Adipose-targeted SWELL1 deletion exacerbates obesity- and age-related nonalcoholic fatty liver disease

Adipose SWELL1 deletion limits adipose depot expansion with overnutrition. We showed previously that mice with adipose SWELL1-specific deletion (Adiponectin-Cre SWELL1fl/fl; Adipo KO) are indistinguishable with respect to adiposity compared to the wild-type (WT; SWELL1fl/fl) controls when raised on a regular chow diet (18% kcal fat; 59% kcal carbohydrate) (28, 29) for 12 weeks. However, when raised on a high-fat diet (HFD; 60% kcal fat), Adipo-KO mice develop significantly less adiposity based on body composition and adipose tissue mass, driven by reduction in adipocyte size (28). In this study, we raised WT control and Adipo-KO mice on either high-fat, high-sucrose (HFHS, 58% kcal fat, 18% sucrose) diet or Gubra Amylin NASH (GAN, 40% kcal fat, 40% kcal carbohydrate) diet to more closely mimic diet-induced NAFLD/NASH models (30, 31). Similar to our previous findings in regular HFD-fed mice (28), Adipo-KO mice raised on HFHS diet for 27 weeks gained less weight than WT mice (Figure 1A), which was driven by an approximately 20% reduction in total fat mass (Figure 1B), with no change in lean mass as assessed by NMR (Figure 1B). This reduction in fat mass is largely attributed to a 42% decrease in epididymal white adipose tissue (eWAT) fat pad weights while the inguinal white adipose tissue (iWAT) fat content remained unchanged (Figure 1, C and D). Adipo-KO mice raised on a GAN diet for 23–25 weeks also gained less weight (Figure 1E), driven by a 38% reduction in total fat mass (Figure 1F), with no change in lean mass (Figure 1F), as assessed by EchoMRI. In contrast to HFHS diet, this reduction in fat mass in GAN-fed Adipo-KO mice was associated with a 52% and 45% reduction in both eWAT and iWAT depots, respectively (Figure 1G), compared with controls. The disparities in adipose depot size in mice raised on different diets may be attributed to dietary composition (fat, carbohydrates), as this is known to induce obesity at variable rates (32, 33).

Adipose SWELL1 supports adipose tissue expansion and maintenance.Figure 1

Adipose SWELL1 supports adipose tissue expansion and maintenance. (A) Total body weight of WT (n = 10) and Adipo KO (n = 8) mice (males) fed with high-fat/high-sucrose (HFHS) diet for 27 weeks. (B) Body composition for total fat and lean mass estimated by NMR from mice in A at 21-week time point of HFHS diet. (C) Representative images of epididymal (eWAT) and inguinal (iWAT) fat pads isolated from mice in A. (D) Total fat pad weights and ratio of fat pad over body weight of WT and Adipo-KO mice from A. (E) Total body weight of WT (n = 8) and Adipo-KO (n = 9) mice (males) fed with Gubra Amylin NASH (GAN) diet for 23–25 weeks. (F) Body composition for total fat and lean mass estimated by EchoMRI from mice in E at 22-week time point of GAN diet. (G) Total fat pad weights and ratio of fat pad over body weight of WT and Adipo-KO mice from E. Data are represented as mean ± SEM. Two-tailed unpaired t test was used in A, B, and DG where *, **, and *** represent P < 0.05, P < 0.01, and P < 0.001, respectively.

