Intercellular Mitochondria Transfer to Macrophages Regulates White Adipose Tissue Homeostasis and Is Impaired in Obesity

Abstract

Recent studies suggest that mitochondria can be transferred between cells to support the survival of metabolically compromised cells. However, whether intercellular mitochondria transfer occurs in white adipose tissue (WAT) or regulates metabolic homeostasis in vivo remains unknown. We found that macrophages acquire mitochondria from neighboring adipocytes in vivo and that this process defines a transcriptionally distinct macrophage subpopulation. A genome-wide CRISPR-Cas9 knockout screen revealed that mitochondria uptake depends on heparan sulfates (HS). High-fat diet (HFD)-induced obese mice exhibit lower HS levels on WAT macrophages and decreased intercellular mitochondria transfer from adipocytes to macrophages. Deletion of the HS biosynthetic gene Ext1 in myeloid cells decreases mitochondria uptake by WAT macrophages, increases WAT mass, lowers energy expenditure, and exacerbates HFD-induced obesity in vivo. Collectively, this study suggests that adipocytes and macrophages employ intercellular mitochondria transfer as a mechanism of immunometabolic crosstalk that regulates metabolic homeostasis and is impaired in obesity.

Keywords: beige adipose tissue; brown adipose tissue; horizontal mitochondria transfer; immunometabolism; intercellular mitochondria transfer; macrophage; metabolism; mitochondria; obesity; white adipose tissue.

Figure 1.. Macrophages acquire mitochondria from other cell types in vivo.

(A) Experimental design. (B) Frequency of live donor-derived WT immune cells that are mtD2⁺ in recipient epididymal white adipose tissue (eWAT). Pre-gated on live CD45.1⁺ CD45.2⁻ cells. (C) Cell types represented within the population of WT donor cells that are mtD2⁺. (D) Frequencies of WT donor-derived macrophages that are mtD2⁺ in eWAT. Pre-gated on live CD45.1⁺ CD45.2⁻ CD64⁺ F4/80⁺ macrophages in eWAT. (E) Frequencies of the indicated donor-derived immune cells that are mtD2⁺ in eWAT. (F) Frequencies of donor-derived macrophages that are mtD2+ in eWAT compared to inguinal (i)WAT and brown adipose tissue (BAT). Data represent 2-4 independent experiments with n=4-8 mice/group combined and are expressed as mean ± standard error of the mean (SEM). Student’s t-tests (B and D), one-way ANOVA with Fisher’s post-hoc test compared to macrophages (E), and two-way ANOVA with Fisher’s post-hoc test (F), *P<0.05, **P<0.01, ***P<0.001. See also Supplemental Figure S1.

Figure 2.. Intravital 2-photon microscopy reveals mitochondria transfer in white adipose tissue in vivo.

(A) Experimental design. (B) Low power magnification showing host adipocytes (dashed circular line), host-derived mtD2⁺ mitochondria (green, arrows), and a wildtype CMTMR-labelled macrophage (red). Scale bar, 10 μm. (C) High power magnification Z-stack of donor macrophage, showing yellow host-derived mitochondria within a donor macrophage (mac). Scale bar, 5 μm. (D) Three-dimensional reconstruction and positional mapping of host cell-derived mtD2⁺ mitochondria in a donor macrophage. Horizontal and vertical lines show coordinates. (E) Time lapse imaging revealing a mitochondria uptake event in WAT in vivo. Representative of n=4 mice from 2 independent experiments with n=2 mtD2 controls without adoptive transfer.

Figure 3.. Adipocytes transfer mitochondria to macrophages in white adipose tissue in vivo, defining a distinct macrophage subpopulation.

(A) Frequencies and (B) numbers of live CD45⁺ CD11b⁺ CD64⁺ macrophages per gram of eWAT from control (Con, n=9) and adipocyte-specific mitochondria reporter mice (MitoFat, n=7). (C) Wildtype eWAT macrophages (mac) were labelled with CMTMR and adoptively transfer into MitoFat mice for intravital 2-photon microscopy of host eWAT. Three-dimensional reconstruction (left) and positional mapping (right) of donor macrophage containing adipocyte-derived mitochondria (arrows). (D) Gating strategy to define macrophages that have (mtD2+) or have not (mtD2-) acquired mitochondria from adipocytes in vivo in MitoFat mice (n=5/group) for comparison of fluorescence intensity of (E) the charge-independent mitochondrial mass indicator MitoID-Red, (F) the charge-dependent mitochondria dye MitoTracker Red CMX Rosamine (CMXRos) normalized to MitoID-Red, (G) the charge-dependent dye MitoSOX Red normalized to CMXRos. (H) Comparison of mtD2+ and mtD2− eWAT macrophage phagocytosis of 1 μm polyred latex beads (n=6/group). For E-H, lines connect paired data points obtained from the same sample. (I-L) Messenger RNA sequencing of mtD2+ and mtD2− macrophages from eWAT of MitoFat mice (n=4). (I) Principal component analysis with dashed lines connecting paired samples from the same mouse and shaded ellipses defining 95% confidence intervals for each group. (J) Volcano plot of detected transcripts. (K-M) Gene Set Enrichment Analyses showing enrichment of the HIF-1α/TF pathway (K) and de-enrichment of genes associated with electron transport (L) and collagen synthesis (M). Data expressed as mean ± standard error of the mean. Student’s t-test (A-B) and paired t-test (E-H), *P<0.05, ***P<0.001. See also Supplemental Figure S2.

Figure 4.. Genome-wide CRISPR knockout screen identifies heparan sulfates as essential for mitochondria uptake by macrophages.

(A) Experimental design of genome-wide CRISPR-Cas9 knockout screen. (B) STARS Score analyses identifies 23 enriched genes with false discovery rate (FDR) <0.05. (C) List of 23 enriched genes, of which red genes are required for heparan sulfate biosynthesis. (D) GO Term enrichment analysis of the 23 genes identified by the CRISPR screen. (E) Schematic relating the 13 heparan sulfate synthesis genes enriched in the CRISPR screen, tracing the entire HS biosynthesis pathway unsupervised. (F) Surface heparan sulfate levels, (G) mtD2 mitochondria uptake, and (H) 1 μm red latex bead phagocytosis in 13 clonally selected BV2-Cas9 cell lines, with sgRNA targeting the indicated genes. For F-H, n=3 independent experiments, one-way ANOVA with Dunnett post-hoc test comparing against the sgRNA to Gfp control, *P<0.05, **P<0.01, ***P<0.001. (I) Mitochondria uptake after pre-treating BV2 cells with heparanases I-III (5 U/mL each), n=3/group. (J) Mitochondria uptake by BV2 cells after pre-incubation of mtD2 mitochondria with 1 mg/mL heparan sulfate proteoglycan (HSPG), n=4/group. (K) Frequencies of mtD2⁺ macrophages in eWAT from MitoFat mice after 7 days of treatment with PBS or heparin (5 mg/kg, n=3/group). For I- K, Student’s t-test, ***P<0.001. *P<0.05, **P<0.01, ***P<0.001. For F-K, data are expressed as mean ± SEM. See also Supplemental Figure S4.

Figure 5.. Intercellular mitochondria transfer to macrophages is impaired in obesity and following stimulation with IFN-γ and LPS.

(A-B) Male 6-week-old MitoFat mice were fed a chow or high fat diet (HFD) for 12 weeks, and live CD45⁺ SiglecF⁻ CD11b⁺ CD64⁺ macrophages in eWAT macrophages were identified by flow cytometry. The proportion of mtD2+ macrophages are shown. N=6-8/group from 3 independent experiments. (C) Frequency of WT recipient macrophages that are mtD2+ after WAT explant co-culture with congenically disparate mtD2 donor WAT. C, chow diet; H, high fat diet; n=5/group from 2 independent experiments combined. (D) Relative surface heparan sulfate levels on eWAT macrophages in mice fed a chow or HFD for 10-12 weeks, representative of n=6-8 mice/group pooled from 2 independent experiments. For A-D, Student’s t-test, ***P<0.001. (E) Frequencies of mtD2+ cells in CD206⁺ CD11c⁻ and CD206⁻ CD11c⁺ macrophages from eWAT of MitoFat mice fed a chow (n=4) or HFD (n=3) for 12 weeks. (F) Pearson linear regression correlating the proportion of mtD2+ eWAT macrophages defined in panel E to their relative HS levels. Grey, chow; black, HFD. (G-I) BV2 cells were stimulated with phosphate buffered saline (PBS), 20 ng/mL interleukin (IL-4), or 10 ng/mL interferon (IFN)-γ plus 1 ng/mL lipopolysaccharide (LPS) for 24hr before co-culturing with 1 μm polyred latex beads and purified mtD2 mitochondria. Frequencies of cells that took up (G) beads and (H) mitochondria. (I) mRNA expression of the indicated HS biosynthesis genes in BV2 cells. For G-I, n=4 independent experiments, one-way ANOVA with Fisher’s LSD post-hoc test. *P<0.05, **P<0.01, ***P<0.001. Data are represented as mean ± SEM. See also Supplemental Figure S5.

Figure 6.. Genetic deletion of Ext1 in myeloid cells impairs mitochondria transfer to macrophages and is associated with increased fat mass and glucose intolerance.

(A) Frequencies and (B) numbers per gram of eWAT macrophages in Ext1^ΔLyz2 mice (n=15) and Ext1^F/F littermate controls (n=14) at steady state. Pre-gated on live CD45⁺ SiglecF⁻ cells. (C) Relative surface heparan sulfate levels on eWAT macrophages in Ext1^ΔLyz2 mice (n=6) and Ext1^F/F littermate controls (n=6). (D) Frequencies of Ext1^ΔLyz2 (n=7) and Ext1^F/F (n=8) eWAT recipient macrophages that are mtD2+ after eWAT explant co-culture with CD45.1 mtD2 donor WAT. (E) Body weight, (F) absolute eWAT mass, and (G) relative eWAT mass normalized to body weight in Ext1^ΔLyz2 mice (n=21) and Ext1^F/F littermate controls (n=24). (H) Whole body lean and fat masses, (I) adiposity, (J) glucose tolerance tests, and (K) insulin tolerance tests in 5-6-month-old Ext1^ΔLyz2 mice (n=8) and Ext1^F/F littermate controls (n=11). For A-I, Student’s t-test. For J-K, two-way ANOVA with repeated measures with Fisher’s LSD post-hoc test. *P<0.05, **P<0.01, ***P<0.001. Data are represented as mean ± SEM.

Figure 7.. Deletion of Ext1 in myeloid cells is associated with decreased energy expenditure and increased susceptibility to diet-induced obesity.

(A-E) Metabolic cage analyses of Ext1^ΔLyz2 mice (n=19) and Ext1^F/F littermate controls (n=17). Energy expenditure expressed as heat produced per kg/body weight per hr at (A) 5.5 min resolution and (B) on average for light vs dark phases. (C) Respiratory quotient (RQ) during the light vs dark phases. (D) Food intake and (E) physical activity levels. (F) Body weight, (G) indicated adipose tissue depot masses, and (H) glucose tolerance tests of 5-6-month old Ext1^ΔLyz2 mice (n=14) and Ext1^F/F littermate controls (n=12) fed a HFD for 9 weeks. For A-C and H, two-way ANOVA with repeated measures and LSD post-hoc test. For D-F, Student’s t-test. For G, two-way ANOVA with LSD post-hoc test. *P<0.05, **P<0.01. Data are represented as mean ± SEM.