Foreword

Mismatch repair defect (dMMR) refers to the lack of expression of at least one of the four mismatch repair (MMR) proteins due to three main causes: germline, systemic, and hypermethylation of the promoter region of the MLH1 gene. Lynch syndrome (LS), also known as hereditary nonpolyposis colorectal cancer (HNPCC), is an autosomal dominant hereditary disease, which is usually caused by embryogenic mutation of MMR gene. MLH1 methylation-induced deletions of the MLH1 protein are more common than mutations in the MMR gene. Multiple guidelines have recommended that in colorectal cancer (CRC) and endometrial cancer (EC), abnormal MLH1 immunohistochemistry (IHC) results should be detected by MMR gene embryology or methylation of MLH1 in tumor tissue, which indicates sporadic tumors ^[1–3].

Constitutive epigenetic mutation

The screening procedures described above used MLH1 gene methylation testing, excluding sporadic cases from the MMR gene embryology testing. Although it can reduce the cost of screening and improve patient compliance, rare cases of constitutive MLH1 epigenetic mutation/methylation are ignored. Constitutive apparent mutations are characterized by promoter methylation and transcriptional inactivation of individual alleles in normal tissues and may be detected in samples of peripheral blood lymphocytes (PBL, mesoderm), normal colonic mucosa (endoderm), muscle (mesoderm), and oral mucosa (ectoderm) ^[4]. The constitutive apparent mutations are divided into primary and secondary. The primary apparent mutations have no obvious linkage sequence changes, and are reversible during meiosis. They are occasionally transmitted to the next generation in a non-Mendelian mode and rarely associated with a clear family history of cancer. Secondary apparent mutations are caused by cis-acting changes in DNA sequence and follow the Mendelian genetic pattern of transmission ^[4–7]. The constitutive apparent mutations may be chimeric, and the methylation levels are heterogeneous among different alleles and tissues ^[4,5].

Constitutive apparent mutations in the MLH1 and MSH2 genes have been identified as a rare etiology of LS. The MSH2 apparent mutation is a typical secondary apparent mutation resulting from the deletion of the 3' terminal embryo line of the EPCAM gene located 17 kb upstream of MSH2 ^[6]. The molecular mechanism of the constitutive MLH1 apparent mutation is complex, and it occurs in 2–3% of suspected Lynch families with mutation negativity ^[7]. Cases of constitutive methylation presenting with early onset and/or multiple primary tumors consistent with MLH1-LS; It can also result in tumor MLH1 deletions, MSI-H, and MLH1 methylation that overlap with common sporadic cases ^[8].

Cohort study on MLH1 apparent mutation ^[8]

Two CRC cohorts, Columbus(n=1566) and OCCPI(n=3310), were included in a retrospective study in Ohio, USA, based on (1)MLH1 deletion and/or MSI-H, (2)MLH1 methylation, and (3) availability of PBL or whole blood DNA. Blood DNA samples were analyzed using both pyrosequencing and methylation specific multiplex fluorescent PCR(MS-qPCR) and either test generated a signal above the limit of detection which was considered positive for methylation.

In the Columbus cohort, 95 methylation studies were successful, with pyrosequencing results fully consistent with MS-qPCR. Constitutive MLH1 methylation was detected in four patients (4.2%) at the ages of 34, 36, 52, and 74 years at the time of diagnosis, all of whom had an initial CRC, and only one patient (Columbus-65) had a family history. Approximately one-half of the alleles in our patient, Columbus-6, were methylated, while the other 3 patients had low levels of methylation suggestive of chimeric methylation.

Methylation detected in two cohorts of blood samples

In the OCCPI cohort, methylation was successfully detected in 281 cases, and the results of pyrosequencing were completely consistent with those of MS-qPCR. Constitutive MLH1 methylation was detected in four patients (1.4%) aged 20, 34, 50, and 55 years, all of whom had the first episode of CRC and all had an unclear family history. Approximately half of the alleles in 3 patients were methylated, with low levels (4.2%) of chimeric methylation in OCCPI-1 blood and C-93G > A heterozygosity in the promoter region. Their tumor and paired normal colorectal mucosa (NCM) revealed similarly low levels (2.6%) of chimeric MLH1 methylation detected by NCM and high levels (60.8%) in the tumor. Sequencing of the MLH1 promoter revealed loss of heterozygosity (LOH) for the unmethylated "A" allele at c.-93g > A in the tumor as a "second strike" of tumorigenesis.

