Characterization of the Interaction of Neuropathy Target Esterase with the Endoplasmic Reticulum and Lipid Droplets

Abstract

: Neuropathy target esterase (NTE) is an endoplasmic reticulum (ER)-localized phospholipase that deacylates phosphatidylcholine (PC) and lysophosphatidylcholine (LPC). Loss-of-function mutations in the human NTE gene have been associated with a spectrum of neurodegenerative disorders such as hereditary spastic paraplegia, ataxia and chorioretinal dystrophy. Despite this, little is known about structure-function relationships between NTE protein domains, enzymatic activity and the interaction with cellular organelles. In the current study we show that the C-terminal region of NTE forms a catalytically active domain that exhibits high affinity for lipid droplets (LDs), cellular storage organelles for triacylglycerol (TAG), which have been recently implicated in the progression of neurodegenerative diseases. Ectopic expression of the C domain in cultured cells decreases cellular PC, elevates TAG and induces LD clustering. LD interactions of NTE are inhibited by default by a non-enzymatic regulatory (R) region with three putative nucleotide monophosphate binding sites. Together with a N-terminal TMD the R region promotes proper distribution of the catalytic C-terminal region to the ER network. Taken together, our data indicate that NTE may exhibit dynamic interactions with the ER and LDs depending on the interplay of its functional regions. Mutations that disrupt this interplay may contribute to NTE-associated disorders by affecting NTE positioning.

Keywords: PNPLA6; endoplasmic reticulum; lipid droplet; lysophospholipase; neuropathy target esterase; triacylglycerol.

Figure 1

Functional contribution of the N-terminal transmembrane (TM) domain and regulatory (R)-region to ER targeting of neuropathy target esterase (NTE). (A) Domain architecture of NTE and the variants used in this experiment. (B) Subcellular distribution of NTE-GFP, NTETM-GFP and NTER-GFP in COS-7 cells. COS-7 cells were co-transfected with NTE-GFP, NTETM-GFP, NTER-GFP and the ER marker ER-DsRed as indicated on each panel for 48 h and then visualized live by confocal microscopy. Scale bar, 10 μm. Figures are representative of three separate experiments. (C) Subcellular distribution of NTE-GFP, NTETM-GFP and NTER-GFP in transfected mammalian cells. After transfection for 48 h, cells were harvested, homogenized and fractionated into membrane (mem) and cytosolic (cyto) fractions. The cytosolic and membrane fractions were further subjected to Western blotting analysis with an anti-GFP antibody. Migration of molecular mass standard proteins is indicated left of the figure.

Figure 2

NTE esterase domain (NEST) localizes to LDs. (A) Domain architecture of a NEST and a S966A NEST mutant (mtNEST)construct used in this experiment. (B) NEST-GFP partially localizes to LDs in COS-7 cells without OA-loading. COS-7 were co-transfected with NEST-GFP and the ER marker, ER-DsRed or transfected alone with NEST-GFP or mtNEST-GFP as indicated. Within 48 h, living cells were imaged by confocal microscopy. To detect the distribution of NEST-GFP, mtNEST-GFP and LDs, COS-7 cells were fixed and incubated with LipidTOX™ Deep Red to stain LDs and then visualized by confocal fluorescence microscopy. (C) Increased fatty acid flux stimulates NEST binding to LDs. COS-7cells expressing NEST-GFP or mtNEST-GFP were treated with OA overnight. LDs were labeled by LipidTOX™ Deep Red or by co-expression of PLIN2-mCherry. Colocalization of NEST-GFP, mtNEST-GFP and LDs was visualized by confocal microscopy. Scale bar = 10 μm. Figures are representative of three separate experiments.

Figure 3

Functional contribution of NTE protein domains to LD targeting. (A) Domain architecture of NTE and its variants used in this experiment. (B) Subcellular distribution of NTE variants and LDs. COS-7 cells were transfected with NTE-GFP or truncated NTE-GFP mutants, incubated with FAs to induce LD formation and analyzed by confocal fluorescence microscopy. LDs were visualized using the neutral lipid stain HSC LipidTOX™ Deep Red with PLIN2-mCherry. Bar size: 10 µm. (C) Subcellular distribution of NEST-GFP, ∆R-NTE-GFP and ER. NEST-GFP or ∆R-NTE-GFP was expressed in COS-7 cells loaded with OA or not as indicated. The ER was marked by co-expression of ER-DsRed. Images were acquired by confocal fluorescence microscopy. Scale bar, 10 µm. (D) Subcellular distribution of ∆R-NTE-GFP in COS-7 cells in the absence or presence of OA. After transfection for 24 h, cells were incubated in the absence or presence of OA for 24 h before being subjected to subcellular fractionation. The soluble (S) and particulate (P) fractions were subjected to immunoblotting analysis using antibodies against GFP and Calnexin.

Figure 4

NEST but not NTE co-localizes with the pre-LD marker HPos. COS-7 cells co-expressing GFP, NTE-GFP or NEST-GFP and HPos-mCherry were starved for 24 h. NEST-GFP is present in HPos positive puncta upon starvation as indicated by white arrowheads. NTE does not co-localize with pre-LDs. Scale bar = 10 µm. Figures are representative of three separate experiments.

Figure 5

Neither the patatin domain nor putative TM regions in NEST localize to LDs. (A) Schematic overview of the patatin domain and putative TM domains in NEST tagged with GFP, NTEPP-GFP and NESTTM-GFP. (B) Subcellular localization of NEST truncation variants and LDs. COS-7 cells expressing NTEP-GFP or NESTTM-GFP were loaded with OA for 16 h and then fixed and incubated with LipidTOX Deep Red to stain LDs. Colocalization of proteins and the LDs was visualized by confocal laser scanning microscopy. Scale bar = 10 µm. Figures are representative of three separate experiments.

Figure 6

The lysophospholipase activities of neuropathy target esterase (NTE) and its truncated mutants. Lysates of COS-7 cells overexpressing GFP, NEST-GFP, ΔR-NTE-GFP or NTE-GFP in the absence or presence of OA were incubated with LPC as lipid substrate and hydrolytic activities were determined by measuring the released NEFA. GFP served as negative control. Assays were linear with time and protein amount. The Western blot shows comparable protein expression of GFP, NEST-GFP, ΔR-NTE-GFP and NTE-GFP. Normalization was performed by densitometric analyses of the respective expression level detected with an antibody towards GFP. “+” and “-” indicated OA loading or not respectively. Data are presented as means ± SD and are representative of at least three experiments. Asterisk denotes p values: * p < 0.05, n = 3.

Figure 7

LD morphology and lipid levels upon knockdown of NTE and overexpression of NEST in human neuroblastoma cells. SH-SY5Y control cells, NTE-knockdown cells (SH/NTE-shRNA) and NEST-expressing cells (SH/NEST) were incubated with OA overnight. (A) LDs were labeled by LipidTOX Deep Red and visualized by confocal laser scanning microscopy. Scale bar = 10 µm. Figures are representative of three separate experiments. (B) and (C) Triacylglycerol (TG) and PC levels were measured in control cells (SY5Y), NTE-knockdown cells (NTE shRNA) and NEST-overexpressing cells (NEST) incubated in the absence (−OA) or presence (+OA) of exogenous OA. Data are presented as means ± SD and are representative of at least three experiments. Asterisk indicates p values: * p < 0.05, n = 3.

Figure 8

Summary of the subcellular localization of NTE constructs. Localization to the ER or LDs is described as positive (+), whereas failure to localization as negative (−). “+/−” indicates partial localization to the ER. For comprehensiveness, the results of a previous study were included in this summary [5].