Function of Mismatch Repair Genes
Mismatch repair (MMR) genes are involved in numerous cellular functions including: (1) repairing DNA synthesis errors; (2) repairing double-strand DNA breaks; (3) apoptosis; (4) antirecombination; and, surprisingly, (5) destabilization of DNA. These responsibilities make MMR proteins extremely important in the basic maintenance of the genetic material, the regulation of the cellular cycle, and in the last instance, development of an effective immune system. When MMR is lost or defective there is a decrease in apoptosis, an increase in cell survival, and a potential increase in damage-induced mutagenesis. This can provide a selective growth advantage to the cell, thus causing an increased susceptibility to tissue-specific cancers.
MMR Genes and Human Disease
The role of mismatch repair genes in human disease was first elucidated in 1993 when Fishel et al. and Leach et al. independently showed that mutations in the human homologue of the bacteria mutS gene (MSH2) predispose individuals to Lynch syndrome.
Shortly thereafter, a colorectal cancer susceptibility gene was identified on chromosome 3p by Bronner et al. and Papadopoulos et al. Both groups identified mutations in the mutL homologue, MLH1.
It wasn’t until 1997 that mutations in MSH6 were demonstrated to cause Lynch syndrome (Miyaki et al., 1997; Akiyama et al., 1997).
More recently, mutations in PMS2, have been shown to be an important cause of Lynch syndrome (Worthly et al., 2005; Hendriks et al, 2006)
It is estimated that 10-15% of CRC patients have a family history of CRC (Vasen, 2005). The most common inherited form of CRC is Lynch syndrome. This is an autosomal dominant cancer susceptibility syndrome characterized by predominantly right-sided (proximal) colon cancer. Additionally, tumours often develop in a variety of other sites, including the extra-colonic gastrointestinal tract, genitourinary tract, endometrium, ovaries, biliary tract, pancreas and brain (Lynch et al., 1993). All cancers associated with the syndrome occur at a young age of onset. As well, multiple tumours are often found in patients. Lynch syndrome accounts for between 1-5% of all CRC cases, depending on the population studied and the methods used to determine HNPCC status (Samowitz et al., 2001; Aaltonen et al., 1998).
Clinical criteria to identify Lynch syndrome families were established in 1990 by the International Collaborative Group on Hereditary Non-Polyposis Colorectal Cancer (ICG-HNPCC) and are now termed the Amsterdam criteria I (AC I) (Vasen et al., 1991). Subsequently, due to concern that these criteria were too restrictive, they were expanded to include extra-colonic tumours. These are known as Amsterdam criteria II (AC II) (Vasen et al., 1999). More recently, groups studying Lynch syndrome have observed that many families appear to be at high risk for hereditary cancer, but did not fulfill the criterion that at least one family member had developed cancer before 50 years of age. Therefore, they modified the age and cancer criteria of the Amsterdam classifications, creating more relaxed criteria (Woods et al., 2005; Quehenberger et al., 2005; Hampel et al., 2005).
Unfortunately, mutations are identified in only slightly more than 50% of families that are considered high risk. This paucity of identifiable deleterious alterations in these cases could be due to a number of factors: (1) the MMR genes are not being screened in a comprehensive manner; (2) not all the MMR genes are being screened; (3) mutations in non-MMR genes are the cause of the cancer in the families; (4) some families may have a chance clustering of cancers or have developed cancers primarily by environmental factors.
Screening of MMR Genes
The primary function of DNA MMR is to correct mismatches during DNA replication, thus maintaining genetic stability. One of the most recognized molecular events of defective MMR is the instability of repetitive DNA sequences called microsatellites. The resulting mutator phenotype is called microsatellite instability (MSI). MSI is tested for in the Lynch syndrome spectrum of tumours suspected of harbouring deleterious MMR gene variants.
