A Novel STK4 Mutation Impairs T Cell Immunity Through Dysregulation of Cytokine-Induced Adhesion and Chemotaxis Genes

Abstract

Purpose: Human serine/threonine kinase 4 (STK4) deficiency is a rare, autosomal recessive genetic disorder leading to combined immunodeficiency; however, the extent to which immune signaling and host defense are impaired is unclear. We assessed the functional consequences of a novel, homozygous nonsense STK4 mutation (NM_006282.2:c.871C > T, p.Arg291*) identified in a pediatric patient by comparing his innate and adaptive cell-mediated and humoral immune responses with those of three heterozygous relatives and unrelated controls.

Methods: The genetic etiology was verified by whole genome and Sanger sequencing. STK4 gene and protein expression was measured by quantitative RT-PCR and immunoblotting, respectively. Cellular abnormalities were assessed by high-throughput RT-RCR, RNA-Seq, ELISA, and flow cytometry. Antibody responses were assessed by ELISA and phage immunoprecipitation-sequencing.

Results: The patient exhibited partial loss of STK4 expression and complete loss of STK4 function combined with recurrent viral and bacterial infections, notably persistent Epstein-Barr virus viremia and pulmonary tuberculosis. Cellular and molecular analyses revealed abnormal fractions of T cell subsets, plasmacytoid dendritic cells, and NK cells. The transcriptional responses of the patient's whole blood and PBMC samples indicated dysregulated interferon signaling, impaired T cell immunity, and increased T cell apoptosis as well as impaired regulation of cytokine-induced adhesion and leukocyte chemotaxis genes. Nonetheless, the patient had detectable vaccine-specific antibodies and IgG responses to various pathogens, consistent with a normal CD19 + B cell fraction, albeit with a distinctive antibody repertoire, largely driven by herpes virus antigens.

Conclusion: Patients with STK4 deficiency can exhibit broad impairment of immune function extending beyond lymphoid cells.

Keywords: Antibody repertoire; Combined immunodeficiency; Human serine/threonine kinase 4 (STK4) deficiency; Interferon; T cell lymphopenia; Transcriptomics.

Fig. 1

Identification of a homozygous STK4 gene mutation in a patient from consanguineous parents. A Pedigree and segregation of the STK4 gene mutation. The patient (P) is homozygous for the mutation. Question marks (?) indicate individuals whose genetic status could not be evaluated. B Electropherograms of partial sequences of STK4 corresponding to the mutation in a healthy control (bottom), patient (up), and a STK4^wt/mut relative (middle) representative of all three healthy family members. The reference vs. altered nucleotide position is indicated by a black arrow. C Schematic illustration of the protein encoded by the STK4 gene, with domain boundaries and other features retrieved from the UniProtKB (www.uniprot.org) (entry Q13043). Blue arrows indicate previously reported variants [–, –13, 20]. The variant identified in P is indicated in red. CC, coiled coil domain; SARAH, Sav/Rassf/Hpo domain (IPR024205). D Data from the gnomAD database were used to plot minor allele frequency (MAF) against the Combined annotation-dependent depletion (CADD) score values of all known variants in STK4 and the variant identified in P. E Western blot analysis of STK4 protein expression in PBMC-derived T lymphocytes from P, two STK4^wt/mut heterozygous relatives (R1 and R2) and two unrelated STK4^wt/wt, healthy controls (C1 and C2); a-tubulin and b-actin antibodies were used as controls

Fig. 2

Leukocyte subsets in the STK4-deficient patient, his parents and sibling, and one unrelated healthy control. For all experiments, subjects are presented in the following order from left to right: Unrelated control, the patient’s three relatives, and the patient (P). A Frequency of B (CD3⁻CD19⁺) and T (CD3⁺CD19⁻) lymphocytes among CD45⁺ lymphocytes. B Frequency of T lymphocytes (CD3⁺) and NK cell immunophenotyping, showing the frequency of CD56^bright (CD3⁺CD56^bright) and CD56^dim (CD3⁺CD56^dim) NK cells among CD45⁺ lymphocytes. C Frequency of cytotoxic (CD3⁺CD8⁺) and helper (CD3⁺CD4⁺) T lymphocytes among the CD3⁺ lymphocyte subset. D Frequency of PD-1⁺ T lymphocytes (CD4⁺PD-1⁺) among the CD4⁺ T cell subset. E Frequency of naïve (CD45RA⁺CCR7⁺), central memory (CD45RA⁻CCR7⁺), effector memory (CD45RA⁻CCR7⁻) and effector memory cells re-expressing CD45RA (T_EMRA) (CD45RA⁺CCR7⁻) cells among the CD4⁺ T cell compartment. F Frequency of CD27⁺ and CD28⁺ T helper subsets within the CD4⁺ compartment. G Frequency of myeloid dendritic cells (mDCs) (CD123⁻CD11c⁺) and plasmacytoid dendritic cells (pDCs) (CD123⁺CD11c⁻) among the CD45⁺HLA-DR⁺CD3⁻CD14⁻CD19⁻, CD20⁻CD56⁻ dendritic cell population

