ACK1 and BRK non-receptor tyrosine kinase deficiencies are associated with familial systemic lupus and involved in efferocytosis

Abstract

Systemic lupus erythematosus (SLE) is an autoimmune disease, the pathophysiology and genetic basis of which are incompletely understood. Using a forward genetic screen in multiplex families with SLE, we identified an association between SLE and compound heterozygous deleterious variants in the non-receptor tyrosine kinases (NRTKs) ACK1 and BRK. Experimental blockade of ACK1 or BRK increased circulating autoantibodies in vivo in mice and exacerbated glomerular IgG deposits in an SLE mouse model. Mechanistically, NRTKs regulate activation, migration, and proliferation of immune cells. We found that the patients' ACK1 and BRK variants impair efferocytosis, the MERTK-mediated anti-inflammatory response to apoptotic cells, in human induced pluripotent stem cell (hiPSC)-derived macrophages, which may contribute to SLE pathogenesis. Overall, our data suggest that ACK1 and BRK deficiencies are associated with human SLE and impair efferocytosis in macrophages.

Keywords: ACK1; BRK; genetic diseases; genetics; genomics; human; immunology; inflammation; mouse; systemic lupus.

Figure 1.. NRTK compound heterozygous missense variants in two multiplex families with SLE.

(A, B) Pedigrees and Sanger re-sequencing of DNA from patients and healthy relatives of kindred 1 (A) carrying K161Q and A156T ACK1 mutations and kindred 2 (B) carrying G257A and G321R BRK mutations. Individuals with SLE are indicated by black shapes; deceased individuals are shown by a diagonal bar; shapes with thick outline indicate the members analyzed by WES; squares indicate males, circles indicate females, and hexagons indicate generation I or II individuals with undisclosed sex for confidentiality. Black: Guanine, green: Adenine, red: Thymidine, blue: Cytosine. Arrows indicate nucleotide substitutions. Text indicates amino-acid substitutions. (C) Domain architecture(top panel) of ACK1 and BRK, with indicated mutations. SH2, Src homology 2; SH3, Src homology 3; Kinase, tyrosine kinase domain; C, Cdc42 binding domain; PR, Proline rich domain; SAM, Sterile α motif. Alignment of kinase domains (bottom panel) from ACK1 and BRK orthologs. Arrows indicate positions of mutations and stars indicate the amino acids conserved throughout species. (D) Three dimensional (3D) structures of ACK1 and BRK. Top: the crystal structure of ACK1 in a complex with AMP-PCP (PDB ID: 1U54). The mutated residues (A156 and K161) are shown in red, and the nucleotide analog is in green. Bottom: the crystal structure of BRK in a complex with the ATP-competitive inhibitor dasatinib (PDB ID: 5H2U). The mutated residues (G257 and G321) are shown in red, and dasatinib is in green.

Figure 1—figure supplement 1.. Candidate genes identified by WES analysis in family 1 and family 2.

(A) Candidate genes identified in family 1. For a complete penetrance, three models of inheritance were applied: homozygous, X-linked, and compound heterozygous. Non-synonymous coding mutation with MAF <0.01 were selected and reported for each model of inheritance. The red box shows the ACK1 mutants identified. Total MAF and maximal MAF are reported. Amino acid positions are based on transcript ENST00000333602.11. (B) Candidate genes identified in family 2. For a complete penetrance, two models of inheritance were applied: homozygous and compound heterozygous. Non-synonymous coding mutation with MAF <0.01 were selected and reported for each model of inheritance. The red box shows the PTK6 mutants identified. Total MAF and maximal MAF are reported. Amino acid positions are based on transcript ENST00000542869.2. (C) ACK1 transcripts with NM number reported in Ensembl. ACK1 mutant positions in the respective transcripts are reported. (D) ACK1 and BRK mutant alleles frequency across different ethnic subgroups. Analysis of 27 GWAS studies of SLE (https://www.gwascentral.org/) found no common variants in the close vicinity of the ACK1 (TNK2) or BRK (PTK6) genes with a p-value lower than 5×10^–8, the threshold of significance for GWAS.

