While characterizing the new SAMD9 and SAMD9L - related bone marrow failure syndromes, we found that different germline pathogenic mutations can occur at the same amino acid position (Sahoo, Nat, Medicine 2021). Here, we explore this phenomenon by studying a family with triple allelic mosaicism in SAMD9L and two mutational permutations of the same amino acid in different children. We applied genomic methods (exome sequencing, Sanger, error-corrected sequencing, FISH, single-cell DNA sequencing (scDNAseq), ddPCR, and PacBio haplotype sequencing), developed stable mutant HEK293 cell lines, created V1512M and V1512Lmutant inducible pluripotent stem cells (iPSC) and ortholog mutant mouse models with Samd9l heterozygous (V1507Mhet, V1507Lhet) and homozygous (V1507Lhom) mutations using CRISPR/Cas9 engineering.

In a family with SAMD9L-related bone marrow failure, the affected girl carried heterozygous c.4534G>A (V1512M) mutation, while her brother was heterozygous for c.4534G>T (V1512L) mutation. Both mutations were germline and confirmed to arise from the same haplotype. The child with V1512M also carried somatic rescue events (del7q, acquired SAMD9L mutation) and had a more severe clinical phenotype. The asymptomatic mother who had normal diploid karyotype was a triple-allelic mosaic. The SAMD9L wildtype (WT) allele was present at ~50% frequency in all tissues, while both mutant alleles present in the remaining 50% of reads "competed" in their allelic distribution depending on the tissue origin (hematopoietic cells: V1512M ~9% and V1512L ~41%, hair follicles: both at ~25%). ScDNAseq revealed 3 separate clonal populations in blood, all with biallelic configuration: SAMD9L V1512M in 14%, V1512L in 68%, and WT (with acquired uniparental isodisomy 7q) in 18% of cells. Because her parents were SAMD9L WT, one of the mutations likely arose de novo and underwent failed rescue attempt during embryogenesis leading to a second mutation. Overexpression studies in HEK293 cells demonstrated significant growth inhibition of mutants (V1512M more severe than V1512L) compared to WT. Hematopoietic progenitor cells (HPC) derived from mutated iPSC had severely compromised proliferative capacity, and differentiation of V1512M and V1512L HPC yielded significantly fewer CD235+CD71+ erythroid cells and severely decreased CD45+CD33+ myeloid cells; in both lineages, V1512M mutation had the most severe phenotype.

We also observed divergent mutational phenotypes in mice: two-thirds (13/19) of the Samd9l V1507Mhet pups died before weaning and the vast majority died within the first 2 weeks of life. On the contrary, V1507Lhet mice had overall survival equal to WT mice, while V1507Lhom mice had a trend of decreased survival with 40% of mice dying by 19 weeks. At baseline, Samd9l V1507Mhet mice showed severe growth retardation with multisystem involvement (kidney, ovary, and alimentary tract), while V1507Lhet and V1507Lhom mice had no such abnormalities. The histology of BM and spleen in >7 weeks old V1507Mhet mice showed decreased erythroid and increased myeloid precursors in BM with extramedullary hematopoiesis. Significant multi-lineage peripheral blood (PB) cytopenia were observed in V1507Mhet mice compared to WT littermates. V1507Lhet mice had normal PB counts and BM phenotype comparable to WT, while V1507Lhom mice had isolated cytopenia and reduced numbers of BM erythroid and B-cells, and an increase in myeloid cells. Next, to induce replicative stress in V1507Lhom model, we performed non-competitive BM transplantation (BMT) into lethally irradiated B6.SJL/J WT recipients. At 12 weeks post BMT, there was a significant reduction of total V1507Lhom cell numbers with a decrease in B-cells, increase in myeloid cells, and no difference in T-cells. For V1507Mhet mice, we performed competitive BMT (because of the severe baseline phenotype), revealing significant changes post BMT, like V1507Lhom transplants, yet more severe and arising earlier. Altogether, our data suggest that Samd9l V1507Mhet and V1507Lhom cells have lower erythroid and B-cell output and a shift towards myelopoiesis.

To our knowledge, this is the first report of germline triple-allelic mosaicism with bi-allelic segregation in humans. Our study establishes how different mutational permutations of the same amino acid exert divergent phenotypic effects in humans, mice, and cellular models of SAMD9L bone marrow failure disorder.

No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution