Programming human pluripotent stem cells into white and brown adipocytes

Abstract

The utility of human pluripotent stem cells is dependent on efficient differentiation protocols that convert these cells into relevant adult cell types. Here we report the robust and efficient differentiation of human pluripotent stem cells into white or brown adipocytes. We found that inducible expression of PPARG2 alone or combined with CEBPB and/or PRDM16 in mesenchymal progenitor cells derived from pluripotent stem cells programmed their development towards a white or brown adipocyte cell fate with efficiencies of 85%-90%. These adipocytes retained their identity independent of transgene expression, could be maintained in culture for several weeks, expressed mature markers and had mature functional properties such as lipid catabolism and insulin-responsiveness. When transplanted into mice, the programmed cells gave rise to ectopic fat pads with the morphological and functional characteristics of white or brown adipose tissue. These results indicate that the cells could be used to faithfully model human disease.

Figure 1

Experimental scheme and characterization of hPSC-derived MPCs. (a) Experimental scheme for the differentiation of ADSVCs and hPSCs into white and brown adipocytes. hPSCs were differentiated as embryoid bodies and then re-plated and passaged to generate MPCs. MPCs and ADSVCs were either not infected (unprogrammed) or transduced (programmed) with lentivirus constitutively expressing the lenti-rtTA M2 domain and virus carrying the inducible cDNA transgenes PPARG2 (lenti-PPARG2), CEBPB (lenti-CEBPB) and PRDM16 (lenti-PRDM16). All differentiated cells were cultured in media containing adipogenic factors and doxycycline for 14–16 days and maintained without doxycycline until 21 days before analysis. Four hPSC lines were used—two hESC lines, HUES 8 and 9, and two induced pluripotent stem cell (iPSC) clones generated by reprogramming BJ fibroblasts with modified ribonucleic acids, BJ RiPSC. (b) Top, table showing the results of the flow cytometry analysis of the hPSC lines BJ RiPS #1.1 and HUES 9 as well as MPCs derived from these lines. Cells were stained for the surface antigens Stro1, CD105, CD73, CD44, CD29 and CD4. The numbers represent the percentage of positive cells. Where appropriate, positive stainings were distinguished as high- or low-expression groups. Bottom, flow cytometry results for the surface antigens Stro1, CD105, CD73, CD44 and CD29 presented as histograms: BJ RiPS #1.1 p15 (blue), BJ RiPS #1.1 MPC p6 (green), HUES 9 p30 (orange) and HUES 9 MPC p7 (violet). The x axis indicates the relative fluorescence intensity from 10 to 100.000 on a logarithmic scale. The y axis represents the percentage of cells. (c) Doxycycline-inducible expression of PPARG2 and eGFP. Top, BJ RiPS MPCs transduced with lenti-rtTA and doxycycline-inducible eGFP virus and cultured either in the presence (+DOX; top panels) or absence (−DOX; bottom); Bottom, RT–qPCR assays for viral PPARG2 cDNA expression normalized to HPRT: BJ RiPS MPCs transduced with lenti-rtTA only (control); BJ RiPS MPCs transduced with lenti-rtTA and lenti-PP ARG2 cultured in the absence (−DOX) or presence (+DOX) of doxycycline.

Figure 2

Programming hPSC-derived MPCs with PPARG2, PPARG2–CEBPB or PPARG2–CEBPB–PRDM16 generates white and brown adipocytes. (a) Efficiency of white adipocyte differentiation. Top, table showing the efficiency of white adipocyte formation from cells in adipogenic media alone (untransduced) or PPARG2-programmed differentiated cells (PPARG2) as determined by the ratio of Hoechst-positive and CEBPA-positive nuclei. Bottom, representative images of Hoechst-stained (top left panel) and CEBPA-stained (bottom left panel) nuclei from HUES 9 MPCs programmed with PPARG2 (×100 magnification). The right panels illustrate the threshold assignment of positive nuclei and counting carried out by the ImageJ analysis software. (b) Programming hPSCs into white adipocytes. HUES-8-derived MPCs were differentiated with adipogenic media alone (top panels; untransduced) or in combination with exogenous PPARG2 expression (bottom panels; PPARG2). From left to right: bright-field images illustrating the morphology of immature (top panel) and mature (lower panel) white adipocytes; fluorescence micrographs of corresponding immunostains with antibodies against the adipocyte marker protein FABP4 (red) and the neutral lipid dye BODIPY (green); all cells were co-stained with Hoechst (blue) to identify nuclei (×100 magnification). (c) hPSC-derived white adipocytes express endogenous CEBPA and PPARG2. BJ-RiPSC-derived MPCs were differentiated and programmed with exogenous PPARG2 expression for 16 days and, 5 days after withdrawal of doxycycline, stained with antibodies against CEBPA (red; lower left panel) or PPARG2 (red; lower right panel). All cells were also stained with the neutral lipid dye BODIPY (green, both lower panels, ×100 magnification). The upper panels show corresponding bright-field images. (d) hPSC-derived brown adipocytes express UCP1 and can be efficiently labelled with MitoTracker. BJ-RiPSC-derived MPCs were differentiated with adipogenic media alone (top panels; untransduced) or programmed with either exogenous PPARG2 + CEBPB (middle panel) or exogenous PPARG2 + CEBPB + PRDM16 (lower panel) expression for 14 days and, 7 days after withdrawal of doxycycline, labelled with MitoTracker (red) or stained with antibodies against UCP1 (green). The left panels show corresponding bright-field images (all images ×200 magnification).

