Induction of TFEB promotes Kupffer cell survival and reduces lipid accumulation in MASLD

Abstract

Background: Kupffer cells (KCs) are the tissue-resident macrophages of the liver, where they serve a critical role in maintaining liver tissue homeostasis and act as a filter for circulation. The composition of hepatic macrophages changes during metabolic dysfunction-associated liver disease (MASLD), with the loss of resident KCs being a hallmark of disease progression. The mechanism(s) and consequences of KC death in metabolic liver disease have yet to be defined. Transcription factor EB (TFEB) is a master regulator of lysosome function and lipid metabolism, which has been shown to protect macrophages from lipid stress in atherosclerosis. We hypothesized that TFEB would improve KC fitness in MASLD.

Methods: To investigate the potential beneficial effect of TFEB induction in KCs, we created a transgenic mouse in which TFEB was overexpressed specifically in KCs and evaluated its impact on disease pathogenesis in high-fat, high-sucrose (HFHS) and choline-deficient diet models of MASLD.

Results: We found that TFEB induction protected KCs from cell death in both models of MASLD. KC preservation through TFEB induction reduced liver steatosis with HFHS diet via mechanisms that were dependent on macrophage lysosomal lipolysis and mitochondrial fatty acid oxidation. Fibrosis was unchanged in choline-deficient diet studies. TFEB protected KCs from cell death by diminishing oxidative stress and reducing ferroptosis through a mechanism that involved enhanced NADPH levels.

Conclusions: TFEB induction promotes KC fitness upon lipid stress during MASLD. Preservation of lipid-adapted KCs demonstrates beneficial effects against liver steatosis and protects portal filtration during MASLD.

Keywords: de novo lipogenesis; ferroptosis; lipid peroxidation; macrophages; oxidative stress.

Graphical abstract

FIGURE 1

Overexpression of TFEB in KCs enhances lipid uptake and metabolic gene expression. (A) Constructs to generate KC-specific TFEB-overexpressing mice, or KC^Tfeb. (B) Schematics for isolating KCs and BMDMs from KC^Cre and KC^Tfeb animals for gene expression analyses. (C) Gene expression of Tfeb in KCs and BMDMs isolated from KC^Cre and KC^Tfeb mice (n=3/group; technical duplicate for BMDMs). (D) Gene expression of selected targets of TFEB in WT-KCs or TFEB-KCs (n=3/group). (E–H) Characteristics of 8–10-week-old KC^Cre and KC^Tfeb mice. (E) Liver weights and (F) flow cytometric quantification of different macrophage and monocyte populations (n=4–5/group). Circles represent males; triangles represent females. (G) Representative immunofluorescence images of livers showing overall macrophage distribution. Green: F4/80; red: LYVE1; blue: DAPI. Scale bar=100 µm. (H) Representative immunofluorescence images of livers showing KC-specific markers with LSEC-lined sinusoids. Green: TIM4; red: LYVE1; white: CLEC4F; blue: DAPI. Scale bar=50 µm. Insert scale bar=20 µm. (I) Mean fluorescent intensity (MFI) of CD36 in WT or TFEB-KCs quantified by flow cytometry (n=3–5/group). (J) MFI of BODIPY-C₁₆ signal in WT or TFEB-KCs quantified by flow cytometry (n=3/group). Ctrl=no BODIPY-C₁₆ controls. Data represent individual biological replicates and are presented as means±SEM. p-values were calculated using (C) 2-way ANOVA, followed by multiple t tests, (D–F, I) unpaired 2-tailed Student t tests, and (J) one-way ANOVA followed by multiple t tests. NS=not significant, *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001. Abbreviations: BMDMs, bone marrow–derived macrophages; FMO, Fluorescence minus one; KC, Kupffer cell; TFEB, transcription factor EB; WT, wild type.

