A Novel Non-Coding Variant in DCLRE1C Results in Deregulated Splicing and Induces SCID Through the Generation of a Truncated ARTEMIS Protein That Fails to Support V(D)J Recombination and DNA Damage Repair

Abstract

Severe Combined Immune Deficiency (SCID) is a primary deficiency of the immune system in which opportunistic and recurring infections are often fatal during neonatal or infant life. SCID is caused by an increasing number of genetic defects that induce an abrogation of T lymphocyte development or function in which B and NK cells might be affected as well. Because of the increased availability and usage of next-generation sequencing (NGS), many novel variants in SCID genes are being identified and cause a heterogeneous disease spectrum. However, the molecular and functional implications of these new variants, of which some are non-coding, are often not characterized in detail. Using targeted NGS, we identified a novel homozygous c.465-1G>C splice acceptor site variant in the DCLRE1C gene in a T^-B^-NK⁺ SCID patient and fully characterized the molecular and functional impact. By performing a minigene splicing reporter assay, we revealed deregulated splicing of the DCLRE1C transcript since a cryptic splice acceptor in exon 7 was employed. This induced a frameshift and the generation of a p.Arg155Serfs*15 premature termination codon (PTC) within all DCLRE1C splice variants, resulting in the absence of full-length ARTEMIS protein. Consistently, a V(D)J recombination assay and a G0 micronucleus assay demonstrated the inability of the predicted mutant ARTEMIS protein to perform V(D)J recombination and DNA damage repair, respectively. Together, these experiments molecularly and functionally clarify how a newly identified c.465-1G>C variant in the DCLRE1C gene is responsible for inducing SCID. In a clinical context, this demonstrates how the experimental validation of new gene variants, that are identified by NGS, can facilitate the diagnosis of SCID which can be vital for implementing appropriate therapies.

Keywords: ARTEMIS; DNA damage repair; NGS; SCID; V(D)J recombination.

Figure 1

Flow cytometry analysis of PBMCs of a SCID patient shows a lack of T and B lymphocytes (A) Flow cytometry analysis of PBMCs of six healthy children and the index patient (III-2). Bars indicate the mean of the corresponding populations with error bars as the standard deviation. HC, Healthy Control; PBMC, Peripheral Blood Mononuclear Cells. (B–F) Pseudocolor plots of CD3⁺CD4⁺ and CD3⁺CD8β⁺ T cells, CD19⁺HLADR⁺ B cells, CD56⁺ NK cells and CD14⁺CD4⁺ monocytes in both a representative healthy control and the index patient. Numbers indicate the percentage of the gated populations. HC, Healthy Control. (G) Flow cytometry analysis of the CD56 expression level on peripheral NK cells, corresponding to the pseudocolor plots in (E). Numbers indicate the geometric mean fluorescence intensity (MFI) of CD56 expression.

Figure 2

Genetic analyses reveals a novel c.465-1G>C non-coding variant of the DCLRE1C gene (A) Sanger electropherogram of both the index patient and parents covering the region of the c.465-1G>C variant in the DCLRE1C gene. The location of the novel variant is indicated by the red arrow. (B) Pedigree of the affected family showing the prevalence of the variant across multiple generations. The index patient is depicted as a black circle and indicated by the black arrow. The genotype of every available family member is denoted. M, mutation; WT, wild-type. (C) Schematic representation of the canonical DCLRE1C transcript and the canonical ARTEMIS protein. The DCLRE1C transcript contains 14 coding exons. The ARTEMIS protein contains 3 protein domains consisting of 692 amino acids. The location of the variant is indicated by the red arrow.

Figure 3

Molecular study of the DCLRE1C gene and ARTEMIS protein reveals an alternative splicing mechanism leading to a frameshift mutation and a PTC. (A) Schematic representation of the generation of the pSplice Express vector containing the patient DCLRE1C minigene with the c.465-1G>C variant. The minigene containing exons 6, 7, 8 and corresponding introns is cloned in between the two rat insulin exons by XhoI and NotI compatible overhangs. The location of the variant is indicated by the red arrow. (B) cDNA Sanger sequencing of the minigenes of both control and patient, expressed by host HEK cells. Sanger electropherograms of both minigenes are depicted and the loss of the AG dinucleotide at the start of exon 7 is indicated by the red arrow. (C) Schematic representation of the pre-mRNA, mRNA and ARTEMIS protein containing the altered sequence resulting from the alternative splicing. On the pre-mRNA level, the new cryptic splice acceptor site in exon 7 is depicted in blue. The canonical splicing pattern is shown by a solid black line whereas the alternative splicing mechanism is indicated by a dashed red line. The location of the c.465-1G>C variant is depicted in red. On the mRNA and protein level, the altered sequence that results in a frameshift and a PTC is shown in red. The PTC is depicted as an asterisk on the truncated ARTEMIS protein (*). (D) Schematic representation of the truncated ARTEMIS protein, as predicted by the alternative splicing event. (E) Western blot analysis for the ARTEMIS protein (upper panel) and for beta-ACTIN (lower panel) on total thymus, HC fibroblast and patient fibroblast lysates (n=3). HC, Healthy Control. (F) RT-PCR of DCLRE1C cDNA of 3 independent fibroblast controls and patient fibroblasts. The upper panel shows the 5’ RT-PCR of DCLRE1C cDNA, with 3 bands each representing a different DCLRE1C transcript corresponding to the transcripts shown in Supplemental Figure 2 . The middle panel shows the 3’ RT-PCR of DCLRE1C cDNA. The lower panel depicts the RT-PCR of ACTB as a positive control. HC, Healthy Control; PTC, Premature Termination Codon.

Figure 4

In vitro V(D)J recombination assay shows the inability of the truncated ARTEMIS protein to contribute to V(D)J recombination. (A) Flow cytometry analysis of the EGFP expression of transfected healthy control and patient fibroblasts with the corresponding constructs as indicated above the pseudocolor plots. EGFP expression is a direct read-out of V(D)J recombination of the substrate. Numbers indicate the EGFP⁺ percentage of the corresponding gated population. As a positive control, recombined substrates are provided (‘Flipped control’). Plots shown are representative for all replicates (n=3). HC, Healthy Control. (B) Graph showing the EGFP⁺ percentage of transfected control and SCID fibroblasts with the corresponding constructs. Bars indicate the mean of the 3 independent experiments that are shown by the individual data points and error bars indicating the standard deviation (n=3). ***indicates p < 0.001 and *indicates p < 0.05 resulting from a two-tailed paired t-test. HC, Healthy Control.

Figure 5

Truncated patient-derived ARTEMIS protein is unable to contribute to DNA repair in a G0 micronucleus assay. (A) Schematic representation of the G0 micronucleus assay. Fibroblasts are irradiated at the G0 phase of the cell cycle before addition of cytochalasin B to block cytokinesis. Upon nuclear division, the micronuclei (MN) resulting from acentric chromosomal fragments can be counted relative to the amount of binucleate cells (BN). (B) Graph showing the relative abundance of micronuclei compared to binuclei in 3 independent healthy control and patient fibroblast lines. Cells are irradiated with increasing intensities of Gray (Gy). Bars indicate the mean of the 3 independent experiments that are shown by the individual data points and error bars indicating the standard deviation (n=3). **indicates p < 0.01 and * indicates p < 0.05 resulting from a two-tailed paired t-test. HC, Healthy Control.