The clearest molecular evidence for the genetic instability hypothesis comes from the elucidation of the genes causing hereditary nonpolyposis colon cancer (HNPCC). Patients with HNPCC have an increased risk of tumor development over the course of their lifetime but do not display the widespread changes in the at-risk tissue cellular architecture characteristic of other inherited tumor syndromes, for instance, the thousands of polyps in familial adenomatous polyposis (FAP). Instead, they typically develop a single advanced primary tumor at an atypically young age (see Chap. 32 on HNPCC).
Indeed, the tissue specificity is even more of a mystery upon consideration of the underlying genetic defect in these patients. They inherit a mutation in one of the mismatch repair (MMR) genes, such as MSH2, MLH1, PMS1, PMS2, or GTBP(MSH6).4-11 Unlike classical tumor suppressor genes such as p53 or Rb, the MMR genes do not directly affect the growth or death of a tumor cell.12 Experimentally, this distinction can be seen upon reintroduction of a MMR gene into a tumor cell that has two mutant copies. In contrast to reintroduction of a classical tumor suppressor, there is no effect on the tumor cell growth or death.13
Instead, the loss of MMR genes imbues these tumors with an elevated nucleotide mutation rate—2 to 3 orders of magnitude higher than normal cells or MMR-proficient cancers of the same cell type14-16 . Thus, there is an increased rate of mutation at oncogene and tumor suppressor loci throughout the tumor cell genome. This link between isolated cellular genetic instability and organism-wide tumorigenesis is strong evidence for the genetic instability hypothesis, as sequence instability alone is able to drive the autosomal dominant inheritance of colorectal neoplasia in these families (Fig. 22-1).
Pathways to genetic instability. Different types of genetic instability require a different number of mutational “hits” in order to engender the respective instability phenotype. In a heterozygote with one defective nucleotide excision repair (NER) allele (step 1), inactivation of the normal allele (step 2) does not immediately lead to mutations. It additionally requires exposure to an environmental agent (i.e., ultraviolet light) (step 3) to create large numbers of mutations (NER-related instability; NIN). In contrast, in a heterozygote with one defective mismatch repair (MMR) allele (step 1), all that is required to begin to develop mutations at a high rate (microsatellite instability; MIN) is the inactivation of the normal allele inherited from the unaffected parent (step 2). Cell fusion and other experiments suggest that chromosomal instability (CIN) can have a dominant quality.29 One example of a gene that can be mutated in a dominant negative manner to cause CIN is hBub1, a component of the mitotic spindle checkpoint (MSC).30 It apparently requires only a single mutational “hit” of such a gene to engender the CIN phenotype.
Defects of another major DNA repair system have been documented in tumors as well. Nucleotide-excision repair (NER) is responsible for repairing damage caused by many exogenous mutagens.17 Mutations in one of several different NER genes result in xeroderma pigmentosum and related disorders. Patients with these autosomal recessive, inherited diseases develop numerous skin tumors in sun-exposed areas (see also Chap. 28 on xeroderma pigmentosum).18,19 Surprisingly, skin tumors represent the major tumor type to which patients with NER defects are susceptible, and the incidence of internal cancers in these patients is not raised to the same degree20-22 . The simplest explanation for these results is that ultraviolet light is the major mutagen that results in NER-correctable DNA damage to which humans are exposed (Fig. 22-1).20-22
However, there is a more pervasive genomic abnormality of sporadic tumors. Virtually all epithelial solid tumors have structural and numeric chromosome variation—they are aneuploid.23 The only exceptions are, strikingly, the MMR deficient tumors, which remain diploid throughout tumor progression. Such observations have led to the suggestion that cancers require instability either at the sequence level or at the chromosomal level, but not generally at both levels.24 This logic would posit that one form of instability is sufficient to drive tumorigenesis. Consistent with this hypothesis, aneuploidy can be found in the earliest neoplastic lesions, such as benign adenomas of the colon, and the accumulation of aneuploid cells is a classic finding in advanced stages of tumorigenesis.25-28
Recent analysis has shown that cancer cell aneuploidy is a reflection of an underlying chromosomal instability (CIN).29 Quantitative studies of aneuploid tumor cell divisions have demonstrated that the chromosomal abnormalities in these cells are the result of an intrinsic segregation instability.30 This observation gave rise to the proposal that CIN could be considered the primary class of instability required for neoplastic progression in the majority of tumors.
The causes of CIN underlying the widespread aneuploid phenotype are just beginning to be investigated. One theory posits that karyotypic instability in cancer is a truism; it is a natural side effect of the malignant transformation process, driven by the preceding mutations in growth-controlling oncogenes and tumor suppressor genes such as ras or p53.
However, the existence of karyotypically stable MMR-deficient tumors argues against a causal role for classical oncogenes and tumor suppressors in the CIN phenotype. These tumors have mutations in the same oncogenes and tumor suppressor genes as CIN tumors, and have similar stage-specific growth and progression characteristics, but are not aneuploid. These cases prove that the mutant genes driving advanced neoplastic progression do not inevitability generate or require aneuploidy.
At the other end of the spectrum, a different theory proposes that aneuploidy is not caused by specific genetic alterations but instead results from the altered cellular architecture that ensues whenever an abnormal chromosome complement is present within cells. Thus, a chance abnormal division in an otherwise normal cell gives rise to a karyotypically abnormal daughter cell with a selective growth advantage compared to its neighbors. The abnormal number of chromosomes in this cell destabilizes the segregation machinery, auto-catalyzing chromosome missegregation and further aneuploidy. Aneuploidy begets aneuploidy.31
There is some recent evidence to support an alternative to these hypotheses. Perhaps aneuploid tumors sustain an early mutational event in a chromosome stability gene that drives chromosomal instability (Fig. 22-1). In some tumors, CIN has been proposed to be driven by mutations in mitotic checkpoint genes.30,32,33 For the majority of tumors, however, the molecular basis of CIN is not known yet. This is an area of active investigation. Many of the known inherited tumor suppressor genes seem to play an important role in genome stability but their mechanistic relationship to genome instability is poorly understood (see chapters 29, 30, 47 and 37 on ATM, BLM, BRCA1/BRCA2 and p53, respectively). It will be interesting to see how many instability genes can be shown to be altered in sporadic cancers.