Mutation detection in the probands and family members
The proband in Case 1 (II:1, Figure 2A) was a 33-year-old female afflicted with congenital orofacial clefts. No discernible phenotypic abnormalities were noted in her parents or spouse. Following the conception of a fetus with cleft lip and palate, the couple opted to terminate the pregnancy. Trio-WES of the proband and her parents revealed the presence of a de novo variant c.784C>T (p.Gln262*) in IRF6 (NM_006147.3). Sanger sequencing of the family members confirmed the de novo variant (Figure 2B). The IRF6 gene, which encodes interferon regulatory transcription factor 6, is implicated in Van der Woude Syndrome (VWS), an autosomal dominantly inherited developmental disorder characterized by cleft lip and/or palate. The IRF6 c.784C>T variant was reported in a patient with VWS [17]. According to the recommendation on sequence variants interpretation by the 2015 ACMG guidelines [13], the IRF6 c.784C>T variant would be classified as “pathogenic”, meeting the criteria PVS1+PS2+PM2_Supporting, and was recognized as the disease-causing factor of the proband.
The proband in Case 2 (II:1, Figure 2C) was a 32-year-old female with premature ovarian failure. Her follicle-stimulating hormone (FSH) level was 59.32 mIU/mL, and anti-mullerian hormone (AMH) level was 0.052ng/ml. She was revealed as a premutation carrier of FMR1 with the CGG repeats of 29 and 90. The FMR1 CGG repeats of the proband’s father, mother and husband were 90/Y, 29/29, and 24/Y, respectively (Figure 2C, D).
In Case 3, the couple remained unaffected, yet they had given birth to two boys afflicted with hypohidrotic ectodermal dysplasia (HED). The little son had passed away. A hemizygous missense variant c.1045G>A (p.Ala349Thr) of the EDA gene (NM_001399.5) was found in the proband (II:1, Figure 2E) by WES. Defects in the EDA gene cause X-linked HED which is characterized by hypotrichosis, hypohidrosis, and hypodontia. The EDA c.1045G>A (p.Ala349Thr) variant had been reported in several patients with X-linked HED [18-21] and was interpretated as “pathogenic” by ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/, Variation ID: 11040). Sanger sequencing of the family members revealed that the proband’s mother carrying the heterozygous variant of EDA c.1045G>A (Figure 2E, F).
Direct mutation detection in PBs
A total of 15 PB1s/PB2s from 4 PGT-M cycles were biopsied in the three couples. The mutations were detected in the MDA products of PBs by Sanger sequencing for Case 1 and 3, GC-rich PCR and TP-PCR for Case 2. The mutation carrier statuses of the corresponding embryos were deduced by the test results of PB1s and PB2s. The detection chromatograms were shown in Figure 3 and the results were presented in Table 1.
In Case 1, 6 oocytes were retrieved in one cycle. PB1 and PB2 of each oocyte were biopsied and examined by Sanger sequencing (Figure 3A, Table 1). In oocyte 1, only the wildtype genotype was detected in PB1 (F1-1-PB1) and PB2 (F1-1-PB2), suggesting that the corresponding embryo carried the mutant allele. The F1-1-PB1 should contain one wildtype allele and one mutant allele. The absence of the mutant allele in F1-1-PB1 suggests that the mutant allele was dropped out during the detection. PBs from oocyte 3 and 5 had similar test results with PBs in oocyte 1, with the exception of the absence of allele drop-out (ADO) in the detection of PB1s (F1-3-PB1 and F1-5-PB1). In oocyte 2 and 6, homozygous mutant genotype was detected in the PB1s (F1-2-PB1 and F1-6-PB1) and the wildtype was detected in the PB2s (F1-2-PB2 and F1-6-PB2). The mutation carrier statuses of these embryos were uncertain due to the possibility of ADO during the detection of the PB1s. The ADO rate for detecting genetic markers in PB1 was reported to be approximately 5.9%-9.6% [22]. Given the relatively low ADO rate, the corresponding embryos of oocyte 2 and 6 probably inherited the wildtype allele. A similar approach was taken to assess the mutation carrier status of the embryo from oocyte 4, suggesting that it likely carried the mutant allele. Haplotype analysis was conducted to determine whether the Sanger sequencing results of the PB1s (F1-2-PB1, F1-4-PB1, and F1-6-PB1) had been affected by ADO.