Fig. 3

Microbial exposure profile and antiviral antibody repertoire in the STK4-deficient patient. A Antibody profile in the STK4^−/− patient (P), his STK4^WT/− family members (R1, R2, and R3), and two unrelated STK4^WT/WT controls (C1 and C2). Pooled human plasma was used for intravenous immunoglobulin therapy (IVIg); human IgG-depleted serum (IgG-depl.) and mock-IP samples served as additional controls. All samples were assayed in duplicate, and the results are derived from one experiment. Heatmap shows species-specific adjusted score values, which served as a quantitative measure of the number of antibody specificities targeting a given microbial species. B Bar plot depicting, for each sample shown in A, the number of species for which peptides were significantly enriched by PhIP-Seq (i.e., at least one antibody specificity was detected) (light blue) and the number of species for which the adjusted virus score values passed the significance cut-off (i.e., the sample was considered seropositive for that given species) (dark blue). C Principal component (PC) analysis of the -log₁₀(P-values) of significantly enriched peptides for each sample as shown in A. D Scatter plot showing the contribution of the significantly enriched peptides to PC1 and PC2 in the patient’s sample. Peptides depicted in color correspond to species for which more than two peptides were enriched and had a delta (PC1-PC2) in excess of 70th percentile (top)

Fig. 4

Unique gene expression signature in whole blood samples from the STK4-deficient patient following in vitro stimulation. Heatmap showing the log₂-transformed fold change values (log2FC) of the differentially expressed genes (DEGs) among the 180 target genes for which transcriptional responses of the patient’s (P) whole blood samples to in vitro stimulation showed a variance of |log2FC|> 1 compared to those of the other family members (R1, R2, and R3) and an unrelated control (C1). Gene-stimuli pairs are grouped according to the functional annotation of the gene cluster as described previously [16]. Results were obtained from one experiment. The target genes represented 60 functionally annotated transcriptional modules (i.e., sets of co-expressed genes), with each module represented by three target genes. Full gene names and functional annotation are detailed in Supplementary Tables S3 and S4

Fig. 5

Gene enrichment analyses of IFN-a/IFN-b or PMA/ionomycin-responsive genes in PBMCs obtained from the STK4^−/− patient (P), his STK4^WT/− family members (R1, R2, and R3), and three unrelated STK4^WT/WT controls (C1, C2, and C3). Results were generated from one mRNA-Seq experiment. Each condition was assayed in duplicate. A Heatmap shows functionally grouped GO and KEGG/BioCarta pathway annotation networks (ClueGO) (P < 0.05 (BH correction); FDR < 0.05) that encompass either STK4 (red font), or at least two genes from the STK4-interacting gene set (black font) (see Supplementary Figure S3C for a representation of STK4 interaction partners). Percentage of associated genes shown as a color gradient. Circle sizes represent the adjusted P-values for the gene enrichment analyses. B Heatmap shows the log₂-transformed fold change values (log2FC) for genes that are part of selected gene networks shown in A. Upregulation and downregulation of genes is shown as a red and blue gradient, respectively. Genes that are part of the human interferome network [19] are labeled in green font; STK4 is labeled in blue font. C Analysis of regulatory effects (IPA) of IFN-a/IFN-b-responsive genes that are dysregulated in the patient’s PBMCs (P < 0.05 (Fishers exact t-test); z-score > 2; consistency score 18.98). The top panel of nodes in the network graph depicts the predicted upstream regulators; the middle panel depicts the selected gene set, and the bottom panel depicts the best matching downstream effect. Solid cyan edges depict indirect relationships between nodes of the network and STK4 (Ingenuity Knowledge Base, Qiagen). D Scatter plot shows the log2FC values of the analyzed genes in the patient versus the mean FC values of the unrelated controls after stimulation with IFN-a/IFN-b. Orange symbols indicate responsive genes for which regulation was considered different in the patient compared to the controls (ratio < 0 or > 2). Genes that form part of the regulatory network in C are annotated. E Heatmap shows the results of a canonical pathway comparison analysis (IPA). The color gradient depicts z-score values. Pathways with P < 0.05 and z-score > 2 were considered significantly regulated. Red and blue indicate activated and repressed pathways, respectively. Only canonical pathways that were found to be differentially activated/repressed in the patient relative to the control subjects are shown