Adipose SWELL1 ablation augments lipolysis via hormone sensitive lipase activation. Our previous studies revealed that impaired adipocyte growth observed in Adipo-KO mice is associated with disrupted insulin/PI3K/AKT2 signaling (28). Impaired insulin/PI3K/AKT2 signaling is predicted to not only reduce lipogenesis, but also increase lipolysis, and adipocyte contraction in Adipo-KO mice, under conditions of reduced caloric intake. To directly test this, we performed a weight loss experiment where obese mice raised on HFD for 6–9 months were switched to regular chow (RC) diet for 4 weeks, and body composition was measured by NMR each week to noninvasively track reductions in fat mass. During the first 3 weeks, Adipo-KO mice exhibited a larger proportionate decrease in adiposity than WT mice (Figure 2A), consistent with increased adipose depot contraction, and increased lipolysis. To assess the extent of lipolysis, we measured plasma nonesterified free fatty acids (NEFAs) and glycerol in WT and Adipo-KO mice raised on a GAN diet for approximately 22 weeks. Despite the fact that Adipo-KO mice developed less adiposity (Figure 2, B and C), the absolute values of NEFAs and glycerol were similar to WT mice (Supplemental Figure 1, A and B; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.154940DS1), reflecting higher plasma NEFA and glycerol concentrations per unit fat (Figure 2D), as reported in CD36-KO mice (34). These elevated levels of plasma NEFA and glycerol per gram of adipose suggest increased lipolysis in the Adipo-KO mice than WT mice. Curiously, both CD36, a protein that facilitates transport of long chain fatty acids, and perilipin 1 (PLIN), a lipid droplet binding protein (34, 35), were found to be reduced in eWAT of the Adipo-KO mice compared with the WT (Figure 2E).

Adipose SWELL1 ablation augments lipolysis and hormone sensitive lipase actFigure 2

Adipose SWELL1 ablation augments lipolysis and hormone sensitive lipase activation. (A) Percentage decrease in total fat in high-fat (HF) diet (6–9 months) WT and Adipo-KO mice (males) after switching to regular chow (RC) diet for 4 weeks estimated by NMR. (B) Total body weight of WT (n = 7) and Adipo-KO (n = 8) mice (males) fed with GAN diet for 22 weeks. (C) Body composition for total fat and lean mass estimated by EchoMRI from mice in B at 22-week time point of GAN diet. (D) Random plasma NEFAs and glycerol normalized to total fat mass of WT and Adipo-KO mice in B. (E) mRNA expression of CD36 and PLIN relative to GAPDH in eWAT isolated from WT and Adipo-KO (n = 8 each, males) mice on GAN diet for 22–25 weeks. (F and G) Ex vivo lipolysis assay measuring NEFA (F) and glycerol (G) released from eWAT of WT (n = 9 experimental replicates from 3 mice) and Adipo-KO mice (n = 8 experimental replicates from 3 mice) at 30, 60, 90, and 120 minutes under basal, unstimulated conditions. (H and I) Ex vivo lipolysis assay measuring NEFA (H) released from WT and Adipo-KO (n = 11–12 experimental replicates from 2 mice/group) mice on HFD for 5 weeks at 30, 60, 90, and 120 minutes upon stimulation with 50 nM isoproterenol and the corresponding rate of NEFA production (I). (J) Western blots for protein levels of SWELL1, p-HSL, total HSL, and β-actin from primary adipocytes isolated from SWELL1fl/fl mice transduced with either Ad-GFP or Ad-Cre-GFP to generate WT and SWELL1-KO primary adipocytes and treated with either vehicle or 100 nM isoproterenol for 15 minutes and the corresponding densitometry analysis. Data are represented as mean ± SEM. Two-tailed unpaired t test was used in AE and I. Two-way ANOVA was used in FH. One-way ANOVA was used in J. *, **, and *** represent P < 0.05, P < 0.01, and P < 0.001, respectively.

To more directly test lipolysis, we performed an ex vivo lipolysis assay by measuring NEFA and glycerol release over time from eWAT dissected from WT and Adipo-KO mice. We found basal NEFA and glycerol release to be consistently increased from eWAT of Adipo-KO mice compared with WT mice at all time points examined (Figure 2, F and G). Moreover, isoproterenol stimulated lipolysis more robustly in eWAT isolated from Adipo-KO versus WT mice raised on HFD, based on NEFA concentrations in media over time (Figure 2H) and the rate of NEFA production (Figure 2I). As isoproterenol stimulates lipolysis by activating hormone sensitive lipase (HSL), we measured HSL phosphorylation (p-HSL, Ser660) in response to stimulation in WT and Adipo-KO primary adipocytes, in order to examine the underlying mechanism of lipolysis. In this in vitro culture system, consistent with the ex vivo lipolysis results from eWAT (Figure 2, H and I), we observed a striking ~3-fold increase in p-HSL in Adipo-KO primary adipocytes upon stimulation with isoproterenol (Figure 2J).

