MEF2C regulates NK cell effector functions through control of lipid metabolism

Abstract

Natural killer (NK) cells are a critical first line of defense against viral infection. Rare mutations in a small subset of transcription factors can result in decreased NK cell numbers and function in humans, with an associated increased susceptibility to viral infection. However, our understanding of the specific transcription factors governing mature human NK cell function is limited. Here we use a non-viral CRISPR-Cas9 knockout screen targeting genes encoding 31 transcription factors differentially expressed during human NK cell development. We identify myocyte enhancer factor 2C (MEF2C) as a master regulator of human NK cell functionality ex vivo. MEF2C-haploinsufficient patients and mice displayed defects in NK cell development and effector function, with an increased susceptibility to viral infection. Mechanistically, MEF2C was required for an interleukin (IL)-2- and IL-15-mediated increase in lipid content through regulation of sterol regulatory element-binding protein (SREBP) pathways. Supplementation with oleic acid restored MEF2C-deficient and MEF2C-haploinsufficient patient NK cell cytotoxic function. Therefore, MEF2C is a critical orchestrator of NK cell antiviral immunity by regulating SREBP-mediated lipid metabolism.

Extended Data Fig. 1 |. Functional CRISPR cRNP screen in primary human NK cells identifies positive regulators of human NK cell function.

(a) Differentially expressed transcription factors between human NK cell developmental subsets. (b) Primary human NK cells are isolated from fresh human PBMCs via negative magnetic bead selection. Following expansion in IL-2/15 for 9 days, cells are electroporated with CRISPR-Cas9 RNP complexes. CRISPR-edited NK cells are further expanded for 6 days before functional and flow cytometric analyses. (c) Left, quantification of percent IFN-γ⁺ in TRAC^cRNP or TCF7^cRNP NK cells after 16 h stimulation with IL-2, IL-15, K562 cells, and IL-12 and/or IL-18. Right, specific lysis of K562 cells after 16 h coculture with IL-2 and IL-15 at indicated effector:target ratios. (d) Left, density of viable TRAC^cRNP or MYC^cRNP NK cells 6 days after cRNP editing expanded with IL-2 and IL-15. Right, quantification of percent IFN-γ⁺ in TRAC^cRNP or MYC^cRNP NK cells after 16 h stimulation with IL-2, IL-15, K562 cells, and IL-12 and/or IL-18. (e) Density of viable TRAC^cRNP or ZEB2^cRNP NK cells 6 days after cRNP editing expanded with IL-2 and IL-15. (f) Quantification of percent IFN-γ⁺ in TRAC^cRNP or RORC^cRNP NK cells after 16 h stimulation with IL-2, IL-15, K562 cells, and IL-12 and/or IL-18. (g) Representative gating strategy for peripheral human NK cells. Data are representative of n = 6–8 independent donors presented as individual paired donors. *p < 0.05, **p < 0.01 by two-sided paired t test. Specific p-values are as follows: c percent IFN-γ⁺ NT = 0.1686, IL-18 = 0.0200, IL-12 = 0.0136, IL-12/18 = 0.8445, % specific lysis 1:8 = 0.0399, 1:4 = 0.0366; d cells/mL = 0.0040, percent IFN-γ⁺ NT = 0.0769, IL-18 = 0.0128, IL-12 = 0.0244, IL-12/18 = 0.4241; e = 0.0216, f = 0.0456.

Extended Data Fig. 2 |. MEF2C is required for human NK cell function without impacting fitness.

(a) Immunoblot showing MEF2C protein expression in TRAC^cRNP or MEF2C^cRNP NK cells 6 days after CRISPR cRNP editing. (b) Indel percentage by CRISPR cRNP editing in TRAC^cRNP or MEF2C^cRNP NK cells 6 days after editing. Sanger sequencing results were analyzed using SYNTHEGO ICE analysis software. (c) Quantification of annexin⁺ early apoptotic or annexin⁺PI⁺ late apoptotic TRAC^cRNP or MEF2C^cRNP NK cells 6 days after editing. (d) MFI of BCL2, BIM, or ratio of BCL2/BIM MFI in TRAC^cRNP or MEF2C^cRNP NK cells 6 days after editing. (e) Specific lysis of A375 human melanoma cells after 16 h coculture with TRAC^cRNP or MEF2C^cRNP NK cells in the presence of IL-2/15. (f) MFI of perforin in TRAC^cRNP or MEF2C^cRNP NK cells 6 days after editing. Data are representative of n = 4⁻7 independent donors presented as individual paired donors. *p < 0.05, **p < 0.01 by two-sided paired t test. Specific p-values are as follows: c = 0.0361, 0.0204; d BCL2 = 0.0047, BIM = 0.2848, BCL2/BIM = 0.1384; e = 0.0405; f = 0.9390.

Extended Data Fig. 3 |. MEF2C haploinsufficiency disrupts CD56^dim NK cells without impacting CD56^bri or other circulating immune populations.

(a) Schematic of pathogenic point mutations in MCHS patients. (b) Expression of MEF2C in FACS-sorted peripheral human immune cells based on DICE database data. (c) Frequency of non-NK cell immune populations in peripheral blood of healthy donor control or MCHS patients. (d) Percent IFN-γ⁺ (above) and IFN-γ MFI of cytokine-producing cells (below) of total (left) or CD56^bri (right) healthy donor control or MCHS patient NK cells stimulated for 16 h with IL-2, IL-15, K562 cells, and IL-12 after 5 d expansion in IL-2/15. (e) MEF2C transcript expression in NK cell maturation subsets. (f-g) MFI of perforin (f) or GzmB (g) in healthy donor or MCHS patient NK cells by maturation subset. (h) MFI of GzmB in healthy donor or MCHS patient CD56^dim NK cells. (i) Schematic showing adenine base editor (ABE8e) mediated generation of MEF2C point mutation. (j) Assessment of point mutation frequency at targeted base using MEF2C-targeting sgRNA in conjunction with electroporation of ABE8e mRNA in healthy primary human NK cells. Point mutation rate was evaluated by Sanger sequencing and analysis using the EditR package on day 6 post ABE8e electroporation in culture with IL-2/15. (c) Left, gated on CD3⁺CD14⁻ T cells, CD3⁻CD14⁺ monocytes, and CD3⁻CD19⁺ B cells. Center, gated on CD3⁺CD14⁻CD4⁺ or CD3⁺CD14⁻CD8⁺ cells. Right, gated on CD3⁻CD14⁺CD16^hi classical, CD3⁻CD14⁺CD16^int intermediate, or CD3⁻CD14⁺CD16^lo non-classical monocytes. (d,f-h) Gated on CD3⁻CD56⁺ cells or CD3⁻CD56^dimCD16⁺ cells. Data represent mean ± SEM. Data are representative of (c,d,f-h) n = 5–8 independent healthy donors alongside n = 2 MCHS patients each sampled two independent times, (e) n = 3⁻4 independent donors, or (j) n = 6 independent donors. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 by two-sided Student’s t test. Specific p-values are as follows: d CD56^bri percent IFN-γ⁺ NT = 0.0010, IL-18 = 0.0029; g = 0.0198; h = 0.0039; j < 0.0001.

Extended Data Fig. 4 |. Mef2c haploinsufficiency impairs antiviral immunity.

(a) Schematic of WT:Mef2c^+/− mixed bone marrow chimeric (mBMC) mice or WT and Mef2c^+/− single bone marrow chimeric (sBMC) mice. (b) Percent WT or Mef2c^+/− bone marrow-derived NK cells on D0 or D42 post bone marrow transplant of WT:Mef2c^+/− mBMC mice. (c) Representative contour plots (left) and frequencies (right) of peripheral NK cell subsets of wild-type:Mef2c^+/− mixed bone marrow chimeric (mBMC) mice after 4 weeks engraftment. (d-e) Percent IFN-γ⁺ (left) and IFN-γ MFI of cytokine-producing cells (right) of splenic NK cells from WT:Mef2c^+/− mBMC mice stimulated ex vivo with IL-15 and IL-12 (d) or IL-18 (e) stratified by maturation subset. (f) Percent IFN-γ⁺ (left) and IFN-γ MFI of cytokine-producing cells (right) of splenic NK cells from WT:Mef2c^+/− mBMC mice on D1.5 post MCMV. (g) Schematic of adoptive transfer of WT:Mef2c^+/− mBMC splenocytes. (h) Schematic of adoptive transfer of CRISPR edited Rosa26^cRNP or Mef2c^cRNP NK cells. (i) Representative contour plots (left) and frequency (right) of Rosa26^cRNP or Mef2c^cRNP Ly49H⁺KLRG1⁺ mouse NK cells on D0 and D7 post MCMV. (j) Percent IFN-γ⁺ of Rosa26^cRNP or Mef2c^cRNP NK cells stimulated for 4 h with IL-12 or IL-18. (k) Percent specific lysis of MC38 β2 M^−/− by Rosa26^cRNP or Mef2c^cRNP NK cells at 1:1 E:T. (l) Representative gating strategy for splenic mouse NK cells. (b-e,j) Gated on CD3⁻TCRβ⁻NK1.1⁺ cells. (f,i) Gated on CD3⁻TCRβ⁻NK1.1⁺Ly49H⁺KLRG1⁺ cells. Data shown as mean ± SEM, paired WT and Mef2c^+/− NK cells from the same mBMC mouse, or paired Rosa26^cRNP and Mef2c^cRNP NK cells from the same mouse where applicable. Data representative of at least 2 independent experiments. Data are representative of (b,c) n = 23, (d,e) n = 11, (f) n = 10, (i,j) n = 4, and (k) n = 7 mice. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 by two-sided paired t test. Specific p-values are as follows: b < 0.0001; c DN = 0.0020, DP < 0.0001, CD11b SP < 0.0001; d CD11b SP percent IFN-γ⁺ = 0.0144, CD11b SP IFN-γ MFI = 0.0469; e CD11b SP percent IFN-γ⁺ = 0.0447; f CD11b SP percent IFN-γ⁺ = 0.0032; DP IFN-γ MFI = 0.0409, CD11b SP IFN-γ MFI = 0.0016; i < 0.0001; j IL-12 percent IFN-γ⁺ = 0.0484, IL-18 percent IFN-γ⁺ = 0.0466; k = 0.0009.

Extended Data Fig. 5 |. MEF2C is required for cytokine-activated metabolic reprogramming.

(a) MFI of CD25/IL-2Rα (left) or CD122/IL-2Rβ (right) of peripheral blood NK cells from WT or Mef2c^+/− sBMC mice on D0 or D1.5 of MCMV infection. (b) MFI of CD25/IL-2Rα (left) or CD122/IL-2Rβ (right) of TRAC^cRNP or MEF2C^cRNP human NK cells 6 days post CRISPR edit. (c) MFI of pSTAT5 (left) or pSTAT1 (center) and representative histogram of pSTAT1 expression (right) in splenic NK cells from WT or Mef2c^+/− sBMC mice on D1.5 of MCMV infection. FMO, fluorescence minus one control. (d) MFI of pSTAT5 (left) and representative histogram (right) of TRAC^cRNP or MEF2C^cRNP human NK cells 6 days post CRISPR edit. FMO, fluorescence minus one control. (e) Maximal respiration of TRAC^cRNP or MEF2C^cRNP human NK cells 6 days post CRISPR edit evaluated by Seahorse extracellular flux assay. (f) Metabolic dependencies (left) and translation rate (right) of TRAC^cRNP or MEF2C^cRNP human NK cells 6 days post CRISPR edit measured by SCENITH. Gluc dep, glucose dependence; Mito dep, mitochondrial dependence; Glyc cap, glycolytic capacity; FAO/AAO cap, fatty acid oxidation/amino acid oxidation capacity; Co, control; DG, 2-deoxyglucose; O, oligomycin. (g) Metabolic dependencies of TRAC^cRNP or MEF2C^cRNP human NK cells 6 days post CRISPR edit measured by SCENITH, stratified by maturation subset. (a) Gated on CD3⁻TCRβ⁻NK1.1⁺ cells. (b,f,g) Gated on CD3⁻CD56⁺ cells. Data represent (a) n = 5–7 mice per group, (b,d) n = 3 paired independent donors, (c) n = 3–4 mice per group, (e) n = 5, or (f,g) n = 4–6 paired independent donors. *p < 0.05, **p < 0.01 by two-sided paired t test. Specific p-values are as follows: f mito dep = 0.0091, glyc cap = 0.0091, Co puro MFI = 0.0140, O puro MFI = 0.0040; g CD16⁻ = 0.0492, 0.0492; CD16⁺CD57⁻ = 0.0491, 0.0322, 0.0322, 0.0491.

Extended Data Fig. 6 |. MEF2C maintains SREBP activity in mouse and human NK cells.

