Anti-angiogenic pigment epithelium-derived factor regulates hepatocyte triglyceride content through adipose triglyceride lipase (ATGL)

Journal of Hepatology 48 (2008) 471–478 www.elsevier.com/locate/jhep

Anti-angiogenic pigment epithelium-derived factor regulates hepatocyte triglyceride content through adipose triglyceride lipase (ATGL)q Chuhan Chung2,3, Jennifer A. Doll1, Arijeet K. Gattu2,3, Christine Shugrue2,3, Mona Cornwell1, Philip Fitchev1, Susan E. Crawford1,* 1

Department of Pathology, Northwestern University Feinberg School of Medicine, W127, 300 E. Superior, Tarry 7-753, Chicago, IL 60611, USA 2 Department of Medicine, Section of Digestive Diseases, Yale University School of Medicine, New Haven, CT, USA 3 VA Connecticut Healthcare System, West Haven, CT, USA

Background/Aims: Anti-angiogenic pigment epithelium-derived factor (PEDF) is a 50 kDa secreted glycoprotein that is highly expressed in hepatocytes. Adipose triglyceride lipase (ATGL), a novel lipase critical for triglyceride metabolism, is a receptor for PEDF. We postulated that hepatocyte triglyceride metabolism was dependent on interactions between PEDF and ATGL, and loss of PEDF would impair mobilization of triglycerides in the liver. Methods: Immunoprecipitation studies were performed in PEDF null and control hepatocytes with recombinant PEDF (rPEDF) as bait. Immunofluorescent microscopy was used to localize ATGL. Triglyceride content was analyzed in hepatocytes and in whole liver with and without rPEDF. ATGL was blocked using an inhibitor, (R)-bromoenol lactone. Results: PEDF co-immunoprecipitated with ATGL in hepatic and HCC lysates. All PEDF deficient livers demonstrated steatosis. Triglyceride content was significantly increased in PEDF null livers compared to wildtype (p < 0.05) and in isolated hepatocytes (p < 0.01). Treatment of PEDF null hepatocytes with rPEDF decreased TG content (p < 0.05) and this activity was dependent on ATGL. Conclusions: Our results identify a novel role for PEDF in hepatic triglyceride homeostasis through binding to ATGL and demonstrate that rPEDF and ATGL localize to adiposomes in hepatocytes. Dysregulation of this pathway may be one mechanism underlying fatty liver disease.  2007 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. Keywords: PEDF; ATGL; Adiposome; Triglyceride; Liver

1. Introduction Hepatic steatosis is a common clinical entity that can be associated with pathologic effects on the liver. These Received 16 August 2007; received in revised form 18 October 2007; accepted 22 October 2007; available online 26 November 2007 Associate Editor: C.P. Day q The authors declare that they do not have anything to disclose regarding funding from industries or conflict of interest with respect to this manuscript. NIH funded study. * Corresponding author. Tel.: +1 312 503 2844; fax: +1 312 503 2843. E-mail address: [email protected] (S.E. Crawford).

range from abnormal liver function tests to end stage liver failure and hepatocellular carcinoma (HCC) [1,2]. The accumulation of fat in the liver may also impair hepatic insulin sensitivity and is characteristic of the Type II Diabetes Mellitus and the Metabolic Syndrome [3]. Recently, adipose triglyceride lipase (ATGL) was identified as a novel lipase in adipose tissue that catalyzes the initial step in triglyceride (TG) hydrolysis [4]. ATGL was discovered after several groups noted that hormone sensitive lipase (HSL) null animals were lean and that organs from these animals contained predominately diglycerides [5]. This indicated that HSL was rate limiting for the hydrolysis of diglycerides rather than

0168-8278/$32.00  2007 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jhep.2007.10.012

472

C. Chung et al. / Journal of Hepatology 48 (2008) 471–478

triglycerides. Furthermore, adipose tissue from HSL null animals retained triglyceride lipase activity implicating an unidentified lipase with affinity for triglycerides [6,7]. This enzyme was subsequently identified by three separate groups and variously named ATGL, desnutrin or iPLA2f [4,8,9]. An evolutionarily conserved enzyme, ATGL contains a signature nucleotide-binding ((G/ A)XGXXG) and a lipase consensus (GXSXG) sequence motif [9]. Notably, the ATGL null animals were reported to have elevated triglyceride content in nearly every organ including the liver [10]. Since its discovery, ATGL activity has been confirmed in human adipose tissue and single nucleotide polymorphisms in the ATGL gene have been implicated as a potential risk factor for the development of Type II Diabetes Mellitus [11,12]. A recent report noted that ATGL is a receptor for pigment epithelium-derived factor (PEDF) in retinal epithelial cells and, using a cell-free system, demonstrated that this interaction induced lipase activity [13]. PEDF is a 50 kDa hypoxia-sensitive secreted glycoprotein with multiple functions in several organ systems [14–16]. First discovered in retinal epithelial cells, PEDF was found to be a critical survival factor for neuronal cells in culture [17]. Dawson et al. identified PEDF as the most potent endogenous angiogenesis inhibitor [16], and in target organs affected by diabetes such as the eye, PEDF expression is decreased [15]. In the normal adult liver, PEDF is highly expressed in hepatocytes [18]; however, the function of PEDF in the liver is not well understood. Hepatic tumors express lower PEDF levels and replenishing the liver with PEDF by injection or expression vectors exerted anti-tumor activity resulting in decreased tumor growth by blocking neovascularization [19]. Here, we provide evidence to support an important liver-specific function of PEDF. PEDF avidly binds hepatocyte- and HCC-derived ATGL and this interaction is critical in regulating TG content. Similar to ATGL deficient mice [6], deletion of PEDF in a murine model results in hepatic steatosis and TG accumulation. Restoration of PEDF to PEDF null hepatocytes ameliorates the steatotic phenotype, an effect blocked by ATGL inhibition. These findings demonstrate that PEDF is a novel regulator of hepatocyte TG metabolism through ATGL.

2. Materials and methods 2.1. PEDF-null mice PEDF-deficient mice were generated and genotyped as described [20]. Livers were harvested from adult wildtype (n = 65) and compared to adult PEDF null (n = 48) mice. Animals received humane care with unrestricted access to food and water. A separate group of animals (n = 8) received a slightly higher fat (15%; n = 8) diet and were com-

pared to mice on normal chow (5% fat; n = 8). Age-matched wildtype and PEDF null animals were used in all experiments. Procedures were approved by the Institutional Animal Care and Use Committee of Northwestern University and VA Connecticut Healthcare System.

2.2. Cell cultures, rPEDF production and hepatocyte and adiposome isolation Human HCC cell lines, Hep3b-2.1-7 and HepG2, were grown at 37 C in 5% CO2 in MEM with 10% FCS, non-essential amino acids (Cellgro), sodium pyruvate, penicillin (100 U/ml), and streptomycin (100 lg/ml) (Gibco). Oleate was obtained from Sigma. HEK cells, transfected with the full length human PEDF cDNA, with a 6X-Histidine tag at the C-terminus, cloned into a hygromycin-resistance expression vector, pCEP4 (Invitrogen), were grown in DMEM with 10% FCS and penicillin/streptomycin. Hygromycin was added (400 lg/ml) for selection. Recombinant PEDF (rPEDF), containing a His-tag, was purified as previously described [16]. Isolated wildtype and PEDF null hepatocytes from 6- to 8-weekold mice were prepared as described [21]. Briefly, mouse livers were perfused with Hanks A and then Hanks B medium containing 0.05% collagenase (Boehringer Mannheim Biochemicals) and 0.8 U trypsin inhibitor (Sigma). Livers were excised, minced, passed through serial nylon mesh filters. Cells were washed and plated (1 · 106 per 35 mm collagen I coated plates, BD Biosciences) in Williams’ Medium E with 10% FCS, 10 mM HEPES, 2 mM L-glutamine, 1 lM dexamethasone, 4 mg/L insulin and penicillin/streptomycin. Cells were incubated at 37 C overnight and used 24–48 h after isolation. Medium was changed daily. Adiposomes were isolated from WT and PEDF null livers as described [22].

