All cells package neutral lipids in discrete storage droplets that are characterized by unique surface proteins. In adipose cells, the enormous droplets contain triacylglycerols, the primary bodily energy stores, whereas in steroidogenic cells, much smaller droplets contain cholesteryl esters, the precursors for steroid hormone synthesis. In the general cell population, available fatty acids are captured in even smaller droplets and are used eventually as energy sources or for membrane remodeling. The adipocyte remains our primary model system, and we focus on the processes whereby hormones regulate both the packaging and hydrolysis of stored neutral lipids. Obesity is characterized by a surfeit of stored adipose lipids, and the wasting syndrome, cachexia, is associated with a marked loss of these important energy stores. Moreover, type 2 diabetes (NIDDM, adult onset) is characterized by elevated plasma fatty acids, which are thought to result from unchecked lipolysis in adipose cells. Current evidence indicates that suppression of this lipolytic response ameliorates many of the symptoms associated with type 2 diabetes. Our fundamental approach to the questions, outlined below, involves manipulation of cultured cell systems using molecular biological tools complemented with biochemical and cellular biological techniques. We also employ transgenesis and knock out technology. One important goal is to dissect the molecular events subsequent to stimulation of adipose cells by lipolytic and antilipolytic hormones, such as epinephrine and insulin, respectively. Stimulation by catecholamines involves activation of adenylyl cyclase, elevation of cAMP, and activation protein kinase A. Hormone-sensitive lipase, the rate-limiting enzyme of lipolysis, is phosphorylated by protein kinase A. Although the structure and PKA phosphorylation sites of HSL are known, there is little information on the process whereby cytoplasmic HSL gains access to its substrate, the triacylglycerols housed within the lipid storage droplets. We have found that activated HSL rapidly translocates and adheres to the surface of lipid droplets, but nothing is known of its cytoplasmic location in unstimulated cells, the translocation process, or the target locus on the droplet surface. A number of approaches are aimed at dissecting this process, including (1) construction of a fusion library for identifying possible HSL binding proteins; (2) use of retroviral systems to introduce HSL constructs that encode a variety of mutagenized forms, especially those in which the various phosphorylation sites are mutated either singly or in combination; and (3) identification of lipolytic inhibitors that act downstream of PKA activation. To this end, in collaboration with a biotech company we have established a screening system that has identified compounds that are both strongly antilipolytic and exhibit considerable efficacy in reversing a number of symptoms in diabetic animal models. Such compounds are not only useful in our biochemical research but also represent potential therapeutic agents. An important, but largely overlooked, component in the lipolytic equation is the lipid droplet. Previously thought to be merely an amorphous accumulation of neutral lipids, we have discovered a family of proteins, termed perilipins, that are found exclusively at lipid droplet surfaces. Perilipin is a single copy gene that gives rise, by alternative splicing, to three isoforms, A, B, and C. The perilipin gene is expressed most strongly in adipose cells where the A and B isoforms coat the triacylglycerol-containing droplets. The gene is expressed also in steroidogenic cells where perilipins A and C coat the cholesteryl ester-containing droplets. Like adipocytes, steroidogenic cells use a cAMP stimulated process and an HSL-like enzyme to release cholesterol, which serves as a substrate for steroid hormone synthesis. Perilipins are polyyyphosphorylated by PKA and, thus, their occurrence in only those cells in which lipolysis is mediated by increased cAMP points to a role for perilipin in the process of lipid breakdown. We have found that a related protein, adipose differentiation?related protein (ADRP), also termed adipophilin, coats the lipid droplet in all other cells. However, expression of perilipin in fibroblastic cells leads to the disappearance of ADRP, after which the lipid droplets acquire a coating of perilipin. The non-phosphorylated perilipin exerts a protective effect and suppresses lipolysis in such cells. Upon activation of PKA and subsequent phosphorylation of perilipin, a robust lipolysis ensues, which is due solely to perilipin phosphorylation, since the fibroblastic cells contain no PKA-mediated lipases. A further manifestation of the protective effect of non-phosphorylated perilipin is the normal deposition of fat reserves in adipose tissue. We found that the perilipin null mouse had a 70% decrease adipose tissue, but are of normal weight and have similar caloric intake as wild type mice. Oddly, despite their greatly diminished adipose tissue, these animals have elevated plasma leptin values. The perilipin null animal also provided important clues on perilipin function. As expected, the adipocytes from these animals exhibit elevated basal lipolysis. Surprisingly, their adipocytes were also refractory to lipolytic stimulation, and we subsequently found that perilipin was required to elicit the PKA-mediated translocation of HSL from the cytosol to the surface of the lipid droplet. We have also shown that HSL translocation also requires phosphorylation of the enzyme at one of its C-terminal PKA sites. Thus, stimulated lipolysis is a concerted reaction requiring PKA phosphorylation of both perilipin and HSL. While the specialized functions of lipid deposition in adipocytes and steroidogenic cells are our primary focus, we have explored the function of ADRP in lung, where its expression level is second only to the expression in adipose tissue. In lung, ADRP is expressed in lipofibroblasts, cells that capture lipid from the serum and pass these lipids on to type 2 epithelial cells, where they are incorporated into surfactant phospholipids. ADRP is highly expressed in the lipofibroblasts and may have a role in the transfer of lipids from theses cell to the type 2 epithelial cells. In addition to the perilipins and ADRP, we have identified a number of related genes in Drosophila melanogaster and Dictyostilium discoidium, plus additional mammalian genes. When fused to GFP all of the proteins encoded by these genes target to lipid droplets when expressed in mammalian CHO fibroblasts. In addition to their sequence homologies, similarities in gene structures indicate that the mammalian genes derive from an ancient gene family, and it is likely that all of these proteins will be found to have a role in lipid metabolism. We have designated this family of proteins (perilipin, ADRP, TIP-47 plus those from distantly-related proteins as the PAT family of proteins) Indeed, very recent publications from other groups show that the related drosophila proteins play a role in lipid metabolism. Currently structural information on most PAT proteins is lacking, with one notable excpetion, TIP-47. This protein exist in both lipid droplet-bound and in freely soluble form. We managed to crystallize a major segment of TIP-47, and this domain shares considerable structural similarity with the C-termina domain of ApoC. Notably, this is the domain of ApoC that binds to its cell surface recptor, which raises the intriguing possibility that the PAT proteins, espcially ADRP, may mediate cell-to-cell transfer of lipid droplet components via a cell surface binding event.
Showing the most recent 10 out of 19 publications
