Nanomedicine targeting PPAR in adipose tissue macrophages improves lipid metabolism and obesity-induced metabolic dysfunction

Abstract

Excess body fat leads to an overabundance of adipose tissue macrophages (AT MΦs) with altered phenotypes that play pathogenic roles in obesity comorbidities including diabetes and cancer. Peroxisome proliferator-activated receptors (PPARs) are leading targets to modulate AT MΦ phenotype. Here, we developed a dextran-based nanomedicine that delivers PPARα/γ agonists to AT MΦs and improves obesity and diabetic phenotypes in vivo. Within 1 week of treatment, AT MΦs decreased and became lipid laden, while extracellular vesicles secreted from AT decreased and reduced in lipid content. Within 2 weeks, glucose tolerance returned to levels of lean controls, followed by weight loss and reduced food intake. After 4 weeks, AT browning and amelioration of hepatic steatosis were evident. The physiological shifts were reproducible in three rodent models of obesity, spanning sexes and gonadal status. Effects were enhanced for the targeted nanomedicine compared with free drugs at equivalent doses, supporting the hypothesis that targeted PPAR activation in AT MΦs benefits systemic metabolism.

Fig. 1.. Dextran-conjugated PPAR agonist D-PPAR targets MΦs in AT in vivo and alters gene expression of metabolic regulators in MΦs in vitro.

(A) Dextran distribution by PET (color) overlay with x-ray computed tomography (CT; black and white). A male DIO mouse was imaged 1, 4, or 24 hours after left-side intraperitoneal injection of 30-nm dextran-NOTA-⁶⁴Cu radiochelate. PET/CT coronal planes point show distribution in left perirenal (PR) and left gonadal (GD) AT. 3D, three-dimensional. (B) Tissue distribution of four sizes of dextran-NOTA-⁶⁴Cu 24 hours after left-side intraperitoneal injection in male DIO mice by gamma well counting. Tissues include heart, blood, liver, subcutaneous AT (SubQ), and visceral AT depots: left and right PR, left and right GD, and mesenteric (Mes) (n = 4 to 6). Data for 3.5-nm dextran were previously published (33). (C) D-PPAR [30-nm tetramethyl rhodamine isothiocyanate (TRITC) conjugate] is specific to CD11b⁺ immune cells from AT in DIO mice by flow cytometry. (D) D-PPAR drug release mechanism. (E) Hydrodynamic diameter of dextran and D-PPAR by dynamic light scattering. (F) Drug release from D-PPAR in phosphate-buffered saline (PBS) at 37°C. (G) Metabolic gene expression in RAW264.7 MΦs polarized to the M1 phenotype by lipopolysaccharide (LPS) and interferon-γ (IFN-γ) shows up-regulation of Sirt1, Cd36, Ucp2, Ucp3, and Ifnβ after treatment with D-PPAR or F-PPAR after 24 hours. Expression is measured by reverse transcription quantitative polymerase chain reaction (RT-qPCR) normalized to the housekeeping gene Tbp by the ΔΔCt method. Letters indicate P < 0.05 for mean differences by one-way analysis of variance (ANOVA) with Tukey’s post hoc test. Asterisks are included as a redundant notation of letter conventions that indicate groups with significant differences, i.e., “a” indicates significant difference from “b” but neither “a” nor “b” is significantly different from “ab” (97). Data are presented as mean ± SEM (n = 3). Rel., relative.

Fig. 2.. Treatment of male DIO mice with D-PPAR reduces body weight, reduces adiposity, and improves metabolic phenotype.

The study design is shown in (A). Compared to other groups, treatment of DIO mice by intraperitoneal injection of D-PPAR led to significant (B) loss in body weight after 2 weeks, (C) reduced cumulative food intake over 4 weeks, (D) loss of adipose tissue weight in subcutaneous (SC), PR, and mesenteric (MS) fat pads after 4 weeks, and (E) improved glucose tolerance by week 2, which preceded weight loss. (F) Treatment with D-PPAR significantly improved hepatic steatosis scores after 4 weeks, with (G) representative liver images showing reduced numbers and sizes of lipid droplets. Scale bars, 100 μm. Blinded scoring metrics for liver are described in Materials and Methods. D-PPAR treatment over 4 weeks led to (H) moderate reduction in liver triglyceride (TG) but no impact on (I) circulating TG or (J) liver toxicity marker ratio of aspartate transaminase to alanine transaminase (AST/ALT). Data are presented as mean ± SEM. Letters denote significance of P < 0.05 by one-way ANOVA with Tukey’s post hoc test or Kruskal-Wallis test (n = 7 to 8). See Fig. 1G caption for additional description of statistical notations.

Fig. 3.. D-PPAR treatment of male DIO mice induces AT browning.

