Patatin-like phospholipase domain–containing protein 3 promotes transfer of essential fatty acids from triglycerides to phospholipids in hepatic lipid droplets - PubMed (original) (raw)

Patatin-like phospholipase domain–containing protein 3 promotes transfer of essential fatty acids from triglycerides to phospholipids in hepatic lipid droplets

Matthew A Mitsche et al. J Biol Chem. 2018.

Erratum in

Abstract

Fatty liver disease (FLD) is a burgeoning health problem. A missense variant (I148M) in patatin-like phospholipase domain-containing protein 3 (PNPLA3) confers susceptibility to FLD, although the mechanism is not known. To glean first insights into the physiological function of PNPLA3, we performed detailed lipidomic profiling of liver lysates and lipid droplets (LDs) from WT and _Pnpla3_-/- (KO) mice and from knock-in (ki) mice expressing either the 148M variant (IM-ki mice) or a variant (S47A) that renders the protein catalytically inactive (SA-ki mice). The four strains differed in composition of very-long-chain polyunsaturated fatty acids (vLCPUFA) in hepatic LDs. In the LDs of IM-ki mice, vLCPUFAs were depleted from triglycerides and enriched in phospholipids. Conversely, vLCPUFAs were enriched in triglycerides and depleted from phospholipids in SA-ki and Pnpla3_-/- mice. Release of vLCPUFAs from hepatic LDs incubated e_x vivo was increased in droplets from IM-ki mice and decreased from droplets isolated from _Pnpla3_-/- and SA-ki mice relative to those of WT mice. Thus, the physiological role of PNPLA3 appears to be to remodel triglycerides and phospholipids in LDs, perhaps to accommodate changes in LD size in response to feeding. Because SA-ki and IM-ki both cause FLD and yet have opposite effects on the lipidomic profile of LDs, we conclude that the FLD associated with genetic variation in PNPLA3 is not related to the enzyme's role in remodeling LD lipids.

Keywords: hepatic steatosis; lipase; phosphatidylglycerol; phospholipid; polyunsaturated fatty acid (PUFA); triglyceride.

© 2018 Mitsche et al.

PubMed Disclaimer

Conflict of interest statement

The authors declare that they have no conflicts of interest with the contents of this article

Figures

Figure 1.

Figure 1.

Hepatic lipid composition of genetically modified PNPLA3 mice. A, liver TG content of WT (white), PNPLA3 I148M knock-in (IM-ki; red), PNPLA3 S47A knock-in (SA-ki; blue), and PNPLA3 knockout (KO; black) male mice (n = 4–6/group, 12 weeks old). In this and all other experiments, mice were entrained for 3 days by fasting from 6 p.m. to 8 a.m. and fed a HSD from 8 a.m. to 6 p.m. and killed 4 h into the last refeeding cycle. TGs were measured using enzymatic assays. B, total hepatic FA composition of genetically modified male mice fed a HSD for 4 weeks (n = 4–6/group, 12 weeks old). Liver lipids were hydrolyzed and derivatized with trimethylsilane and then measured by GC-flame ionization detection. The data are expressed per mg of liver. The experiment was repeated twice, and the results were similar. Error bars, S.D. **, p < 0.01; ***, p < 0.001.

Figure 2.

Figure 2.

Hepatic TG composition in genetically modified PNPLA3 mice. A, the number of carbons (left) and number of double bonds (right) in hepatic TGs of IM-ki (red), SA-ki (blue), and KO (black) mice relative to their WT counterparts. Male mice were fed a HSD for 4 weeks (n = 6–8/group, 14 weeks old) and then sacrificed after 4 h of refeeding. Hepatic TGs were analyzed by direct infusion lipidomics. TGs were grouped based on number of FA carbons (left) or double bonds (right), as determined by the MS/MS signal from untargeted lipidomics analysis. Values were normalized to the total lipid signal and expressed relative to the corresponding value in WT mice. The data are plotted on a log-normal scale. B, selected TGs were measured by LC-MS/MS analysis. Shown are representative examples of abundant TGs (top) and vLCPUFA-containing TGs (bottom) in livers of PNPLA3 genetically modified mice. Mass pairs for identification of abundant TG were 876.7 and 577.6, 848.7 and 577.6, and 874.7 and 577.6 Da for the TGs shown left to right, respectively, and 898.8 and 577.7, 924.8 and 577.7, and 922.8 and 577.7 for the vLCPUFA-containing TGs shown left to right, respectively. Underlined FAs indicate the neutral loss of the measured species. Values were normalized to the sum of all measured TG signals. Liver TG analysis was repeated twice with similar results. Error bars, S.D. *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Figure 3.

