Characterization of desnutrin functional domains: critical residues for triacylglycerol hydrolysis in cultured cells

Abstract

Murine desnutrin/human ATGL is a triacylglycerol (TAG) hydrolase with a predicted catalytic dyad within an alpha-beta hydrolase fold in the N-terminal region. In humans, mutations resulting in C-terminal truncation cause neutral lipid storage disease with myopathy. To identify critical functional domains, we measured TAG breakdown in cultured cells by mutated or truncated desnutrin. In vitro, C-terminally truncated desnutrin displayed an even higher apparent V(max) than the full-length form without changes in K(m), which may be explained by our finding of an interaction between the C- and N-terminal domains. In live cells, however, C-terminally truncated adenoviral desnutrin had lower TAG hydrolase activity. We investigated a role for the phosphorylation of C-terminal S406 and S430 residues but found that these were not necessary for TAG breakdown or lipid droplet localization in cells. The predicted N-terminal active sites, S47 and D166, were both critical for TAG hydrolysis in live cells and in vitro. We also identified two overlapping N-terminal motifs that predict lipid substrate binding domains, a glycine-rich motif (underlined) and an amphipathic alpha-helix (bold) within amino acid residues 10-24 (ISFAGCGFLGVYHIG). G14, F17, L18, and V20, but not G16 and G19, were important for TAG hydrolysis, suggesting a potential role for the amphipathic alpha-helix in TAG binding. This study identifies for the first time critical sites in the N-terminal region of desnutrin and reveals the requirement of the C-terminal region for TAG hydrolysis in cultured cells.

Fig. 1.

Desnutrin mutants and truncations. A: Kyte-Doolittle hydropathy plot (window size = 19) with red line designating the cutoff for potential transmembrane domains. B: Schematic illustrating desnutrin mutants and truncations, shown in alignment with hydropathy plot. Red = patatin-like domain/α-β hydrolase fold. Yellow = putative hydrophobic span. Green = GFP tag. Blue = unstructured region.

Fig. 2.

Desnutrin C-terminal deletion mutants have increased catalytic activity in vitro. A: Immunoblot of full-length and truncated desnutrin expressed in COS-7 cells. B: Substrate velocity plot showing in vitro [³H]triolein hydrolase activity of COS-7 cell lysates overexpressing desnutrin-GFP (DES-GFP) or C-terminal deletion mutants of desnutrin-GFP truncated at W367 (W367×-GFP) or Q289 (Q289×-GFP). Lysates from cells expressing GFP alone were taken as a baseline measure at each substrate concentration and subtracted from test values to derive net TAG hydrolase activity that was expressed per mg of lysate protein (n = 3). C–E: Double reciprocal plots for desnutrin-GFP and C-terminal truncations of desnutrin-GFP. F: K_m and apparent V_max derived from double reciprocal plots. G: Immunoblot showing the interaction of N-terminally and C-terminally truncated desnutrin. Lysates from COS-7 cells cotransfected with plasmids expressing the HA-tagged N-terminal region of desnutrin (amino acids 1–246) and the FLAG-tagged C-terminal region of desnutrin (amino acids 237–486) were immunoprecipitated with IgG (negative control) or with anti-FLAG (top) or anti-HA (bottom) antibodies, subjected to SDS-PAGE, and probed to determine the presence of the interacting partner. Upper and lower nonspecific bands are IgG heavy and light chains, respectively.

Fig. 3.

Desnutrin C-terminal deletion mutants lack TAG hydrolase activity in cells. Immunoblots of full-length GFP-tagged desnutrin (DES-GFP) or GFP-tagged truncated desnutrin (W367×-GFP and Q289×-GFP) in COS-7 cells (A) or 3T3-L1CARΔ (C). TAG hydrolysis measured by ¹⁴C-TAG content remaining in COS-7 cells (B) or 3T3-L1CARΔ adipocytes (D) pulse labeled with [U-¹⁴C]palmitate, washed, and then infected with GFP, desnutrin-GFP, or GFP-tagged desnutrin C-terminal deletion mutants prior to chase with unlabeled medium for 60 h. Values are means ± SEM from two to three separate experiments. *P < 0.05,**P < 0.01, and ***P < 0.001.

Fig. 4.

Phosphorylation of desnutrin at S406 and S430 is not essential for activity or localization. Confocal micrographs showing diffuse localization of GFP (control) or prominent lipid droplet localization of GFP-tagged desnutrin, S406A, and S430A desnutrin mutants in COS-7 cells (A) or 3T3-L1CARΔ adipocytes (B). TAG hydrolysis by desnutrin-GFP or serine to alanine mutants of desnutrin-GFP measured in live 293FT cells (C). Immunoblots show expression of desnutrin and mutant desnutrin and GAPDH (control). Values are means ± SEM from a representative experiment. **P < 0.01 versus GFP control.

Fig. 5.

The proximal N-terminal region contains the S47-D166 catalytic dyad as well as a putative lipid substrate-binding domain. A: Confocal micrograph showing lipid droplet localization of GFP-tagged desnutrin lacking the N-terminal domain (M237-C486-GFP). B: TAG hydrolysis by S47A and D166A desnutrin mutants is decreased to control levels in live 293FT cells. C: Sequence alignment of the proximal N-terminal regions of murine and human desnutrin. Underlined glycines (G) comprise the glycine-rich motif (XhXhGXGXXG), while letters in bold or italics are part of a predicted amphipathic α-helix (FLXVXXXn), where X is any amino acid except proline, h is a hydrophobic residue, and n is any nonpolar amino acid. D: TAG hydrolysis by G14A, G16A, and G19A mutants of desnutrin in live 293FT cells. E: TAG hydrolysis by F17A, L18A, and V20A desnutrin mutants in live 293FT cells. Values are expressed as a percentage of pcDNA3.1 control vector. F: In vitro TAG hydrolase activity of wild-type and mutant desnutrin assayed in COS-7 cell lysates using 100 μM ³H-triolein. Values are means ± SEM from two to four separate experiments. Columns with different superscripts are significantly different, ^abcP < 0.01, *P < 0.05, **P < 0.01, and ***P < 0.001.