Figure 1—figure supplement 2.. Haplotype member’s family 2.

SNPs close to the G321R mutation are reported for the three patients, the healthy sister, the healthy mother and deduced for the healthy father.

Figure 1—figure supplement 3.. SLE-causing genes.

(A) List of genes already described as disease-causing genes in Lupus. (B) Synonymous and non-synonymous mutations with a MAF <0.01 in the coding region of known SLE-causing genes identified in the 10 kindreds. Only mutations segregating with disease in the kindreds were selected.

Figure 2.. ACK1 and BRK mutations are null and hypomorph alleles.

(A) Immunoprecipitation (IP) kinase assay. ACK1 (left panel) was immunoprecipitated from 293T cells expressing Flag-tagged ACK1 wild type (WT), ACK1 A156T, or ACK1 K161Q with anti-Flag Ab. Immunoprecipitated proteins were used for duplicate in vitro kinase reactions with WASP synthetic peptide. Samples of the immunoprecipitates were also probed with anti-Flag Ab. BRK (right panel) was immunoprecipitated as above from 293T cells expressing Flag-tagged BRK WT and mutants with anti-Flag Ab. Kinase reactions were performed with peptide AEEEIYGEFEAKKKG, and represented as above, and samples of the immunoprecipitates probed with anti-Flag Ab. n=2 per condition. p-values were calculated using an Anova test (Tukey’s multiple comparisons test). (B) Western blot analyses of lysates from 293T cells expressing (left panel) Flag-tagged WT or mutant forms ACK1 probed with anti-ACK1 Tyr(P)²⁸⁴ (PY284), anti-Flag and anti-tubulin antibodies (Ab), and expressing (right panel) Flag-tagged WT or mutant BRK probed with anti-BRK Tyr(P)³⁴² (PY342), anti-Flag and anti-tubulin antibodies. For BRK, 293T cells were starved overnight, and stimulated with 100 ng/ml EGF for 10 min. The lysate from WT BRK indicated as low was from cells transfected with one-tenth the amount of WT DNA. (C) Western blot analyses of lysates from 293T cells expressing ACK1-Flag or BRK-Flag treated with AIM100 or Cpd4f and probed with anti-ACK1 Tyr(P)284 (PY284) or anti-BRK Tyr(P)342 (PY342) and anti-Flag antibodies. (D) May-Grunwald-Giemsa staining of iPSC-derived macrophages from familial controls and ACK1 and BRK patients. Scale bar 10 μm, 100 X objective. Representative images from over 50 observed cells per line. (E) Immunoprecipitation (IP) kinase assay in patients’ macrophages. (Left panel) ACK1 was immunoprecipitated from BRK^WT/G321R, ACK1^WT/K161Q and ACK1^A156T/K161Q iPSC-derived macrophages with anti-ACK1 Ab. The immunoprecipitated proteins were used in duplicate for in vitro kinase reactions with WASP synthetic peptide. Samples of the immunoprecipitates were also probed with anti-ACK1 Ab and anti-tubulin Ab. (Right panel) BRK was immunoprecipitated from ACK1^WT/K161Q, ACK1^A156T/K161Q, BRK^WT/G321R and BRK^G257A/G321R iPSCs-derived macrophages with anti-BRK Ab. The immunoprecipitated proteins were probed with anti-BRK Tyr(P)³⁴² (PY342) and anti-BRK antibodies.

Figure 2—figure supplement 1.. Generation and characterization of control and patient derived iPSCs and iPSC-macrophages.

