Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Dec 19;16(12):e3000091.
doi: 10.1371/journal.pbio.3000091. eCollection 2018 Dec.

Non-proteolytic Ubiquitin Modification of PPARγ by Smurf1 Protects the Liver From Steatosis

Affiliations
Free PMC article

Non-proteolytic Ubiquitin Modification of PPARγ by Smurf1 Protects the Liver From Steatosis

Kun Zhu et al. PLoS Biol. .
Free PMC article

Abstract

Nonalcoholic fatty liver disease (NAFLD) is characterized by abnormal accumulation of triglycerides (TG) in the liver and other metabolic syndrome symptoms, but its molecular genetic causes are not completely understood. Here, we show that mice deficient for ubiquitin ligase (E3) Smad ubiquitin regulatory factor 1 (Smurf1) spontaneously develop hepatic steatosis as they age and exhibit the exacerbated phenotype under a high-fat diet (HFD). Our data indicate that loss of Smurf1 up-regulates the expression of peroxisome proliferator-activated receptor γ (PPARγ) and its target genes involved in lipid synthesis and fatty acid uptake. We further show that PPARγ is a direct substrate of Smurf1-mediated non-proteolytic lysine 63 (K63)-linked ubiquitin modification that suppresses its transcriptional activity, and treatment of Smurf1-deficient mice with a PPARγ antagonist, GW9662, completely reversed the lipid accumulation in the liver. Finally, we demonstrate an inverse correlation of low SMURF1 expression to high body mass index (BMI) values in human patients, thus revealing a new role of SMURF1 in NAFLD pathogenesis.

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Fig 1
Fig 1. Smurf1KO mice in the mixed BL background developed hepatic steatosis.
(A) HE and Oil Red-O staining of liver sections of WT, SF1KO, and SF2KO mice from the BL background at the age of 9–12 months. Bar = 40 μm. (B) Histological scores of steatosis of mice in (A). For each group, n = 15 (7 males and 8 females). Scores: 0, no steatosis; 1, minimal; 2, mild; 3, moderate; 4, severe. (C) Liver TG, CHO, and FFA content of the above male mice (n = 7 per group). (D) BW, eWAT/BW ratio, and Liver/BW ratio of the above male mice (n = 7 per group). (E) Liver ALT and AST activities of the above mice (n = 10; 5 males, 5 females per group). (F) Glucose and (G) insulin tolerance tests in male mice at the age of 9–12 months (n = 8 per group). All data are presented as mean ± SD; statistical significance of difference is indicated as *p < 0.05, **p < 0.01, ***p < 0.001. Original raw data can be found in S1 Data. ALT, alanine transaminase; AST, aspartate transaminase; BL, mixed black Swiss × 129/SvEv background; BW, body weight; CHO, total cholesterol; eWAT/BW, epididymal WAT weight to body weight; FFA, free fatty acid; HE, hematoxylin–eosin; KO, knockout; Liver/BW, liver weight to body weight; ns, not significant; Smurf, Smad ubiquitin regulatory factor; SF1KO, Smurf1 KO; SF2KO, Smurf2 KO; TG, triglyceride; WAT, white adipose tissue; WT, wild-type.
Fig 2
Fig 2. Inverse correlation of low Smurf1 expression to high fat accumulation.
(A–C) Loss of Smurf1 exacerbates HFD-induced steatosis in mice. (A) BW, Fat/Lean ratio, Liver/BW ratio, and histological scores of steatosis in male WT and SF1KO mice from BL background reared on either a ND (n = 7 per group) or HFD (n = 8 per group), beginning at 10–12 weeks of age for 8 weeks. Liver sections were scored as in Fig 1B. (B) HE staining of representative liver sections of the above BL mice at the end of diet treatment. Bar = 100 μm. (C) Liver TG levels of BL mice from (A). Data are presented as mean ± SD; statistical significance of differences is indicated as *p < 0.05, **p < 0.01, ***p < 0.001. Original raw data can be found in S1 Data. (D–E) Low SMURF1 expression in human livers is inversely correlated to high BMI values. (D) Human liver tissues from the LCI cohort with low levels of SMURF1 expression show a significantly higher BMI than those with high SMURF1 expression. Non-tumor liver samples from the LCI cohort were grouped into SMURF1 High (top 25%, n = 61) and SMURF1 Low (bottom 25%, n = 59) groups. (E) Inverse correlation between SMURF1 low and BMI high was also observed in non-tumor liver samples from the TCGA-LIHC cohort. Non-tumor liver samples (n = 37) from the TCGA-LIHC cohort were grouped into SMURF1 High (above the median, n = 18) and SMURF1 Low (below the median, n = 19) groups. Data are presented using box and whisker plot; centerline represents the median, whiskers represent 10th–90th percentile. Nonparametic t tests between the two groups were performed; p-values are indicated in the graph. BL, mixed black Swiss × 129/SvEv background; BMI, body mass index; BW, body weight; Fat/Lean, fat mass to lean mass; HE, hematoxylin–eosin; HFD, high-fat diet; KO, knockout; LCI, Liver Cancer Institute; Liver/BW, liver weight to body weight; ND, normal diet; ns, not significant; SF1KO, Smurf1 KO; Smurf, Smad ubiquitin regulatory factor; TG, triglyceride; TCGA-LIHC, the cancer genome atlas-liver hepatocellular carcinoma; WT, wild-type.
Fig 3
Fig 3. Up-regulation of PPARγ and other relevant lipid metabolic pathways associated with Smurf1 loss.
(A) Venn diagram of genes that were differentially expressed (fold > 1.5, FDR < 0.1) in Smurf1KO and Smurf2KO livers from the BL background relative to their WT counterparts at 11 months of age. For a detailed list, see S3 Table. (B) Heat map of a group of lipid metabolism–related genes that were differentially expressed in Smurf1KO livers and the KEGG pathway analysis of differentially expressed genes from BL-Smurf1KO livers. Enrichment score (p-value (−log10)) is indicated on the x-axis. (C) qRT-PCR analyses of Pparγ1, Pparγ2, and Pparα in livers from BL-WT, Smurf1KO, and Smurf2KO mice (n = 7 per group). (D) Western blot showing the increase of the PPARγ protein in livers of BL-Smurf1KO mice. Quantitation of PPARγ expression against loading control Hsc70 is shown at the bottom. (E) qRT-PCR analyses showing the increase of total Pparγ mRNA in livers and epididymal WAT but not skeletal muscle of BL-Smurf1KO mice, n = 7 per group. (F) qRT-PCR analyses showing up-regulation of a group of PPARγ target genes in livers of BL-Smurf1KO mice (n = 7 per group). Data from qRT-PCR are presented using box and whisker plot showing all points; centerline represents the median. All mice were analyzed at 9–11 months of age, and statistical significance of difference between WT and Smurf1KO is indicated as *p < 0.05, **p < 0.01, and ***p < 0.001. Original raw data can be found in S1 Data. BL, mixed black Swiss × 129/SvEv background; FDR, false discover rate; Hsc70, heat shock 70 kDa protein 8; KEGG, Kyoto Encyclopedia of Genes and Genomes; KO, knockout; ns, nonspecific band; PPAR, peroxisome proliferator-activated receptor; qRT-PCR, quantitative real-time PCR; SF1KO, Smurf1 KO; SF2KO, Smurf2 KO; Smurf, Smad ubiquitin regulatory factor; WAT, white adipose tissue; WT, wild-type.
Fig 4
Fig 4. Smurf1 regulates fatty acid uptake and lipid synthesis through PPARγ.