2.3. Immunoblotting and immunoprecipitation Antibodies against PEDF [16], ATGL (Cayman Chemicals) and adipose differentiation-related protein (ADRP; kind gift of Dr. B. Chang, Baylor School of Medicine) were used to assess protein expression and interactions by Western blotting. Protein content was determined by Bradford analysis. Forty micrograms of protein was electrophoresed on a 10% SDS–polyacrylamide gel as described [20], and transferred onto PVDF membranes. After blocking overnight (5% milk solution), immunoblotting was performed using anti-PEDF or anti-ATGL. Blots were washed in TBS + 0.1% Tween, incubated with a peroxidase-conjugated donkey anti-rabbit IgG, and peroxidase detected by a chemiluminescence assay (Pierce). Membranes were stained with Ponceau S or immunoblotted versus b-actin (Sigma) to confirm equal protein loading. Densitometry, using NIH software ImageJ, was used to compare signal intensities between samples. Immunoprecipitation (IP) was performed by incubating rPEDF 200 ng or BSA (negative control) with 150 lg of lysate from wildtype or PEDF null hepatocytes, or HCC cells, overnight. Precipitation was performed by adding 40 ll Ni–NTA beads after washing in PEDF binding buffer (20 mM NaH2PO4, 500 mM NaCl, pH 7.8). Beads were spun down and supernatant saved to assess pull-down efficiency. Beads were washed at least three times, resuspended in sample buffer, electrophoresed on 7.5% SDS–polyacrylamide gels, immunoblotted for ATGL and PEDF, and assessed by densitometry as described above.

2.4. Triglyceride determination Qualitative histologic TG content analysis was assessed using Oil Red O (ORO) staining. Quantitative TG measurements were performed using an enzymatic method (Diagnostic Chemicals Ltd.) after lipid extraction using a modified Folch technique [3]. Briefly, 200 mg of liver tissue or isolated hepatocytes from wildtype and PEDF null mice was washed with ice cold PBS and homogenized with 2:1 chloroform:methanol. After agitation and centrifugation, the organic phase was collected and dried under nitrogen gas. After reconstitution in chloroform, TG content was measured using an L-a-glycerol phosphate oxidase system. Additional experiments were validated using a commercial TG kit (Stanbio Triglyceride Liquicolor). Samples were normalized for protein content.

C. Chung et al. / Journal of Hepatology 48 (2008) 471–478

a

b

WT 40

ATGL

473

80 µg

rPEDF 200 400 ng

PEDF Oleate

c Control

0.4

d + +

1.0 mM

ATGL PEDF

PEDF-/– + + –

+ +

WT – +

+ –

lysate rPEDF

ATGL PEDF

β -actin

e

BSA post-IP rPEDF post-IP PEDF IP supernat. IP supernat. control

f

% ATGL after IP

ATGL PEDF

Fig. 1. ATGL expressed by murine hepatocytes and human HCC cells binds avidly to PEDF. (a) Immunoblotting shows ATGL expression by wildtype and PEDF null hepatocytes, as well as by two human HCC cells. (b) Endogenous PEDF levels are shown in wildtype hepatocyte lysates by immunoblot. rPEDF is shown as a control. (c) By immunoblot, ATGL levels increased while PEDF levels decreased with oleate treatment in HepG2 HCC cells. (d) Recombinant PEDF (rPEDF) was able to preferentially immunoprecipitate ATGL from PEDF null lysates compared to those from wildtype hepatocytes. This suggests rPEDF is unable to displace the avid binding affinity between native PEDF and ATGL. (e) ATGL from HepG2 lysates was immunoprecipitated using rPEDF. No ATGL was obtained using an equivalent amount of the negative control, BSA. (f) The relative amounts of ATGL extracted by rPEDF in (e) were assessed by densitometry of the post-immunoprecipitation supernatants by densitometry. PEDF bound to nearly 70% of ATGL. [This figure appears in colour on the web.] Wildtype and PEDF null hepatocytes were isolated, incubated for 24 h, and treated with rPEDF (500 ng/ml) 2 h prior to TG measurements. In a separate set of experiments, rPEDF treatment was performed in serum-free media due to recent reports that demonstrate PEDF’s binding affinity to serum proteins [23]. Hepatocytes were treated with (R)-BEL, an irreversible submicromolar ATGL inhibitor that blocks 90% of TG lipase activity at 2 lM [9], 4–6 h before TG measurements. To assess responsiveness to rPEDF after removal of (R)-BEL, PEDF null hepatocytes treated with (R)-BEL were washed, re-fed media without inhibitor, incubated for 24 h, treated with rPEDF and TG content assessed.

555 (Molecular Probes) were used as secondary antibodies at 1:500. DAPI was used for nuclear localization. BODIPY 493/503 (2 lg/ml; Molecular Probes) was used to localize intracellular neutral lipid vesicles. Controls were incubated in secondary antibody and DAPI only.

2.6. Statistical analysis p-Values were calculated, assuming equal sample variance, using two-tailed Student’s t-tests, on Prism software, with p < 0.05 deemed to be statistically significant. Values were stated as means ± standard error.

