Screening for Mutations in the Germline

Question 1

Why do we want to find cancer mutations? Who may benefit?

Answer 1

1) . Implications for patient:
- May have risk of other cancers - surgery/screening
- Treatment options - e.g. PARP inhibitors
- BRCA1/2 - breast and ovarian cancer
- HNPCC - colorectal and endometrial cancer risk

2) . Risk to relatives:
- Can offer predictive testing
- Prophylactic surgery/screening
- Relief/less screening (morbidity/financial)

3) . Where does the process start?
- ‘worried-well’ with FH may go to GP and then get referred to clinical genetics dept. Could also get referred to a family history clinic for triage.
- Affected individual with FH and early onset cancer may get referred to clinical genetics.

Question 2

How do we decide which families are going to benefit from mutation testing?

Answer 2

• Need a large enough family history of relevant tumours - e.g. breast/ovarian (BRCA1/2), colorectal/endometrial (HNPCC).
• Age of onset will be low if it is an inherited cancer.
• There are formulae to calculate the likelihood that an individual has a cancer mutation - funding cut offs.
• Phenotype may give indication e.g. FAP and multiple GI polyps.
• Biomarkers of tumours from individual/family that suggest that there is a predisposition. BRCA1 results in triple negative tumours, HNPCC gives microsatellite instability and loss of antibody staining in IHC.

Question 3

Discuss Familial Adenomatous Polyposis (FAP).

Answer 3

• In FAP there are usually >100-1000 polyps in the colon and rectum by the second decade of life.
• May also be extracolonic features.
• APC gene - AD - genotype-phenotype correlation - mutations in the first 4 exons tend to give rise to attenuated polyposis where you have <100 polyps. Starts to give rise to difficulties in identifying them as individuals who may have mutations in the APC gene.
• Two common 5bp deletions account for about 15-20% of APC mutations but all the rest are scattered throughout the gene. You just need to sequence the whole gene. Usually truncating mutations. 80% are point mutations. 7-12% are large genomic deletions.
• MYH-associated polyposis - 15-200 colonic polyps by age 50 -Autosomal recessive - MYH gene - 2 mutations account for 82% mutant alleles in UK caucasian populations.
• There is an overlap between the MYH and APC associated polyposis phenotypes.

Question 4

Are there any common mutations in the APC gene seen in FAP?

Answer 4

• Two common 5bp deletions account for about 15-20% of APC mutations but all the rest are scattered throughout the gene.
• You just need to sequence the whole gene.
• Usually truncating mutations.
• 80% are point mutations.
• 7-12% are large genomic deletions.

Question 5

What locus heterogeneity do we get with HNPCC?

Answer 5

• 4 genes where mutations may occur. The 2 major genes are MLH1 and MSH2 which are responsible for about 90% of the mutations. MSH6 is responsible for about 10% and PMS2 a few %.
• If you need to screen all 4 genes you need to look through 60 exons.
• Locus heterogeneity is not so much a problem now that we can make NGS gene panels.

Question 6

What locus heterogeneity do we get with familial breast and ovarian cancer?

Answer 6

• BRCA1 and BRCA2 genes - 63 exons to search through.

- Locus heterogeneity is not so much a problem now that we can make NGS gene panels.

Question 7

How does allele heterogeneity affect our detection of cancer mutations?

Answer 7

• There is heterogeneity of mutations across different populations - some cancer mutations will be far more common in certain ethnic groups etc.
• There are no obvious major mutations for HNPCC, BRCA1/2 in caucasian UK populations so we need to sequence the whole genes.
• BRCA1/2 there are founder mutations in other populations - e.g. Iceland there are common founder mutations. In Ashkenazy Jewish populations there are 3 mutations which are high pop. frequency so BPGs advise testing in all predictive tests.

Ashkenazy BRCA mutations

• 2 in BRCA1, 1 in BRCA 2.
• BRCA1 185delAG (old nom) - 1% Ashk Jewish Women - 20% early onset breast cancer.
• BRCA2 6174delT (old nom) - 1-1.5% Ashk Jews - 8% early onset breast cancer.

Question 8

What common BRCA mutations are seen?