Adipose SWELL1 deletion predisposes to developing NAFLD with overnutrition. Adipose inflammation associated with increased lipolysis and elevated circulating free fatty acids (FFAs) are prone to ectopic lipid deposition in peripheral tissues. Indeed, male Adipo-KO mice raised on HFHS diet were noted to have a significant increase in both total (32%) and normalized (46%) liver mass (Figure 3, A and B). Female Adipo-KO mice showed no differences in these parameters under the same experimental conditions (Supplemental Figure 2, A–C). This may reflect sex differences in the propensity of male versus female mice to develop obesity in response to different diets (36). Histological examination of livers (HFHS diet) from Adipo-KO mice compared with WT revealed the increase in liver mass to be associated with a 43% increase in hepatic steatosis (Figure 3, C–E), including both macro- and microvesicular steatosis (Figure 3D). Consistent with increased hepatic steatosis, peroxisome proliferator–activated receptor γ (PPARG) expression, a driver of hepatic steatosis (37, 38), was induced 1.6-fold in Adipo-KO livers (Figure 3F), while other hepatic lipogenic genes, such as adipocyte protein 2 (AP2), sterol regulatory element binding protein-1a/1c (SREBP1A, SREBP1C), acetyl-CoA carboxylase 1 (ACC1), stearoyl-CoA desaturase 1 (SCD1), and fatty acid synthase (FAS), remained unchanged (Supplemental Figure 3A). Moreover, the Adipo-KO livers had more foci of mononuclear inflammatory cells and developed greater lobular inflammation compared with WT upon histological scoring (Figure 3D and Supplemental Figure 3B). Correspondingly, CD68 expression, a marker of macrophage and monocyte lineage (39, 40), was increased 2-fold in Adipo-KO livers (Figure 3G). Other inflammatory genes, TLR4, TNFA, TGFB, IL1B, and IL6, were unaltered (Supplemental Figure 3C). Similar to the phenotype observed on HFHS diet, Adipo-KO mice (males) raised on GAN diet exhibited significant increases (36%) in normalized liver mass (Figure 3H). Moreover, the livers from Adipo-KO mice exhibited an approximately 10-fold increase in hepatic diacylglycerides (DAGs; Figure 3I), but no increase in hepatic triacylglyceride (TAG; Supplemental Figure 3D) level was observed compared with the WT livers. Histological examination (Figure 3J) revealed increased hepatic steatosis (Figure 3, J and K) and hepatocellular hypertrophy (Figure 3, J and L) in Adipo-KO mice compared with WT mice raised on a GAN diet.

Adipose SWELL1 deletion predisposes to developing NAFLD with overnutrition.Figure 3

