Genetic defects in lipid storage: lipodystrophies
Human lipodystrophies are the clinical manifestations of total (congenital generalized lipodystrophy, CGL) or partial (familial partial lipodystrophy, FPL) loss of body fat (see Garg & Agarwal, 2009; Huang-Doran et al, 2010; Vigouroux et al, 2011). Lipodystrophies cause severe changes of whole-body energy metabolism and are commonly associated with insulin resistance, hepatic steatosis, hypertension and other metabolic dysfunctions. Inability to store TGs in white adipose tissue gives rise to lipid storage in other tissues and tissue lipotoxicity. Lack of white adipose tissue leads to leptin deficiency and associated metabolic defects, such as insulin resistance. Many defects of severe lipodystrophies can be corrected by leptin supplementation (Oral et al, 2002; Shimomura et al, 1999).
Numerous gene defects cause CGL or FPL (reviewed in Garg, 2011). Here we review lipodystrophies with established connections to LD biology. Many of the lipodystrophy genes encode TG synthesis or storage proteins, and some act at the LD surface and are involved in LD formation and regulation. Still others regulate adipogenesis, thereby affecting LDs indirectly.
Two CGL loci encode proteins that regulate de novo TG synthesis: 1-acylglycerol-3-phosphate O-acyltransferase 2 (AGPAT2) and lipin1 (LPIN1). AGPAT2 catalyzes the formation of phosphatidic acid from lyso-phosphatidic acid and fatty acyl-CoA. Lipin1, a phosphatidic acid phosphatase, removes the phosphate group from phosphatidic acid to form diacylglycerol, the direct precursor of TG. AGPAT2 mutations account for ∼50% of CGL (Van Maldergem et al, 2002). So far, none of the other AGPAT isoforms appears to be implicated in CGL. This suggests a specialized role for AGPAT2, at least in adipocytes. In contrast to AGPAT2, LPIN1 mutations have been found to be associated with lipodystrophy only in mouse models; there are no known human LPIN1 mutations causing lipodystrophy.
How AGPAT2 and LPIN1 mutations cause lipodystrophy is uncertain. Mutation of either enzyme causes reduced TG levels as well as accumulation of lipid synthesis intermediates, and phospholipids in tissues. For example, both Lipin1 and Agpat2 knockout mice have increased phosphatidic acid and lyso-phosphatidic acid levels in the adipose tissue (Gale et al, 2006; Peterfy et al, 2001). One model posits that lipid synthesis products, such as phosphatidic acid or lyso-phosphatidic acid, accumulate and inhibit adipocyte differentiation, perhaps by influencing PPARγ signalling (Gale et al, 2006; Yang et al, 2011).
Mutations of Berardinelli–Seip congenital lipodystrophy 2 gene (BSCL2/SEIPIN) also lead to severe CGL. BSCL2 mutations in lipodystrophy patients result mostly in null alleles (Agarwal et al, 2004). BSCL2 mutations exhibit more severe lipodystrophy and metabolic alterations than AGPAT2 mutations (Van Maldergem et al, 2002).
How BSCL2 deficiency causes lipodystrophy remains unclear. The BSCL2 protein seipin is an ER protein, embedded in the membrane by a hairpin-type hydrophobic sequence. BSCL2 is expressed in adipocytes and up-regulated during adipocyte differentiation in 3T3L1 cells (Chen et al, 2009). BSCL2 knockdown inhibits adipocyte differentiation and suppresses PPARγ expression in this cell line (Chen et al, 2009). Moreover, it causes strong alterations in LD size, distribution, and number (usually fewer LDs) in multiple cell types and organisms, leading to markedly altered LD morphology (Fei et al, 2008; Szymanski et al, 2007; Tian et al, 2011). The yeast seipin orthologue forms oligomers and localizes near LDs (Binns et al, 2010; Lundin et al, 2006; Szymanski et al, 2007), suggesting a role in LD formation. However, seipin might also function in lipid biosynthesis pathways: its knockout alters cell phospholipids in yeast and Drosophila (Fei et al, 2011; Tian et al, 2011) and fatty acid composition in lymphoblastoid cell-lines, the latter from reduced desaturase activity (Boutet et al, 2009). In yeast and Drosophila, seipin knockout leads to accumulation of phosphatidic acid, which might induce LD fusion and alter LD morphology (Fei et al, 2011). Phosphatidic acid excess may also influence signalling pathways for adipocyte differentiation. In Drosophila, BSCL2/seipin depletion leads to loss of fat body lipids and to ectopic TG accumulation (Tian et al, 2011).