Answer 8

Ashkenazy BRCA mutations

• 2 in BRCA1, 1 in BRCA 2.
• BRCA1 185delAG (old nom) - 1% Ashk Jewish Women - 20% early onset breast cancer.
• BRCA2 6174delT (old nom) - 1-1.5% Ashk Jews - 8% early onset breast cancer.

There are no obvious major mutations for BRCA1/2 in caucasian UK populations so we need to sequence the whole genes.

Question 9

Describe the BRCA1/2 genes.

Answer 9

• Both large genes.
• Both involved in DNA repair by homologous recombination and integrity of the genome.
• More Alu repeats are seen in BRCA1 introns than seen in BRCA2 - therefore a larger proportion of mutations in BRCA1 will be by whole exon deletions and duplications than in BRCA2 - still MLPA both. Alu repeats can cause unequal crossover.

Question 10

In which gene are more whole exon deletions likely to occur and why - BRCA1 or BRCA2?

Answer 10

• More Alu repeats are seen in BRCA1 introns than seen in BRCA2 - therefore a larger proportion of mutations in BRCA1 will be by whole exon deletions and duplications than in BRCA2 - still MLPA both.
• Alu repeats can cause unequal crossover.
• Do still see whole exon deletions in BRCA 2.

Question 11

What is the risk of a female BRCA1 mutation carrier developing cancer?

Answer 11

• BRCA1 - 65% risk of breast cancer, 39% risk of ovarian cancer by age 70.

Question 12

What is the risk of a female BRCA2 mutation carrier developing cancer?

Answer 12

• BRCA2 - 45% risk of breast cancer, 11% risk of ovarian cancer by age 70.

Question 13

Other than breast and ovarian cancer what other cancer risks may be increased by BRCA mutations?

Answer 13

Increased risk of other cancers:

• Increased risk of prostate and pancreatic cancer in BRCA2.
• Increased risk of prostate, pancreatic, endometrial, cervical cancer in BRCA1.

Question 14

What may we find when looking for mutations in BRCA1/2 genes? What do we need to do with this information?

Answer 14

• May find numerous different mutations.
• The problems come when interpreting mis-sense changes of unknown significance.
• Need to go through variant assessment process to classify previously unseen variants - conservation? amino-acid effect? is the change in a functionally important part of the protein? any functional tests? Segregation studies.
• There are some known pathogenic mis-sense mutations.
• Most BRCA2 truncating mutations are pathogenic but stop codon in last exon of BRCA2 is actually a benign polymorphism.
• Sometimes individuals may have both a BRCA1 and 2 mutation.

Question 15

What can be done to confirm whether a variant is altering splicing?

Answer 15

Can check blood RNA.

Question 16

What’s the very basic process for BRCA testing?

Answer 16

1) . Sequence (Sanger panels or NGS).
2) . Sequence analysis.
3) . Variant analysis and interpretation.

Question 17

How do the characteristics of HNPCC tumours help inform testing for them and decide what family members may need most frequent monitoring?

Answer 17

• HNPCC is due to mutations in one of the 4 MMR proteins.
• All HNPCC tumours have microsatellite instability.
• MS instability is a good indicator that someone might have HNPCC. Can be used as a screening tool before you look for mutations in those genes.
• If you can say that an individual from a family with a FH of CRC does not have MSI then you can justify reducing their screening = reduced risk = reduced cost.
• Can also test the actual proteins using IHC to inform which MMR protein the mutation might be in.
• Most labs test individuals that have a certain family history of colorectal cancer. Do this as a pre test and increase mutation pick up rates.

Question 18

What % of all colorectal cancers are due to HNPCC?

Answer 18

• Only about 1-2% of all colorectal cancers are due to HNPCC.

Question 19

What is the inheritance pattern of HNPCC? What causes HNPCC?

Answer 19

• Inherited in an autosomal dominant fashion.

- Caused by mutations in the mismatch repair genes.

Question 20

What other cancer cancer do women with HNPCC have a high risk of developing?

Answer 20

• Women with HNPCC have a 40% risk of endometrial cancer by age 70.
• Good reason for identifying the mutation in individuals suffering from CRC.

Question 21

What is the risk of individuals with HNPCC developing colon cancer?