(A) Western blots showing that knockdown of Smurf1 but not Smurf2 in Hep3B and AML12 cells increased PPARγ protein level. (B) qRT-PCR analyses showing that knockdown of Smurf1 but not Smurf2 increased Pparγ mRNA level in AML12 cells. (C) qRT-PCR analyses showing that knockdown of Smurf1 in AML12 cells increased expression of Fabp1, Cd36, Acacb, and Apoc3 in a PPARγ-dependent manner. (D) Fatty acid uptake in AML12 cells as measured by 3H-palmitate incorporation (n = 3). (E) Lipid synthesis in AML12 cells as measured by incorporation of 3H-acetate into lipid (n = 3). (F) In vivo fatty acid uptake after intraperitoneal injection of BODIPY-FL-C16. The BODIPY-FL-C16 accumulation in the liver, epididymal WAT, and skeletal muscle was normalized to tissue weight (n = 8 per group). (G) Lipogenesis in primary hepatocytes as measured by the incorporation of 3H-acetate into lipid (n = 6 per group). Data are presented as mean ± SD; statistical significance of difference is indicated as *p < 0.05, **p < 0.01, ***p < 0.001. Original raw data can be found in S1 Data. BODIPY-FL-C16, 4,4-Difluoro-5,7-Dimethyl-4-Bora-3a,4a-Diaza-s-Indacene-3-Hexadecanoic Acid; eWAT, epididymal WAT; HSC70, heat shock cognate 71 kDa protein; NS, non-silencing control; PPAR, peroxisome proliferator-activated receptor; qRT-PCR, quantitative real-time PCR; RFU, relative fluorescence units; Rxr, retinoid x receptor; SF1KO, Smurf1 KO; siNS, non-silencing control siRNA; Smurf, Smad ubiquitin regulatory factor; WAT, white adipose tissue; WT, wild-type.
Fig 5
Fig 5. Smurf1 catalyzes non-proteolytic ubiquitination of PPARγ.
(A) Co-immunoprecipitation showing interaction between endogenous Smurf1 and PPARγ in AML12 cells. *nonspecific band. (B) PY motif in PPARγ contributes to the interaction between Smurf1 and PPARγ. Myc-tagged Smurf1, Flag-tagged PPARγ2, and its ΔPY mutant were transfected into AML12 cells as indicated. WCL were immunoprecipitated with Flag-M2 beads and followed by western blot analyses. (C) Smurf1 but not Smurf2 promotes polyubiquitination of PPARγ1 and PPARγ2. Flag-PPARγ were immunoprecipitated from transfected AML12 and resolved by SDS-PAGE. Western blot analyses were carried out to detect HA-ubiquitin (top) and Flag-PPARγ1 or -γ2 (second panel) in the precipitates. The levels of total HA-Ub, Flag-PPARγ, Myc-Smurfs, and endogenous Hsc70 (loading control) in the WCL were also analyzed and are shown in the bottom panels. *nonspecific band. (D) E3 ligase activity of Smurf1 is required for Smurf1-mediated polyubiquitination of PPARγ. Flag-PPARγ2 and WT Myc-Smurf1 and its mutant Myc-Smurf1(CA) were transfected into the Smurf1KO MEFs along with HA-Ub. Ubiquitination of PPARγ2 was analyzed by western blot after Flag-M2 immunoprecipitation, as in C. (E) In vitro ubiquitination assay using recombinant proteins showing that PPARγ is a direct substrate of Smurf1-mediated polyubiquitination. (F) Smurf1 induces K63-linked polyubiquitination of PPARγ. Purified ubiquitin with no lysine residue (K0) or with single lysine residue at indicated position was used in the in vitro ubiquitination assay. E3, ubiquitin ligase; HA, human influenza hemagglutinin; Hsc70, heat shock cognate 71 kDa protein; IB, immunoblot; IP, immunoprecipitation; K, lysine; KO, knockout; MEF, mouse embryonic fibroblast; PPAR, peroxisome proliferator-activated receptor; PY, PPxY; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; Smurf, Smad ubiquitin regulatory factor; Ub, ubiquitin; WCL, whole cell lysate; WT, wild-type.
Fig 6
Fig 6. Smurf1 inhibits PPARγ transcriptional activity.
(A) Smurf1 inhibits PPARγ-induced transcriptional activity in AML12 cells. Relative luciferease activities were measured 1 day after transfection. Data are presented as mean ± SD; statistical significance of differences is indicated by *p < 0.05, **p < 0.01, ***p < 0.001. Expression of transfected Smurf1 and PPARγ in these cells is shown at right. (B) E3 ligase activity of Smurf1 is required for its inhibition of PPARγ transcriptional activity. Luciferase activities were measured and showed as above. Expression of transfected Smurf1 and PPARγ in these cells are shown at right. (C) ChIP analyses of PPARγ binding to its own or Fabp1 promoter in AML12 cells after transfecting the plasmids as indicated. (D) ChIP analyses of PPARγ binding to its own or Fabp1 promoter in liver tissues from WT and Smurf1KO mice (n = 8 per group). Data are presented as mean ± SD; statistical significance of difference is indicated by *p < 0.05, **p < 0.01, ***p< 0.001. Original raw data can be found in S1 Data. ChIP, chromatin immunoprecipitation; E3, ubiquitin ligase; Hsc70, heat shock cognate 71 kDa protein; IgG, Immunoglobulin G; KO, knockout; PPAR, peroxisome proliferator-activated receptor; PPRE-Luc, PPAR response element-luciferase reporter; SF1KO, Smurf1 KO; Smurf, Smad ubiquitin regulatory factor; WT, wild-type.
Fig 7
Fig 7. Treatment with PPAR antagonist GW9662 protects BL-Smurf1KO mice from steatosis.
(A) BW of male mice from the BL background before (age 7–9 months) and after 60 days (age 9–11 months) of GW9662 treatment (n = 8 per group). (B) Fat/Lean and Liver/BW ratios and histological score of steatosis of the above mice after GW9662 treatment, compared with those of untreated control mice at the same age (n = 8 per group). (C) Representative HE-stained sections of the above mice. Bar = 100 μm. (D) Serum TG and CHO contents of the above mice, n = 8 per group. (E) Liver TG, CHO, and FFA contents of the above GW9662-treated mice. (F) qRT-PCR analysis of expression of Pparγ2 and relevant lipid metabolism genes in the livers of GW9662-treated mice. Data are presented as mean ± SD; statistical significance of difference is indicated by *p < 0.05, **p < 0.01, ***p < 0.001. Original raw data can be found in S1 Data. BL, mixed black Swiss × 129/SvEv background; BW, body weight; CHO, total cholesterol; Ctrl, control; Fat/Lean, fat mass to lean mass; FFA, free fatty acid; GW, GW9662; HE, hematoxylin–eosin; KO, knockout; Liver/BW, liver weight to BW; ns, not significant; PPAR, peroxisome proliferator-activated receptor; qRT-PCR, quantitative real-time PCR; SF1KO, Smurf1 KO; Smurf, Smad ubiquitin regulatory factor; TG, triglyceride; WT, wild-type.