FIGURE 2

TFEB induction maintains KC numbers and reduces MdM accumulation in MASLD. (A–M) STD versus HFHS diet experiments with 8–10-week-old male KC^Cre and KC^Tfeb littermate mice (n=10–24/group). (A) Schematic of experiment. (B) Final body and organ weights of STD-fed or HFHS diet-fed mice. (C) Flow cytometric quantification of live, single CD45⁺ cells, (D) total liver macrophages (CD45⁺F4/80^hiCD11b^int). (E) Representative flow plots of liver macrophages in HFHS livers. (F) Flow cytometric quantification of KCs (TIM4⁺). (G) Percentage of KCs within total liver macrophages. (H) Changes in KC numbers comparing HFHS-fed KC^Cre and KC^Tfeb littermates with respective STD-fed mice. (I) Representative immunofluorescence images of the livers of HFHS diet-fed mice stained with TIM4 to identify KCs. Green: TIM4; blue: DAPI. Scale bar=50 µm. (J) Flow cytometric quantification of MdMs (TIM4⁻) and (K) MdM subsets, MoKCs (TIM4⁻VSIG4⁺), and LAMs (TIM4⁻VSIG4⁻). (L) Percentages of MoKCs among total TIM4⁻ MdMs. (M) Flow cytometric quantification of Ly6C^hi monocytes (F4/80^loCD11b⁺MHCII⁻Ly6C⁺). (N–R) KC^Cre and KC^Tfeb littermates were fed a 10-week fibrogenic CDAA diet (n=9/group). Circles represent males and triangles represent females. (N) Final body and liver weight. (O) Flow cytometric quantification of CLEC2⁺ liver macrophages (CLEC2⁺F4/80^hiCD11b^int) and (P) KCs (TIM4⁺). (Q) Percentage of KCs in CLEC2⁺ macrophages. (R) Quantification of MdMs (CLEC2⁺TIM4⁻). (S) Relative expression of Tfeb in FACS-purified KCs (TIM4⁺VSIG4⁺) recruited MdMs (CD45⁺F480^hi CD11b^hi/int MHCII⁺ Ly6C⁻ CLEC2^+/−CD11c⁻TIM4⁻VSIG4⁻) flow-sorted from 4-week CDAA diet-fed mice (n=3–4/group). (T) Body weights, liver weights, quantification of KC number and percentages, and quantification of MdMs in Mac^Cre and Mac^Tfeb littermates fed 10 weeks of CDAA diet (n=6-7/group). Circles represent males and triangles represent females. Data represent individual biological replicates and are presented as means±SEM. p-values were calculated using (B–D, F, G, J-K, M) 2-way ANOVA followed by multiple t tests and (H, L, N–T) 2-tailed unpaired t tests. NS=not significant, *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001. Abbreviations: CDAA, choline-deficient amino acid-defined; HFHS, high-fat, high-sucrose, and high-cholesterol; KC, Kupffer cell; LAM, lipid-associated macrophages; MdM, monocyte-derived macrophage; MoKCs, monocyte-derived KCs; MASLD, metabolic dysfunction–associated steatotic liver diseases; STD, standard diet; TFEB, transcription factor EB.

FIGURE 3

TFEB induction in KCs reduces hepatic steatosis. (A–D) Liver steatosis phenotyping for 16-week HFHS diet-fed KC^Cre and KC^Tfeb mice (n=13–21/group). (A) Representative H&E images of livers. (B) Macrovesicular and (C) microvesicular steatosis scores and percentages as determined by a blinded pathologist. (D) Liver TAG measurement. (E) Plasma ALT measurement from 16-week HFHS diet-fed KC^Cre and KC^Tfeb mice (n=12–15/group). (F) qPCR gene expression analyses of selected lipid metabolic genes in whole liver tissues of KC^Cre and KC^Tfeb mice after 16 weeks of STD or HFHS feeding (n=5–7/group). Shown are fold changes relative to STD-fed KC^Cre. (G) Heatmap of normalized (within individual TAG species) relative abundance of 45 TAG species from hepatic tissue, accompanied by quantification of selected species with statistical significance (n=3–6/group). Data represent individual biological replicates and are presented as means±SEM. p-values were calculated using (B–F) unpaired 2-tailed Student t tests, and (H) 2-way ANOVA followed by multiple t tests. NS=not significant, *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001. Abbreviations: HFHS, high-fat, high-sucrose, and high-cholesterol; H&E, hematoxylin and eosin; KCs, Kupffer cells; qPCR, quantitative polymerase chain reaction; STD, standard diet; TAG, triglyceride; TFEB, transcription factor EB.

FIGURE 4

MASLD augments TFEB-mediated lysosomal and lipid metabolic transcriptional programs in KCs. (A) Schematic of experiments to isolate KCs from different conditions for bulk RNAseq (n=2–4/group). (B, C) (B) Venn diagram comparing significantly upregulated DEGs and (C) KEGG and Reactome pathway analyses of the 70 commonly upregulated genes, among which 13 unique genes contributed to the pathway analyses in HFHS TFEB-KCs versus WT-KCs and STD TFEB-KCs versus WT-KCs. (D–F) (D) Volcano plots, (E) KEGG pathways analysis, (F–H) heatmaps of DEGs in TFEB-KCs versus WT-KCs after HFHS feeding. Log₂fold change >0, p-value <0.05. Abbreviations: DEG, differentially expressed gene; HFHS, high-fat, high-sucrose, and high-cholesterol; KCs, Kupffer cells; KEGG, Kyoto Encyclopedia of Genes and Genomes; MASLD, metabolic dysfunction–associated steatotic liver diseases; STD, standard diet; TFEB, transcription factor EB; WT, wild type.

