaries, which together account for about 75% of the observed internal malignancies in MTS. Furthermore, breast, hematological, head and neck, and small intestinal malignancies have been described in combination with sebaceous gland tumors [29]. Because a subset of patients appeared to have hereditary nonpolyposis colorectal cancer (HNPCC), a relation between HNPCC and MTS was first suggested in 1981 [30]. HNPCC is characterized by an autosomal dominantly inherited predisposition to the development of colorectal cancer or specific extracolonic cancers, such as endometrial or urothelial carcinomas. HNPCC is caused by an inherited germ-line mutation in one allele of mismatch repair (MMR) genes. The MMR system repairs small errors including those affecting repeat sequences of the DNA (microsatellites), which occur during replication. Consequently, MMR deficiency results in microsatellite instability (MSI). Carcinomas of HNPCC patients show MSI [31]. Molecular genetic studies in MTS patients have shown MSI in both sebaceous gland tumors and colorectal cancer [32]. In addition, germ-line mutations in the MMR genes MSH2 and MLH1 have been described in MTS patients, further indicating that MTS might be an expression variant of HNPCC [33].

Of note, absence of MSI was found in 31% [34] and 54% [32] of sebaceous gland carcinomas of MTS patients, suggesting that another molecular genetic mechanism might lead to the MTS phenotype. Based on clinical and genetic differences, these two groups of investigators suggested different subgroups of MTS patients. The data points to two variants of MTS: I) MSI-positive variant, that we could call "caretaker variant" and that shares its pathophysiology and genetic cause with HNPCC, characterized by early age colorectal carcinoma and a strong family history of at least colorectal carcinoma; and II) MSI-negative variant, with late onset of cancer and a less pronounced family history, although the possibility that some cases of Muir-Torre syndrome may occur through the accidental coincidence of sebaceous gland neoplasm and internal malignancy could also explain the missing family history in some cases.

Interestingly, tumors developed by our Fhit +/- mice do not show MSI; thus, it is unlikely that the mouse syndrome involves MMR deficiency. All of the tumors were Fhit-negative when analyzed by immunohistochemical detection of Fhit protein expression, while the normal epithelial cells were Fhit-positive; consequently, loss of Fhit expression plays a role in murine MTS-like disease. Sebaceous tumor sections from two human MTS cases were also analyzed for expression of human Fhit. Fhit protein was detected in normal human sebaceous gland. The protein was not expressed, however, in two human sebaceous tumors from one case, but was expressed in the sebaceous tumor from the other case [26]. Furthermore, although NMBA treatment increases the frequency of occurrence of sebaceous and gastric tumors, sebaceous and lymphoid tumors and gastric papillomas do spontaneously occur with later onset (Fig. 1 and unpublished data) even though the full spectrum of tumors that develop spontaneously in Fhit-deficient mice is not yet completely known.

Two sets of mice with MTS phenotype are now available: Fhit-deficient mice with a mutation in a gatekeeper TSG and Msh2-deficient mice with a mutation in a caretaker TSG. Msh2-deficient mice developed lymphomas and intestinal tumors with high frequency. Some animals (7%) developed a variety of skin neoplasms, with sebaceous tumors and keratoacanthoma-like squamous neoplasms, analogous to the MTS. Differently from the Fhit-deficient mice, that cutaneously developed only tumors of sebaceous glands with a 9-fold higher frequency [26], Msh2-deficient mice showed MSI in all tumor types, but rarely in normal tissues [35]. Possibly, crossing Fhit-deficient with Msh2-deficient mice could lead to increased frequency of sebaceous and other tumors with the involvement of two different pathways (Table I).

Toward an apoptotic function of Fhit in the skin?

It has been shown that Fhit protein is an Ap3A (diadenosine triphosphate) hydrolase that cleaves Ap3A into adenosine 5'-diphosphate and AMP [36]. The tumor-suppressing function of Fhit does not depend on cleavage of Ap3A [21]. In fact a mutant Fhit protein, in which the middle histidine of the histidine triad was changed to asparagines, lost the enzymatic activity and still suppressed tumorigenicity. This observation indicates that the ability to cleave Ap3A is not required for tumor suppression [21]. Successive experiments indicate that the mutant Fhit protein binds Ap3A as well as the wild-type protein, which suggests that the Ap3A bound form of Fhit may be the active suppressor [37].

The Fhit gene was also cloned on the Drosophila salivary gland chromosome 3. The sequence of the Fhit gene and its encoded protein revealed, however, that the fly Fhit protein has a stretch of 314 amino acids added to the amino terminus [38]. By taking advantage of this information, human and mouse homologs of the Drosophila DNA stretch were cloned and a gene homologous to bacterial and plant nitrilases (Nit) was discovered. This gene, designated NIT 1, is independent of FHIT on human and mouse chromosome, but is fused with Fhit in Drosophila and the worm Caenorhabditis elegans, presumably displaying dual enzymatic activities. Other proteins with multiple enzymatic activities have been discovered that are chimeric in one species and encoded by two or more different genes in other species. Biology teaches us that, in these cases, those genes are involved in the same biochemical pathway. Thus, Fhit and Nit should act together. Interestingly a coordinate expression of Nit and Fhit has been observed in different mouse and human tissues [38].

Cancer cells that are deficient in Fhit are defective in programmed cell death [39, 40] but the mechanism of action of Fhit in apoptosis is still unclear. It is known that interferons are signals for cell cycle arrest and programmed cell death, induce accumulation of Ap3A [41], and are possible regulators of the Fhit activity. Induction of gene transcription is an essential part of the cellular response to interferons. One of the genes stimulated by interferons alpha and gamma in cultured cells is an enzyme involved in protein synthesis, the tryptophanyl-tRNA synthetase [42]. Ap3A formation in response to interferons is catalysed by an excessive amount of this tRNA synthetase [41].

Brenner et al. [43] proposed a model for Fhit-Ap3A and Nit function in proapoptotic tumor suppression. According to this model, signals like interferons cause tRNA synthetase to produce Ap3A rather than deliver amino acids to tRNA and Fhit-Ap3A complexes would then activate a proapoptotic activity of Nit proteins. Recently it has also been reported that the treatment with interferon alpha and retinoids seems to be of promise to prevent tumor development in human MTS [44].

In conclusion, if human and mouse MTS cases arise through similar mechanisms and if Fhit inactivation is a frequent pathway to MTS, we suggest that Fhit protein expression could underlie a proapoptotic mechanism, defective in this syndrome, and our Fhit-deficient mice could be an animal model to investigate the additional steps involved in human MTS and to identify other important suppressor and modifier genes involved in the development of this cancer syndrome. Studies are also needed to clarify the involvement of FHIT in at least some of human MSI-negative MTS cases. The future challenge is a better understanding of the physiologic role of Fhit and the consequences of its inactivation in the regulation of cell growth and apoptosis that, in turn, could be translated into a better treatment of Fhit-negative tumors and also the elimination or reduction of precancerous lesions, as a preventive measure.

REFERENCES

1. Yuspa SH. Overview of carcinogenesis: past, present, and future. Carcinogenesis 2000; 21: 341-4.

2. Klausner RD. Studying cancer in the mouse. Oncogene 1999; 18: 5249-52.

3. Cohen AJ, Li FP, Berg S, Marchetto DJ, Tsai S, Jacobs SC, Brown RS. Hereditary renal-cell carcinoma associated with a chromosomal translocation. N Engl J Med 1979; 301: 592-5.

4. Wang N, Perkins KL. Involvement of band 3p14 in t(3;8) hereditary renal carcinoma. Cancer Genet Cytogenet 1984; 11: 479-81.

5. Ohta M, Inoue H, Cotticelli MG, Kastury K, Baffa R, Palazzo J, Siprashvili Z, Mori M, McCue P, Druck T, Croce CM, Huebner K. The FHIT gene, spanning the chromosome 3p14.2 fragile site and renal carcinoma-associated t(3;8) breakpoint, is abnormal in digestive tract cancers. Cell 1996; 84: 587-97.

6. Mimori K, Druck T, Inoue H, Alder H, Berk L, Mori M, Huebner K, Croce CM. Cancer-specific chromosome alterations in the constitutive fragile region FRA3B. Proc Natl Acad Sci USA 1999; 96: 7456-61.

7. Brenner C, Garrison P, Gilmour J, Peisach D, Ringe D, Petsko GA, Lowenstein JM. Crystal structures of Hint demonstrate that histidine triad proteins are GalT-related nucleotide-binding proteins. Nat Struct Biol 1997; 4: 231-8.

8. Popescu NC. Chromosome fragility and instability in human cancer. Crit Rev Oncog 1994; 5: 121-40.

9. Yunis JJ, Soreng AL. Constitutive fragile sites and cancers. Science 1984; 226: 1199-204.

10. Druck T, Hadaczek P, Fu TB, Ohta M, Siprashvili Z, Baffa R, Negrini M, Kastury K, Veronese ML, Rosen D, Rothstein J, McCue P, Cotticelli MG, Inoue H, Croce CM, Huebner K. Structure and expression of the human FHIT gene in normal and tumor cells. Cancer Res 1997; 57: 504-12.

11. Sozzi G, Veronese ML, Negrini M, Baffa R, Cotticelli MG, Inoue H, Tornielli S, Pilotti S, DeGregorio L, Pastorino U, Pierotti MA, Ohta M, Huebner K, Croce CM. The FHIT gene at 3p14.2 is abnormal in lung cancer. Cell 1996; 85: 17-26.

12. Hendricks DT, Taylor R, Reed M, Birrer MJ. FHIT gene expression in human ovarian, endometrial, and cervical cancer cell lines. Cancer Res 1997; 57: 2112-5.

13. Greenspan DL, Connolly DC, Wu R, Lei RY, Vogelstein JT, Kim YT, Mok JE, Munoz N, Bosch FX, Shah K, Cho KR. Loss of FHIT expression in cervical carcinoma cell lines and primary tumors. Cancer Res 1997; 57: 4692-8.

14. Baffa R, Veronese ML, Santoro R, Mandes B, Palazzo JP, Rugge M, Santoro E, Croce CM, Huebner K. Loss of FHIT expression in gastric carcinoma. Cancer Res 1998; 58: 4708-14.

15. Simon B, Bartsch D, Barth P, Prasnikar N, Munch K, Blum A, Arnold R, Goke B. Frequent abnormalities of the putative tumor suppressor gene FHIT at 3p14.2 in pancreatic carcinoma cell lines. Cancer Res 1998; 58: 1583-7.

16. Sorio C, Baron A, Orlandini S, Zamboni G, Pederzoli P, Huebner K, Scarpa A. The FHIT gene is expressed in pancreatic ductular cells and is altered in pancreatic cancers. Cancer Res 1999; 59: 1308-14.

17. Tanaka H, Shimada Y, Harada H, Shinoda M, Hatooka S, Imamura M, Ishizaki K. Methylation of the 5' CpG island of the FHIT gene is closely associated with transcriptional inactivation in esophageal squamous cell carcinomas. Cancer Res 1998; 58: 3429-34.

18. Hadaczek P, Siprashvili Z, Markiewski M, Domagala W, Druck T, McCue PA, Pekarsky Y, Ohta M, Huebner K, Lubinski J. Absence or reduction of Fhit expression in most clear cell renal carcinomas. Cancer Res 1998; 58: 2946-51.

19. Ingvarsson S, Agnarsson BA, Sigbjornsdottir BI, Kononen J, Kallioniemi OP, Barkardottir RB, Kovatich AJ, Schwarting R, Hauck WW, Huebner K, McCue PA. Reduced Fhit expression in sporadic and BRCA2-linked breast carcinomas. Cancer Res 1999; 59: 2682-9.

20. Van Heerden WF, Swart TJ, van Heerden MB, van Rensenburg EJ, Engelbrecht S, Dreyer L, Huebner K. Immunohistochemical evaluation of Fhit protein expression in oral squamous cell. J Oral Pathol Med 1999; 28: 433-7.

21. Siprashvili Z, Sozzi G, Barnes LD, McCue P, Robinson AK, Eryomin V, Sard L, Tagliabue E, Greco A, Fusetti L, Schwartz G, Pierotti MA, Croce CM, Huebner K. Replacement of Fhit in cancer cells suppresses tumorigenicity. Proc Natl Acad Sci USA 1997; 94: 13771-6.

22. Sozzi G, Alder H, Tornielli S, Corletto V, Baffa R, Veronese ML, Negrini M, Pilotti S, Pierotti MA, Huebner K, Croce CM. Aberrant FHIT transcripts in Merkel cell carcinoma. Cancer Res 1996; 56: 2472-4.

23. Sikkink SK, Rehman I, Rees JL. Deletion mapping of chromosome 3p and 13q and preliminary analysis of the FHIT gene in human nonmelanoma skin cancer. J Invest Dermatol 1997; 109: 801-5.

24. Pekarsky Y, Druck T, Cotticelli MG, Ohta M, Shou J, Mendrola J, Montgomery JC, Buchberg AM, Siracusa LD, Manenti G, Fong LY, Dragani TA, Croce CM, Huebner K. The murine Fhit locus: isolation, characterization, and expression in normal and tumor cells. Cancer Res 1998; 58: 3401-8.

25. Glover TW, Hoge AW, Miller DE, Ascara-Wilke JE, Adam AN, Dagenais SL, Wilke CM, Dierick HA, Beer DG. The murine Fhit gene is highly similar to its human orthologue and maps to a common fragile site region. Cancer Res 1998; 58: 3409-14.

26. Fong LYY, Fidanza V, Zanesi N, Lock LF, Siracusa LD, Mancini R, Siprashvili Z, Ottey M, Martin SE, Druck T, McCue PA, Croce CM, Huebner K. Muir-Torre-like syndrome in Fhit-deficient mice. Proc Natl Acad Sci USA 2000; 97: 4742-7.

27. Muir EG, Bell AJY, Barlow KA. Multiple primary carcinomata of the colon, duodenum, and larynx associated with kerato-acanthomata of the face. Br J Surg 1967; 54: 191-5.