In Case 2, three oocytes were retrieved from two cycles. The CGG mutation of FMR1 was detected in PBs by GC-rich PCR and TP-PCR (Figure 3B, Table 1). In the first cycle, a solitary oocyte was retrieved. Only the expanded CGG allele was detected in the PB1 (F2-C1-1-PB1) by TP-PCR, suggesting that it carried the homozygous FMR1 CGG mutant allele and no ADO occurred during the detection. Because it was unlikely that the expanded allele was kept while the unexpanded allele was dropped out. The failure of GC-rich PCR to directly detect the CGG repeat in PB1 (F2-C1-1-PB1) was primarily attributed to the large size of the expanded allele, which hindered amplification by the MDA technique. The unexpanded CGG allele was detected in the PB2 (F2-C1-1-PB2) by both GC-rich PCR and TP-PCR. These results suggested that the corresponding embryo inherited the normal FMR1 allele. In the second cycle, two oocytes were retrieved. Analysis of the detection results of PB1s and PB2s (Figure 3B and Table 1) suggested that the embryo from oocyte 1 (F2-C2-1) exhibited a normal allele, whereas the embryo from oocyte 2 (F2-C2-2) displayed an expanded allele. The detection failure of PB2 in oocyte 1 may have been due to DNA degradation during the biopsy or amplification failure in the MDA procedure, but this did not impact the deduction made.
In Case 3, a total of six oocytes were retrieved in one cycle. Similar detection and analysis methods were applied (Figure 3C, Table 1). The embryos from oocytes 2 and 5 were determined to carry the mutant allele, while the mutation carrier statuses of the remaining embryos were indeterminate. Specifically, the embryos from oocytes 1, 4, and 6 were likely to have inherited the normal allele, while the embryo from oocyte 3 was likely to carry the mutant allele, assuming absence of ADO. Haplotype analysis was conducted on the corresponding PB1s (F3-1-PB1, F3-3-PB1, F3-4-PB1, and F3-6-PB1) to assess the occurrence of any ADO.
Haplotype analysis of family members and PBs
In Case 1, the proband and the specific PBs (PB1s and PB2s from oocyte 2, 4 and 6) were selected to conduct linkage analysis. The SNP haplotypes of PB1s and PB2s from oocyte 2 and 4 were presented in Table 2. The haplotypes of PB1 and PB2 in oocyte 6 were found to be identical to those in oocyte 2 (data not shown due to redundancy). The haplotypes of PBs were found to be consistent with the results obtained from direct mutation detection. These findings indicated that the embryos from oocyte 2 and 6 inherited the wildtype allele, whereas the embryo from oocyte 4 was carrying the IRF6 c.784C>T mutant allele.
In Case 2, the proband and her father were selected for linkage analysis in order to determine the SNP haplotype surrounding the FMR1 mutation. Haplotyping of the corresponding PBs of low-risk embryos was conducted to validate the mutation detection findings in the PBs (Table 1 and 3). The consistency between the haplotypes and the direct mutation detection results of PBs confirmed the low pathogenicity of the embryos from oocytes F2-C1-1 and F2-C2-1.
In Case 3, the proband and his mother were chosen for linkage analysis to investigate the SNP haplotype flanking the EDA c.1045G>A mutation. The specific PBs (PB1s and PB2s from oocyte 1, 3, 4 and 6) were haplotyped to determine the mutation carrier statuses of the embryos (Table 1 and 4). The haplotypes of PB1 and PB2 in oocyte 6 were found to be identical to those in oocyte 4 (data not shown due to redundancy). The haplotypes of the PBs were consistent with the results of mutation detection and deduced that embryos from oocyte 1, 4, and 6 inherited the wildtype allele, whereas the emrbyo from oocyte 3 carried the EDA c.1045G>A mutation.
Embryo selection and transfer
The mutation-free embryos deduced by the direct mutation detection and haplotype analysis were selected to transfer. Considering the morphology of these embryos, Case 1 chose the embryo from oocyte 2, Case 2 chose the embryo from oocyte 1 in the second cycle, Case 3 chose the embryo from oocyte 1, to transfer at the first time. The couples in Case 1 and 2 were clinically pregnant, while the couple in Case 3 were not.
Validation of the PGT-M results
In Case 1, no cleft lip or palate was detected in the fetus by ultrasound methods throughout the pregnancy. A healthy baby girl was born and the IRF6 c.784C>T mutation was not detected in the genomic DNA from the umbilical cord blood at birth (data not shown). In Case 2, the amniocentesis of the couple was performed at 18 weeks of the gestation. The PGT-M result was reconfirmed by the mutation detection of genomic DNA in amnion cells (data not shown). A healthy baby girl was achieved after a normal period of gestation. The Case 3 was waiting for another embryo transfer cycle.