Deficiency of the membrane protein caveolin 1 also causes lipodystrophy. A homozygous nonsense mutation with complete loss of CAVEOLIN1 (CAV1) expression causes CGL, and heterozygous mutations were found in patients with atypical partial lipodystrophy (Cao et al, 2008; Kim et al, 2008). Caveolin 1 is an essential organizer of caveolae, specialized cholesterol-rich microdomains in the plasma membrane that form invaginations. Caveolin 1 also localizes to LDs (Ostermeyer et al, 2001) and is highly expressed in adipocytes (Lisanti & Razani, 2001). A related protein, cavin 1, another component of caveolae (Hill et al, 2008), was found to be mutated in a patient with lipodystrophy (Matsuo et al, 2010; Shastry et al, 2010). Cavins interact with caveolin, and loss of cavin 1 leads to loss of caveolae and caveolin 1 mislocalization (Liu et al, 2008). Since mutations in CAV1 and cavin 1 profoundly affect whole-body TG storage, caveolae may be important in lipid storage (Pilch & Liu, 2011). Caveolae are thought to function in the cellular response to mechanical stress, endocytosis, transcytosis, fatty acid uptake, LD formation, and lipid trafficking (Parton & Bastiani, 2010). The relationship between caveolae and LDs is still unclear.
Mutations in two LD proteins, perilipin1 and CIDEC, have been reported to cause FPL. A homozygous CIDEC mutation in a FPL patient led to a truncated protein that does not target LDs (Rubio-Cabezas et al, 2009) and cannot induce formation of unilocular LDs in adipocytes. Instead multiple small LDs form, similar to the murine knockout phenotype (Nishino et al, 2008). Two frame-shift mutations in the C-terminus of perilipin1 are associated with FPL and insulin resistance. Unlike wild-type perilipin1, the mutated perlipin1 fails to bind to and inhibit comparative gene identification-58 (CGI-58), leading to constitutive ATGL activation by CGI-58, resulting in increased basal lipolysis (Gandotra et al, 2011). While those mutations account for FPL in very few individuals, most FPL cases are caused by mutation of lamin A (LMNA). As a component of the nuclear lamina, LMNA is important for maintaining the nuclear envelope functions, loss of which leads to premature cell death and loss of adipocytes (Garg, 2004). This might also happen in FPL patients carrying mutations of ZMPSTE24, encoding a metalloprotease required for processing lamin A (Agarwal et al, 2003). In several other FPL patients, mutations in the genes for transcription factor PPARγ and AKT2, a factor involved in insulin signalling, were reported (George et al, 2004; Savage et al, 2003). Those mutations lead to defective adipocyte differentiation and thereby disturb TG storage. It is unclear why certain mutations cause FPL rather than CGL. Perhaps FPL is caused by deficient TG storage in existing adipocytes, and CGL is due to failure to form adipocytes.
In contrast to genetic inherited forms of lipodystrophies, acquired forms due to autoimmune diseases or drug treatment are much more common (see Garg, 2004). Protease inhibitors used to treat HIV infection are the most frequent cause of acquired lipodystrophy. The mechanism of the pathogenesis is not well understood. HIV protease inhibitors change the expression and localization of transcription factors, such as PPARγ or SREBP, which mediate adipocyte differentiation (Caron et al, 2001). Moreover, they increase basal and stimulated lipolysis in adipocytes, as shown in 3T3-L1 cells. This effect is mediated by a decrease in perilipin levels on the LDs by increased lysosomal perilipin degradation (Adler-Wailes et al, 2005).
Rapid loss of triglyceride stores: cancer cachexia
Cachexia, a complex metabolic syndrome, is common in cancer patients, particularly gastrointestinal, prostate and lung cancer (Tisdale, 2005). Unlike lipodystrophy, which features chronic deficiency of adipose tissue, cachexia is an acute wasting disease. Lipid metabolism is fundamentally changed, leading to dramatically reduced body weight, caused early by a loss of adipose tissue and later by atrophy of skeletal muscle (Bing & Trayhurn, 2009). Those changes are associated with poor response to chemotherapy and high mortality: 15–20% of cancer deaths are caused by cachexia (Tisdale, 2002). Increased lipolysis is a key factor in cachexia, and cachexia patients show elevated blood glycerol and fatty acids (Shaw & Wolfe, 1987).
The role of LDs in cachexia is beginning to be unravelled. Recent advances reveal that ATGL, and not HSL as previously thought, mediates increased lipolysis in cachexia (Das et al, 2011). Atgl knockout mice are completely protected from adipose tissue wasting and had no increased lipolysis after inducing cachexia, despite high levels of lipid-mobilizing factors, such as zinc-alpha-2-glycoprotein 1 (AZGP1), tumour necrosis factor α (TNF-α), or interleukin-1 (Das et al, 2011). Thus, inflammatory and lipolytic mediators that activate ATGL, potentially secreted by the tumour, might cause uncontrolled loss of adipose tissue in cachexia. Intriguingly, skeletal muscle loss was also absent in Atgl knockout mice.