Figure 3.

FA composition of hepatic CEs and REs. A, levels of hepatic CEs (all CEs) and most abundant polyunsaturated CEs in PNPLA3(WT) (white), IM-ki (red), SA-ki (blue), and KO mice (black). Male mice (n = 6–8/group, 14 weeks old) were fed a HSD for 4 weeks and killed 4 h after refeeding. CEs were analyzed by direct infusion lipidomics and normalized to total lipid signal. Liver CE analysis was repeated twice with similar results. B, total REs (all REs) and the most abundant REs were measured in male mice fed a HSD for 4 weeks (n = 4–6/group, 12 weeks old). REs were measured using LC-MS/MS with RE17:0 as an internal standard. Total REs were determined by summing the intensity of all mass features containing an MS2 fragment of 269.2 Da with an MS1 mass corresponding to an RE. Values were normalized to the internal standard of RE17:0. The experiment was repeated once, and the results were similar. Error bars, S.D.

Figure 4.

Figure 4.

PL composition of liver lysates. Male mice (n = 6–8/group, 14 weeks old) were fed a HSD for 4 weeks and killed after 4 h of refeeding. Total hepatic PL composition was determined by direct infusion lipidomics in WT (white), IM-ki (red), SA-ki (blue), and PNPLA3 KO mice (black). PLs were identified by their negative ion. The reported vLCPUFA-containing PLs were identified based on the MS2 value. Similar results were observed in positive mode. All values were normalized to the sum of the positive and negative signal from direct infusion lipidomics. The experiment was repeated twice with similar results. Error bars, S.D.

Figure 5.

Figure 5.

Lipidomic comparisons of lipid droplets. Top, lipid composition of LDs from a Bligh–Dyer extract compared with the middle and top phase of a three-phase extraction based on direct infusion lipidomics analysis. Lipids were identified based on their tandem mass pairs. Lipid features that were not identified were classified as unknown lipids. Reported values are means of livers (n = 4). -Fold change of KO mice relative to WT mice LDs were plotted against the -fold change in SA-ki to WT mice. Hepatic LDs were prepared from male mice on a HSD for 4 weeks (n = 5–7/group, 14 weeks old) after 4 h of refeeding. Each point represents the -fold change in KO mice on the x axis and SA-KI mice on the y axis. Values were normalized to the sum of all lipid signals. The relationship between IM-ki and SA-ki or KO mice relative to WT is also shown. Black points had a p value < 0.05 in the IM-ki mice. The analysis was repeated once with similar results.

Figure 6.

Figure 6.

Composition of PLs of hepatic LDs by untargeted lipidomics. Male mice (n = 5–7/group, 14 weeks old) of the indicated genotype were fed a HSD for 4 weeks and refed for 4 h before tissues were collected. Four PL classes (PCs, PEs, PSs, and PGs) were analyzed by direct infusion lipidomics (see “Experimental procedures”). The total content of each PL class was calculated by summing all unambiguous signals obtained in negative mode. Individual PLs were identified by their negative ion with the MS2 of the vLCPUFA. Similar results were observed in positive mode. All values were normalized to the sum of the positive and negative mode signal, where the majority of the signal was TGs. Analysis was repeated once with similar results. Error bars, S.D. *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Figure 7.

Figure 7.

PL composition of hepatic LDs by targeted LC-MS/MS. PL composition of hepatic LDs from WT (white), IM-ki (red), SA-ki (blue), and PNPLA3 KO mice (black) was analyzed by targeted LC-MS/MS. Male mice (n = 5–7/group, 14 weeks old) were fed a HSD for 4 weeks and refed for 4 h before collecting tissues. PCs (A), PGs, and BMPs (B) were targeted for analysis using LC-MS/MS. Individual PGs were identified by their negative ion with the MS2 of the vLCPUFA. Values were normalized to an abundant PL in the same class, as indicated by the denominator on the y axis. Analysis was repeated once, and the results were similar. Error bars, S.D.; *, p < 0.05; **, p < 0.01.

Figure 8.

Figure 8.

Synthesis of TG and PL in in HuH7 cells overexpressing PNPLA3. HuH7 cells were infected with a GFP adenoviruses (GFP, dotted line), PNPLA3(WT) (green), or PNPLA3(148M) (red). After 2 days, palmitate-_d_31 and arachidonate-_d_8 were conjugated with BSA and added to the medium at a final concentration of 1 μ

m

each. Cells were then harvested at the indicated time points (n = 3 dishes/time point). Expression of PNPLA3 constructs was analyzed by immunoblotting using a V5 antibody. Lipids were extracted, and the incorporation of the isotope-labeled FAs into TGs and PCs was measured by direct infusion lipidomics. PCs were identified by the 184-Da MS2 fragment in positive mode. TGs were identified based on their FA neutral loss. Values were normalized to the PC and TG standards in the SPLASH standard mixture. The experiment was repeated twice with similar results. Error bars, S.D. *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Figure 9.

