Five individuals (two males and three females, age range: 31–40 years) from two unrelated Swedish families were studied. Family A (NCI-457) consisted of three affected siblings (AII:1, AII:2 and AII:3, Fig. 1A) while family B consisted of two affected siblings (BII:1 and BII:2, Fig. 1A). Neither family had a history of TBD-related diseases, however early graying of hair was reported for BI:1. The father (BI:2) in family B unexpectedly passed away due to a myocardial infarction at the age of 36 years. All five affected individuals presented symptoms consistent with a TBD, including macrocytic anemia with bone marrow hypocellularity (n = 5/5) managed with erythropoietin and blood transfusions, fibrotic interstitial lung disease or restrictive spirometric pattern (n = 4/5), graying of hair before age 20 (n = 4/5), and low-trauma fractures (n = 2/5). Moreover, symptoms consistent with Coats plus, such as exudative retinopathy (n = 5/5 bilateral in 4/5), cutaneous telangiectasias (n = 4/5), and GI bleeding due to angiodysplasia (n = 2/5) were noted. BII:1 was treated with tamoxifen for GI angiodysplasia with some effect. A brain MRI performed in BII:1 at age 29 revealed cysts and calcifications. All affected individuals in family A had moderate to severe kidney failure. Kidney biopsy performed on AII:1 and AII:3 showed glomerulosclerosis with mesangiolysis and glomerular microaneurysms. Both individuals who had undergone pregnancy (AII:1 and BII:2) experienced preeclampsia (AII:1: gestational week [GW] 29 and BII:2: GW26), a condition that has been described in up to 24% of pregnancies of women with TBDs [18]. Moreover, both affected individuals in family B were born during GW37 after induced labor due to maternal pre-eclampsia. Somatic investigations of bone marrow samples revealed a small clone with acquired activating TERT promoter mutations in two individuals. Three out of the five affected individuals are alive, to date; AII:2 died at age 31 due to complications from allogeneic hematopoietic stem cell transplantation, while BII:1 died at age 33 due to respiratory insufficiency. An overview of clinical manifestations in the five affected individuals is provided in Table 1.
A Pedigrees of families A and B with three and two affected individuals, respectively. In family A the parents were confirmed to be heterozygous carriers of each POLA2 variant, while in family B the mother was confirmed as heterozygous carrier of the complex rearrangement involving POLA2, and the father could not be tested as he was deceased. B Analysis of relative telomere length (RTL) by qPCR in peripheral blood leukocytes collected from family A (red symbols) and family B (blue symbols). Filled and unfilled symbols represent compound heterozygous and heterozygous genotypes, respectively. Squares and circles represent males and females, respectively. The reference percentiles were determined from telomere length analysis of blood leukocytes from 283 healthy subjects (0–78 years of age). The curves shown depict the first, 10th, 50th, 90th, and 99th normal percentiles at each age, where 50th represents the mean. C Sanger sequencing of POLA2 (NM_002689.4) sequence variants performed on peripheral blood-derived DNA from all individuals tested in family A [c.287 T > C, p.(Ile96Thr) and c.1271 C > T, p.(Pro424Leu)] and (D) B [c.287 T > C, p.(Ile96Thr)] in family B. E Screenshot from Integrative Genome Viewer (IGV) over the complex rearrangement involving exon 1 of POLA2 in family B. F Sanger sequencing of breakpoint junctions (BpJ) 1 and 2 of the complex rearrangement involving exon 1 of POLA2 in BII:1. The HGVS nomenclature for the complex rearrangement is NC_000011.9:g.[65,023,295_65,024,239del;65,024,240_65,026,040inv;65,026,041_65,031,189del]]. The red box in BpJ1 highlights a 27-bp microhomology sequence found at both breakpoints. Arrows indicates specific genomic positions at the breakpoints.
Relative telomere length was estimated using quantitative polymerase chain reaction (qPCR) analysis from blood leukocytes and showed reduced telomeres in all affected individuals (Fig. 1B). Short telomeres were confirmed by flow-FISH in family A (Supplementary Fig. 1).
Overall, family history and clinical presentation were suggestive of an autosomal recessive TBD. WGS performed on AII:3 and BII:1 did not reveal any disease-causing variants in known genes for TBDs (Supplementary Table 2). Further filtering for rare deleterious variants detected two heterozygous missense variants in the POLA2 (NM_002689.4) gene, c.287 T > C, p.(Ile96Thr) and c.1271 C > T, p.(Pro424Leu) in patient AII:3. The same missense variant p.(Ile96Thr) was also identified in heterozygous state in BII:1, together with a heterozygous intragenic structural variant (deletion-inversion-deletion, 7.8 kb in total) of POLA2, resulting in deletion of the POLA2 5’ terminus and exon 1 (NC_000011.9:g.[65,023,295_65,024,239del;65,024,240_65,026,040inv;65,026,041_65,031,189del]).
Segregation analysis performed in both families confirmed compound heterozygosity for the POLA2 variants in all affected individuals (Fig. 1C–F). The POLA2 missense variant p.(Ile96Thr), shared between the two families, was detected in a heterozygous state in 12 additional individuals in a local clinical database of 11,002 individuals who underwent clinical exome or genome sequencing at Karolinska University Hospital. The variant is also present in a heterozygous state in 51 individuals in the gnomAD database (minor allele frequency [MAF]: 0.003181%; gnomAD v.4). Based on data from gnomAD v.2.1.1, where Swedish ancestry is specified, this variant was present in 22 individuals of which 18 of them were of Swedish origin. The MAF for this variant in gnomAD v.2.1.1 was therefore 0.00887% for the whole cohort and 0.06899% in the Swedish population. The other missense variant, p.(Pro424Leu) has ten heterozygous carriers in the gnomAD database v4 (MAF 0.0006841%). Intragenic structural variants were rare in gnomAD v.4 and no identical event to the one identified in BII:1 was found in gnomAD nor in our local clinical database. The identified missense variants are located in different domains of POLA2, yet both show conservation across species (Supplementary Fig. 2A–C). Both variants are predicted as deleterious by in-silico prediction tools (Supplementary Table 3). The variant p.(Ile96Thr) is located in the N-terminal domain of POLA2 which has been proposed to represent a site of protein-protein interaction, potentially with CTC1 [19]. No large differences were noted when modeling the interaction surface with AlphaFold (Supplementary Fig. 2F). However, overlay of AlphaFold-generated models for p.(Ile96Thr) revealed a shortened alpha-helix compared to wild-type POLA2 (Supplementary Fig. 2G). The variant p.(Pro424Leu) is instead located in an inner core of the protein; clash modeling suggests a steric clash is introduced by p.(Pro424Leu) compared to a hypothetical p.(Pro424Ala) substitution that is predicted benign by AlphaMissense (Supplementary Fig. 2D, E). The missense variants were not classified according to the guidelines from the American College of Medical Genetics and Genomics, as these guidelines are not intended to fulfill the needs of the research community in its effort to identify new disease-causing genes [20]. We thereafter asked whether the POLA2 p.(Ile96Thr) variant could impact telomere length maintenance in human cells using CRISPR/Cas9 genome engineering in 293 T cells. Guide RNAs were designed targeting the human POLA2 locus on exon 3 at amino acid 96, complexed with Cas9, and electroporated alongside HDR templates into 293 T cells (Fig. 2A). Electroporated cells were subcloned and genotyped. We successfully engineered three subclones with knock-in of p.(Ile96Thr) in POLA2 exon 3, at a frequency likely reflective of HDR at one of three loci in this triploid cell line as determined by amplicon sequencing. Of note, in all three subclones, one allele remained wild-type and the third allele was disrupted by out-of-frame insertion/deletion (WT/KO/KI). Three subclones with no alterations at POLA2 exon 3 (WT/WT/WT) and three subclones with disrupting mutations on one allele (WT/WT/KO) were used as controls. We examined telomere length in all subclones by TRF Southern blot four weeks after engineering. We found that all three subclones carrying the POLA2 p.(Ile96Thr) mutation showed significant telomere shortening with a median length of ~4 kb, compared to wild-type subclones (WT/WT/WT) showing a median length of ~6 kb, similar to unmanipulated 293 T cells and WT/WT/KO controls (Fig. 2B, C).
A Schematic of the Cas9/crRNA-targeting sites in human POLA2 locus on exon 3 at amino acid 96. Exons are shown as blue boxes. The crRNA-targeting sequence is labeled as Guide sequence, and the protospacer-adjacent motif (PAM) sequence is labeled as PAM. The homology arms specific to the gene target are flanking the nucleotide to be changed (Ile>Thr or ATT > ACT on +strand labeled in red) and HDR silent mutation (Gln>Gln or CAA > CAG on +strand labeled in green). B Terminal Restriction Fragment Southern blot showing telomere lengths four weeks after CRISPR/Cas9 genetic modification of the triploid POLA2 locus in 293 T cells, indicating wild-type (WT), monoallelic knock-out (KO), and p.Ile96Thr knock-in (KI) alleles. WT/WT/WT represents clones with no editing of the loci (KI score 0–1%); WT/KO/KI represents POLA2 p.Ile96Thr on one allele (KI score of 24–35%), and one disrupted allele (KO); and WT/WT/KO indicates POLA2 disruption on one allele (KO score of 28–35%). C Quantitation of telomere length in (B). Statistical analysis was done by one-way ANOVA followed by Dunnett’s multiple-comparisons test. ****P < 0.0001, ns: not significant.
- SEO Powered Content & PR Distribution. Get Amplified Today.
- PlatoData.Network Vertical Generative Ai. Empower Yourself. Access Here.
- PlatoAiStream. Web3 Intelligence. Knowledge Amplified. Access Here.
- PlatoESG. Carbon, CleanTech, Energy, Environment, Solar, Waste Management. Access Here.
- PlatoHealth. Biotech and Clinical Trials Intelligence. Access Here.
- Source: https://www.nature.com/articles/s41431-024-01722-8