Similar articles

See all similar articles

Cited by 3 articles

References

    1. Byrne CD, Targher G. NAFLD: a multisystem disease. Journal of hepatology. 2015;62(1 Suppl):S47–64. Epub 2015/04/29. 10.1016/j.jhep.2014.12.012 . - DOI - PubMed
    1. Charrez B, Qiao L, Hebbard L. Hepatocellular carcinoma and non-alcoholic steatohepatitis: The state of play. World journal of gastroenterology. 2016;22(8):2494–502. Epub 2016/03/05. 10.3748/wjg.v22.i8.2494 . - DOI - PMC - PubMed
    1. Kallwitz ER, McLachlan A, Cotler SJ. Role of peroxisome proliferators-activated receptors in the pathogenesis and treatment of nonalcoholic fatty liver disease. World journal of gastroenterology. 2008;14(1):22–8. Epub 2008/01/08. 10.3748/wjg.14.22 - DOI - PMC - PubMed
    1. Tailleux A, Wouters K, Staels B. Roles of PPARs in NAFLD: potential therapeutic targets. Biochimica et biophysica acta. 2012;1821(5):809–18. Epub 2011/11/08. 10.1016/j.bbalip.2011.10.016 . - DOI - PubMed
    1. Zardi EM, Navarini L, Sambataro G, Piccinni P, Sambataro FM, Spina C, et al. Hepatic PPARs: their role in liver physiology, fibrosis and treatment. Curr Med Chem. 2013;20(27):3370–96. Epub 2013/06/12. . - PubMed

Publication types

MeSH terms

Grant support

This research was supported by the Intramural Research Program of the NIH, National Cancer Institute, Center for Cancer Research. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Feedback