(a) Heatmap showing differentially expressed genes from RNA-seq performed on human and mouse control and MEF2C knockout NK cells. (b) Gene Ontology pathway analysis of downregulated genes conserved between human MEF2C^cRNP and mouse Mef2c^cRNP NK cells compared to control edited cells ranked by FDR. (c) Gene Ontology pathway analysis of downregulated genes conserved between human MEF2C^cRNP and mouse Mef2c^cRNP NK cells compared to control edited cells ranked by FDR, excluding mitosis and cell division-related pathways. (d) Heat maps showing changes in gene expression of canonical SREBP pathway genes separated by biological replicate with hierarchical clustering of genes. (e) Transcript expression of SREBF1 and SCD1 in TRAC^cRNP or MEF2C^cRNP human NK cells 6 days post CRISPR edit. (f) MFI of BODIPY 493/503 of total splenic NK cells from naive and D1.5 MCMV-infected mice. (g-h) MFI of LDLR (g) or BODIPY 493/503 (h) in TRAC^cRNP or LDLR^cRNP human NK cells 6 days post CRISPR edit. (i) Proposed model of MEF2C-directed lipid metabolic reprogramming driving NK cell effector function in response to cytokine stimulation and mTORc1 activation. Data represent mean ± SEM or individual paired donors where applicable. (f) Gated on naïve CD3⁻TCRβ⁻NK1.1⁺ cells or D1.5 CD3⁻TCRβ⁻NK1.1⁺Ly49H⁺KLRG1⁺ cells. (g-h) Gated on CD3⁻CD56⁺ cells. Data are representative of (a-d) n = 6 mice and n = 3 independent donors, (e) n = 6⁻7 independent donors, (f) n = 13 naive and 10 D1.5 mice, or (g-h) n = 6 independent donors. *p < 0.05, **p < 0.01 by two-sided paired t test or Student’s t test. Specific p-values are as follows: e = 0.0108, 0.0421; f = 0.0092; g = 0.0032, h = 0.0485.

Fig. 1 |. MEF2C is required for human NK cell proliferation and effector function.

a, Overview of knockouts of transcription factors regulating NK cell function. b, Density of viable TRAC^cRNP or MEF2C^cRNP NK cells 6 d after cRNP editing, expanded with IL-2 and IL-15. c, Left: representative histograms showing dilution of CTV in TRAC^cRNP or MEF2C^cRNP NK cells on day 6 after cRNP editing. Right: frequency of cells undergoing more than two cell divisions. d, Frequency of Ki-67⁺ cells on day 6 after cRNP editing. e, Representative contour plots (left) and quantification of percent IFN-γ⁺ (center) and IFN-γ mean fluorescence intensity (MFI) of cytokine-producing cells (right) of TRAC^cRNP or MEF2C^cRNP NK cells after 16 h of stimulation with IL-2, IL-15, K562 cells and IL-12 and/or IL-18. NS, not significant; NT, no treatment. f, TNF-α MFI of NK cells stimulated for 16 h with IL-2, IL-15, K562 cells, IL-12 and IL-18. g, Specific lysis of K562 cells by edited NK cells after 16 h of co-culture with IL-2 and IL-15 at the indicated effector:target (E:T) ratios. h, Representative histograms (left) and quantification (right) of GzmB expression in edited NK cells. i, Representative histogram (left) and quantification (right) of CD107a expression in edited NK cells cultured for 4 h with IL-2, IL-15, K562 cells, brefeldin A and monensin in the presence of anti-CD107a–PE antibody. c–f,h,i, Gated on CD56⁺CD3⁻ cells. Data are representative of n = 6–11 independent donors presented as individual paired donors and are shown as mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001 by two-sided paired t-test. Specific P values are as follows. For b, 0.0012. For c, 0.0056. For d, 0.0177. For e, percent IFN-γ⁺, NT (0.8166), IL-18 (0.0018), IL-12 (0.0003), IL-12 and IL-18 (0.0042); IFN-γ MFI, IL-18 (0.0019), IL-12 (0.0033), IL-12 and IL-18 (0.0032). For f, NT, 0.0262; IL-12 and IL-18, 0.0333. For g, 1:4, 0.0031; 1:2, 0.0120; 1:1, 0.0041. For h, 0.0490. For i, percent CD107a⁺, 0.0250; CD107a MFI, 0.0290.