(A) Gene expression in AT of DIO animals treated with D-PPAR for 4 weeks, showing shifts in genes associated with AT metabolic regulation. Expression is measured by RT-qPCR and normalized to the mean of two housekeeping genes, Ppia and Rps13, by the ΔΔCt method. Exp., expression. (B) Representative hematoxylin and eosin (H&E)–stained AT micrographs from DIO mice showing MΦ infiltration and CLSs (red arrows) in control obese mice and features of AT browning in D-PPAR–treated mice, including eosinophilic microenvironments (yellow arrows) and adipocyte shrinkage. Scale bars, 100 μm. (C) Representative fluorescence micrographs of whole-mount AT stained with MitoTracker (red), 4,4-difluoro-4-bora-3a,4a-diaza-s-indacene (BODIPY, green), and Hoechst (blue). Scale bars, 50 μm. (D) Percentage of area of AT stained with MitoTracker indicating density of mitochondria. (E) UCP1 and (F) UCP2 increase in AT with D-PPAR treatment, indicating browning of AT. O, obese; F, F-PPAR; D, D-PPAR. (G) PPARγ coactivator 1α (PGC-1α) is nonsignificantly up-regulated in AT with D-PPAR treatment (P = 0.263). Western blot insets show example images from full blots shown in fig. S5; β-actin controls are the same in (F) and (G) but different from those in (E). (H and I) Treatment of adipocytes in vitro with osteopontin (OPN; protein product of Spp1) led to (H) enhanced mitochondrial respiration and (I) reduced fat content. Different letters indicate significance of P < 0.05 by one-way ANOVA with Tukey’s post hoc test (n = 6 to 8). See Fig. 1G legend for additional description of statistical notation. RFU, relative fluorescence units. Abs., absorbance.

Fig. 4.. D-PPAR decreases AT MΦ numbers, increases MΦ lipid accumulation, and reduces lipid-laden EV secretion from AT in male DIO mice.

MΦ populations are reduced in visceral AT after 1-week treatment with D-PPAR, including (A) CD45⁺CD11b⁺, (B) CD45⁺CD11b⁺Ly6C⁺, (C) CD45⁺CD11b⁺Ly6C⁻F4/80⁺, and (D) CD45⁺CD11b⁺Ly6C⁻F4/80⁺CD9⁺ by flow cytometry. (E) Flow cytometry gating strategy for leukocytes (CD45), myeloid cells (CD11b), and myeloid subtype markers Ly6C, F4/80, and CD9. Full gating strategy is in fig. S8A and described in Materials and Methods. FSC-A, forward scattering area. PE, phycoerythrin. (F to I) Neutral lipid content in MΦ subtypes following F-PPAR or D-PPAR treatment measured by mean fluorescence intensity (MFI) of BODIPY by flow cytometry. Values correspond to each cell population in panels [(A) to (D)]. Example dataset is in fig. S8A and described in Materials and Methods. (J) Neutral lipid content in SVF cells treated with D-PPAR in vitro, showing confocal fluorescence micrograph of BODIPY dye (green) with nuclear stain (blue). Quantification is in fig. S6. (K) Confocal fluorescence micrograph of whole-mount AT of mice treated with TRITC–D-PPAR, showing colocalization of AT MΦs (red), TRITC (white), and small lipid droplets (BODIPY; green). Representative image from untreated control is shown in fig. S9. (L) TRITC–D-PPAR uptake corresponds to lipid-rich cells (BODIPY) in SVF CD45⁺CD11b⁺ populations measured by flow cytometry. Results are from the same experiment as in (K). Example dataset is in fig. S8B and described in Materials and Methods. (M) One-week D-PPAR treatment reduces EV secretion by AT resected from DIO mice via nanoparticle tracking analysis (NTA; scattering). (N) EVs did not differ in size. (O and P) F-PPAR and D-PPAR reduced lipid-laden EVs (NTA; BODIPY fluorescence). Different letters denote significance of P < 0.05 by one-way ANOVA with Tukey’s post hoc test (n = 3 to 4). See Fig. 1G legend for additional description of statistical notation.

Fig. 5.. Male, female, and OVX female DIO mice demonstrate similar metabolic improvements with lower dose D-PPAR.

(A) Study design. F, female; M, male. Reductions in (B) body weight and (C) food intake were most apparent in OVX mice. BW, body weight. (D) Adipocyte size is significantly reduced in PR AT of male mice and less so in female and OVX mice, with representative micrographs shown in (E). Scale bars, 100 μm. Micrographs for female and OVX mice are shown in fig. S12. Adipocyte size was calculated as described in Materials and Methods. Reductions in (F) glucose tolerance and (G) liver TG levels were also most evident in male mice, while (H) liver steatosis scores were significantly improved across all three models. Blinded scoring metrics for liver are described in Materials and Methods. (I) Representative H&E-stained liver section micrographs. Scale bars, 100 μm. Data are presented as mean ± SEM. Letters within each sex denote significance of P < 0.05 by one-way ANOVA with Tukey’s post hoc test or Kruskal-Wallis test (n = 7 to 10). See Fig. 1G caption for additional description of statistical notations.