28. Torre D. Society transactions: New York Dermatological Society, Oct 24, 1967 (multiple sebaceous tumors). Arch Dermatol 1968; 98: 549-51.

29. Schwartze RA, Torre DP. The Muir-Torre syndrome: a 25-year retrospect. J Am Acad Dermatol 1995; 33: 90-104.

30. Lynch HT, Lynch PM, Pester J, Fusaro RM. The cancer family syndrome: rare cutaneous phenotypic linkage of Torre's syndrome. Arch Intern Med 1981; 141: 607-11.

31. Lengauer C, Kinzler KW, Vogelstein B. Genetic instabilities in human cancers. Nature 1998; 396: 643-9.

32. Honchel R, Halling KC, Schaid DJ, Pittelkow M, Thibodeau SN. Microsatellite instability in Muir-Torre syndrome. Cancer Res 1994; 54: 1159-63.

33. Kruse R, Rutten A, Lamberti C, Hosseiny-Malayeri HR, Wang Y, Ruelfs C, Jungck M, Mathiak M, Ruzicka T, Hartschuh W, Bisceglia M, Friedl W, Propping P. Muir-Torre phenotype has a frequency of DNA mismatch-repair-gene mutations similar to that in hereditary nonpolyposis colorectal cancer families defined by the Amsterdam criteria. Am J Hum Genet 1998; 63: 63-70.

34. Entius MM, Keller JJ, Drillenburg P, Kuypers KC, Giardiello FM, Offerhaus GJA. Microsatellite instability and expression of hMLH-1 and hMSH-2 in sebaceous gland carcinomas as markers for Muir-Torre syndrome. Clin Cancer Res 2000; 6: 1784-9.

35. Reitmar AH, Redston M, Cai JC, Chuang TCY, Bjerknes M, Cheng H, Hay K, Gallinger S, Bapat B, Mak TW. Spontaneous intestinal carcinomas and skin neoplasms in Msh2-deficient mice. Cancer Res 1996; 56: 3842-9.

36. Barnes LD, Garrison PN, Siprashvili Z, Guranowski A, Robinson AK, Ingram SW, Croce CM, Ohta M, Huebner K. Fhit, a putative tumor suppressor in humans, is a dinucleoside 5',5'''-P-1,P-3-triphosphate hydrolase. Biochemistry 1996; 35: 11529-35.

37. Pace HC, Garrison PN, Robinson AK, Barnes LD, Draganescu A, Rosler A, Blackburn GM, Siprashvili Z, Croce CM, Huebner K, Brenner C. Genetic, biochemical, and crystallographic characterization of Fhit-substrate complexes as the active signaling form of Fhit. Proc Natl Acad Sci USA 1998; 95: 5484-9.

38. Pekarsky Y, Campiglio M, Siprashvili Z, Druck T, Sedkov Y, Tillib S, Draganescu A, Wermuth P, Rothman JH, Huebner K, Buchberg AM, Mazo A, Brenner C, Croce CM. Nitrilase and Fhit homologs are encoded as fusion proteins in Drosophila melanogaster and Caenorhabditis elegans. Proc Natl Acad Sci USA 1998; 95: 8744-9.

39. Ji L, Fang B, Yen N, Fong K, Minna JD, Roth JA. Induction of apoptosis and inhibition of tumorigenicity and tumor growth by adenovirus vector-mediated fragile histidine triad (FHIT) gene overexpression. Cancer Res 1999; 59: 3333-9.

40. Sard L, Accornero P, Tornielli S, Delia D, Bunone G, Campiglio M, Colombo MP, Gramegna M, Croce CM, Pierotti MA, Sozzi G. The tumor-suppressor gene FHIT is involved in the regulation of apoptosis and in cell cycle control. Proc Natl Acad Sci USA 1999; 96: 8489-92.

41. Vartanian A, Narovlyansky A, Amchenkova A, Turpaev K, Kisselev L. Interferons induce accumulation of diadenosine triphosphate (Ap3A) in human cultured cells. FEBS Letters 1996; 381: 32-4.

42. Fleckner J, Rasmussen HH, Justesen J. Human interferon gamma potently induces the synthesis of a 55-kDa protein (gamma 2) highly homologous to rabbit peptide chain release factor and bovine tryptophanyl-tRNA synthetase. Proc Natl Acad Sci USA 1991; 88: 11520-4.

43. Brenner C, Bieganowski P, Pace HC, Huebner K. The histidine triad superfamily of nucleotide-binding proteins. J Cell Physiol 1999; 181: 179-87.

44. Graefe T, Wollina U, Schulz HJ, Burgdorf W. Muir-Torre syndrome. Treatment with isotretinoin and interferon alpha-2a can prevent tumor development. Dermatology 2000; 200: 331-3.