Fig. 2 |. Patients with MCHS present with a functional NK cell deficiency.

a,b, Representative contour plots (a) and quantification (b) showing NK cell maturation in freshly isolated PBMCs from healthy donor controls or patients with MCHS. c, Representative contour plots (left) and quantification of percent IFN-γ⁺ cells (center) and IFN-γ MFI of cytokine-producing cells (right) of CD56^dim healthy donor control or patient NK cells stimulated for 16 h with IL-2, IL-15, K562 cells and IL-12 after 5 d of expansion with IL-2 and IL-15. d, Specific lysis of K562 cells by healthy donor control or patient NK cells co-cultured for 16 h with K562 cells, IL-2 and IL-15 after 5 d of expansion with IL-2 and IL-15. e, Representative histogram of CTV dilution in control or base-edited (BE) human NK cells 6 d after base editing. f, Density of viable control or base-edited NK cells 6 d after base editing. g, Specific lysis of K562 cells by control or base-edited NK cells co-cultured as in d. h, Representative contour plots (left) and quantification of percent IFN-γ⁺ cells (center) and IFN-γ MFI of cytokine-producing cells (right) of control or base-edited NK cells stimulated as in c. a,b,e,f,h, Gated on CD56⁺CD3⁻ cells. c, Gated on CD56^dimCD16⁺CD56⁺CD3⁻ cells. Data represent mean ± s.e.m. or individual paired donors. Data are representative of n = 14 (b), n = 6 (c,d) and n = 7 (e–h) independent healthy donors alongside n = 2 patients with MCHS, each sampled two independent times where applicable. *P < 0.05 by two-sided paired t-test or Student’s t-test. Specific P values are as follows. For b, CD56^briCD16⁻ control versus patient, <0.0001; healthy children versus patient, 0.0138; CD56^briCD16^int control versus patient, <0.0001; healthy children versus patient, 0.0027; CD56^dim control versus patient, <0.0001; healthy children versus patient, 0.0019. For c, percent IFN-γ⁺, NT (0.6644), IL-12 (0.0044); IFN-γ MFI, NT (0.3327), IL-12 (0.0276). For d, 1:2, 0.0020; 1:4, 0.0031. For f, 0.0173. For g, 1:8, 0.0133; 1:4, 0.0146; 1:2, 0.0052; 1:1, 0.021. For h, percent IFN-γ⁺, NT (0.5197), IL-12 (0.0323); IFN-γ MFI (0.0177).

Fig. 3 |. MEF2C haploinsufficiency disrupts antiviral immunity.

a, Percent IFN-γ⁺ (left) and IFN-γ MFI of total NK cells (right) of WT or Mef2c^+/− splenic NK cells stimulated ex vivo for 4 h with IL-15, brefeldin A, monensin, IL-12 and/or IL-18. b, Percent CD107a⁺ WT or Mef2c^+/− bone marrow-derived splenic NK cells stimulated ex vivo with anti-Ly49H antibody for 4 h with IL-15, brefeldin A and monensin. c, Percent change in body weight from day 0 of WT or Mef2c^+/− sBMC mice infected with a sublethal dose of MCMV. d, Kaplan–Meier survival curves of WT or Mef2c^+/− sBMC mice infected with a sublethal dose of MCMV. e, Percent IFN-γ⁺ (left) and IFN-γ MFI of cytokine-producing cells (right) of WT or Mef2c^+/− splenic NK cells in WT:Mef2c^+/− mBMC mice on day 1.5 after MCMV infection. f, GzmB MFI of WT or Mef2c^+/− bone marrow-derived splenic NK cells in WT:Mef2c^+/− mBMC mice on day 1.5 after MCMV infection. g, Representative contour plots showing splenic NK cell ratios on days 0 and 7 (left) and quantification on day 7 (right) in peripheral organs of WT or Mef2c^+/− NK cells cotransferred into male Klra8^−/− hosts and infected with MCMV. a,b, Gated on CD3⁻TCR-β⁻NK1.1⁺ cells. e–g, Gated on CD3⁻TCR-β⁻NK1.1⁺Ly49H⁺KLRG1⁺ cells. Data are representative of at least two independent experiments. Data represent mean ± s.e.m. or paired WT and Mef2c^+/− bone marrow-derived cells from the same mBMC mouse where applicable. Data are representative of n = 11 mice (a,b), n = 23 mice (c,d), n = 10 mice (e,f) and n = 4 (g) mice. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 by two-sided paired t-test (a,b,e–g), two-sided Student’s t-test (c) or Mantel–Cox test (d). Specific P values are as follows. For a, percent IFN-γ⁺, NT (0.0008), IL-15 (0.0109), IL-12 and IL-15 (0.0074), IL-18 and IL-15 (0.0074); IFN-γ MFI, NT (0.0003), IL-15 (0.0723), IL-12 and IL-15 (0.0042), IL-18 and IL-15 (0.0167). For b, NT, 0.5830; Ly49H, 0.0408. For c, 0.0176, 0.0114. For d, 0.0037. For e, percent IFN-γ⁺, <0.0001; IFN-γ MFI, 0.0002. For f, 0.0138. For g, blood, 0.0009; spleen, 0.0016; liver, 0.0004; lung, 0.0125.

Fig. 4 |. MEF2C is required for IL-15- and mTORC1-induced metabolic reprogramming.

a, Immunoblot showing MEF2C and β-actin loading control protein levels in naive human NK cells or cells stimulated with IL-2 and IL-15 for 72 h (left) with quantification (right). b, Immunoblot showing MEF2C and β-actin loading control protein levels in human NK cells stimulated with IL-2 and IL-15 for 72 h alone or with the mitogen-activated protein kinase kinase (MEK) inhibitor (MEKi) AZD6244 (50 μM) or the STAT5 inhibitor (STAT5i) CAS 285986–31-4 (100 μM) (left) with quantification (right). DMSO, dimethylsulfoxide. c, Immunoblot showing MEF2C and β-actin loading control protein levels in human NK cells stimulated with IL-2 and IL-15 for 72 h alone or with the PI3K inhibitor (PI3Ki) LY294002 (50 μM) or the mTORC1 inhibitor rapamycin (rapa; 20 nM) (left) with quantification (right). d, Representative histograms (left) and MFI of phosphorylated (p)AKT (center) and pS6 (right) in TRAC^cRNP or MEF2C^cRNP human NK cells 6 d after CRISPR editing. FMO, fluorescence minus one control. e, OCR and ECAR of TRAC^cRNP and MEF2C^cRNP NK cells measured by the Seahorse extracellular flux assay. O, oligomycin; F, carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP); R/A, rotenone and antimycin A. f, Quantification of basal ECAR and the ratio of OCR to ECAR from e, normalized to the rate per 10,000 cells. Data represent mean ± s.e.m. of technical replicates (e) or n = 6 (a), n = 4 (MEK inhibitor) or n = 6 (STAT5 inhibitor) (b), n = 5 (c), n = 8 (d) or n = 5 (e,f) paired independent donors. *P < 0.05, **P < 0.01 by two-sided paired t-test. Specific P values are as follows. For a, 0.0164. For b, 0.0429. For c, DMSO versus PI3Ki, 0.0061; DMSO versus rapamycin, 0.0224. For e, 0.0327. For f, ECAR, 0.0327; OCR:ECAR, 0.0462.

Fig. 5 |. MEF2C promotes cytokine-activated SREBP signaling and increased lipid content in NK cells.

a, Volcano plots displaying significant DEGs in cRNP-edited human (left) or mouse (right) NK cells compared with non-targeting controls with labeled genes of interest. For human NK cells, RNA samples were from paired TRAC^cRNP and MEF2C^cRNP NK cells from three independent donors. For mouse NK cells, RNA was extracted in triplicate from isolated NK cells from two mice pooled and edited with Rosa26^cRNP or Mef2c^cRNP. b, GSEA of MEF2C^cRNP human NK cells compared with the TRAC^cRNP control. P_adj, adjusted P value. c, RNA-seq on human and mouse control and MEF2C- and Mef2c-knockout NK cells identified 184 conserved DEGs. d, Heatmap showing changes in gene expression of canonical SREBP pathway genes with hierarchical clustering of genes. e, Representative histograms (left) and MFI of BODIPY 493/503 staining (right) in TRAC^cRNP or MEF2C^cRNP NK cells on day 6 after cRNP editing and expansion with IL-2 and IL-15. f, MFI of BODIPY 493/503 in healthy control or MCHS patient NK cells immediately after isolation or after 5 d of expansion with IL-2 and IL-15. g, BODIPY 493/503 MFI of WT or Mef2c^+/− bone marrow-derived splenic NK cells under uninfected conditions (left) or on day 1.5 after MCMV infection (right) in WT:Mef2c^+/− mBMC mice. e,f, Gated on CD3⁻CD56⁺ cells. g, Gated on naive CD3⁻TCR-β⁻NK1.1⁺ or day 1.5 CD3⁻TCR-β⁻NK1.1⁺Ly49H⁺KLRG1⁺ cells. Data are representative of at least two independent experiments. Data represent mean ± s.e.m. or individual paired donors where applicable. Data are representative of n = 6 mice and n = 3 independent donors (a–d), n = 7 independent donors (e), n = 3–8 healthy donors and n = 2 patients with MCHS each sampled two independent times (f) and n = 11 naive and n = 10 MCMV-infected mice at day 1.5 (g). *P < 0.05, **P < 0.01 by two-sided paired t-test (e,g) or two-sided Student’s t-test (f). Specific P values are as follows: 0.0026 (e); day 0 (0.7548), day 5 (0.0019) (f); naive (0.3860), day 1.5 (0.0190) (g).