Figure 3

Programmed adipocytes exhibit a mature white or brown adipocyte gene expression profile. (a) hPSC-derived white adipocytes express mature marker genes. RT–qPCR assays were carried out for adipocyte marker genes PPARG2, CEBPA, FABP4, ADIPOQ, HSL and LPL. The expression values represent three biological replicates and are shown relative to HPRT expression in each sample. White bars represent cells that were not exposed to adipogenic media (undifferentiated); grey bars represent cells that were exposed to adipogenic media but not transduced with lenti-PPARG2 (−PPARG2); black bars represent cells that were exposed to adipogenic media and transduced with lenti-PPARG2 (+PPARG2). P values represent two-tailed Student’s t-tests between − PPARG2 and +PPARG2 expression values for each cell line. n = 3, *P < 0.05, **P < 0.01 (s.d.). P values shown under each gene name represent results of analysis of variance among all expression values for −PPARG2 and +PPARG2 cell lines. (b) Comparison of hPSC-derived white adipocytes and brown adipocytes. RT–qPCR assays were carried out for a range of white or brown adipocyte marker genes, PPARG2, PGC1a, FABP4, ADIPOQ, HSL, LPL, CYC1, ELOVL3 and UCP1. Expression values represent three biological replicates and are shown relative to HPRT expression and relative to the lenti-PPARG2 condition set as 1. White bars represent cells that were differentiated with adipogenic media alone (untransduced); black bars represent cells that were exposed to adipogenic media and transduced with lenti-PPARG2 (+PPARG2); grey bars represent cells that were exposed to adipogenic media and transduced with lenti-PPARG2 and lenti-CEBPB (+PPARG2–CEBPB, light grey bars) or with lenti-PPARG2, lenti-CEBPB and lenti-PRDM16 (+PPARG2–CEBPB–PRDM16, dark grey bars) respectively. All experiments were carried out with BJ-RiPSC-derived MPCs. P values represent two-tailed Student’s t-tests between the PPARG2 and PPARG2–CEBPB or PPARG2–CEBPB–PRDM16 set-ups respectively. Values for each cell line, n = 3, *P < 0.05, **P < 0.01 (s.d.).

Figure 4

Global transcriptional analysis confirms the identity of programmed hPSC-derived white and brown adipocytes. (a) ADSVCs, HUES 9 MPCs and BJ RiPS MPCs either not exposed to adipogenic media (undifferentiated) or cultured with adipogenic media and transduced with lenti-PPARG2 (+PPARG2) were compared with primary adipocytes using Affymetix 1.0 ST microarrays. Shown is a row-centred heat map of hierarchical clustering carried out on the 2,136 differentially expressed genes at a 5% FDR. Probe sets are coloured according to the average expression level across all samples, with blue denoting a lower expression level and red denoting a higher expression level. (b) Row-centred heat map and hierarchical cluster of an adipo-specific gene panel across the same samples as in a. Probe sets are coloured according to the average expression level across all samples, with blue denoting a lower expression level and red denoting a higher expression level. (c) Row-centred heat map hierarchical cluster of a brown and white adipo-specific gene panel across the same samples as in a and b (white adipocytes, primary white adipose tissue and undifferentiated MPCs) in addition to PPARG2–CEBPB–PRDM16-programmed MPCs (brown adipocytes) on the Agilent G3 human GE array platform. Probe sets are coloured according to the average expression level across all samples, with blue denoting a lower expression level and red denoting a higher expression level.