(A) Schematic representation of the process of reprograming peripheral blood mononuclear cells (PBMCs) into induced pluripotent stem cells (iPSCs). See Materials and methods for details. (B) Sanger sequencing of reprogrammed iPSC lines from unrelated wild type (WT) control (C12), familial controls (1-II-3 ACK1^WT/K161Q and 2-II-3 BRK^WT/G321R), and SLE patients (1-III-1 ACK1^A156T/K161Q and 2-III-3 BRK^G257A/G321R). Gray bars indicate positions of nucleotide substitution. (C) Nucleic acid and predicted amino acid sequences for both gene copies of TNK2(AKC1) or PTK6(BRK) in iPSC lines from C12 and SLE families. Variations from WT are indicated in red. (D) Karyotypes of iPSC lines from C12 and SLE families. Chromosome analysis was performed on a minimum of 20 DAPI-banded metaphases. (E) Flow cytometry analysis of surface receptor expression on iPSC-derived macrophages from C12, familial controls, and SLE patients iPSC lines. Histograms show fluorescence intensity for indicated antibodies (red) and FMO controls (grey). (F) Bar plots show percent viability based on DAPI staining of macrophages derived from C12, familial controls, and SLE patients iPSC lines. n≥12, from three independent experiments. (G) RT-QPCR analysis of mRNA for BRK, ACK1, TIM4, and MERTK. mRNA expression was normalized to GAPDH (2^-ΔCt; see Materials and methods). n≥3, from one to two independent experiments. p-Values were obtained using an Anova test with Tukey’s correction for multiple comparisons.

Figure 2—figure supplement 2.. Homozygote mutant alleles reported in public database gnomAD with MAF >0.005 do not affect the kinase activity of ACK1.

(A, B) ACK1 and BRK homozygous mutants reported in gnomAD database (A) CADD score (y axis) plotted against minor allele frequency (MAF, x axis) for the mutations found in our patients and homozygous ACK1 missense and LOF variations described in the gnomAD database. The black line corresponds to the mutation significance cutoff (MSC). ACK1^A156T and ACK1^K161Q missense variations are annotated and shown as red diamonds. Variations with MAF >0.005 (right side dotted line) and CADD/MSC high (>1, above black line) have been tested biochemically and shown as filled black circle. (B) CADD score (y axis) plotted against minor allele frequency (MAF, x axis) for the mutations found in our patients and homozygous BRK missense and LOF variations described in the gnomAD database. The black line corresponds to MSC. BRK^G257A and BRK^G321R missense mutants are annotated and shown as red diamonds. (C) ACK1^R99W, ACK1^R877H, ACK1^R1038H and ACK1^V890M don’t impair ACK1 kinase activity. Immunoprecipitation kinase assay (on the left). ACK1 was immunoprecipitated from HEK293T cells expressing Flag-tagged ACK1 (WT and mutants) with anti-Flag Ab. The immunoprecipitated proteins were used in duplicate in vitro for kinase reactions with WASP synthetic peptide. Western blot analysis (on the right). Lysates from 293T cells expressing Flag-tagged WT or mutant forms (R99W, R877H and R1038H) of ACK1 were probed with anti-ACK1 Tyr(P)²⁸⁴ (PY284), anti-Flag and anti-tubulin antibodies (Ab). (D) Table indicates f parameter value for negative selection in the human genome from SnIPRE (Eilertson et al., 2012), lofTool (Fadista et al., 2017), evoTol (Rackham et al., 2015), and CoNeS (Rapaport et al., 2021) metrics, as well as intraspecies metrics from RVIS (Petrovski et al., 2013), LOEUF (Karczewski et al., 2020), and pLI / pRec (Lek et al., 2016).

Figure 3.. ACK1 and BRK blockade induces autoimmunity in mice.