Figure 9.

Release of FAs from hepatic LDs incubated ex vivo. A, effect of ATGListatin on release of oleate (18:1) and arachidonate (20:4) from hepatic LDs. LDs were isolated from WT male mice (n = 6/group, 15 weeks old) fed a HSD for 4 weeks. After a 12-h overnight fast, mice were refed for 4 h and then killed. Livers were harvested, and hepatic LDs were isolated and diluted to a total volume of 5 ml. Aliquots (50 μl) of LDs were transferred to a glass 96-well plate containing 250 μl of isolated liver cytosol, ATP-Mg (1 m

m

), and 0.5% albumin with or without 1 μ

m

ATGListatin solution and warmed to 37 °C. For each liver, LDs were added to 24 wells with or without ATGListatin. The plate was then lightly vortexed and incubated at 37°C. At the indicated times, LDs were disrupted by the addition of 300 μl of methanol and stored at 4°C until they were extracted and analyzed by direct infusion TOF-MS. Values were normalized to the amount of FA in the medium at the 0-h time point and expressed as -fold change over baseline. The experiment was repeated with similar results. B, release of FAs by hepatic LDs from WT (white), IM-ki (red), SA-ki (blue), and PNPLA3 KO (black) male mice (n = 6/group, 15 weeks old). Mice were fed a HSD for 4 weeks, fasted overnight, and killed after 4 h of refeeding. Hepatic LDs were isolated and incubated ex vivo as described in A above, except that no ATGListatin was added. Values were normalized to the amount of FA in the medium at the 0-h time point and expressed as -fold change over baseline. The experiment was repeated once, and the results were similar. Each point represents the average of three mice, each of which was assayed in triplicate. Error bars, S.D. *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Figure 10.

Figure 10.

Models of PNPLA3 function in FA remodeling. PNPLA3 may act as a transacylase, transferring vLCPUFA directly from TG to lyso-PL. Alternatively, PNPLA3 may function as a vLCPUFA-specific TG hydrolase in a remodeling pathway for LD PL analogous to the Lands cycle for membrane PL. PNPLA3(148M) expression causes an increase in TG and PL remodeling, whereas PNPLA3(47A) has an opposite effect because it has no enzyme activity.

References

    1. Romeo S., Kozlitina J., Xing C., Pertsemlidis A., Cox D., Pennacchio L. A., Boerwinkle E., Cohen J. C., and Hobbs H. H. (2008) Genetic variation in PNPLA3 confers susceptibility to nonalcoholic fatty liver disease. Nat. Genet. 40, 1461–1465 10.1038/ng.257 - DOI - PMC - PubMed
    1. Tian C., Stokowski R. P., Kershenobich D., Ballinger D. G., and Hinds D. A. (2010) Variant in PNPLA3 is associated with alcoholic liver disease. Nat. Genet. 42, 21–23 10.1038/ng.488 - DOI - PubMed
    1. Sookoian S., Castaño G. O., Burgueño A. L., Gianotti T. F., Rosselli M. S., and Pirola C. J. (2009) A nonsynonymous gene variant in the adiponutrin gene is associated with nonalcoholic fatty liver disease severity. J. Lipid Res. 50, 2111–2116 10.1194/jlr.P900013-JLR200 - DOI - PMC - PubMed
    1. Rotman Y., Koh C., Zmuda J. M., Kleiner D. E., Liang T. J., and NASH CRN (2010) The association of genetic variability in patatin-like phospholipase domain-containing protein 3 (PNPLA3) with histological severity of nonalcoholic fatty liver disease. Hepatology 52, 894–903 10.1002/hep.23759 - DOI - PMC - PubMed
    1. Falleti E., Fabris C., Cmet S., Cussigh A., Bitetto D., Fontanini E., Fornasiere E., Bignulin S., Fumolo E., Bignulin E., Pirisi M., and Toniutto P. (2011) PNPLA3 rs738409C/G polymorphism in cirrhosis: relationship with the aetiology of liver disease and hepatocellular carcinoma occurrence. Liver Int. 31, 1137–1143 10.1111/j.1478-3231.2011.02534.x - DOI - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources