Table 2

Data of all MLH1 promoter variants

NomenclatureDetected in categoryrsAllelic frequencyClassificationPrediction in Alamut, conservation, TFBScDNA expressionOur classifications due to our results and data from literatureRemark
MLH1 c.-63_−58delinsCACGAGGCACGAGCACGA1x CEMNovelNot in ExAC or 1000 GenomesLOVD NCsCiV highly positive, nine TFBS lost MLH1 and EPM2AIP1 monoallelic, LRRFIP2 biallelicFour by ACMG: PM2+PS3, three in InSiGHT and Liu et al 45 (lacking in vitro functional assay/segregation with disease), Liu et al: secondary epimutation suspected
MLH1 and EPM2AIP1 allelic loss in expression, absence in controls
(cosegregation of CEM and promoter variant in two family members is not a criterion yet, and only one MSI-H, MLH1-deficient CRC to reach class 4)
Δ
EPM2AIP1 c.7A>G p.Met3Val
(MLH1 c.-477T>C)
1x H1Drs
746415556
ExAC: 0.000009, in East Asia
0.00012, no homozygotes; not in 1000 Genomes
LOVD NCBenign, splice-neutral, CiV positive, two TFBS lost MLH1 n.i., EPM2AIP1 biallelicThree in InSiGHT, ACMG and Liu et al
-No data
EPM2AIP1 c.-102T>C
(MLH1 c.-369A>G)
1x CIMPNovelNot in ExAC or 1000 GenomesLOVD NCNo CiV, one TBFS lostno cDNAThree in InSiGHT, ACMG and Liu et al
-No data
MLH1 c.-230G>C1x C-H1Prs
587782631
Not in ExAC or 1000 GenomesLOVD NC, class 3 in ClinVarNo CiV, one TBFS lostno cDNAThree in InSiGHT, ACMG and Liu et al
–No data (one H1P case)
MLH1 c.-33T>G1x LSrs
201247839
ExAC: 0.00025, no homozygotes;
1000 Genomes: 0.0002
LOVD NCSplice-neutral, CiV negative, one TBFS lostno cDNAThree in InSiGHT and Liu et al (allelic phase of pathogenic variant unknown), ACMG: BP2+BP5+BS3=1 (allelic phase not needed), Liu et al: consult InSiGHT database
–No data, patient with LS with additional pathogenic MLH1 variant and allelic phase unknown, AF
Δ
MLH1 c.-42C>T1x H1Drs
41285097
ExAC: 0.000025, no homozygotes;
1000 Genomes: 0.0002
LOVD class 3CiV highly positive, four TFBS lost MLH1 and EPM2AIP1 biallelicTwo in InSiGHT and Liu et al (additional argument lacking; 3 CRC MSS/lack of cosegregation), ACMG: BS1+BS3=1
- AF, MLH1 cDNA no functional defect in our case, but contradictory results: in literature 37% reduced expression in luciferase essay
(cosegregation with late-onset CRC)23 25 26
Δ n
R?
EPM2AIP1 c.123C>G p.=
(MLH1 c.-593G>C)
1x C-H1Prs
34566456
ExAC: 0.0052, Africa 0.058, 18 homozygotes; 1000 genomes: 0.02, Africa 0.08LOVD NCSplice-neutral, CiV negative, no TFBS lost MLH1 n.i., EPM2AIP1 biallelicTwo in InSiGHT and Liu et al (founder mutation not excluded, or additional argument in combination with cDNA no functional defect), ACMG: BS1+BS2=1,
- AF in Africa, 18 homozygotes
Δ
MLH1 c.-28A>G1x C-H1Prs
56198082
ExAC: 0.002, in Finland 0.009, one homozygote;
1000 Genomes: 0.0008
LOVD class 3, also in ClinVarCiV negative, one TBFS lost MLH1 biallelic, n.i. EPM2AIP1 One in InSiGHT, ACMG: BS1+BS2+BS3, and Liu et al
– AF, one homozygote, in >3 cases with H1P, MLH1 cDNA no functional defect23 48 49
R
MLH1 c.-7C>T for variant c.[-28A>G;-7C>T]1x C-H1P, 2x C-nLSrs
104894994
ExAC: 0.0015, in Finland 0.0087, one homozygote;
1000 Genomes: 0.0004
LOVD class 3, also in ClinVarNo CiF, two TBFS lost MLH1 and EPM2AIP1 biallelicOne in InSiGHT, ACMG: BS1+BS2+BS3, and Liu et al
– AF, one homozygote, in >3 cases with H1P, MLH1 cDNA no functional defect in our case, but in literature reduced expression of the variant allele to 28%–33% in cDNA without other variants detectable, in vitro assays performed27 50
n
R?
EPM2AIP1 c.-202G>C
(MLH1 c.-269C>G)
1x CEM, 1x H1D, 1x CIMP, 4x C-H1P, 1x C-nLSrs
35032294
1000 Genomes: 0.002LOVD class 2, also in ClinVar, benign in InvitaeCiV negative, one TBFS lost MLH1 and EPM2AIP1 biallelicOne in InSiGHT and ACMG: BS1+BS3, but Liu et al: secondary epimutation suspected (due to 1 CEM case with this variant)
– AF, in >3 cases with H1P, MLH1 cDNA no functional defect, confirmed by Zavodna et al 51
Δ
R
  • Summary of data of all MLH1 promoter variants detected in our study including category of patient or control in which the variant was detected, rs if known, the allelic frequency (AF) in ExAC or 1000 Genomes database, existence of homozygotes, variant classification in InSiGHT-LOVD or other database, in silico predictions in Alamut including splicing, the level of nucleotide conservation in vertebrates (CiV) from UCSC and loss of transcription factor binding sites (TFBS) predicted by ALGGEN-PROMO. The expression was investigated for all variants with cDNA available, and ’biallelic' indicates that the transcript showed a heterozygous variant in the transcript with normal allelic distribution, n.i. indicates cDNA analysis was ’not informative' due to a lack of heterozygous variants in the transcript. All variants are regarded as class 3 by default. Our classification of the variants applying the InSiGHT43 and ACMG44 guidelines and proposal by Liu et al 45 due to our results are given including the arguments and literature. ACMG arguments in detail were: PS3: abrogated mRNA expression, PM2: absence in controls, BS1: allele frequency higher than expected for disorder, BS2: observed in a health adult, BS3: in vivo functional studies (here: cDNA analyses) show no damaging effect on splicing, BP2: observed in trans or in cis with a pathogenic variant in any inheritance pattern, BP5: variant found in a case with an alternate molecular basis for disease (= control group). In brackets, additional information not sufficient as arguments are provided. In the remark, a discordance in classification between InSiGHT and ACMG is depicted by Δ, contradictory results in literature are marked with ↔ and suggested reclassification by R.

  • CRC, colorectal cancer; CEM, constitutional MLH1 epimutation; H1P, MLH1-proficient colorectal cancer.