NCM and Tumor Test Results in OCCPI-1 Patients

The age stratification shows a high rate of constitutive MLH1 methylation in young patients. In the Columbus and OCCPI cohorts, the constituent methylation rates, defined by age ≤55 years, were 75%(3/4) and 23.5%(4/17), respectively, with the majority of cases detectable. Although the data are incomplete, all 4 patients with constitutive methylation with BRAF V600E tumor test results were BRAF wild-type. Consistent with previous reports, patients with constitutive methylation have a higher frequency of BRAF wildtype. If BRAF V600E testing is used instead of MLH1 methylation as a screening test for patients with MLH1 deletions, patients with constitutive MLH1 methylation will be further tested for MMR embryogenesis and may produce negative/clinically insignificant test results.

Detection of constitutive MLH1 methylation by age in the Columbus and OCCPI cohorts

Representative studies

In addition to the cohort studies cited in the NCCN guidelines above, there have been a number of previous studies or case reports regarding constitutive MLH1 apparent mutations, both primary and secondary. We selected a few representative studies.

Study I ^[6]:

A French study using a combination of large fragment PCR and high-throughput sequencing (NGS) studied 4 families with inherited epigenetic mutations and 10 patients without confirmed epigenetic mutations. Family 1 detected the insertion of AluYc sequence 6 bp upstream of the 3' end of exon 1 of MLh1, measuring 345 nucleotides in length. Such insertion would result in the destruction of the MLH1 gene, providing a new splice donor site, and the resulting transcript encoding the truncated protein. Alleles with Alu insertion are co-isolated from methylation in the family.

Family 1 showed autosomal dominant inheritance of the constitutive MLH1 epigenetic mutation, with exon 1 sequence inserted into the AluYc sequence

The proband of Family 2 was CRC, and the first-degree relative had no history of cancer and showed absent expression of isolated proteins of MSI-H and PMS2, but no mutation was detected by PMS2 gene sequencing. Methylation of the MLH1 promoter was detected in our patient's tumor and PBL DNA, and apparent mutations were identified in six asymptomatic relatives. Using multiplex ligation-dependent amplification (MLPA) and comparative genomic hybridization array (CGH-array), we found a large fragment repeat consisting of the EPM2AIP1 gene and exons MLH1 1-6, with a repeat fragment size of about 29.54 kb and flanking the Alu element, co-isolated from methylation.

Family 2 showed autosomal dominant inheritance of the constitutive MLH1 apparent mutation, as well as the inclusion of both EPM2AIP1 and exons 1-6 of MLH1.

The new mutation c.116+106 g > A in intron MLh1 was detected in family 3, which was co-isolated from methylation in the family. One of the 10 patients (P4) with no confirmed apparent mutation was also a carrier of the mutation, but no apparent mutation or C.116+106G > A mutation was detected in either the diseased or non-diseased family members. The synonym mutation of MLh1 C.27G > A is detected in Family 4, and it has been reported in the previous family registration for colon cancer (C-CFR) multicenter study ^[5]. The mutation was also detected in an unproven patient (P5), but the haplotype was different from Family 4. In addition, the c.-167delA mutation in the MLH1 5'UTR region was associated with methylation in an unproven patient (P3).

Studies also confirmed the theory that gene silencing preceded DNA methylation, and alleles with lower expression levels were more likely to be methylated. Some mutations themselves result in complete allelic inactivation, independent of epigenetic silencing, and promoter methylation may appear as a redundant mechanism of inactivation. Methylation can serve as a stabilizing mechanism for transcriptional silencing, preventing cell transcription of nonfunctional alleles. In addition, some mutations that have been shown to have an effect on splice function have no associated promoter hypermethylation, which may be related to the effect of sequence changes on binding to epigenetic modifying elements.

Study II ^[9]:

In an Italian study using techniques such as NGS and MS-MLPA, 61 MLH1 missing tumors from 56 patients (all BRAF of CRC were wild-type), four tumor samples had very high methylation levels (greater than 0.9), and the blood DNA of all four patients identified constitutive MLH1 methylation.

In familial F-5, both members P-09 and P-10 carry the MLH1 c.-168_c.116+713del blastocyst deletion, which develops three tumors (2 CRCs and 1 EC) with the same somatic methylation pattern. Both patients were constitutive carriers of MLH1 methylation, demonstrating that large deletions in the MLH1 promoter region caused epigenetic silencing of the gene and were defined as secondary epigenetic mutations.

Family F-5 with secondary apparent mutation

In familial F-6, P-38 was a young female patient with CRC who carried the PMS2 P794S embryogenic indeterminate clinical significance (VUS) mutation with high levels of MLH1 methylation detected in tissues and in blood. This was defined by the investigator as a primary apparent mutation in MLH1. Three years later, our patient developed complex endometrial hyperplasia with MLH1 defects.

Detection of F-6 and P-38 in Primary Apparent Mutation Families

Guide recommendations

NCCN hereditary/familial high risk assessment: The Guidelines for colorectal cancer, endometrial cancer and gastric cancer recommend the following for the detection of constitutive MLH1 apparent mutation ^[1]:

Ø The patient's constitutive MLH1 apparent mutation is a rare exception. Patients with > 1 tumor with early onset CRC(≤55 years), no BRAF V600E mutation, absence of MLH1 in IHC, no embryogenic MLH1 pathogenic/suspected pathogenic mutation, or hypermethylation of the MLH1 promoter at any age are considered for referral to an institution with expertise in genetic testing for MLH1 methylation testing.

Ø Evaluation of apparent mutations in constituent MLH1 The methylation status of the MLH1 promoter region was detected using blood or other normal tissue samples.

Summary

Constitutive MLH1 methylation, one of the non-negligible mechanisms of suspected LS tumors, has a certain frequency in patients with missing MLH1 and no detectable MMR sequence mutation, 4.2% in the Columbus cohort, 1.4% in the OCCPI cohort ^[8], and 3.8% (16/416) in the C-CFR study ^[4]. Overall, however, constitutive MLH1 methylation is extremely rare in an unselected CRC, with 0.26 in the Columbus cohort (4/1566) and 0.12% in the OCCPI cohort (4/3310) ^[8], without prejudice to guidelines' recommendations that MLH1 methylation is suggestive of a more probable sporadic tumor.

For CRC and EC with missing MLH1 protein expression, methylation detection of the MLH1 gene promoter helps to clarify the MMR and MSI status of the tumor and rule out the risk of Lynch syndrome. In addition, MLH1 methylation can also indicate the prognosis of EC and the selection of adjuvant treatment options ^[10–11], and has a supplementary role in indicating the molecular typing ^[12]. The MLH1 gene methylation detection kit of SpaceGen employs MS-qPCR technology, and samples with methylation frequency as low as 1% can be detected in one-step operation after transformation. SpaceGen is committed to providing the most innovative products and services for the individualized and precise medical detection of tumors, and continuously updates the existing products.

References

[1] NCCN Genetic/familial high risk assessment: colorectal cancer, endometrial cancer, gastric cancer 2024 v3

[2] CSCO colorectal cancer diagnosis and treatment guidelines 2024

[3] CSCO endometrial cancer diagnosis and treatment guidelines 2024

[4] Cancer Med. 2018 Feb; 7(2):433-444.

[5] Genet Med. 2013 Jan; 15(1):25-35.

[6] Genet Med. 2018 Dec; 20(12):1589-1599.

[7] Fam Cancer. 2016 Jul; 15(3):385-93.

[8] J Natl Compr Canc Netw. 2023 Jul; 21(7):743-752.e11.

[9] Genes (Basel). 2023 Nov 9; 14(11):2060.

[10] J Gynecol Oncol. 2021 Nov; 32(6):e79.

[11] Cancer. 2022 Mar 15; 128(6):1206-1218.

[12] Chinese journal of obstetrics and gynecology in october 2023, vol. 58, no. 10