{"paper_id":"09fe3a73-911e-4a1e-aa1f-7266adef77f0","body_text":"Acute deletion of PLIN5 in brown adipocytes causes mitochondrial dysfunction \nand cold intolerance  \n \nVioleta I. Gallardo Montejano1, Chaofeng Yang1, Hannah Hurtado1, and Perry E. \nBickel1* \n1. Division of Endocrinology, Department of Internal Medicine, The University of Texas \nSouthwestern Medical Center, 5323 Harry Hines Blvd., Dallas, Texas, 75390-8857, \nUSA \n* Correspondence:  perry.bickel@utsouthwestern.edu \nDivision of Endocrinology, Department of Internal Medicine, The \nUniversity of Texas Southwestern Medical Center, 5323 Harry \nHines Blvd., Dallas, Texas, 75390-8857, USA \n214-648-3492 \n \nShort title: Acute deletion of BAT PLIN5 causes cold intolerance \n \nFunding Sources: This work was supported by the National Institutes of Health [grant \nnumber R01DK115875 (P.E.B); Electron Microscopy Core for the use of JEOL 1400 \nPlus microscope 1S10OD021685-01A1 to Katherine Luby-Phelps \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nAbstract  \nCold exposure of mice is associated with adaptive molecular and organellar changes in \nbrown adipose tissue (BAT) that promote thermogenesis to defend body temperature. \nWe previously reported that the lipid droplet protein Perilipin 5 (PLIN5) robustly \nincreases in BAT during acute exposure of mice to cold. We demonstrated that chronic \ninduction of BAT PLIN5 within the physiological range in mice housed at room \ntemperature mimics the effects of cold exposure in terms of increased thermogenic \ngene expression in BAT, increased BAT mitochondria cristae packing, and increased \nuncoupled mitochondrial respiration. Additionally, BAT PLIN5 overexpression led to \nhealthy remodeling of inguinal white adipose tissue with improved systemic glucose \ntolerance and reduced diet-induced hepatic steatosis. Conversely, PLIN5 constitutive \ndeletion in brown adipose tissue resulted in decreased BAT thermogenic gene \nexpression and in BAT mitochondrial dysfunction but did not lead to cold intolerance or \nchanges in glucose tolerance. We hypothesized that preserved cold tolerance despite \nchronic deficiency of PLIN5 in BAT was the result of compensatory white adipose tissue \n(WAT) beiging, as suggested by the observed increase in thermogenic gene expression \nin inguinal WAT (iWAT).  To test this hypothesis, we developed a mouse model of \ndoxycycline-inducible, acute deficiency of PLIN5 in BAT of adult mice (BiKOPLIN5 \nmice).  After 7 days of doxycycline treatment and housing at 6 °C, PLIN5 was \nsignificantly reduced in the BAT of BiKOPLIN5 mice compared with littermate control \nmice but was unchanged in the iWAT of these experimental groups. Under these \nconditions, thermogenic gene expression was reduced significantly in the BAT of \nBiKOPLIN5 mice compared to Control mice, as were mitochondrial cristae density and \nuncoupled BAT mitochondrial respiration. These effects of acute PLIN5 deficiency in \nBAT were associated with cold intolerance, which was consistent with the observed \nfailure in iWAT of thermogenic gene expression to increase beyond that of Controls. \nThese findings clarify the essential role of BAT PLIN5 in the physiological adaptive \nresponses of mice to cold ambient temperature. \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nKeywords: \nPerilipin 5; Lipid Droplet; adaptive thermogenesis; brown adipose tissue, mitochondria \n \nIntroduction  \nPLIN5 is a lipid droplet protein that is expressed in highly oxidative tissues such as \nheart, oxidative skeletal muscle, fasted liver, and brown adipose tissue (BAT) [1-3]. \nUnder basal conditions, PLIN5 resides on the surface of lipid droplets and in the cytosol. \nUpon activation of the β-adrenergic-protein kinase A (PKA) pathway by catecholamines, \nMouse PLIN5 is phosphorylated on serine 155 and exerts distinct functions on the lipid \ndroplet and in the nucleus. On the lipid droplet surface PLIN5 coordinates the activation \nof adipose triglyceride lipase (ATGL) to hydrolyze the triglycerides sequestered within \ndroplets, [4]. Once phosphorylated PLIN5 also enriches in the nucleus where it \npromotes the SIRT1-PGC1α gene program to augment mitochondrial biogenesis and \nfunction [5]. Additionally, PLIN5 is proposed to associate with mitochondria to form a \nphysical tether between lipid droplets and mitochondria [6] [7]. \n \nOver the past few years, we have investigated the role of PLIN5 in BAT, based on its \npotential to promote mitochondrial oxidative capacity in that tissue. First, to explore the \nnormal physiological role of BAT PLIN5 in the context adaptive thermogenesis, we \nhoused C57BL/6 mice at 6 °C for intervals of up to 2 weeks. We reported that PLIN5 \nprotein and mRNA increased significantly in BAT during cold housing and reached a \nmaximum at 48 h [8]. Conversely, PLIN5 mRNA and protein levels were suppressed in \nmouse BAT after housing at thermoneutrality (30 °C) for 48 h. To dissect the function of \nincreasing PLIN5 expression in BAT independent of cold challenge, we created a \nmouse model with doxycycline-inducible expression of a Plin5 transgene. We found that \na 4- to 5-fold increase in PLIN5 protein in mice housed at room temperature was \nassociated with increased mitochondria cristae density, mitochondrial DNA, and \nmitochondrial respiratory capacity, as well as other markers of mitochondrial biogenesis \nand function. These changes in BAT mitochondria coincided with improved acute cold \ntolerance and chronic cold acclimation, increased systemic glucose tolerance and \ninsulin sensitivity, decreased high fat diet-induced hepatic steatosis, and healthy \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\ninguinal white adipose tissue (iWAT) remodeling [8]. As might be predicted from these \ngain-of-function experiments, we reported that constitutive knockout Plin5 in BAT via a \nUcp1-Cre allele [8] resulted in dysfunctional mitochondria in BAT at room temperature \nand dysmorphic mitochondria with dramatically reduced cristae density in the BAT of \ncold-exposed mice. However, mice with constitutive knockout of Plin5 in BAT did not \nexhibit cold or glucose intolerance [8] likely explained by the induction of beiging genes \nin iWAT with compensatory beige adipocyte thermogenesis. Herein, we report our \nfindings in a mouse model of acute disruption of the Plin5 gene specifically in BAT that \nwe created to avoid the potential metabolic compensation of iWAT beiging that may \nhave obscured loss-of-function systemic phenotypes of BAT Plin5 constitutive gene \nknockout. \n \nMaterials and Methods \nAnimal studies \nWe performed all animal experiments with approval from the University of Texas \nSouthwestern Medical Center (UTSWMC) Institutional Animal Care and Use \nCommittee, and all experiments were performed in adherence to the guidelines of \nNational Research Council, 2011, Guide for the Care and Use of Laboratory Animals: \nEighth Edition [9]. \n \nFor all experiments presented in this study, we used male or female mice as indicated \non a C57BL/6J background.  We housed mice in a conventional animal facility at 23 °C \nin a 12-h light/dark cycle with free access to food and water, unless otherwise indicated \nin the text, figure legends, or “Methods”. For controlled temperature experiments, we \nhoused mice in a thermoneutrality box at 30 °C (Powers Scientific Inc., Model # \nRIS70SD) or cold box (Powers Scientific Inc., Model # RIS70SD) at 6 °C. Mouse \neuthanasia was by isoflurane anesthesia followed by cervical dislocation. \n \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nGeneration of mouse lines \nBKOPLIN5 strain \nWe used this line in this manuscript only in Figure 6 and the characterization of this line \nwas reported previously by us [8]. This line is a constitutive knockout line specific for \nPlin5 gene deletion in BAT. To create a conditional Plin5 allele in mice, two loxP sites \nwere introduced flanking exons 3–8 of the Plin5 gene (NM_025874.3). An FRT-PKG-\nNeo-FRT cassette [10] followed the loxP site flanking exon 8, which generated a \nknockout-first allele. BAC injection of the targeting construct and homologous \nrecombination in C57BL/6J ES cells was performed by the UTSW Transgenic Core. The \ncorrect ES cell clones were screened and verified by Southern blot. The founder was \nbackcrossed to the C57BL/6J strain. The knockout-first allele was crossed with flp mice \n(JAX 009086) to remove the Neo cassette and generate the floxed line, Plin5loxp/loxp. The \nfinal knockout allele (deletion of exons 3–8) was generated by crossing Plin5loxp/loxp mice \nwith Ucp1-Cre mice to generate the BKOPLIN5 strain. The Ucp1-Cre mouse line \n(B6.FVB-Tg (Ucp1-Cre)1Evdr/J) was generated by the Evan Rosen Lab [11] and \nobtained from Jackson Laboratories Stock # 024670. \n \nBiKOPLIN5 strain \nThis line is a doxycycline-inducible PLIN5 knockout specific for BAT.  To create the \nBiKOPLIN5, we used our previously described Plin5loxp/loxp mice described above. We \nthen crossed this line with transgenic mice expressing the “tet-on” transcription factor \nrtTA under the control of the Ucp1 gene promoter (UCP1-rtTA), which was generously \nprovided by Philipp Scherer [12]. Finally, we crossed this line (Plin5 loxp/loxp; Ucp1-rtTA) \nwith a tet-responsive Cre-recombinase (TRE-Cre) line that can be activated with rtTA in \nthe presence of doxycycline, which was obtained from Jackson Laboratories \n(RRID:IMSR_JAX:006234) and described previously [13]. \n \nDiets and timeline for experiments \nFor the BKOPLIN5 mice, we used chow diet (Teklad, 2016). For BiKOPLIN5 mice, we \nused special diets that contained 600 mg of Dox/kg diet and are referred to as Chow \n(S4107, BioServ) or HFD (60% high fat, S7067, BioServ). For the BiKOPLIN5 mice \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nexperiments we performed the experiments after 7 days or 21 days on Dox diet and at \nthe indicated housing temperature.  \n \nCold housing and cold tolerance test \nFor cold housing we single-housed the mice in a cold box (Powers Scientific Inc., Model \n# RIS70SD) at 6 °C. Mice had free access to food and water except during the 8-hour \ncold tolerance test (see below). We housed mice for 7 days or 21 days as described in \nthe Figures and Figure legends. For cold tolerance we followed the method previously \ndescribed [8] with minor modification as follows: we measured body temperature using \nan implantable temperature transponder (IPTT300, BioMedic Data Systems Inc, \nSeaford, DE) inserted subcutaneously in the dorsal side (back) of the mice, but \npositioned outside from the interscapular BAT region. We used the manufacturer’s \nneedle assembly under general anesthesia with isoflurane via precision vaporizer. To \nallow recovery, we performed temperature experiments 3 days after the transponder \ninsertion. After recovery we single-housed the mice, moved them to the cold box, and \nstarted 600 mg Dox chow diet. Mice were then housed for 7 days at 6 °C with free \naccess to food and water. On day 8 we fasted the mice at 8 a.m., and we measured \nbody temperature using a temperature reader (DAS-8007-IUS, BioMedic Data Systems \nInc, Seaford DE) every two hours for 8 hours.  \n \nOral glucose tolerance test (OGTT) \nFor OGTT we followed a previously described method [8] with minor modifications, as \nfollows. We performed OGTT after 7 days of cold housing and Dox 600 mg/kg HFD diet. \nOn day 8, we fasted mice for 5 h. We administered 2.5 g of glucose/kg of body weight \nby oral gavage and collected tail blood at the indicated time points for measurement of \nglucose. For glucose measurement, we used a Contour Next EZ glucometer (Bayer \nHealthCare LLC). \n \nRNA extraction, qPCR and primers \nFor RNA extraction and qPCR, we followed the same method as previously described \n[8]. Briefly, for RNA extraction we homogenized ~50 mg of tissue in 1 ml of QIAzol \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\n(Qiagen, Cat #79306) using a Tissue Lyser II and stainless-steel beads (Qiagen, \n69989). After homogenization, we centrifuged the samples at 14,000 × g for 5 min and \nthen removed the floating fat layer from the top by pipetting; we then added 200 ul \nchloroform and centrifuged the