(A) Heatmaps comparing the levels of IgG autoantibodies detected in serum of mice treated with inhibitors. Heatmaps show autoantigen microarray panels performed on serum from 4-month-old BALB/cByJ female mice which received a weekly intra-peritoneal injection of DMSO (vehicle, 20 µl/mice), AIM100 (25 mg/kg in 20 µl), or Cpd4f (20 mg/kg in 20 µl) since the age of 5 weeks. Top panel depicts results for mice that did not receive a pristane injection. Bottom panel represents results for the top differentially produced auto-antibodies in inhibitor treated or control mice that received a Pristane injection at the age of 5 weeks. Plotted values represent Ab Scores (Log₂ [antigen net fluorescence intensity (NFI) x signal-to-noise ratio (SNR) +1]). Heatmap columns represent serum analysis of independent mice (n=4–5 for each of the 3 conditions). Heatmap rows sorted top to bottom starting with most significantly increased Ab Score in Cpd4f and AIM100 mice in comparison to DMSO-treated mice. p-Values were calculated using a Wilcoxon matched-pairs signed rank tests. Hierarchical clustering is based on one minus Pearson correlation with complete linkage method. K-means clustering is based on Euclidean distance, with two clusters, with 10,000 maximum iterations. (B,C) Immunofluorescence for mouse IgG on kidney sections. Representative micrographs (B) displaying glomeruli on kidney sections from 4-month-old BALB/cByJ female mice treated as in (A) and stained with Hoechst 33342, anti-mouse IgG, and anti-mouse Podoplanin antibody. In the quantification plot (C) each symbol represents the IgG mean fluorescence intensity (MFI) in a single glomerulus, of mice treated with designated inhibitors, in the presence or absence of pristane. Approximately 250 glomeruli we analyzed per section/mouse (>95% of all glomeruli in an entire longitudinal kidney section). n=4–5 mice per condition. p-Values were obtained using a Kruskal-Wallis test with multiple comparisons.

Figure 3—figure supplement 1.. Extended analysis of ACK1 and BRK inhibitor treated mice.

(A) Heatmaps showing a comparison of serum autoantibodies detected in 4-month-old BALB/cByJ female mice. Mice received an intraperitoneal pristane injection and a weekly intraperitoneal injection of DMSO (vehicle, 20 µl/mice), AIM100 (25 mg/kg in 20 µl), or Cpd4f (20 mg/kg in 20 µl/) since the age of 5 weeks. Plotted values represent Ab Scores (Log₂ [antigen net fluorescence intensity (NFI) x signal-to-noise ratio (SNR) +1]). Heatmap columns represent serum analysis of independent mice (n=4–5 for each of the 3 conditions). Heatmap rows sorted top to bottom starting with most significantly increased Ab Score in Cpd4f and AIM100 mice in comparison to DMSO-treated mice. p-Values were calculated using a Wilcoxon matched-pairs signed rank tests. (B) Additional immunofluorescence images for mouse IgG on kidney sections. Representative images of glomeruli displaying the average observed signal on kidney sections from 4-month-old BALB/cByJ female mice which were left untreated or received a single intraperitoneal injection of pristane, and received a weekly intraperitoneal injection of DMSO (vehicle, 20 µµ/mice), AIM100 (25 mg/kg in 20 ul), or Cpd4f (20 mg/kg in 20 µl/) since the age of 5 weeks. Kidney sections were stained with Hoechst 33342, anti-mouse IgG, and anti-mouse Podoplanin antibody. Representative images from 250 glomeruli analyzed per kidney section/mouse (>95% of all glomeruli in an entire longitudinal kidney section) where n=4–5 mice per condition.

Figure 4.. ACK1 and BRK kinase deficiency disrupts the anti-inflammatory response driven by apoptotic cells in macrophages.