Answer 21

Males with HNPCC have a 70% risk of colon cancer by age 70, females 30%

Question 22

What is the difference in the phenotype of HNPCC if the mutation is in the MSH6 gene?

Answer 22

• MHS6 families - later onset, more endometrial cancer.

- The presentation of HNPCC can be an indication of which genes there may be a mutation in.

Question 23

What are EPCAM mutations in HNPCC?

Answer 23

• EPCAM mutations are a very small % of the mutations in MSH2 (probable <1%).
• Was found that a del in ex8 and 9 in the EPCAM gene upstream of MSH2 removed the EPCAM RNA stop signal (poly-adenylation signal).
• By removing the PolyA signal you get readthrough of the RNA polymerase into the MSH2 gene downstream. This results in the methylation of the promoter of the MSH2 gene thus turning it off. The net result is essentially the same as if you had a mutation in the MSH2 gene.
• This effect only occurs in the particular tissues where the EPCAM gene is turned on - epithelial tissues (not blood).
• Important EPCAM del, exon 9 is in the MLPA kit used to test for whole exon deletions in MLH1 and MSH2.

Question 24

What % of MSH2 mutations in HNPCC are the result of an EPCAM deletion? What exon in EPCAM is deleted?

Answer 24

• <1% of MSH2 mutations are the result of deletions in exon 9 in the upstream EPCAM gene.

Question 25

What is the main indicator of MLH1 and MSH1 mutations in tumours?

Answer 25

• MSI in tumours?

Question 26

Are all cases of MSI in tumours due to HNPCC?

Answer 26

• It is not simple a case of MSI in tumours being a definite indication of MlH1 and MSH2 mutations in HNPCC. Sporadic cases of microsatellite instability can confound diagnosis.

Question 27

What is mismatch repair and what proteins are involved in the process?

Answer 27

• DNA mismatch is occurs when you get an inappropriate base change.
• 4 proteins are involved in mismatch repair, MLH1, MSH2, MSH6, and PMS2.
• The 4 proteins occur in 2 pairs - MLH1 is paired with PMS2, MSH2 is paired with MSH6.
• MSH2 and MLH1 are the proteins from each pair that bind the DNA. Therefore if you get mutations in MSH2 or MLH1 you will lose these proteins AND their partners from acting on the DNA strand (as the partners are only attached via these bound proteins). However, if you get mutations in MSH6 or PMS2 then the main partner protein can still bind the DNA stand and won’t be lost.

Question 28

What pairs do the mismatch repair proteins occur in?

Answer 28

• 4 proteins are involved in mismatch repair, MLH1, MSH2, MSH6, and PMS2.
• The 4 proteins occur in 2 pairs - MLH1 is paired with PMS2, MSH2 is paired with MSH6.
• MSH2 and MLH1 are the proteins from each pair that bind the DNA. Therefore if you get mutations in MSH2 or MLH1 you will lose these proteins AND their partners from acting on the DNA strand (as the partners are only attached via these bound proteins). However, if you get mutations in MSH6 or PMS2 then the main partner protein can still bind the DNA stand and won’t be lost.

Question 29

What is microsatellite instability?

Answer 29

• Microsatellites are repeats of small numbers of bases at various places in the genome - on average about every thousand bases.
• Genetic instability results in subtle sequence alterations within microsatellites - termed microsatellite instability.
• In HNPCC, MSI is due to a defect in the ability of cells to repair nucleotide mismatched during DNA replication, due to a mutation in a MMR gene.
• MSI is also seen in other types of tumours.
• Most HNPCC adenocarcinomas show MSI (doesn’t always appear right at the beginning of tumour progression but not all the adenomas).
• However, 15-20% of sporadic CRC tumours also show MSI. Testing for MSI isn’t an all or nothing test for HNPCC, it just alters your risk.
• In practice we see extra alleles for the microsatellite on the QF-PCR for the tumour compared with the normal tissue from an individual with a tumour displaying MSI.
• IHC of MMR proteins can also be used to identify which MMR protein is missing and therefore which MMR gene the mutation is likely to be in. However, the downside to using IHC for diagnosis is that you will miss approximately 5% of HNPCC tumours - possibly because there might be missense changes where the protein is there but it is not functioning correctly.
• There are some reports that some MSH6 mutations do not give rise to MSIH tumours. May be much more sensitive to use mononucleotide polymorphisms to look for MSI.