The Bethesda guidelines were established in 1996 to determine which cases should be processed for further mutation analyses of the mismatch repair genes and to help in determining clinical management (Boland et al., 1998). A panel of five microsatellites was recommended as a reference panel for screening. Three MSI designations are used to indicate the level of microsatellite instability: (1) MSI-High (MSI-H), in cases when >30% of tested loci are unstable; MSI-Low (MSI-L), when <30% of tested loci are unstable; and microsatellite stability (MSS), when no tested loci are unstable. In most instances, MSI-L and MSS tumours lack familial MMR gene mutations. These guidelines were revised in 2002 to clarify a number of issues concerning the original Bethesda criteria and to further aid in the identification of families for additional testing (Umar et al., 2004).
Also, immunohistochemistry (IHC) can be used to determine the presence of the MMR proteins. If there is an absence or noticeably lower levels of MMR protein in the tumour tissue compared to normal tissue, subsequent molecular tests should be performed to determine the exact germline alteration. This is a more specific test for MMR deficiency, as compared to MSI testing, because it can pinpoint the exact gene that is mutated. Currently, MSH2, MLH1 and MSH6 are tested for by IHC regularly and testing for PMS2 is becoming increasingly prevalent. Also, IHC is a faster and more cost efficient test, although it is more subjective an assessment than MSI testing.
Once an MMR gene is identified that may harbour a deleterious alteration in a Lynch syndrome kindred, targeted mutational analyses can commence. Sometimes this is a two stage approach whereby mutations are identified indirectly by single stranded conformational polymorphism (SSCP) analysis or denaturing high performance liquid chromatography (dHPLC) and then altered DNA segments are sequenced. However, because of the large number of polymorphisms occurring in some of the MMR genes, direct sequencing of the genes is often immediately performed.
Approximately 90% of mutations identified in Lynch syndrome families are found in MLH1 and MSH2. The majority of the remaining 10% are found in MSH6. However, the significance of PMS2 has not been well assessed due to the lack of testing of the gene or protein, in relation to Lynch syndrome. It is very difficult to screen PMS2 for genomic alterations because of the high frequency of pseudogenes nearby on chromosome 7p, and antibodies to PMS2 were not readily available until recently.
What’s in a Name?
In 1913, Warthin, a pathologist at the University of Michigan, described a cancer family which had tumours occurring predominantly in the stomach, colon and uterus (Warthin, 19131). In this same study he also reported other ‘cancer fraternities’ he had recognized. A number of follow-up studies by Warthin and his colleagues further described these ‘cancer fraternities’ (Warthin, 19252; Hauser and Weller, 19363). Then in 1966, Lynch identified two large families that were phenotypically similar to those studied by Warthin (Lynch et al., 1966). By the late 1970s, the clinical description of these families, and those like it, were referred to as ‘cancer family syndrome’.
‘Lynch syndrome’ was first used in 1984 by Boland and Troncale to further refine the name to suggest that not all organs were involve. At this point, ‘Lynch syndrome I’ and ‘Lynch syndrome II’ were coined to distinguish families having only colorectal tumours from those having tumours in other organs, respectively. Henry Lynch himself first used the term hereditary non-polyposis colorectal cancer (HNPCC) in 1985, primarily to ensure that this clinical entity was not confused with familial adenomatous polyposis (FAP). Unfortunately, HNPCC also infers that only the colorectum is affected.
It wasn’t until 1993 that the molecular mechanisms causing this syndrome began to be elucidated. Thus, to this point, only the phenotypic aspects of the disease were considered when coining new terms to describe the syndrome. However, more recently, many in the field have suggested that Lynch syndrome is the most appropriate term and should denote those families which have a known MMR deficiency identified by MSI, IHC and/or by mutational analyses. It is now becoming apparent that other genetic mechanisms are the cause of hereditary colorectal cancer, as a number of studies have identified kindreds which fulfill Amsterdam I criteria but do not have a mismatch repair deficiency (Lindor et al, 2005; Woods et al., 2005; Mueller-Koch et al., 2005). Such families have been termed “Familial Colorectal Cancer – Type X’.
- Warthin AS. Heredity with reference to carcinoma. Arch Intern Med 1913; 12:546-55.
- Warthin AS. The further study of a cancer family. J Cancer Res 1925; 9:279-86.
- Hauser IJ, & Weller CV. A further report on the cancer family of Warthin. Am J Cancer 1936; 27:434-49.
Updated March 2006