In Drosophila, the masses and sheets of adipose tissue that are distributed throughout the fly are collectively called the fat body. Like mammalian adipocytes, insect fat body cells provide the major energy reserve of the animal organism. Both cell types accumulate triacylglycerols (TAG) in intracellular lipid droplets; this finding suggests that the strategy of energy storage as well as the machinery and the control to achieve fat storage might be evolutionarily conserved. Studies addressing the control of lipid-based energy homeostasis of mammals identified proteins of the PAT domain family, such as Perilipin, which reside on lipid droplets. Perilipin knockout mice are lean and resistant to diet-induced obesity. Conversely, Perilipin expression in preadipocyte tissue culture increases lipid storage by reducing the rate of TAG hydrolysis. Factors that mediate corresponding processes in invertebrates are still unknown. The function of Lsd2, one of only two PAT domain-encoding genes in the Drosophila genome has been analyzed. Lsd2 acts in a Perilipin-like manner, suggesting that components regulating homeostasis of lipid-based energy storage at the lipid droplet membrane are evolutionarily conserved. Lsd2 is predominantly expressed in tissues engaged in high levels of lipid metabolism, the fat body and the germ line of females. Ultrastructural analysis in the germ line shows that Lsd2 localizes to the surface of lipid droplets. Mutant adults have a reduced level of neutral lipid content compared to wild type, showing that Lsd2 is required for normal lipid storage. Ovaries from Lsd2 mutant females exhibit an abnormal pattern of accumulation of neutral lipids from mid-oogenesis, which results in reduced deposition of lipids in the egg. Consistent with its expression in the female germ line, Lsd2 is shown to be a maternal effect gene that is required for normal embryogenesis (Gronke, 2003; Teixeira; 2003).

Lipids are a major form of energy storage in animals. They are stored in the intracellular neutral lipid droplets of specialized tissues such as adipose tissue in mammals and the fat body in insects. Although initially found in fat-related tissues, lipid droplets are organelles present in many, if not all, cell types. Hence, understanding the role of lipid metabolism at the level of the cell and whole organisms requires identification of the molecular mechanisms governing the biogenesis, trafficking and turnover of lipid droplets (Teixeira, 2003 and references therein).

Lipid droplets are formed by a unique monolayer of amphipatic phospholipids surrounding a central hydrophobic core of neutral lipids, mainly consisting of triacylglycerol (TAG) and sterol esters. Two mammalian proteins have been studied for their property to specifically localize at the surface of these organelles: Perilipin and ADRP (adipocyte differentiation-related protein also known as adipophilin) (Blanchette-Mackie, 1995; Brasaemle, 1997b; Greenberg, 1991). In addition to this particular property, ADRP and Perilipin also show sequence similarity, especially in their N-terminal region where they are ~40% identical (Lu, 2001). This N-terminal region, also present in another mammalian protein––TIP47, has been termed PAT domain (Lu, 2001). TIP47 was originally identified as a protein required for the transport of mannose 6-phosphate receptors from endosomes to the trans-Golgi network. Although initially controversial, the association of TIP47 to lipid droplets was recently verified (Miura, 2002). The presence of a PAT domain correlates with the ability of proteins to localize to lipid droplets, although it has been recently shown not to be absolutely required (Garcia, 2003: McManaman, 2003; Targett-Adams, 2003). Proteins with a PAT domain have been found in a wide variety of species, including Drosophila, and together form the PAT family (Lu, 2001; Teixeira, 2003 and references therein).

Perilipin and ADRP were originally identified as genes highly expressed in adipose tissue (Greenberg, 1991; Jiang, 1992). Further studies have shown that Perilipin expression is restricted to differentiated adipocytes and steroidogenic cells (Greenberg, 1993; Servetnick, 1995), whereas ADRP is more ubiquitously expressed (Brasaemle, 1997b; Heid, 1998). In cultured cells, it has been shown that the ectopic expression of Perilipin or ADRP increases the capacity of cells to take up long fatty acids from the medium and to accumulate neutral lipids (Brasaemle, 2000; Gao, 1999; Imamura, 2002; Souza, 2002). Reciprocally, the addition of fatty acids to the culture medium stimulates neutral lipid accumulation in cells and increases intracellular levels of Perilipin or ADRP (Brasaemle, 1997a: Gao, 2000; Souza, 2002). The reciprocal regulation of Perilipin/ADRP and neutral lipids suggests that these two proteins have a role in lipid metabolism regulation (Teixeira, 2003 and references therein).

The function of Perilipin in vivo has been analyzed in Perilipin-deficient mice (Martinez-Botas, 2000; Tansey, 2001). These mice are viable, fertile and have normal size and weight. However, they have a reduced adipose tissue mass and are more muscular than controls. They are resistant to induced obesity and have a higher metabolic rate. These phenotypes are explained by the observed increase in basal lipolysis activity. This analysis together with data from cell culture experiments ( [Brasaemle, 2000; Souza, 1998) led to the proposal that Perilipin has a protective role against lipases. The viability of Perilipin mutant mice could be explained by a partial compensation by other mammalian PAT-members. The identification of the in vivo role of the other members of this family awaits the generation of mutants (Teixeira, 2003).

Two Drosophila melanogaster members of the PAT family, Lsd1 and Lsd2, have been identified by BLAST search (Lu, 2001). Both are equally related to any of the three mammalian members of the PAT family (Miura, 2002). Ectopically expressed GFP-tagged Lsd1 or Lsd2 localize to lipid droplets, both in mammalian cell culture and in the fat body of Drosophila (Miura et, 2002). This shows that the targeting to lipid droplets is a feature conserved between Drosophila and mammalian PAT-family members (Teixeira, 2003).