FIGURE 5

TFEB promotes lipid digestion and oxidative phosphorylation in macrophages. (A–F) FACS-purified KCs from HFHS diet-fed KC^Cre and KC^Tfeb mice were subjected to electron microscopy (n=2–3/genotype). (A) Representative electron microscopy images of KCs. Top scale bar=2 µm; bottom scale bar=500 nm. (B) Quantification of the total number and (C) total area of LDs across 30 cells per genotype on HFHS diet. (D) Size distribution of LDs per genotype on HFHS diet. (E) Percentage and (F) count of KCs with lipid-filled vacuoles. (G–J) In vitro experiments using BMDMs derived from Mac^Cre and Mac^Tfeb mice. (G) qPCR gene expression analysis on Tfeb (n=6/group). (H) MFI corrected for baseline fluorescence and (I) representative immunofluorescence images of BMDMs cultured with 250  µM oleic acid conjugated to BSA + 1 µM BODIPY-C₁₆. Green: BODIPY-C₁₆. White: DAPI. Scale bar=20 µm. (J) Cumulative oxygen consumption rate (OCR) for basal and maximum respiration of BMDMs. (K) A graphic representation of potential pathways that TFEB induction could alter in lipid handling. (L–N) FAO, lysosomal lipolysis, and GDF15-deficient KC^Cre or KC^Tfeb mice were fed an HFHS diet for 16 weeks. (L, M) Hepatic TAG species of KC^CreCPT2^fl/fl, KC^TfebCPT2^fl/fl, KC^CreLAL^fl/fl, and KC^TfebLAL^fl/fl mice (n=5/group). (N) Plasma GDF15 in KC^CreGDF15^fl/fl, KC^TfebGDF15^fl/fl mice (n=4–6/group). Data represent (G, L–N) individual biological, or (H–J) technical replicates, and are presented as (B–F) cumulative data or (G, H, J, L, M) means±SEM. p-values were calculated using (G) paired and (H, J, L–N) unpaired 2-tailed Student t tests. NS=not significant, *p<0.05, **p<0.01, and ****p<0.0001. Abbreviations: BMDMs, bone marrow-derived macrophages; BSA, bovine serum albumin; FACS, fluorescence-activated cell sorting; FAO, fatty acid oxidation; HFHS, high-fat, high-sucrose, and high-cholesterol; KCs, Kupffer cells; LP, lipid droplet; MFI, mean fluorescent intensity; qPCR, quantitative polymerase chain reaction; TAG, triglyceride; TFEB, transcription factor EB.

FIGURE 6

TFEB mitigates macrophage death and improves portal filtration. (A) Quantification of Ki67⁺ percentage in KCs after 16 weeks HFHS diet feeding in male KC^Cre and KC^Tfeb mice (n=9/genotype). (B) Schematic of BrdU incorporation experiment where KC^Cre and KC^Tfeb littermates fed a 16-week HFHS diet were intraperitoneally (i.p.) injected with BrdU once a week for 3 weeks, and flow cytometric quantification of BrdU⁺ percentage in cells (n=8–9/genotype). (C, D) MFI of TIM4-BV421 staining on BMDMs isolated from (C) Mac^Cre or Mac^Tfeb mice, and (D) Mac^Tfeb mice loaded with 250 µM oleic acid (OA): BSA for 24 hours, compared with WT-KCs. (E, F) Macrophage depletion experiment in KC^Cre and KC^Tfeb mice. (E) Experimental design. (F) Representative flow plots (pregated on F4/80^hiCD11b^int), average percentage of TIM4⁺ KCs as a percentage of F4/80^hiCD11b^int cells, and quantification of TIM4⁺ KCs by flow cytometry. (G–I) Cx3cr1^CreER-TdT and Cx3cr1^{CreER-TdT-Tfeb} littermates fed HFHS diet and i.p. injected with tamoxifen once a week for 4 weeks before harvest at 16 weeks (n=3–4/group). (G) Schematic of injection regimen, (H) flow cytometric quantification of TdT⁺ percentage in cells, and (I) representative immunofluorescence images of livers. Green: TIM4; red: TdT reporter; blue: DAPI. Scale bar=50 µm. (J) Representative images of TUNEL staining identified in the livers of 16-week HFHS diet-fed KC^Cre and KC^Tfeb mice. Green: TIM4; white: TUNEL; blue: DAPI. Scale bar=20 µm. (K) Male KC^Cre-TdT and KC^Tfeb-TdT mice were placed on HFHS diet for 16 weeks, and cell death was assessed by Zombie Aqua positivity in total KCs or TdT⁺ KCs (n=8–10/genotype). (L) KC^Cre and KC^Tfeb mice were placed on a CDAA diet for 10 weeks, and cell death was assessed by Zombie Aqua positivity in total KCs (n=9/genotype). Circles represent males; triangles represent females. (M–P) KC^Cre and KC^Tfeb male mice were fed HFHS for 16 weeks, and livers were in situ injected with fluorescent beads (n=4–5/group). (M) Schematic of experiment and representative flow histogram of bead signal in KCs. (N) Percentage of KCs with single or multiple bead-positive signals. (O) Representative immunofluorescence images of bead capturing in KCs. Red: beads; green: CLEC4F; blue: DAPI. Scale bar=50 µm. (P) Total beads captured by liver myeloid cells and monocytes per gram of tissue. Values were determined by multiplying cells of interest per gram of tissue by the bead-positive percentage. Data represent individual biological replicates and are presented as means±SEM. p-values were calculated using (A, B, K, L, N) unpaired 2-tailed Student t tests, (C, D) one-way ANOVA followed by multiple t tests, and (F, H, P) 2-way ANOVA followed by multiple t tests. NS=not significant, *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001. Abbreviations: BMDMs, bone marrow-derived macrophages; BSA, bovine serum albumin; CDAA, choline-deficient amino acid-defined; KCs, Kupffer cells; HFHS, high-fat, high-sucrose, and high-cholesterol; MFI, mean fluorescent intensity; TFEB, transcription factor EB; WT, wild type.