Figure 5

Programmed hPSC-derived white adipocytes exhibit mature functional properties. (a) hPSC-derived adipocytes carry out lipolysis. Glycerol was measured in the supernatant of ADSVCs and HUES-9-derived MPCs that were either not exposed to adipogenic media (undifferentiated) or exposed to adipogenic media without (−PPARG2) or with exogenous PPARG2 expression (+PPARG2) followed by either treatment with (+iso) or without (−iso) isoproterenol. The quantity of glycerol released (in micrograms) was normalized to the total amount of protein (in milligrams) for each sample. n = 3, **P < 0.01 (s.e.m.). (b) hPSC-derived adipocytes secrete adiponectin. ELISA for adiponectin in the supernatant of cells exposed to adipogenic media and either not transduced with lenti-PPARG2 (−PPARG2) or transduced with lenti-PPARG2 (+PPARG2). Experiments carried out as biological triplicates, with the exception of BJ RiPS MPCs. n = 3, *P < 0.05, **P < 0.01 (s.d.). (c) Lipid profiling of ADSVC- and hPSC-derived adipocytes. The cellular lipid content of PPARG2-programmed ADSVCs, HUES 9-derived MPCs and BJ-RiPS-derived MPCs was analysed using a tandem mass spectroscopy lipidomics platform and compared with the lipid content of primary adipose tissue. Shown are the relative abundances of several long-chain triacylglyceride species in each cell type. The x axis denotes the total number of carbon atoms in the fatty-acid chains:unsaturated bonds. The y axis represents the relative abundance of each lipid analyte. (d) Attenuation of insulin-induced Ser-473 phosphorylation on AKT by FFAs. BJ-RiPS-MPC-derived adipocytes were treated using either insulin alone, BSA-bound FFAs or with both. Phosphorylation of AKT was determined in the whole-cell lysate by immunoblotting with the phospho-specific AKT (Ser 473) antibody. (e) Glucose uptake in BJ-RiPS-MPC-derived adipocytes was assessed by the transport of [³H]-2-deoxy-D-glucose following insulin stimulation. MPCs were exposed to adipogenic media without (−PPARG2) or with exogenous PPARG2 expression (+PPARG2) followed by either treatment with (+insulin) or without (−insulin) insulin. The quantity of [³H]2-deoxy-D-glucose transported into the cells was normalized to CytoB and the results are shown as c.p.m. n = 3, *P < 0.05, **P < 0.01 (s.e.m.). Uncropped images of blots are shown in Supplementary Fig. S9.

Figure 6

Programmed hPSC-derived brown adipocytes demonstrate mature functional properties. (a) Glycerol release assay with hPSC-derived brown and white adipocytes. Glycerol was measured in the supernatant of HUES-9-derived MPCs that were differentiated with adipogenic media alone (untransduced, white bars), with exogenous PPARG2 expression (+PPARG2, black), with expression of a combination of lenti-PPARG2 and lenti-CEBPB (+PPARG2–CEBPB, light grey bars) or with a combination of lenti-PPARG2, lenti-CEBPB and lenti-PRDM16 (+PPARG2–CEBPB–PRDM16, dark grey bars). After differentiation of the cells, measurements were made at the basal level and after exposure to forskolin (+FSK). The quantity of released glycerol (in micrograms) was normalized to the total amount of protein (in milligrams) for each sample. n = 3, Student’s t-test **P < 0.01 (s.e.m.). (b,c) Comparison of the OCR and ECAR of hPSC-derived brown and white adipocytes. The OCR and ECAR were determined using no cells (brown line) and cells differentiated with adipogenic media alone (untransduced) as controls. Cells in which PPARG2 was exogenously expressed (+PPARG2) are represented with a green graph, cells that were transduced with lenti-PPARG2 and lenti-CEBPB (+PPARG2-CEBPB) are shown with a black line and lenti-PPARG2, lenti-CEBPB and lenti-PRDM16 (+PPARG2–CEBPB–PRDM16)-transduced cells are indicated by a blue line. The OCR and ECAR were measured over time in approximately 5 min intervals. The first two measurements were conducted to establish a baseline rate, followed by three measurements after the addition of oligomycin, an ATPase inhibitor (I). By uncoupling the proton gradient with CCCP, the maximum OCR and ECAR rates were determined over the next three time intervals (II). Finally, at two time points, measurements were conducted after inhibition of the mitochondrial respiratory chain with antimycin (III). All experiments were conducted with BJ-RiPSC-derived MPCs. P values represent two-tailed Student’s t-tests between untransduced and transduced cells. Values for each cell line, n = 4, *P < 0.01, **P < 0.001 (s.e.m.).

Figure 7

Generation of functional hPSC-derived brown and white adipose tissue in vivo. (a) HUES-9-derived MPCs were transduced with lenti-PPARG2 and after 2 weeks of differentiation collected and injected subcutaneously into RAG2; IL2γC double-knockout mice. Four to six weeks after the injection, prominent cell growth was visible at the injection site. This fat pad was collected, sectioned and stained. Top panels, bright-field morphology of the transplant sections (left); zoomed bright-field image of the transplant section (middle); immunohistochemistry overlay of stainings for nuclear marker DAPI (blue), staining with a human-specific nuclei marker MAB1281 (green) and staining with antibody against CEBPA (red; right). Bottom panels, staining with nuclear marker DAPI (left), staining with a human-specific nuclei marker MAB1281 (middle) and staining with antibody against CEBPA (right). (b) HUES-9-derived MPCs were transduced with a combination of lenti-PPARG2, lenti-CEBPB and lenti-PRDM16 and transplanted and collected as described above. Specimens were sectioned and adjacent slides were stained with UCP1 and MAB1261, a human-specific nuclei marker (all images ×200 magnification). (c) Left, FDG intensity. Middle, PET image of mouse skin with transplants from HUES 9 MPC (PPARG2), HUES 9 MPC (PPARG2–CEBPB–PRDM16) and control 3T3-F442A cells. Right, the corresponding bright-field image of mouse skin with transplants from HUES 9 MPC (PPARG2), HUES 9 MPC (PPARG2–CEBPB–PRDM16) and control 3T3-F442A cells.