Question 30

What is the downside of using IHC to diagnose HNPCC in tumours?

Answer 30

IHC of MMR proteins can also be used to identify which MMR protein is missing and therefore which MMR gene the mutation is likely to be in. However, the downside to using IHC for diagnosis is that you will miss approximately 5% of HNPCC tumours - possibly because there might be missense changes where the protein is there but it is not functioning correctly.

Question 31

How much MSI is considered enough to classify a tumour as having MSI?

Answer 31

• If you investigate any tumour in enough depth you will probably find some MSI.
• Early studies tried to define this - would expect 40% of markers to be changed in an MSI positive tumour.
• Some tumours were found to have a lower percentage of changes. - does seem to be something real but no one is quite sure what it means - these would not be referred to as HNPCC tumours.
• Phenomenon of MSI-High vs. MSI-Low - most HNPCC tumours = MSI high.
• Subsequently there have been revised Bethseda criteria recommending using an extra panel of markers to distinguish between low and high.

Question 32

How was the Promega MSI kit developed?

What type of microsatellite nucleotie markers are the most sensitive way of detecting MSI?

Answer 32

• Promega evaluated 266 mono-, di-, and penta- nucleotides to find the most sensitive for MSI.
• Used the Bethseda scoring system.
• Found the mononucleotides were the most sensitive and specific.
• Combined 5 of the mononucleotide markers for the kit.
• Classified 153 tumours using the new system.
• In concordance with other larger studies approximately 5% of the MSI-H tumours had no loss of MMR protein by IHC.
• 43 samples that were classified as MSI-L using a panel of 10 markers were re-classified as MSS using the Promega kit = high specificity for MSI-H tumours.
• Original method based on comparing normal DNA vs tumour DNA.

Question 33

What are the characteristics of MSH6 carriers?

Answer 33

• MSH6 carriers are often characterised by multiple endometrial cancers, a low penetrance of CRCs and an older age of diagnosis.
• Many cancers that arise in MSH6 carriers exhibit an MSI-L phenotype when using dinucleotide microsatellites. May also have to use IHC when testing these families.

Question 34

Would MSI testing alone be sufficient for testing for a HNPCC in a family that was known to carry a MSH6 mutation?

Answer 34

Many cancers that arise in MSH6 carriers exhibit an MSI-L phenotype when using dinucleotide microsatellites. May also have to use IHC when testing these families.

Question 35

What are the main benefits of using IHC in addition to MSI when testing for HNPCC?

Answer 35

• Can give you more confidence that you will detect MSH6 changes because of the supposed lack of sensitivity of MSI.
• Can also usually tell you which gene is mutated based on which MMR protein is absent.

Question 36

15-20% of sporadic, non-HNPCC tumours have MSI. What can we use to help differentiate these from HNPCC tumours?

Answer 36

• There are other mutations which are found in sporadic cancers like the BRAF gene - specific common mutation that is found in sporadic cancer (BRAF-V600E)
• After we find an MSI positive tumour the next thing we would do is test for BRAF-V600E.
• BRAF often used as a pre-screening tool in HNPCC as only BRAF-V600E negative cases need to be screened for mutations in MMR genes.
• Methylation of the hMLH1 promoter is likely to be the cause of most cases of sporadic MSI - it is also found in BRAF-V600E positive cases. MLH1 promoter methylation is likely another marker for sporadic MSI.

Question 37

What is the common BRAF mutation that is found in sporadic CRC?

Answer 37

• BRAF-V600E

Question 38

Is it possible to do MSI with the promega kit without normal tissue?

Answer 38

• No - mononucleotides have low heterozygosity. (i.e. both alleles will have the same number of mononucleotises). If you find that 3 or more of the mononucleotide microsatellite markers in the kit are heterozygous in the tumour sample then you are almost certainly looking at MSI - only a 0.014% of these 3 being heterozygous in the germline/normal sample.
• Some labs have dispensed with the normal material and just base it on detecting heterozygosity in mononucleotide microsatellite markers in tumours.