14-3-3ζ allows for adipogenesis by modulating chromatin accessibility during the early stages of adipocyte differentiation

Abstract

Objective: We previously established the scaffold protein 14-3-3ζ as a critical regulator of adipogenesis and adiposity, but whether 14-3-3ζ exerted its regulatory functions in mature adipocytes or in adipose progenitor cells (APCs) remained unclear.

Methods: To decipher which cell type accounted for 14-3-3ζ-regulated adiposity, adipocyte- (Adipoq14-3-3ζKO) and APC-specific (Pdgfra14-3-3ζKO) 14-3-3ζ knockout mice were generated. To further understand how 14-3-3ζ regulates adipogenesis, Tandem Affinity Purification (TAP)-tagged 14-3-3ζ-expressing 3T3-L1 preadipocytes (TAP-3T3-L1) were generated with CRISPR-Cas9, and affinity proteomics was used to examine how the nuclear 14-3-3ζ interactome changes during the initial stages of adipogenesis. ATAC-seq was used to determine how 14-3-3ζ depletion modulates chromatin accessibility during differentiation.

Results: We show a pivotal role for 14-3-3ζ in APC differentiation, whereby male and female Pdgfra14-3-3ζKO mice displayed impaired or potentiated weight gain, respectively, as well as fat mass. Proteomics revealed that regulators of chromatin remodeling, like DNA methyltransferase 1 (DNMT1) and histone deacetylase 1 (HDAC1), were significantly enriched in the nuclear 14-3-3ζ interactome and their activities were impacted upon 14-3-3ζ depletion. Enhancing DNMT activity with S-Adenosyl methionine rescued the differentiation of 14-3-3ζ-depleted 3T3-L1 cells. ATAC-seq revealed that 14-3-3ζ depletion impacted the accessibility of up to 1,244 chromatin regions corresponding in part to adipogenic genes, promoters, and enhancers during the initial stages of adipogenesis. Finally, 14-3-3ζ-regulated chromatin accessibility correlated with the expression of key adipogenic genes.

Conclusion: Our study establishes 14-3-3ζ as a crucial epigenetic regulator of adipogenesis and highlights the usefulness of deciphering the nuclear 14-3-3ζ interactome to identify novel pro-adipogenic factors and pathways.

Keywords: 14-3-3ζ; Adipogenesis; Adipogenic genes; Chromatin accessibility; Energy homeostasis; Epigenetic regulation.

Figure 1

Adipose tissue expression of Ywhaz is positively correlated with body fat mass and insulin resistance in male and female mice. Correlation of (A–C) male and (D–F) female perigonadal adipose tissue (VAT) gene expression of Ywhaz with HOMA-IR (A and D) and whole body fat mass before (B and E) and after (C and F) high-fat, high-sucrose feeding, using the Hybrid Mouse Diversity Panel (HMDP) resource [19]. (G) Timeline of the experiment performed on 3T3-L1 cells transfected with siCTL or siYwhaz (10 nM each) prior to standard MDI or MDIR differentiation protocols [9,10,87]. After treatments, cells were subjected to Oil Red-O staining (H), measurement of Oil Red-O incorporation level by absorbance at 490 nm (I), and qRT-PCR to measure Pparg2(J), Adipoq(K) and Ywhaz(L) mRNA levels. Significant differences between experimental conditions are indicated by ∗P < 0.05 or #P < 0.05 (calculated by Student’s t-test). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Figure 2

Deletion of 14-3-3ζ in mature adipocytes does not affect adiposity, but impairs glucose tolerance. (A–J) Male and (K–T) female wild type (AdipoqCre+/WT) and Adipoq14-3-3ζKO (AdipoqCre+/Flox) mice were generated by breeding Adipoq-Cre mice with either WT mice or mice harboring floxed alleles of Ywhaz. Animals were fed a chow diet exclusively. (A,K) Inguinal (SAT) and perigonadal (VAT) adipose tissue samples were subjected to qRT-PCR to measure Ywhaz mRNA levels (n = 3–4 per sex per genotype). Mice (n = 4 minimum) were followed for (B and L) body weight gain from 10 to 26 weeks of age, assessed for (C, M) glucose tolerance and (D,N) insulin sensitivity via ip-GTT (2 g kg⁻¹ B.W. d-glucose) and ip-ITT (1.0 U kg-1 humulin® R) at 25 and 26 weeks, respectively. (E,O) H&E-stained microsections (scale bar = 200 μm) from inguinal and perigonadal adipose tissues were analysed for (F,P) white adipocyte area and (G,Q) size distribution with Visiomorph™. (H-J, R-T) Body composition of mice were measured by EchoMRI™ prior to sacrifice. Error bars represent S.E.M. Significant differences between wild type and Adipoq14-3-3ζKO mice are indicated by ∗P < 0.05 (calculated by Student’s t-test).