Fig. 6 |. Fatty acid supplementation restores MEF2C-deficient NK cell cytotoxicity.

a, Representative histograms (left) and MFI (right) of BODIPY C12 uptake in TRAC^cRNP or MEF2C^cRNP NK cells on day 6 after cRNP editing. b, LDLR and Ldlr transcript expression of naive or (3 h) IL-2- and IL-15-activated mouse (left) or human (right) NK cells. c, Ldlr transcript expression of splenic NK cells after MCMV infection. d,e, Representative histograms (d) and percent LDLR⁺ and LDLR MFI (e) of TRAC^cRNP or MEF2C^cRNP human NK cells 6 d after cRNP editing. f, LDLR MFI of naive or (5 d) IL-2- and IL-15-stimulated healthy donor or MCHS patient human NK cells. g, Quantification of percent IFN-γ⁺ (left) and IFN-γ MFI of cytokine-producing cells (right) of TRAC^cRNP or LDLR^cRNP NK cells after 16 h of stimulation with IL-2, IL-15, K562 cells and IL-12 or IL-18. h, Specific lysis of K562 cells by edited NK cells after 16 h of co-culture with IL-2 and IL-15 at the indicated effector:target ratios. i, Specific lysis of K562 cells by untreated healthy donor human NK cells (NT) or cells incubated with 7.5 μg ml⁻¹ methyl-β-cyclodextrin (MβCD) cholesterol for 1 h or 200 μM bovine serum albumin (BSA)-conjugated palmitate or oleate for 24 h after 14 d of expansion with IL-2 and IL-15 at an E:T ratio of 1:2. j, Specific lysis of K562 cells by TRAC^cRNP, MEF2C^cRNP NK cells or MEF2C^cRNP NK cells pretreated with 200 μM BSA-conjugated oleate for 24 h at an E:T ratio of 1:2. k, Specific lysis of K562 cells at the indicated E:T ratios by healthy control cells, MCHS patient NK cells or MCHS patient NK cells pretreated with 200 μM BSA-conjugated oleate for 24 h after 5 d of expansion with IL-2 and IL-15. a,d–g, Gated on CD3⁻CD56⁺ cells. Data are representative of at least two independent experiments. Data represent mean ± s.e.m. or individual paired donors where applicable. Data are representative of n = 7 independent donors (a), n = 3 mice and n = 6 independent donors (b), n = 3 mice (c), n = 6 independent donors (d,e), n = 3–8 healthy donors and n = 2 patients with MCHS each sampled two independent times (e,j), n = 6 (f), n = 7 (g), n = 5 (h) or n = 6 (i) independent donors. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 by two-sided paired t-test or Student’s t-test. Specific P values are as follows. For a, 0.0060. For b, 0.0165. For c, 0.0159, <0.0001, 0.0031. For e, percent LDLR⁺ cells, <0.001; MFI, 0.0002. For f, day 0, 0.8092; day 5, 0.0464. For g, percent IFN-γ⁺ cells, NT (0.9931), IL-18 (0.0147), IL-12 (0.0271); IFN-γ MFI, IL-18 (0.0467). For h, 1:4, 0.0045; 1:2, 0.0014; 1:4, 0.0322. For i, cholesterol, 0.0036; palmitate, 0.0307; oleate, 0.0460. For j, TRAC^cRNP versus MEF2C^cRNP, 0.0191; MEF2C^cRNP versus MEF2C^cRNP and oleate, 0.0016. For k, 1:1, MCHS versus MCHS and oleate, 0.0372; 1:4, healthy control versus MCHS and oleate, 0.0437; 1:4, MCHS versus MCHS and oleate, 0.0229.