(A) Western blot analysis for AKT phosphorylation by ACK1 and BRK. Cell lysates from 293T cells were incubated with anti-AKT. Immunoprecipitated proteins were probed with anti-phosphotyrosine and anti-AKT antibodies. (B) Western blot analysis for STAT3 phosphorylation by ACK1 and BRK. Lysates from 293T cells coexpressing STAT3 and Flag-tagged WT or mutant forms (A156T and K161Q) of ACK1 or mutant forms (G257A and G321R) of BRK were probed with anti-phospho-STAT3 (Tyr705), anti-STAT3 and anti-Flag antibodies. For analysis of BRK, cells were treated with 100 ng/ml EGF for 10 min. (C) RAC activation by WT ACK1 and BRK. Cell lysates from 293T cells expressing WT or mutant forms of ACK1 (left) and lysates from 293T cells expressing WT or mutant forms of BRK (right) were incubated with GST-PAK CRIB sepharose beads, and the level of RAC1 GTP was determined by immunoblotting with anti-Rac1 antibody. Lysates were also probed with anti-Rac1, anti-FLAG and anti-tubulin antibody. For analysis of BRK, 293T cells were cotransfected with CAS and stimulated with 100 ng/ml EGF for 10 min. (D) MERTK increases kinase activity of BRK and ACK1. IP kinase assay. ACK1 (left) was immunoprecipitated from 293T cells co-transfected with Flag-tagged ACK1 WT, ACK1 A156T, or ACK1 K161Q and MERTK with anti-Flag Ab. Immunoprecipitated proteins were used in duplicate in vitro for kinase reactions with WASP synthetic peptide and results represented as pmol phosphate transferred. BRK (right) was immunoprecipitated as above from 293T cells co-transfected with Flag-tagged BRK WT or mutants and MERTK with anti-Flag Ab. Kinase reactions was performed with peptide AEEEIYGEFEAKKKG, and represented as above. p-Values were calculated using an Anova test (Tukey’s multiple comparison test). (E) Regulation of inflammatory response. Significant normalized enrichment scores (NES) for GO ‘positive regulation of acute inflammation’ gene set, GO ‘negative regulation of inflammatory response’ gene set, and GO ‘AKT_UP.V1_UP’ gene set in WT and mutant macrophages, and WT treated with AIM100 (2 µM) or Cpd4f (0.5 µM), exposed to apoptotic cells, with three replicates per experimental condition. Significant enrichment (p-value <0.05 and FDR (q-value) <0.25) are calculated as reported in Materials and methods. (F) Table of the top 10 differentially regulated genes by apoptotic cells in WT macrophages are not differentially expressed in mutant macrophages and WT macrophages treated with AIM100 or Cpd4f (treated as in E). Numbers indicate FDR (q-value). Known target genes of STAT3 and AKT are labeled in blue and red respectively (G) TNF mRNA production by WT macrophages treated with AIM100 (2 µM) 4 hr after exposure to apoptotic cells. n=6, from two independent experiments. (H,I) TNF and IL1β production by macrophages, as measured by ELISA on media collected from mutant and isogenic WT macrophages (C12.1) incubated with mouse apoptotic thymocytes for 90 min, then stimulated with LPS (1 ng/ml) for 18 hr. n≥4, from ≥2 independent experiments. p-Values in H were calculated by Wilcoxon matched-pairs signed rank tests for data that is not normally distributed, while p-values in G and I were calculated using an Anova test with Tukey’s correction for multiple comparisons.

Figure 4—figure supplement 1.. Principal component analysis (PCA) and differentially expressed genes in RNA sequencing datasets.

(A) PCA on all samples included in the RNA sequencing dataset (treated as in Figure 4E). n=2–3 per condition. (B) Table shows the number of significantly (adjusted p-value ≤0.05) upregulated and downregulated genes for specified comparisons.

Figure 4—figure supplement 2.. Generation and characterization of isogenic control and mutant iPSCs and iPSC-macrophages.

(A) Schematic representation of the CRISPR-CAS9 gene editing process to derive C12 isogenic lines with TNK2 or PTK6 mutations. See Materials and methods for details. Abbreviations: single guide RNA (sgRNA); Single stranded donor oligonucleotides (ssODN). (B) Sanger sequencing of CRISPR-CAS9 modified isogenic iPSC lines. Gray bars indicate position of nucleotide substitutions. (C) Nucleic acid and predicted amino acid sequences for both gene copies of TNK2(AKC1) or PTK6(BRK) in CRISPR-CAS9 modified isogenic iPSC lines. Variations from WT are indicated in red. One dash represents one nucleotide deletion. (D) Karyotypes of CRISPR-CAS9 modified isogenic iPSC lines. Chromosome analysis was performed on a minimum of 20 DAPI-banded metaphases. (E) Representative brightfield images of Isogenic WT, BRK and ACK1 mutant macrophages (left panel) or WT macrophages treated with designated inhibitors for 2 hr (right panel). Similar morphology observed in n>5 independent experiments. (F) Bar plots show percent viability based on Hoechst staining of macrophages derived from isogenic iPSC lines, and percent viability of C12.1 WT macrophages after a 120 min treatment with designated inhibitors. n≥3 replicates per experimental condition. (G) Flow cytometry analysis of surface receptor expression on iPSC-derived macrophages from isogenic WT and mutant iPSC lines. Histograms show fluorescence intensity for indicated antibodies (red) and FMO controls (grey).