Figure 3

Deletion of 14-3-3ζ in adipocyte progenitor cells affects adiposity. Male (A–J) and female (K–T) wild type (PdgfraCre+/WT) and Pdgfra14-3-3ζKO (PdgfraCre+/Flox) mice were generated by breeding Pdgfra-Cre mice with either WT mice or mice harboring floxed alleles of Ywhaz and fed a chow diet. (A,K) Inguinal (SAT) and perigonadal (VAT) adipose tissue samples were subjected to qRT-PCR to measure Ywhaz mRNA levels (n = 3–4 per sex per genotype). (B,L) Mice (n = 4 minimum) were followed for body weight gain from 10 to 26 weeks of age, assessed for (C,M) glucose tolerance and (D, N) insulin sensitivity via ip-GTT (2 g kg⁻¹ B.W. d-glucose) and ip-ITT (1.0 U kg-1 Humulin® R) at 25 and 26 weeks, respectively. (E, O) H&E-stained microsections (scale bar = 200 μm) from inguinal and perigonadal adipose tissues were analysed for (F, P) white adipocyte area and (G, Q) size distribution with Visiomorph™. (H-J, R-T) Body composition of mice were measured by EchoMRI™ prior to sacrifice. (U) Experimental procedure for the extraction of SVF from subcutaneous white adipose tissues from male and female Pdgfra14-3-3ζKO and WT mice for plating, culture, and treatment with MDIR, followed by Oil Red-O staining. Differentiated or non-differentiated SVF from male (V and W) and female (X and Y) mice were subjected to Oil Red-O staining (V and Y) and quantification of its accumulation level by absorbance at 490 nm (W and Y). Error bars represent S.E.M. Significant differences between WT and Pdgfra14-3-3ζKO mice are indicated by ∗P < 0.05 (calculated by Student’s t-test). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Figure 4

Key effectors of chromatin accessibility are influenced by 14-3-3ζ as part of its nuclear interactome during the early step of adipogenesis. (A) Schematic outline of how TAP-3T3-L1 cells were used to elucidate the nuclear interactome of 14-3-3ζ during the first 24 and 48 h of adipogenesis (n = 4 for each condition). (B) Venn diagram showing the number of unique and overlapping proteins enriched at 24 h and 48 h post induction. (C,D) GO analysis showing the most enriched Biological Processes attributed to 14-3-3ζ interactome at 24 h (C) and 48 h (D) of adipogenesis. (E,F) Volcano plots of the decreased and enriched proteins in the 14-3-3ζ nuclear interactome. (G–K) Genomic DNA (G), crude nuclear fractions (H), nuclear protein fractions (I,J), and Histone fractions (K) extracted from 3T3-L1 cells transfected for 48 h with siCTL (10 nM) or siYwhaz (10 nM) were respectively used for assessing whole 5-mC (G), DNMT activity (H), HDAC (I) and HAT (J) activity, and histone H3 acetylation level (K), normalized by total protein (n = 3–6 for each condition). (L) Schematic outline of how siCTL- or siYwhaz- (10 nM each) transfected 3T3-L1 cells were subjected to an 8-day MDI differentiation protocol in the presence or absence of either ITSA-1 or S-adenosylmethionine (SAM, 100 μM each) prior to Oil Red-O staining and RNA extraction for qRT-PCR. (M-R) Oil Red-O staining (M,P), Oil Red-O quantification measured by absorbance at 490 nM (O, Q), and mRNA expression level of Pparg2 (O, R) after co-treatment of differentiating 3T3-L1 with ITSA-1 (M–O) or SAM (P–R). Error bars represent S.E.M. Significant differences between conditions are indicated by ∗P < 0.05 (calculated by Student’s t-test). Figure was partly generated with Biorender (A). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Figure 5

14-3-3ζ is involved in chromatin remodeling and accessibility during adipogenesis. (A) Schematic overview of the use of 3T3-L1 preadipocytes for ATAC-seq, 24 h and 48 h post-differentiation with MDI following transfection with siCTL or siYwhaz (10 nM each). (B,C) Venn diagrams showing overlap of (B) down-regulated or (C) up-regulated differentially accessible regions after siYwhaz treatment at 0, 24 and 48 h post-MDI induction. (D,E) Coverage tracks of ATAC-seq signals at the Fabp4 (D) and Adig (E) genes. (F,G) Enhancers (F) and promoters (G) motif enrichment analysis at 48 h post-MDI induction in presence of siYwhaz. Figure was partly generated with Biorender (A).

Figure 6

14-3-3ζ enables accessibility and the transcription of genes involved in adipocyte maturation and function. (A) Over-laying of ATAC-seq results with previous RNA-seq results (GSE60745) to correlate differentially accessible promoter regions with fold-changes in expression of corresponding genes, (B) 24 h and (C) 48 h post-MDI induction in presence of siYwhaz. (D) Experimental outline to assess requirement of 14-3-3ζ for active ATF-dependent expression of adipogenic genes. (E) Quantification of Cebpa, Cebpb, Pparg1, and Pparg2 mRNA levels in NIH-3T3 cells transfected or not with plasmids containing Cebpa, Cebpb, Pparg1, or Pparg2.(F)Ywhaz,(G)Adig, and (H)Fabp4 mRNA expression levels in NIH-3T3 cells successively transfected with siYwhaz (60 nM) and plasmids for Cebpa, Cebpb, Pparg1, and Pparg2 prior to treatment with MDIR for 48h. (I) Graphical illustration of the new proposed mechanism (bright colors) on how 14-3-3ζ regulates adipogenesis via chromatin remodeling in integration with previously found mechanisms (dim colors) [9]. Figure was partly generated with Biorender (A,I). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)