Lsd2 is expressed during all stages of the Drosophila life cycle, and transcripts accumulate in specific spatiotemporal patterns. Northern blot analysis demonstrates a strong enrichment of the 2.4 kb Lsd2 mRNA in early embryos, and this enrichment reflects a maternal contribution, as also visualized by whole-mount in situ hybridization of syncytial blastoderm-staged embryos. Maternal Lsd2 mRNA becomes subsequently degraded, except in germline precursor cells, where the transcripts are enriched up until midembryonic stages. There is transient Lsd2 expression in the amnioserosa and continuous expression in the developing fat body and the anterior midgut, two tissues known to function in lipid storage and nutrient lipid resorption, respectively. During the first and second larval stages, Lsd2 is only moderately expressed. In third instar larvae, the gene is strongly expressed in the fat body, the major TAG storage tissue (Gronke, 2003).

In order to visualize the intracellular localization of LSD2 in vivo, expression of an LSD2-EGFP fusion protein was targeted to the third instar larval fat body by using the Gal4/UAS system in conjunction with a fat body-specific Gal4 driver (FB-Gal4). In living fat body cells of such individuals, LSD2-EGFP is associated with vesicular structures of various sizes. Purification of their fat bodies' intracellular vesicles by density gradient fractionation results in the detection of LSD2-EGFP on lipid droplet surfaces, as identified by Nile red staining. Moreover, Western blot analysis of density-fractionated fat body homogenates of third instar larva shows LSD2 enrichment in the lipid droplet fraction. The spatiotemporal expression patterns and the intracellular localization of LSD2 are therefore consistent with the proposal that the protein plays a regulatory role in global TAG storage by acting at the level of lipid droplets (Gronke, 2003).

Lsd2 mutant flies contain significantly less TAG levels than controls. For simplicity, those individuals are referred to as lean, and those exceeding the TAG levels of controls are referred to as obese. As compared to freshly hatched Lsd2^revKG00149 male individuals that carry a functional Lsd2 allele, the TAG content of Lsd2⁵¹, Lsd2⁴⁰, and Lsd2^KG00149 mutants is reduced by 34.5%, 28%, and 37.2%, respectively. In order to unambiguously establish whether the TAG reduction is caused by the loss of LSD2 function in the fat body, a cDNA-based Lsd2 transgene (UAS-Lsd2) was expressed in response to the FB-Gal4 driver in Lsd2 mutant individuals. Expression of Lsd2 in the fat body reverts the leanness of Lsd2^KG00149 mutant flies, indicating that loss of LSD2 activity is the cause of the mutant phenotype. Thus, LSD2 is an essential component in the regulation of lipid storage in the fly fat body (Gronke, 2003).

In order to test whether LSD2 activity is capable of modulating the TAG level of otherwise wild-type flies, Lsd2 was overexpressed in the fat body. Western blots with proteins extracted from freshly hatched male flies tested with α-LSD2-specific antiserum show gradually increased levels of UAS-Lsd2 transgene-dependent LSD2 activity in the fat body, and these increased levels result in increasingly severe obesity phenotypes. Flies moderately overexpressing LSD2 elevate organismal TAG storage by 28%, whereas strong Lsd2 overexpression causes a TAG storage increase by 48.5% (Adh-Gal4; lane 8) compared to control individuals bearing the noninduced UAS-Lsd2 transgene. These data demonstrate that modulation of LSD2 levels is sufficient to adjust TAG storage in ad libitum fed flies. The obese FB-Gal4:UAS-Lsd2 flies are more starvation resistant than control flies, whereas the lean Lsd2⁴⁰ mutant flies are starvation sensitive. Starvation resistance of Lsd2-overexpressing flies is accompanied by a delayed but complete premortal depletion of the TAG stores. Collectively, these results indicate that Lsd2 activity can adjust TAG storage at an organismal level at times when food is accessible to ensure extended survival when food supply is limiting (Gronke, 2003).

These results provide evidence that Lsd2 of Drosophila is an essential component of the genetic circuitry that controls energy homeostasis at the level of fat storage. The finding that varying the amount of LSD2 causes a dosage-dependent increase of TAG storage, whereas the lack of LSD2 results in lean flies, is reminiscent of results obtained with the vertebrate PAT domain protein Perilipin. This suggests that LSD2 operates in a Perilipin-like manner by modulating the rate of lipolysis. The results also suggest that PAT domain proteins, which are found in higher eukaryotes as diverse as human, fly, and the slime mold Dictyostelium, share an ancestral function in the organismal control of lipid storage homeostasis. The Drosophila flies with Lsd2 lack-of-function and gain-of-function genotypes therefore represent a genetically accessible model system to identify the components and mechanisms underlying the phenomenon of energy homeostasis in order to address questions concerning energy storage disorders (Gronke, 2003).

GENE STRUCTURE

cDNA clone length - 2138

Bases in 5' UTR - 263

Exons - 4

Bases in 3' UTR - 816

PROTEIN STRUCTURE

Amino Acids - 352

Structural Domains

Comparison of the two Drosophila PAT domain-encoding genes, Lsdp1 and Lsd2, with genes present in Anopheles gambiae, Bombyx mori, and Dictyostelium discoideum suggest that insect genomes encode only two PAT domain proteins. In both Drosophila and Dictyostelium, they are associated with intracellular lipid droplets (Miura, 2002). However, they cannot be directly homologized to any of the vertebrate family members. In order to demonstrate a possible conserved function, the Lsd2 gene was characterized and it was asked whether the Lsd2 product participates in the regulation of lipid storage in the fly, as observed with mammalian Perilipin (Gronke, 2003: Teixeira, 2003).

date revised: 25 October 2004

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.