The vascular endothelium of the adipose tissue gives rise to both white and brown fat cells

Abstract

Adipose tissue expansion involves the enlargement of existing adipocytes, the formation of new cells from committed preadipocytes, and the coordinated development of the tissue vascular network. Here we find that murine endothelial cells (ECs) of classic white and brown fat depots share ultrastructural characteristics with pericytes, which are pluripotent and can potentially give rise to preadipocytes. Lineage tracing experiments using the VE-cadherin promoter reveal localization of reporter genes in ECs and also in preadipocytes and adipocytes of white and brown fat depots. Furthermore, capillary sprouts from human adipose tissue, which have predominantly EC characteristics, are found to express Zfp423, a recently identified marker of preadipocyte determination. In response to PPARγ activation, endothelial characteristics of sprouting cells are progressively lost, and cells form structurally and biochemically defined adipocytes. Together these data support an endothelial origin of murine and human adipocytes, suggesting a model for how adipogenesis and angiogenesis are coordinated during adipose tissue expansion.

Figure 1. Murine embryonic and postnatal eWAT morphology

A. eWAT depot at E18 composed of poorly differentiated mesenchymal cells (m = mitosis; some capillaries indicated with asterisk). B. eWAT at P7, where adipocytes appear yellowish in areas with abundant large capillaries (asterisks). C,D. EM micrograph of a vasculo-adipocytic islet showing EC (e; elongated cells), tight junctions (tj), pericytes (p; poorly differentiated cells with glycogen granules and surrounded by a distinct basal membrane), preadipocytes (pa; cells with small lipid droplets, (L)), and glycogen granules in pericyte (arrow). E. EC and pericytes of a capillary wall (asterisk indicates the capillary lumen). F. Enlargement of the black squared area in E revealing EC in endothelial-pericytic position. G. Enlargement of the red squared area in E highlighting typical oblique tight junction (arrow) joining cell in endothelial-pericytic position to adjacent EC. H. Example of cell in endothelial-pericytic position containing abundant glycogen (arrows). I. Example of a "pure" EC containing abundant glycogen (arrow). J. Example of a cell partially associated with the capillary wall (arrowheads) and partially abutting into the interstitial space.

Figure 2. VE-Cadherin lineage tracing in WAT and BAT

A. Immunohistochemical analysis on scWAT from VE-cadherin-Cre/R26Rmice showing the specific expression of VE-cadherin only in EC (brown); to note the negative pericytes indicated by red arrowhead. B, C. In early neonatal eWAT only vasculature is X-gal positive. D, E. scWAT (D) and eWAT (E) from P7 mice revealing X-gal staining (arrows) in developing and mature adipocytes. F. Confocal microscopy of eWAT from VE-cadherin-Cre/eGFPRmice showing a single optical plane of adipocytes (arrows), containing eGFP (green) and perilipin (red). Some eGFP-negative/perilipin-positive adipocytes are also visible (L). G. X-gal positive staining in developing BAT, and in muscle capillaries. H. Enlargement of the squared area in G. I, J. X-gal(I) and UCP-1 (J) colocalization to brown adipocytes (arrows). K-P. eWAT (K, N), scWAT (L, O), and BAT (M, P) adipose tissue from VE-cadherin-CreER^T2−/+/R26R^−/+ mice showing X-gal positive staining in adipocytes and EC. Q–S, Adipose tissues from VE-cadherin-CreER^T2−/−/R26R^−/+ control mice.

Figure 3. Effect of rosiglitazone on angiogenic sprouts originating from human adipose tissue

A. Capillary outgrown after 15 days of culture in the absence of rosiglitazone, indicating areas distal and proximal to the embedded explants. B. Enlargement of area distal to the explants. C. Electron microscopy of area similar to that shown in B, where tight junctions can be seen to connect cells. D. Enlargement of tight junction found between two EC. E. Enlargement of area proximal to the explant. F. Electron microscopy of area similar to that shown in E, revealing lumenized capillaries formed by EC joined by tight junctions. G. Enlargement of area shown in F. H. Capillary outgrowth after 15 days in the presence of rosiglitazone. I. Enlargement of area distal to the explant, revealing lipid droplets in cells interspersed among the capillary sprouts. J. Area proximal to the explant containing cells harboring larger lipid droplets. K. Electron microscopy of lipid-laden cells revealing features of classical white adipocytes, and of EC such as the tight junction in the squared area. L. Enlargement of area outlined in K.

Figure 4. Co-expression of endothelial and adipose cell genes in human and mouse systems

A. Relative mRNA levels of canonical adipocyte (top row) or EC (bottom row) genes in capillary outgrowth from explants grown for 14 days in the absence (−R) or presence (+R) of rosiglitazone. B. Relative mRNA levels of the adipocyte pre-determination marker Zfp423. Plotted are the means and SEM of 6–8 independent experiments. Statistical significance was assessed by non-paired, two tailed student t-test *=p<0.05; **=p< 0.001; C, D. In absence of rosiglitazone cells growing from human adipose explants are mostlyperilipin-negative (green) and vWF-positive (red). E–G. In the presence of rosiglitazone, perilipin coating around lipid droplets is present (E), with majority of cells becoming vWF-negative (F), and about 5% of cells co-expressing perilipin (green) and vWF (red) (G). H, I. Cells growing from adipose explants, that accumulate lipid droplets in the presence of rosiglitazone (I) are also adiponectin-positive (H). J. FACS analysis scheme of cells from SVF of eWAT and BAT. K. Comparison of CD45-CD29+CD24+Sca1+ population and CD45-CD29+CD24+Sca1+CD144+ (ASCM+) between depots. L. Gating strategy for experiments. Plotted in graphs (K) are the means and SEM of 3–4 independent experiments. Data were analyzed using two-tailed student t-tests, **p< 0.005 and *p < 0.00005. Fluorescence-minus-one antibody controls are in Supplemental Figure S4.