2.5. Immunohistochemistry and immunofluorescence Immunohistochemistry of liver sections was performed using a streptavidin–biotin complex immunoperoxidase technique. Sections were deparaffinized, treated to inhibit endogenous peroxidase and subjected to antigen retrieval as described [20]. All tissues were H&E stained. An experienced pathologist (SEC) performed all histological evaluations in a blinded manner. Isolated murine hepatocytes or HepG2 cells were grown on methanol-treated coverslips until 50% confluence. Cells were rinsed with PBS fixed in 2% paraformaldehyde, permeabilized with 0.25% (w/v) saponin for 15 min and blocked in 3% goat serum. Coverslips were incubated with anti-PEDF antibody at 1:500 and anti-ATGL antibody at 1:100. In a separate set of experiments, 500–1000 ng/ml of rPEDF was added to HepG2 cells at intervals from 30 min to 4 h prior to fixation. rPEDF was localized using an anti-His antibody. Alexa 488 and

3. Results 3.1. Murine hepatocyte and human HCC-derived ATGL binds avidly to PEDF ATGL expression levels were evaluated by Western blotting in liver and HCC cells on lysates from wildtype, PEDF null mice and two human HCC cell lines. Murine hepatocytes and human HCC cell lines expressed ATGL (Fig. 1a). Wildtype hepatocytes also expressed low endogenous levels of PEDF (Fig. 1b). The expression

474

C. Chung et al. / Journal of Hepatology 48 (2008) 471–478

of both these proteins is modulated by oleate treatment overnight; ATGL levels increase while PEDF levels decrease (Fig. 1c). To determine the ability of PEDF to bind ATGL, immunoprecipitation of ATGL from hepatocyte lysates was performed in the presence and absence of recombinant PEDF (rPEDF). A signal at 56 kDa corresponding to ATGL was identified in the presence of rPEDF immunoprecipitation from PEDF null hepatocytes (Fig. 1d), whereas, no ATGL signal was detected in the absence of rPEDF. Moreover, binding of ATGL from wildtype hepatocytes with rPEDF was negligible. These data suggest that tight binding between endogenous PEDF and ATGL may mitigate ATGL’s interaction with rPEDF. Alternatively, this may also reflect the relatively higher expression of ATGL within the PEDF null compared to wildtype lysates. Similar to the PEDF null hepatocytes, a human HCC cell line, HepG2, demonstrated immunoprecipitation of ATGL with rPEDF but not with an equivalent amount of BSA (Fig. 1e). Efficiency of ATGL pull-down was assessed by comparing the ATGL signal in equivalent amounts of the post-IP fractions. The presence of rPEDF allowed for efficient extraction of ATGL from hepatic lysates with a threefold decrease in signal in the rPEDF versus BSA supernatants by densitometry (Fig. 1f), indicating avid uptake of ATGL by rPEDF. These results are consistent with another in vitro study that showed high binding affinity between recombinant ATGL and PEDF [13].

3.2. ATGL and rPEDF localize to adiposomes ATGL localizes to lipid droplets or adiposomes in small rings and is an important modulator of adiposome size [24]. To test whether ATGL and PEDF traffic to the adiposome in hepatocytes, co-localization studies were undertaken. Immunofluorescent imaging localized ATGL to the surface of large cytoplasmic vesicles consistent with adiposomes within isolated hepatocytes (Fig. 2a) and to HCC cells (data not shown). Adiposome identity was verified by colocalization with BODIPY 493/503 (Fig. 2a). To further characterize PEDF interactions with ATGL, we treated HepG2 cells with rPEDF and performed immunofluorescent studies using an anti-His antibody. We found rPEDF concentrated in the same discrete circular rings (Fig. 2b) identified as adiposomes by BODIPY staining [24]. Our results suggest that the signaling pathway involving both PEDF and ATGL as regulators of hepatic lipid metabolism likely requires trafficking to adiposomes. 3.3. Hepatic steatosis is evident in the PEDF null animal Examination of PEDF null livers revealed macrovesicular and microvesicular fat accumulation in the hepatocytes as compared to wildtype animals (Fig. 3a and b). The phenotype was apparent as early as one month of age and it progressed with age. When stained with

Fig. 2. ATGL and PEDF localize to adiposomes in liver cells. (a) Staining for ATGL and BODIPY 493/503 in isolated hepatocytes, and merged image, show characteristic localization of ATGL surrounding neutral lipid accumulations. Inset in merged image shows magnified view of ATGL signals surrounding adiposomes. (b) Addition of rPEDF to HepG2 cells and immunofluorescent analysis with an anti-His antibody revealed discrete circular rings (red) consistent with the appearance of adiposomes. Blue-DAPI staining alone is shown as a control.

C. Chung et al. / Journal of Hepatology 48 (2008) 471–478

WT

475

PEDF-/-

ADRP 1.0 3.1 Relative intensity

mg/dl TG/mg protein

1.0

*

0.5

0.0

WT

PEDF null

Fig. 3. Hepatic steatosis is evident in the PEDF null liver. Low power images of H&E staining of liver sections from PEDF null (b) livers show lipid accumulation (arrow) compared to wildtype (a) animals. Oil red O staining of isolated wildtype (c) and PEDF null (d) hepatocytes show greater accumulation of neutral lipids with PEDF deficiency. Immunoblotting for ADRP (e), an adiposome-specific protein that reflects triglyceride content, confirmed this increase in lipid droplets. PEDF null animals (n = 8) challenged with a modestly increased high fat diet (15%) show abundant lipid accumulation, (f) 20·; (g) 40·. Triglyceride content (h) in whole livers was significantly increased (*p < 0.05) in PEDF null versus wildtype animals.

ORO, isolated hepatocytes from PEDF null mice also demonstrated a buildup of neutral lipids compared to wildtype hepatocytes (Fig. 3c and d). Confirming this, ADRP expression, a lipid droplet-specific protein that reflects TG content, was increased threefold in adiposomes isolated from PEDF null livers compared to wildtype (Fig. 3e). Furthermore, challenging PEDF null animals with a diet containing modestly higher fat (15%) resulted in a striking increase in lipid accumulation (Fig. 3f and g). In addition, TG accumulation was pronounced in PEDF null compared to wildtype livers (Fig. 1h), 0.79 ± 0.13 versus 0.37 ± 0.02 mg/dl/mg tissue (n = 3, p < 0.05).

rPEDF treatment negated the effects of rPEDF on TG content (Fig. 4c; p = NS). In contrast, (R)-BEL treatment alone significantly increased TG content in these cells by 4 and 6 h (Fig. 4d; n = 6, p < 0.03). Removal of (R)-BEL restored the response of hepatocytes to rPEDF. PEDF null hepatocytes treated with (R)-BEL and incubated in fresh media without inhibitor for 24 h regained their responsiveness to rPEDF with TG content decreasing to 388 ± 8.2 from 497 ± 23 lg/mg (Fig. 4e; n = 4, p < 0.01). These results indicate that PEDF’s effect on hepatic TG content is ATGL mediated.

3.4. Restoration of PEDF to PEDF null hepatocytes ameliorates hepatic steatosis

4. Discussion

TG measurements in isolated hepatocytes paralleled the findings in total liver. There was significantly more TG content in PEDF null compared to wildtype hepatocytes, 495 ± 33.9 versus 285 ± 13.9 lg/mg of protein (Fig. 4a; n = 4, p < 0.01). Treatment of wildtype hepatocytes with rPEDF had no effect on TG content while PEDF null hepatocytes showed a significant reduction in TG content to 365 ± 22.4 lg/mg of protein (Fig. 4a; n = 4, p < 0.05). In PEDF null hepatocytes switched to serum-free medium, rPEDF treatment resulted in a more significant decrease in TG content (Fig. 4b; n = 6, p = 0.003). Treatment of null hepatocytes with (R)-BEL (2 lM), an ATGL inhibitor, 4–6 h prior to

The functions attributed to PEDF appear divergent and cell type specific. Early studies clearly demonstrated a pro-survival function in neuronal cells [14,17] whereas subsequent reports showed the ability of PEDF to promote apoptosis of activated endothelial cells through the Fas/FasL pathway [25]. Over the past two decades, understanding the signaling pathways and mechanisms responsible for the multiple activities of PEDF has been compromised by the elusive nature of the receptor. Recently, ATGL was implicated as a receptor in retinal epithelial cells [13]. The current study investigated the consequence of PEDF deficiency in the murine liver and demonstrated the importance of the PEDF–ATGL interaction in regulating lipid metabolism. Hepatocyte

476

C. Chung et al. / Journal of Hepatology 48 (2008) 471–478

600 1.0

**

400

200

0

*

0.5

0.0

WT

PEDF null

WT+rPEDF

PEDF null

PEDF null+rPEDF

200

* 1

PEDF null

PEDF null +BEL+rPEDF

*

400

**

200

0

0

0

600

μg TG/mg protein

% PEDF null control

400

PEDF null + rPEDF

**

2

600

μg TG/mg protein

% PEDF null control

μg TG/mg protein

*

PEDF null

4h

6h

+ (R)-BEL

control

+rPEDF

(R)-BEL +rPEDF

(R)-BEL wash-out +rPEDF

Fig. 4. PEDF–ATGL interaction alters TG content. (a) PEDF null hepatocytes had significantly higher TG content than wildtype hepatocytes as normalized per protein content (*p < 0.01). WT hepatocytes treated with rPEDF had no significant change in TG content (p = NS), while addition of rPEDF to null hepatocytes significantly decreased TG content (**p < 0.05). (b) rPEDF treatment of PEDF null hepatocytes under serum-free media conditions resulted in a further decrease of the TG content (*p = 0.0003). (c) Pre-treatment of PEDF null hepatocytes with an ATGL inhibitor, (R)-BEL, prevented the PEDF-mediated decrease in TG content. (d) Treatment of PEDF null hepatocytes with (R)-BEL alone increased TG content (n = 6; *p = 0.028; **p = 0.002). (e) Removal of media containing (R)-BEL and incubation for 24 h restored PEDF null hepatocyte responsiveness to rPEDF as demonstrated by decreased TG content (*p < 0.01).

steatosis was an early finding in the PEDF null liver. Isolated PEDF null hepatocytes revealed an accumulation of neutral lipids and a nearly twofold increase in liver triglyceride content compared to age-matched wildtype animals. These findings are consistent with the relative increase in hepatic triglycerides seen in the ATGL null animals [10]. PEDF null animals, when challenged with a modestly high fat diet, developed diffuse hepatic steatosis pointing to a critical role for PEDF in hepatic lipid homeostasis. PEDF avidly bound ATGL in murine liver and human HCC cells. Moreover, restoration of rPEDF to PEDF null hepatocytes led to a decrease in triglyceride content. A recent clinical study found a significant association between elevated serum PEDF and hyper-triglyceridemia [23]. Although this study did not find a causal relationship, the authors, however, noted that elevated serum PEDF likely represents a counterregulatory mechanism for handling elevated triglycerides [23]. The identification of ATGL as a putative receptor broadens the spectrum of potential PEDF targets, having implications in areas that bridge lipid metabolism and angiogenesis [26]. ATGL catalyzes the initial step in triglyceride hydrolysis and early studies concentrated on its activities in adipose tissue [4,8]. However, the loss of ATGL in a murine model led to accumulation of triglycerides in many organ systems supporting a broader

action for this lipase [10]. Studies investigating the subcellular localization of ATGL discovered an intimate relationship with lipid droplets or adiposomes [24]. Adiposomes were initially regarded as inert storage depots for neutral lipids although emerging data challenge this dogma. Proteomic analysis of lipid droplets isolated from Drosophila embryos disclosed an array of proteins including structural proteins and histones [27]. These findings led this group to conclude that lipid droplets can serve as transient storage depots that have the capacity to engage in membrane transport pathways. ATGL has been shown to localize to the surface of adiposomes, thereby giving the morphological appearance under microscopy as discrete rings. This feature was highlighted when BODIPY stains outlined the central core of the droplet [24]. Here, we provide evidence in hepatocytes to suggest that ATGL has a partner at the surface of the adiposome. PEDF was found to bind ATGL, and, using immunofluorescent studies, PEDF and ATGL co-localized in a pattern consistent with adiposomes. Lipid droplet size and density are dynamic in many cell types and diseases. Both pre-malignant and malignant states in the liver are associated with an increase in lipid droplets. Viral hepatitis, including hepatitis C infection, is commonly associated with lipid accumulation and steatosis. This infection is considered a signifi-

C. Chung et al. / Journal of Hepatology 48 (2008) 471–478

cant risk factor for the development of HCC [28,29]. The presence of lipid droplets within HCC is a well-recognized feature with a recent report indicating marked upregulation of lipogenic enzymes in human HCC specimens [30]. Although the role of ATGL has yet to be investigated in viral hepatitis and HCC, there is some evidence to suggest that mutations in ATGL alter liver function. A mutation of the ATGL adaptor protein, Comparative Gene I dentification-58, leads to a rare autosomal recessive disease called Chanarin–Dorfman syndrome characterized by ichthyosis, mental retardation as well as hepatic steatosis [31]. More recently, mutations in the ATGL gene, itself, were reported in three patients with marked accumulation of neutral lipids [32]. Although liver biopsies were not performed, these patients had varying degrees of hepatomegaly or elevated liver enzymes. Whether alterations in ATGL function occur in more common disease states where hepatocytes accumulate lipids such as NAFLD, alcohol abuse, or viral hepatitis remains to be explored. The newly defined interaction between PEDF and ATGL in this study identifies a regulatory mechanism for modulating triglycerides in the liver and suggests that trafficking to the adiposome could be an essential step in hepatic lipid metabolism. Acknowledgements We thank Dr. Fred Gorelick (Yale University) for thoughtful comments, Kathy Harry (Yale University) for expert technical assistance, and Dr. Rohit Garg (Yale University) for image assistance. This work was funded by an American Liver Foundation grant (C.C.), NIH Liver Center Core P30-DK34989 (C.C.); NIH/NCI Grant R01-CA64239 (to S.E.C.). References [1] Angulo P. Nonalcoholic fatty liver disease. N Engl J Med 2002;346:1221–1231. [2] Marrero JA, Fontana RJ, Su GL, Conjeevaram HS, Emick DM, Lok AS. NAFLD may be a common underlying liver disease in patients with hepatocellular carcinoma in the United States. Hepatology 2002;36:1349–1354. [3] Samuel VT, Liu ZX, Qu X, Elder BD, Bilz S, Befroy D, et al. Mechanism of hepatic insulin resistance in non-alcoholic fatty liver disease. J Biol Chem 2004;279:32345–32353. [4] Zimmermann R, Strauss JG, Haemmerle G, Schoiswohl G, Birner-Gruenberger R, Riederer M, et al. Fat mobilization in adipose tissue is promoted by adipose triglyceride lipase. Science 2004;306:1383–1386. [5] Osuga J, Ishibashi S, Oka T, Yagyu H, Tozawa R, Fujimoto A, et al. Targeted disruption of hormone-sensitive lipase results in male sterility and adipocyte hypertrophy, but not in obesity. Proc Natl Acad Sci USA 2000;97:787–792. [6] Haemmerle G, Zimmermann R, Hayn M, Theussl C, Waeg G, Wagner E, et al. Hormone-sensitive lipase deficiency in mice causes diglyceride accumulation in adipose tissue, muscle, and testis. J Biol Chem 2002;277:4806–4815.

477

[7] Haemmerle G, Zimmermann R, Zechner R. Letting lipids go: hormone-sensitive lipase. Curr Opin Lipidol 2003;14:289–297. [8] Villena JA, Roy S, Sarkadi-Nagy E, Kim KH, Sul HS. Desnutrin, an adipocyte gene encoding a novel patatin domain-containing protein, is induced by fasting and glucocorticoids: ectopic expression of desnutrin increases triglyceride hydrolysis. J Biol Chem 2004;279:47066–47075. [9] Jenkins CM, Mancuso DJ, Yan W, Sims HF, Gibson B, Gross RW. Identification, cloning, expression, and purification of three novel human calcium-independent phospholipase A2 family members possessing triacylglycerol lipase and acylglycerol transacylase activities. J Biol Chem 2004;279:48968–48975. [10] Haemmerle G, Lass A, Zimmermann R, Gorkiewicz G, Meyer C, Rozman J, et al. Defective lipolysis and altered energy metabolism in mice lacking adipose triglyceride lipase. Science 2006;312:734–737. [11] Ryden M, Jocken J, van Harmelen V, Dicker A, Hoffstedt J, Wiren M, et al. Comparative studies of the role of hormone sensitive lipase and adipose triglyceride lipase in human fat cell lipolysis. Am J Physiol Endocrinol Metab 2007;292:E1847–E1855. [12] Schoenborn V, Heid IM, Vollmert C, Lingenhel A, Adams TD, Hopkins PN, et al. The ATGL gene is associated with free fatty acids, triglycerides, and type 2 diabetes. Diabetes 2006;55: 1270–1275. [13] Notari L, Baladron V, Aroca-Aguilar JD, Balko N, Heredia R, Meyer C, et al. Identification of a lipase-linked cell membrane receptor for pigment epithelium-derived factor. J Biol Chem 2006;281:38022–38037. [14] Bouck N. PEDF: anti-angiogenic guardian of ocular function. Trends Mol Med 2002;8:330–334. [15] Chader GJ. PEDF: raising both hopes and questions in controlling angiogenesis. Proc Natl Acad Sci USA 2001;98:2122–2124. [16] Dawson DW, Volpert OV, Gillis P, Crawford SE, Xu H, Benedict W, et al. Pigment epithelium-derived factor: a potent inhibitor of angiogenesis. Science 1999;285:245–248. [17] Tombran-Tink J, Chader GG, Johnson LV. PEDF: a pigment epithelium-derived factor with potent neuronal differentiative activity. Exp Eye Res 1991;53:411–414. [18] Tombran-Tink J, Mazuruk K, Rodriguez IR, Chung D, Linker T, Englander E, et al. Organization, evolutionary conservation, expression and unusual Alu density of the human gene for pigment epithelium-derived factor, a unique neurotrophic serpin. Mol Vis 1996;2:11. [19] Matsumoto K, Ishikawa H, Nishimura D, Hamasaki K, Nakao K, Eguchi K. Antiangiogenic property of pigment epitheliumderived factor in hepatocellular carcinoma. Hepatology 2004;40:252–259. [20] Doll JA, Stellmach VM, Bouck NP, Bergh AR, Lee C, Abramson LP, et al. Pigment epithelium-derived factor regulates the vasculature and mass of the prostate and pancreas. Nat Med 2003;9:774–780. [21] Hirata K, Pusl T, O’Neill AF, Dranoff JA, Nathanson MH. The type II inositol 1,4,5-trisphosphate receptor can trigger Ca2+ waves in rat hepatocytes. Gastroenterology 2002;122:1088–1100. [22] Brasaemle DL, Dolios G, Shapiro L, Wang R. Proteomic analysis of proteins associated with lipid droplets of basal and lipolytically stimulated 3T3-L1 adipocytes. J Biol Chem 2004;279: 46835–46842. [23] Yamagishi S, Adachi H, Abe A, Yashiro T, Enomoto M, Furuki K, et al. Elevated serum levels of pigment epithelium-derived factor in the metabolic syndrome. J Clin Endocrinol Metab 2006;91:2447–2450. [24] Smirnova E, Goldberg EB, Makarova KS, Lin L, Brown WJ, Jackson CL. ATGL has a key role in lipid droplet/adiposome degradation in mammalian cells. EMBO Rep 2006;7:106–113. [25] Volpert OV, Zaichuk T, Zhou W, Reiher F, Ferguson TA, Stuart PM, et al. Inducer-stimulated Fas targets activated endothelium

478

[26] [27]

[28]

[29]

C. Chung et al. / Journal of Hepatology 48 (2008) 471–478 for destruction by anti-angiogenic thrombospondin-1 and pigment epithelium-derived factor. Nat Med 2002;8:349–357. Cao Y. Angiogenesis modulates adipogenesis and obesity. J Clin Invest 2007;117:2362–2368. Cermelli S, Guo Y, Gross SP, Welte MA. The lipid-droplet proteome reveals that droplets are a protein-storage depot. Curr Biol 2006;16:1783–1795. Ohata K, Hamasaki K, Toriyama K, Matsumoto K, Saeki A, Yanagi K, et al. Hepatic steatosis is a risk factor for hepatocellular carcinoma in patients with chronic hepatitis C virus infection. Cancer 2003;97:3036–3043. Pekow JR, Bhan AK, Zheng H, Chung RT. Hepatic steatosis is associated with increased frequency of hepatocellular carcinoma

in patients with hepatitis C-related cirrhosis. Cancer 2007;109:2490–2496. [30] Yahagi N, Shimano H, Hasegawa K, Ohashi K, Matsuzaka T, Najima Y, et al. Co-ordinate activation of lipogenic enzymes in hepatocellular carcinoma. Eur J Cancer 2005;41:1316–1322. [31] Lass A, Zimmermann R, Haemmerle G, Riederer M, Schoiswohl G, Schweiger M, et al. Adipose triglyceride lipase-mediated lipolysis of cellular fat stores is activated by CGI-58 and defective in Chanarin–Dorfman Syndrome. Cell Metab 2006;3:309–319. [32] Fischer J, Lefevre C, Morava E, Mussini JM, Laforet P, NegreSalvayre A, et al. The gene encoding adipose triglyceride lipase (PNPLA2) is mutated in neutral lipid storage disease with myopathy. Nat Genet 2007;39:28–30.