People with the genetic disease Peutz–Jeghers syndrome have many intestinal polyps — benign tissue outgrowths. These seldom become malignant, and the reason may lie in the properties of the affected gene.
Several inherited human diseases are characterized by the formation of polyps in the gut. Polyps are benign outgrowths of tissue with a disordered structure. In some of these diseases, the polyps undergo transformation into malignant gastrointestinal tumours — but this occurs relatively rarely in Peutz–Jeghers syndrome. In an effort to understand why, Bardeesy and colleagues1 (page 162 of this issue) have produced mice carrying mutations in the mouse counterpart of the LKB1 gene, which is affected in the human syndrome. The authors' studies of these animals have thrown up some unexpected findings.
Cancer, a disease characterized by unregulated cell proliferation, is caused by the stepwise accumulation of mutations in certain key genes. Every cancer is different, displaying a unique constellation of genetic changes. Yet one can identify some common mutational targets and molecular themes that lie at the heart of the transformation of a normal cell into a potentially deadly cancer cell. Thus, certain genes such as RAS and p53 are often mutated in different cancers, suggesting that some deregulated cellular processes are common to many, if not all, tumours.
On the other hand, as our understanding of the molecular basis of tumour development increases, the richness of the mechanisms that have evolved to control cell growth is becoming more apparent. For example, studies of inherited mutations in tumour-suppressor genes (which usually keep the brakes on cell growth) show that some cell types are highly susceptible to malignant transformation, whereas others with the same mutations seem to be resistant. In addition, different cancer-predisposing mutations in the same cell type can have very different effects on malignancy.
Hereditary polyposis syndromes illustrate these phenomena2 (Fig. 1). Intestinal polyps are generally composed of two cell types, epithelial and stromal cells; malignant tumours derived from such polyps consist mainly of transformed epithelial cells. Patients with inherited mutations in the APC gene develop familial adenomatous polyposis, which is characterized by numerous epithelial-rich polyps that have a propensity to progress to malignant cancer. By contrast, patients with inherited inactivating mutations in SMAD4 or BMPR1A develop juvenile polyposis, and people with LKB1 mutations develop Peutz–Jeghers syndrome (PJS). Both syndromes are characterized by relatively benign, stromal-rich polyps called hamartomas. In juvenile polyposis, the inactivation of SMAD4 is limited to the stromal cells, so the epithelial cells are influenced only indirectly3 (Fig. 1). But the limited malignant potential of polyps in PJS patients must have a different cause, because LKB1 inactivation occurs in the epithelial, not the stromal, component4. Bardeesy et al.1 now provide the first clues to why LKB1 mutations in the intestinal epithelium produce polyps that rarely become malignant.
To investigate the role of LKB1 in human PJS, several groups5,6,7 have generated mice that lack the relevant mouse gene, Lkb1. In all cases, mice lacking both copies of the gene (homozygous animals) die as embryos. Heterozygous mice (with one deleted and one normal copy of the gene) survive, but develop gastrointestinal hamartomas with features akin to those in patients with PJS or juvenile polyposis. These animals are valuable mouse models of PJS. But the molecular basis for the limited malignant potential of PJS polyps has remained a conundrum, as the early death of the homozygous embryos prevents cell lines from being isolated and studied.
To tackle this issue, Bardeesy et al. generated mice that lacked one copy of Lkb1 entirely and also carried a 'conditional' copy of the gene, which is inactivated only under specific experimental conditions (so the animals could develop normally). The authors isolated mouse embryonic fibroblast cells (MEFs) from these animals, and then switched off the conditional Lkb1 gene in these cells, generating homozygous MEFs in culture.
MEFs are widely used for measuring two hallmarks of cancer: cell immortalization (the ability to multiply indefinitely in culture) and malignant transformation. The propensity of cultured normal MEFs to become immortal is limited by senescence — a programme of cellular 'ageing'. This crucial fail-safe mechanism helps to prevent transformation and is triggered by various stimuli, such as DNA damage, the expression of activated oncogenes (which promote cell growth), and stress induced by the tissue-culture process8.
Bardeesy et al. found that their homozygous MEFs were resistant to tissue-culture-induced senescence, and thereby became immortal, but still underwent senescence induced by DNA damage or oncogenes. This suggests that a lack of Lkb1 renders cells immortal by inhibiting or alleviating the culture-stress-induced activation of senescence. Importantly, however, these immortal MEFs also turned out to be resistant to transformation induced by potent combinations of oncogenes, such as activated RAS and SV40 large T antigen, which readily transform normal cells (Fig. 2). So, although the lack of Lkb1 causes immortalization, the immortalized cells seem to be resistant to subsequent transformation.
So Lkb1 deficiency is a double-edged sword, somehow promoting perpetual cell growth but preventing malignant transformation — and explaining why the polyps in PJS patients develop yet remain benign. But how does Lkb1 loss in the intestinal epithelium promote the development of stromal-rich polyps? Bardeesy et al. examined the genes expressed in the homozygous MEFs and in polyps from heterozygous mice, and detected marked increases in the expression of various genes encoding components of the extracellular matrix and secreted signalling molecules. So perhaps factors such as these, produced by Lkb1-deficient epithelial cells, influence the stromal cells in PJS polyps. In keeping with this idea, the authors observed that homozygous MEFs, and the medium in which these cells were cultured, affected the expression of specific genes in normal MEFs.
An unanswered question is why PJS patients are more likely than usual to develop other types of cancer. This seems odd, given that a lack of LKB1 prevents malignant transformation of intestinal polyps. Does this mean that LKB1-deficient cells are resistant to transformation by some activated oncogenes (such as members of the RAS pathway), but more susceptible to others? The low frequency of activating RAS mutations in tumours from PJS patients certainly hints at this possibility. Alternatively, the order of events may matter. For example, LKB1 inactivation in normal intestinal epithelium might protect those cells from malignant transformation, resulting in benign polyps. But in cell types that have already acquired mutations in other cancer genes, LKB1 loss might promote tumour progression. No doubt Bardeesy and colleagues' mouse model will prove valuable in addressing these and other issues.
Bardeesy, N. et al. Nature 419, 162–167 (2002).
Haggitt, R. C. & Reid, B. J. Am. J. Surg. Pathol. 10, 871–887 (1986).
Kinzler, K. W. & Vogelstein, B. Science 280, 1036–1037 (1998).
Wang, Z. J. et al. J. Pathol. 188, 613–617 (1999).
Ylikorkala, A. et al. Science 293, 1323–1326 (2001).
Miyoshi, H. et al. Cancer Res. 62, 2261–2266 (2002).
Jishage, K. et al. Proc. Natl Acad. Sci. USA 99, 8903–8908 (2002).
Sherr, C. J. Nature Rev. Mol. Cell Biol. 2, 731–737 (2001).
About this article
American Journal of Medical Genetics (2004)
Seminars in Cancer Biology (2004)
The Journal of Pathology (2003)