samples at 14,000 g for 15 min. We collected the \nsupernatant and used an RNA extraction kit (Cat #74104, Qiagen) to obtain RNA. \nDuring RNA purification we used the RNase-Free DNase Set (Cat #79254, Qiagen) for \nDNA digestion. For qPCR we prepared cDNA with iScript kit, (Bio-Rad Cat # 1708891) \nusing 1 μg of RNA and followed the manufacturer’s instructions, using the following \ncycles and temperatures: 5 min at 25 °C, 30 min at 42 °C, 5 min at 85 °C and hold at \n4 °C. After the cDNA preparation, we performed qPCR using Power Sybr green (0.1 μM \nfinal concentration for primers) on Applied Biosystem’s Viaa7 machine. Comparative Ct \nmethod (ΔΔ Ct) was used to analyze all qPCR data. Expression was normalized to that \nof the 18S ribosomal subunit as endogenous control, and the relative expression was \ncalculated in comparison with the reference sample that is indicated in each figure. \nPrimers were designed using Primer Express 3.0.1 (Applied Biosystems) or Primer \nBlast National Center for Biotechnology Information. Primers used for qPCR were as \nfollows:  Plin5, forward GAGGCAGCAACAGGGCTACT, reverse \nCAAAGAGTGTTCATAGGCGAGATG; 18S forward GAG CGA AAG CAT TTG CCA \nAG, reverse GGC ATC GTT TAT GGT CGG AA; Ucp1 forward CCC TGG CAA AAA \nCAG AAG GA, reverse AGC TGA TTT GCC TCT GAA TGC; Dio forward AAG AAG \nCAC CGG AAC CAA GA, reverse GGC GGC AAG GAG AAA CG; Elov3 forward GCC \nAAA CTG AAG CAT CCT AAT CTT, reverse CCC AGA ACC ATC TGC AGA ATC; \nPpargc forward TGC CAT TGT TAA GAC CGA G, reverse TTG GGG TCA TTT GGT \nGAC. \n \nTissue protein lysate preparation, Western Blot and antibodies \nFor protein lysate preparation and Western blot we followed the same methods as \npreviously described [8].We homogenized tissue (~50 mg per sample) in RIPA buffer \ncontaining 50 mM tris(hydroxymethyl)aminomethane, 140 mM sodium chloride, 0.1% \nsodium dodecyl sulfate,1% triton X-100, 0.1% sodium deoxycholate, and 0.5 mM \nethylene glycol-bis(2-aminoethylether)-N,N,N′,N′-tetraacetic acid, using a Tissue Lyser II \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nwith stainless-steel beads (Qiagen, Germany). After homogenization, we centrifuged the \nsamples at 14,000 × g for 10 min at 4 °C to remove cell debris and collected the \nsupernatants. We measured protein concentration using Pierce® 660 nm Protein assay \nreagent (Thermo Scientific, Cat # 22660). We mixed the samples with 2× protein \nsample loading buffer (62.5 mM Tris-HCL, 25% glycerol, 2% SDS, and 0.1% Orange G). \nFor protein electrophoresis, we loaded 20 µg of protein per sample into premade gels \n[CriterionTM TGX TM 4–20% (Bio-Rad, Cat # 5671094) or AnyKD (Bio-Rad, Cat # \n5671124)] and for protein transfer, we used the Criterion Blotter SystemTM with \nnitrocellulose membrane (Bio-Rad, Cat # 1620112). We blocked the membranes post-\ntransfer with 5% nonfat dry milk diluted in Tris-buffered saline, pH 7.4, with 0.1% \nTween-20 (TBS-T) for 1 h, and then incubated with the indicated primary antibody with \n3% BSA diluted in TBS-T (BSA-TBS-T) for 12–16 h at 6 °C. After primary antibody \nincubation, we washed the membranes three times for 5 min each with TBS-T, and then \nincubated with the appropriate secondary antibody from Li-Cor in BSATBS-T for 30 min. \nAfter secondary antibody, we washed membranes three times for 5 min each in TBS-T. \nWe visualized the immunoblotted proteins with the Odyssey CLx near-infrared imaging \nsystem (Li-Cor). \n \nWe used the following primary antibodies: PLIN5 (Progen  Cat # GP31), GDI (Bickel \nLab) [14], Histone H3 (Cell signaling Cat # 9717), Cox4 (Cell signaling Cat# 4850), \nTIM23 (Cell signaling Cat # 34822), TOM20 (Santa Cruz sc-11021), Actin (Cell \nSignaling Cat# 4967). \n \nGlucose and fatty acid uptake \nFor glucose and fatty acid uptake we followed the methods as described previously [15] \nwith minor modifications. We administered by oral gavage 1 mg/g body weight glucose \nwith 20 µCi deoxy-D-glucose, 2-[1-14C]- (Perkin Elmer, Cat # NEC495A00) per mouse \nand 3H-triolein (Perkin Elmer, Cat # NET43100) as tracers. After 1-h, mice were \nsacrificed and tissues of interest were harvested and weighed, small pieces of each \ntissue were cut, weighed, and solubilized using SolvableTM (0.1 ml per 10 mg of tissue) \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nand 200 µl were added to 5 ml of scintillation cocktail in a glass scintillation vial. \nRadioactivity was measured using a scintillation counter (Beckman Coulter, LS6000). \n \nHistology \nWe performed histology as previously described [8]. To obtain mouse tissue samples \nfor histology, we performed cardiac perfusion under ketamine anesthesia. After cardiac \nperfusion with 4% paraformaldehyde in phosphate-buffered saline (PBS), pH 7.4, we \ndissected tissues and fixed with 4% paraformaldehyde solution overnight. The UTSW \nMolecular Pathology Core Facility performed paraffin embedding, sectioning, H&E \nstaining, and Oil red O staining. We acquired bright-field images using a Leica DM \n4000B microscope. \n \nElectron microscopy \nFor electron microscopy we followed the methods as previously described [8]. For \ntransmission electron microscopy sample collection, we performed cardiac perfusion \nunder ketamine anesthesia with a perfusion buffer (4% paraformaldehyde, 1% \nglutaraldehyde, and 0.1 M sodium cacodylate, pH 7.4), and we dissected BAT into 1 mm \npieces that were then fixed with 2.5% glutaraldehyde and 0.1 M sodium cacodylate, pH \n7.4. Further processing of the samples was performed by the UTSW Electron \nMicroscopy Core as follows: tissue samples were rinsed in 0.1 M sodium cacodylate \nbuffer and postfixed in 1% osmium tetroxide and 0.8% potassium ferricyanide in 0.1 M \nsodium cacodylate buffer three times for 3 h at room temperature. After three rinses in \nwater, they were stained en bloc with 4% uranyl acetate in 50% ethanol for 2 h. Next, \nthe samples were dehydrated with increasing concentrations of ethanol, transitioned \ninto resin with propylene oxide, infiltrated with Embed-812 resin, and polymerized in a \n60 °C oven overnight. Blocks were sectioned with a diamond knife (Diatome) on a Leica \nUltracut 7 ultramicrotome (Leica Microsystems), collected onto copper grids, and post \nstained with 2% aqueous uranyl acetate and lead citrate. Images were acquired on a \nJEOL 1400 Plus electron microscope and photographed with a BIOSPR16 camera. \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nQuantification of mitochondria and lipid droplet contact sites \nWe performed quantification of mitochondria and lipid droplet contact sites as previously \ndescribed [8, 16]. For quantification of LD–mitochondria contact sites, we used Image J \nsoftware NIH Ver 1.5.  We quantified mitochondria in contact with LDs by count and \ncontact area as % of mitochondrial perimeter or % of LD perimeter (n = 10 EM fields per \ngenotype at 1000x magnification). \n \nmtDNA quantification \nWe performed mtDNA quantification as previously described [8]. For isolation of cellular \ntotal DNA, 25 mg of tissue were homogenized in PBS (ph. 7.5) using a Tissue Lyser II \nand stainless-steel beads (Qiagen, 69989). Samples were processed with Qiagen’s \nQIAamp DNA Mini Kit (51304) per the kit instructions. Mitochondrial DNA was amplified \nusing primers specific for the mitochondrial cytochrome c oxidase subunit 2 (COX2) \ngene and normalized to genomic DNA by amplification of the ribosomal protein s18 \n(rps18) nuclear gene, using quantitative PCR. We used primers that were previously \ndescribed [17] as follows: Rsp18 forward TGT GTT AGG GGA CTG GTG GAC A and \nreverse CAT CAC CCA CTT ACC CCC AAA A and Cox2 forward ATA ACC GAG TCG \nTTC TGC CAA T and reverse TTT CAG AGC ATT GGC CAT AGA A. \n \nMitochondria isolation \nWe isolated crude and pure mitochondria as described previously [18]. Mice were \neuthanized (Control or BiKOPLIN5) by isoflurane overdose and neck dislocation. We \ndissected BAT and rinsed the tissue in cold buffer containing 225-mM mannitol, 75-mM \nsucrose, 0.5% BSA, 0.5-mM EGTA and 30-mM Tris–HCl pH 7.4. \nWe then minced 50-100 mg of tissue into approximately 1 mm pieces and resuspended \nthe pieces in the