The paralogous genes SAMD9 and SAMD9L, which are located next to each other on human chromosome 7, encode two highly homologous multi-domain proteins [1]. Genes homologous to human SAMD9 and SAMD9L can be traced back to unicellular organisms indicating an ancient function [1]. Phylogenetic analyses uncovered evidence of a duplication event of an ancestral gene giving rise to SAMD9 and SAMD9L, which persist in most vertebrate species. Some species have, however, lost one paralogue, suggesting a degree of functional redundancy [1,2]. Moreover, SAMD9 and SAMD9L have been subject to substantial evolutionary pressure [2], which supports their role at the frontline of host-pathogen interactions.

Initially, SAMD9 and SAMD9L were described as interferon-stimulated genes [[3], [4], [5]]. Their overexpression can restrict proliferation in cell-based assays and tumour xenografts [6,7]. More recently, however, variants in both genes have been associated with severe human diseases. In 2016, heterozygous de novo gain-of-function (GoF) variants in SAMD9 were reported to cause MIRAGE syndrome, a life-threatening congenital disorder with severe early developmental defects [8]. In parallel, heterozygous GoF mutations in SAMD9L were reported as the underlying cause of ataxia-pancytopenia (ATXPC) syndrome with variable clinical penetrance due to a remarkably high incidence of distinct somatic hematopoietic reversions [9,10]. Notably, both syndromes entail a predisposition to myelodysplastic syndrome (MDS) with monosomy 7, which develops due to an adaptation-by-aneuploidy effect with the selective loss of the chromosome 7 copy harbouring the SAMD9 or SAMD9L GoF variant [8,10]. Underscoring the clinical impact of SAMD9 and SAMD9L variants beyond the initially reported MIRAGE and ATXPC syndromes, subsequent studies also revealed a notable prevalence of GoF variants in these genes in larger cohorts of paediatric patients presenting with bone marrow failure or MDS [[11], [12], [13]]. In contrast to inherited disease-causing variants with variable expressivity/penetrance, de novo SAMD9L stop-gain mutations leading to SAMD9L protein truncation downstream of the p-loop NTPase domain cause a clinically distinct severe phenotype in infants referred to as SAMD9L-associated autoinflammatory disease (SAAD) [[14], [15], [16]]. Finally, autosomal recessive variants in SAMD9 have been reported to underly rare cases of normophosphatemic familial tumoral calcinosis (NFTC) [17].

While these clinical observations clearly establish the detrimental effects of pathogenic SAMD9 and SAMD9L variants, they provide little insight into the normal physiological function of SAMD9 and SAMD9L proteins. As SAMD9 and SAMD9L expression is profoundly up-regulated following type I interferon treatment, both homologous genes belong to a group of several hundred interferon-stimulated genes (ISGs) that include key regulators and effectors of cell-intrinsic antiviral responses [18]. Consistently, several studies have established that SAMD9 and SAMD9L can restrict productive viral infections in eukaryotic cells [[19], [20], [21], [22], [23], [24], [25]]. Further supporting that SAMD9/SAMD9L are crucially involved in the eukaryotic cell-intrinsic anti-viral defence, several viruses encode proteins that physically bind and antagonize SAMD9/SAMD9L anti-viral activity [19,23,25,26]. This is best exemplified by host range factors found in poxviruses, e.g. M062 in myxomavirus (MYXV) or C7 and K1 in Vaccinia virus (VACV). When M062 or C7/K1 are experimentally deleted, these viruses cannot replicate in cells permissive to WT virus infection unless host cell expression of SAMD9/SAMD9L is disrupted [19,22,27].

Though well documented, the molecular mechanism(s) underlying SAMD9/SAMD9L anti-proliferative and anti-viral effect(s) are poorly understood. Recent studies revealed a SAMD9/SAMD9L-dependent translational arrest in virus-infected cells [27,28], which is in line with previous observations that SAMD9 can also be recruited into stress granules [22,27,29]. Similarly, pathogenic variants in SAMD9 and SAMD9L also cause a global translational repression in fibroblastic cell lines as well as hematopoietic progenitor cells [15,16,30]. In the case of SAMD9L, this effect is caused both by missense variants clinically associated with ATXPC syndrome or myeloid malignancies and by SAAD-associated stop-gain variants. Ultimately, such variants cause cell cycle arrest and impede cell proliferation but were also reported to promote defects in cellular DNA damage repair and to cause apoptosis in hematopoietic cells [30]. Importantly, while SAMD9/SAMD9L-dependent translational arrest upon virus infection and in the presence of SAMD9/SAMD9L GoF mutations suggests a common underlying mechanism, the precise molecular mechanism(s) whereby SAMD9/SAMD9L inhibits translation are not known.

Until now, the SAMD9/SAMD9L-antagonistic properties of viral host range factors have chiefly been studied in the context of virus-infected cells and assessed by examining virus production. It was only recently explored to what extent pathogenic GoF variants impact the anti-viral properties of SAMD9/SAMD9L [31]. In addition, it is not known whether viral SAMD9/SAMD9L-antagonistic proteins can neutralize anti-proliferative and translation-inhibiting pathogenic GoF variants. Studies addressing this latter question promise important insights into the (intra-)molecular regulation of the multidomain SAMD9/SAMD9L proteins. Here, we used a co-transfection model to experimentally examine the anti-proliferative and translation-inhibiting properties of SAMD9/SAMD9L variants in presence or absence of virus-derived antagonistic proteins. We find that these antagonistic proteins can principally interact with selected pathogenic missense MIRAGE syndrome- or ATXPC-causing SAMD9 or SAMD9L mutations, respectively, and ameliorate their growth-restrictive and translation-inhibiting effect, with VACV K1 exhibiting the strongest potency.