Figure 5.. ACK1 and BRK kinase deficiency alter actin remodeling at the phagocytic cup and modestly decrease engulfment of apoptotic cells in macrophages.

(A) Actin remodeling in macrophages. Schematic and representative images of of F-actin by TIRF microscopy in macrophages of indicated genotype, deposited on PtdSer-coated plates for 20 min. (B) Quantification of actin clearance factor for macrophages of the indicated genotypes. Actin remodeling (actin clearance factor) was calculated as a ratio of F-actin staining intensity at cell border divided by F-actin staining intensity at cell center. The actin clearance factor ratios were normalized to the mean value of WT control. Each replicate indicates actin clearance factor fold change from WT mean in single cells. n>20, from two independent experiments. Red lines denote the mean. (C, D) Actin remodeling quantification (as in A,B) and representative TIRF images of WT macrophages (C12.1 line) pretreated with DMSO, AIM100 (2 µM) or Cpd4f (0.5 µM). n>24, from three independent experiments. p-Values in B-D were obtained using a Mann-Whitney test. (E, F) Uptake of apoptotic cells. (E) Schematic depicts uptake of apoptotic mouse thymocytes treated with Fiz-shFASL and labeled with the pH-sensitive dye pHrodo by iPSC-derived macrophages. Isogenic WT (C12.1 line) and isogenic ACK and BRK point mutant macrophages were incubated with pHrodo-labeled mouse apoptotic thymocytes for 90 min and analyzed by flow cytometry. Graph represents mean pHrodo fluorescence Intensity (MFI) fold change calculated by dividing total pHrodo MFI (610/20 nm) of individual samples by the average MFI of isogenic WT macrophages. n≥3, from three independent experiments. p-Values were obtained using an Anova test with Tukey’s correction for multiple comparisons. (F) Uptake of apoptotic cells as in (E) with WT macrophages (C12.1 line) pretreated with AIM100 (2 µM), R-9b (4 µM), Cpd4f (0.5 µM), or DMSO. n≥8, from ≥4 independent experiments. (G) Uptake of opsonized sheep red blood cells. WT macrophages (C12.1 line) are pretreated as in (F) and incubated with opsonized pHrodo sheep red blood cells for 90 min. Graphs represent mean fluorescence Intensity (MFI) fold change calculated by dividing total pHrodo MFI (610/20 nm) of individual samples by the average MFI of WT macrophages. n≥2, from two independent experiments. p-Values are obtained using an Anova test with Tukey’s correction for multiple comparisons. (H) Schematic representation of ACK1 and BRK proposed function in efferocytosis.

Figure 5—figure supplement 1.. Engulfment of beads, Escherichia coli, and Candida albicans.

(A) Phagocytosis of 2 µm beads. iPSC-macrophages (C12 line) pretreated with DMSO, AIM100 (2 µM), Cpd4f (0.5 µM), or both, were incubated for 60 min at 37°C or 4 °C with red fluorescent 2 µm beads. Quantification of uptake is represented as fold change in mean fluorescence intensity (MFI) (670/30 nm) between DMSO-treated and inhibitor-treated macrophages (left panels). Quantification of uptake as percent red fluorescent (bead) positive macrophages (right panels). n=2 per experimental condition. (B–C) Phagocytosis of E. coli and C. albicans. Phagocytosis quantification of pHrodo labeled E. coli (B) or tdTomato fluorescent C. albicans (C) by iPSC-macrophages (C12 line) pretreated with DMSO, AIM100 (2 µM), Cpd4f (0.5 µM), or both, after a 60 min incubation at 37 °C or 4 °C. MFI fold change determined by dividing total MFI (585/15 nm) of individual samples by the average MFI of DMSO treated macrophages. n=2 per experimental condition. p-Values in all plots were calculated using an Anova test (Tukey’s multiple comparisons test).