Adipose SWELL1 deletion predisposes to developing NAFLD with overnutrition. (A) Representative images of liver dissected from WT (top) and Adipo-KO (bottom) mice (males) fed with HFHS diet for 27 weeks. (B) Total liver mass and ratio of liver mass over body weight of WT and Adipo-KO mice from A. (C) Representative images of H&E-stained liver sections of WT and Adipo-KO mice (Scale bar: 100 μm). (D) Micro- and macrovesicular fat regions along with the mononuclear inflammatory cells (solid black circles) in WT and Adipo-KO mice (20× objective, 200× magnification). (E) Liver steatosis (%area) estimated from H&E-stained liver sections in C of WT (n = 2, 17 ROIs) and Adipo-KO (n = 2, 21 ROIs) mice using ImageJ software (NIH). (F) mRNA expression of PPARG in WT (n = 6) and Adipo-KO (n = 4) livers relative to control GAPDH. (G) mRNA expression of CD68 in WT (n = 6) and Adipo-KO (n = 4) livers relative to control GAPDH. (H) Total liver mass and ratio of liver mass over body weight of WT (n = 8) and Adipo-KO (n = 9) mice (males) fed with GAN diet for 23–25 weeks. (I) Measurement for total diacylglycerides (DAGs) from WT (n = 8) and Adipo-KO (n = 9) livers in H. (J) Representative images of H&E-stained liver sections of GAN diet–fed WT and Adipo-KO mice in H indicating cells that are normal (blue) and with hepatocellular hypertrophy (black outline) (40× objective). (K and L) Cumulative steatosis score (K) derived from macrovesicular, microvesicular, and hepatocellular hypertrophy scores (L) of liver sections from GAN diet–fed WT and Adipo-KO (n = 9 each) mice. Data are represented as mean ± SEM. Two-tailed unpaired t test was used in B, EI, K, and L where *, **, and *** represent P < 0.05, P < 0.01, and P < 0.001, respectively. ROIs, regions of interest.

Adipose SWELL1 depletion alters the hepatic lipid species profile in HFHS-fed mice. Next, we performed a lipidomic screen to measure monoacylglycerides (MAGs), DAGs, TAGs, cholesterol and sphingolipids, and bioactive phospholipids by liquid chromatography/mass spectrometry to examine the lipid species contributing to hepatic steatosis in HFHS-fed Adipo-KO mice. We found that MAGs were unchanged in Adipo-KO livers compared to WT livers (Figure 4A and Supplemental Figure 4A). Remarkably, 18:1/18:1 DAGs (oleic acid incorporated) were increased 64% in Adipo-KO livers compared with WT (Figure 4A and Supplemental Figure 4B), with 18:1-containing DAG lipid species, 16:1/18:1, 18:1/16:0, and 18:0/18:1, showing increasing trends, at 52%, 33%, 37% increased, respectively (Figure 4A and Supplemental Figure 4B), and 16:0/22:6 DAG increased by 76%. All other DAG species were essentially unchanged. Indeed, increases in specifically oleic acid–containing (18:1) DAGs have been shown to drive protein kinase Cε–mediated hepatic insulin resistance in NAFLD (4143) and also stimulate the androgen receptor to drive obesity and NAFLD-associated hepatocellular cancer (44).

Adipose SWELL1 depletion alters the hepatic lipid species profile.Figure 4

Adipose SWELL1 depletion alters the hepatic lipid species profile. (A and B) Relative abundance (A) and total content of hepatic lipid species for MAGs, DAGs and TAGs (B) in Adipo KO compared with WT mice (n = 5 each, males) on HFHS diet for 27 weeks. (C and D) Relative abundance of phosphatidylserine/phosphatidylinositol/phosphatidylglycerol (C) and FFAs (D) in Adipo-KO compared with WT mice (n = 5 each, males) on HFHS diet for 27 weeks. Two-tailed unpaired t test was used in AD where *, **, and *** represent P < 0.05, P < 0.01, and P < 0.001, respectively, and P values are listed when 0.05 < P < 0.1.

Consistent with increases in 18:1/18:1 DAG, 18:1/18:1/18:1 TAG (oleic acid containing) was also markedly increased 71% in Adipo-KO livers compared with WT, as were 18:1/18:2/18:1, 18:1/18:1/18:0, and 16:1/18:1/18:1 TAGs, at 26%, 18%, and 35%, respectively (Figure 4A and Supplemental Figure 4C). Oleic acid–containing TAGs (18:1) are believed to contribute significantly to hepatic steatosis in NAFLD, with dual modulatory functions for the progression of disease (45). Conversely, 16:0- and 16:1-containing TAGs were generally decreased in Adipo-KO livers (Figure 4A and Supplemental Figure 4C), while all other TAG species were essentially unchanged. Though cumulatively, the total MAGs and TAGs were similar in WT and Adipo-KO livers, the DAGs exhibited an approximately 39% increase (P = 0.07) in mice raised on HFHS diet (Figure 4B), similar to that observed in livers of mice raised on GAN diet (Figure 3I). Cholesterol and sphingolipid types including sphingosine, sphinganine, 3-keto-sphinganine, and sphingosine-1-phosphate (S1P) were similar between WT and Adipo KO (Supplemental Figure 4, D and E).

Since dynamic changes in phospholipid (PL) content is another feature in NASH pathogenesis that influences hepatocyte membrane integrity (46), and predisposes to hepatocyte injury, we measured phosphatidylserine (PS), phosphatidylinositol (PI), and phosphatidylglycerol (PG). Of the 7 measured PS subtypes, 38:6, 38:3, and 40:6 were significantly decreased by 34%, 20%, and 25%, respectively, in Adipo-KO livers (Figure 4C and Supplemental Figure 4F). Similarly, PI subtypes 36:4, 34:2, and 36:3 were decreased by 22%, 30%, and 32%, respectively, in Adipo-KO livers (Figure 4C and Supplemental Figure 4G). PG was largely unaltered in WT and Adipo-KO livers, except for 36:4 with a 39% decrease in Adipo KO as compared with WT (Figure 4C and Supplemental Figure 4H). Saturated FFAs were unchanged except for a significant decrease (23%) in myristic acid (14:0) in the Adipo-KO livers (Figure 4D and Supplemental Figure 4I). A lower content (50%) of n-3 polyunsaturated FFAs (n-3 PUFAs) eicosapentaenoic (20:5) and docosapentaenoic acid (22:5) was also observed, while the remaining species were unaltered (Figure 4D and Supplemental Figure 4I). The loss in phospholipids and n-3 FFAs and increases in DAGs and TAGs are consistent with the lipidomic profile of other murine NAFLD models (47, 48).

Adipose SWELL1-KO mice are predisposed to age-related NAFLD. Similar to mice raised on HFHS diet, both male and female aged Adipo-KO (~12–21 months) raised on an RC diet (18% kcal fat; 59% kcal carbohydrate) developed reduced adiposity compared with WT mice, driven largely by a reduction in eWAT, based on both absolute (39% and 54% in females and males, respectively) and normalized fat pad weights (34% and 54% in females and males, respectively; Figure 5, A and D), with no significant reduction in total body weight (Figure 5, B and E). This impairment in age-related adipose tissue expansion in Adipo-KO mice was associated with a mild, but statistically significant, increase in normalized liver mass in female mice (~14%) (Figure 5C) and a trend toward increased liver mass parameters in males (Figure 5F). Combining these data from both sexes reveals significantly lower fat content and higher liver ratio in aged Adipo-KO mice compared with WT mice (Supplemental Figure 5, A and B). Histological examination of livers from aged Adipo-KO females revealed a significant increase in hepatic steatosis (111%) compared with WT controls (Figure 5, G and H). Taken together, these data suggest that maintaining adipose SWELL1 expression, activity, and signaling is metabolically protective against NAFLD, both in the setting of overnutrition and with aging. Indeed, consistent with this notion, adipose SWELL1 protein expression was significantly higher in young mice (2–3 months old, males) as compared with aged mice (~2-fold, Figure 5I) (~18–21 months old, males), indicating that adipose SWELL1 protein declines with age and may contribute to age-related metabolic dysfunction.

Adipose SWELL1 expression protects against age-related NAFLD and declines wFigure 5

Adipose SWELL1 expression protects against age-related NAFLD and declines with aging. (A) Total mass of epididymal (eWAT) and inguinal (iWAT) fat pads and their corresponding ratio of fat pad over body weight, (B) total body weight, (C) total liver mass, and ratio of liver mass over body weight dissected from WT (n = 15) and Adipo-KO (n = 11) female mice ~12 months old on RC diet. (D) Total mass of epididymal and inguinal fat pads and their corresponding ratio of fat pad over body weight, (E) total body weight, (F) total liver mass, and ratio of liver mass over body weight dissected from male WT (n = 6) and Adipo-KO (n = 7) male mice ~18–21 months old on RC diet. (G) Representative images of H&E-stained liver sections of WT and Adipo-KO mice from B (Scale bar: 100 μm). (H) Liver steatosis (%area) estimated from H&E-stained liver sections in G of WT (n = 2, 17 ROIs) and Adipo-KO (n = 2, 20 ROIs) mice using ImageJ software. (I) Representative image of Western blot comparing SWELL1 protein expression in epididymal adipose tissue isolated from young (2–3 months old, males) and aged (18–21 months old, males) WT mice fed with RC diet (left) and its corresponding densitometric ratio (right). Data are represented as mean ± SEM. Two-tailed unpaired t test was used in AF, H, and I where *, P < 0.05, and **, P < 0.01.

Male Adipo-KO mice develop HCC with aging. Chronic NAFLD and inflammation associated with NASH predisposes to HCC and occurs predominantly in males (49). Remarkably, 3/7 of the aged Adipo-KO males had massive tumors on gross inspection, as compared with 1/6 WT males, which had a small tumor (Figure 6, A and B). Further examination of these Adipo-KO livers by H&E staining revealed increased steatosis, particularly in the tumor region (Figure 6C), as compared with the nontumorous region. Gross steatosis and inflammation were similar between WT and Adipo KO (Supplemental Figure 5C). Consistent with HCC, histological features of Adipo-KO livers contained enlarged pleomorphic cells along with eosinophilic hyaline globules (Figure 6, D and E). HCC regions had diagnostic features of malignant hepatocellular proliferation, including nested islands of hepatocytes with peripheral wrapping of flattened endothelial cells along the irregular sinusoidal spaces (Figure 6F). Furthermore, they were devoid of portal structures and included irregularly distributed dilated veins compared with nontumorous regions (Figure 6G). To examine the nature of the tumors, we stained the livers with glutamine synthetase, an enzyme with distinct central perivenular distribution in normal liver that typically has an abnormal distribution in HCC (Figure 6, H and I) (50). Indeed, the aged Adipo-KO livers exhibited an abnormal glutamine synthetase (GS) stain pattern (complete loss) in the tumor region compared with the normal adjacent nontumor region (Figure 6I). The WT group demonstrated normal GS staining (Figure 6H). Taken together, these data reveal that adipose SWELL1 ablation not only impairs healthy adipose depot expansion in the setting of overnutrition (28), and with sedentary aging, but also predisposes to NAFLD, NASH, and consequent HCC associated with longstanding metabolic syndrome.

Male adipose SWELL1-KO mice develop HCC with aging.Figure 6

Male adipose SWELL1-KO mice develop HCC with aging. (A) Representative images of livers isolated from male WT and Adipo-KO mice with tumors indicated by black arrows. (B) Formalin-fixed Adipo-KO liver with massive hepatic tumors. (C and D) H&E-stained liver sections of Adipo-KO mice with indicated nontumor and tumor-containing regions. * indicates enlarged and pleomorphic cells, and black arrows indicate eosinophilic hyaline globules in D. (EG) Liver sections of Adipo-KO mouse with HCC morphological features. * indicates large- and intermediate-sized fat droplets, and black arrows indicate eosinophilic hyaline droplets within the HCC cells in E. Nested islands of HCC are outlined in solid black lines, and the black arrows indicate the endothelial wrapping around these nests with flattened endothelial cells in F. Abnormal veins in HCC regions are indicated with “v,” and normal portal tracts and central veins are indicated as “pt” and “cv” in the nontumorous region in G. (H and I) Representative images of glutamine synthetase staining from a normal WT (H) and Adipo-KO (I) liver with HCC. Each is a 10× eyepiece with a 4×, 10×, 20×, 40× objective; 40× magnification for images in GI; 100× magnification for image in C; 200× magnification for images in D and F; and 400× magnification for image in E.

留言 (0)

沒有登入
gif