FIGURE 7

TFEB-mediated protection of macrophages in vivo and in vitro. (A–C) Flow cytometric quantification of KCs per gram of liver and percentages isolated from various mouse lines fed. (A) HFHS diet for 16 weeks (n=6–18/group) or (B, C) CDAA for 8–10 weeks (n=4–12/group); circles represent males, and the triangles represent females. (D) MFI of BODIPY-C₁₁ staining in KCs isolated from KC^Cre and KC^Tfeb mice fed HFHS for 16 weeks (n=8–10/group). AF=autofluorescence of KCs determined from unstained samples. (E–M) In vitro experiments using BMDMs isolated from Mac^Cre and Mac^Tfeb mice. Flow cytometric percentage of (E) propidium iodide⁺ (PI⁺) signal and (F) oxidized BODIPY-C11⁺ signal in cells treated with 5 µM RSL3±5 µM Fer1 for 24 hours and 3 hours, respectively. Vehicle=DMSO. (G) Cells were stimulated with 20 µM zVAD and 100 ng/mL LPS±20 µM necrostatin-1 (Nec1, necroptosis inhibitor) for 24 hours, and cell death (necroptosis) was measured by PI. Vehicle=DMSO/PBS. (H) Cells were stimulated with 500 ng/mL LPS for 4 hours, then 5 mM ATP for 1 hour, and cell death (pyroptosis) was measured by PI. Vehicle=PBS. (I) MFI of CellRox Deep Red signal in WT-BMDMs or TFEB-BMDMs treated with RSL3±Fer1 for 3 hours. Representative histograms are shown. (J) Cells were stimulated with 2.5 mM H₂O₂ for 2 hours, and cell death was measured by PI. (K) NADP⁺ and NADPH levels in BMDMs. (L, M) Assessment of cytosolic NADPH flux through isotope tracing. (L) Schematic illustration of [3-²H] glucose incorporation into proline as a tracer strategy to assess cytosolic NADPH flux. (M) Relative ratio of proline to glucose-6-phosphate (G6P) with ²H labeling. (N) Schematic of de novo lipogenesis. (O) Gene expression of DNL-related enzymes in BMDMs. (P) Change in oxidized BODIPY-C₁₁ signal between inhibitor-treated and vehicle-treated WT-BMDMs 3 hours post-RSL3 stimulation. Cells were pretreated with respective inhibitors for 16 hours before co-treatment with RSL3. (Q, R) Ex vivo experiments with primary KCs isolated from male KC^Cre and KC^Tfeb mice (n=3–4/genotype). (Q) NADP⁺ and NADPH levels in primary KCs. (R) Gene expression of DNL-related enzymes in primary KCs. Data represent (A–D, Q, R) individual biological replicates, (E–K, M, P) technical replicates, or (O) biological duplicates with technical triplicates, and are presented as means±SEM. p-values were calculated using (A–D, I, K, M, O, Q, R) unpaired 2-tailed Student t tests, (E–H, J) 2-way ANOVA followed by multiple t tests, and (P) 1-way ANOVA followed by multiple t tests. NS=not significant, *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001. Abbreviations: BMDM, bone marrow–derived macrophages; CDAA, choline-deficient amino acid-defined; DNL, de novo lipogenesis; HFHS, high-fat, high-sucrose, and high-cholesterol; KCs, Kupffer cells; LPS, lipopolysaccharide; MFI, mean fluorescent intensity; PI, propidium iodide; SFA, saturated fatty acid; TFEB, transcription factor EB; UFA, unsaturated fatty acid; WT, wild type.