same buffer used for rinsing the BAT (500 l per each 100 mg of \ntissue). We then transferred BAT to a 10 ml glass/Teflon Potter Elvehjem homogenizer \nand homogenized the BAT pieces using a Teflon pestle with eight strokes at 1,500 \nr.p.m. We then we centrifuged the homogenate in a 15 ml polypropylene centrifugation \ntube at 740g for 5 min at 4 °C. We collected the supernatant and centrifuged one more \ntime at 740g for 5 min at 4 °C. We collected the supernatant and centrifuged at \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\n9,000g for 10 min at 4 °C. After this step we discarded the supernatant and carefully \nresuspended the pellet in 1 ml of buffer containing 225-mM mannitol, 75-mM sucrose, \n0.5% BSA and 30-mM Tris–HCl pH 7.4. We then centrifuged the mitochondrial \nsuspension at 10,000g for 10 min at 4 °C. This pellet was crude mitochondria. We \nresuspended the pellet in 50-200 l of buffer containing 250-mM mannitol, 5-mM \nHEPES (pH 7.4) and 0.5-mM EGTA and then measured protein concentration. For \nsubsequent mitochondrial respiration experiments we resuspended these crude \nmitochondria in resuspension buffer (225-mM mannitol, 75-mM sucrose, 0.5% BSA and \n30-mM Tris–HCl pH 7.4) or for Western blotting we added 1X protein loading buffer (Li-\nCor Cat # 928-40004). \n \nFor pure mitochondria isolation we used as starting material the crude mitochondria \nobtained as described above. First, we added 2 ml of buffer containing 225-mM \nmannitol, 25-mM HEPES (pH 7.4), 1-mM EGTA and 30% Percoll (vol/vol) to an \nultracentrifuge tube, then carefully layered 200 l of the crude mitochondria obtained \nabove. Then we layered 1 ml of buffer containing 250-mM mannitol, 5-mM HEPES (pH \n7.4) and 0.5-mM EGTA and centrifuged at 95,000g for 30 mins. After centrifugation we \ncollected the pure mitochondria from the bottom of the tube using a Pasteur pipette. We \nresuspended the pellet in 2 ml of buffer containing 250-mM mannitol, 5-mM HEPES (pH \n7.4) and 0.5-mM EGTA and centrifuged at 6,300g for 10 minutes. We repeated this step \none additional time and discarded the supernatant. Finally, we resuspended the pellet in \n100 l of buffer containing 250-mM mannitol, 5-mM HEPES (pH 7.4) and 0.5-mM \nEGTA. We measured protein concentration, and the mitochondria were used for \nWestern blotting or for protease protection assays. \n \nProtease protection assay \nWe performed protease protection assays using pure mitochondria obtained as \ndescribed above. To the resuspended pellet of pure mitochondria (in buffer containing \n250 mM mannitol, 5 mM HEPES, pH 7.4, and 0.5 mM EGTA), we added Proteinase K \n(100 µg/ml) and incubated the samples on ice for 15 min. We stopped protease \ndigestion by adding Phenylmethylsulfonyl Fluoride PMSF (2 mM) and incubating the \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nsamples on ice for 5 min. We measured protein concentration and further processed the \nsample for Western blot with 1X protein loading buffer (Li-Cor Cat # 928-40004). \n \nMitochondria respiration \nFor assessment of mitochondrial respiration, we used crude mitochondria isolated from \nintrascapular BAT of BiKOPLIN5 mice as described above. To measure oxygen \nconsumption rate, we used the NeoFox apparatus as follows. We suspended \nmitochondria corresponding to 50 g of mitochondrial protein in a buffer containing 250-\nmM mannitol, 5-mM HEPES (pH 7.4) and 0.5-mM EGTA. Oxygen consumption rate \n(OCR) was determined before and after sequential injections of the following \ncompounds pyruvate (5mM), Guanosine 5′-diphosphate sodium (GDP-1 mM), \nAdenosine 5′-diphosphate sodium (ADP-450 M), oligomycin (2 g/ml) and Carbonyl \ncyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP-1 M). We calculated OCR as \nmol/l/min/g of protein. \n \nSoftware \nFor WB band intensity analysis, we used Image Studio Ver. 3.1 (Licor Biosciences). For \nqPCR Ct values analysis, we used Quant Studio Real-time qPCR software Ver. 3.1 \n(Applied Biosystems). For colorimetric microplate assays (protein quantification) we \nused Gen5 Ver 2.01.14 software (Bio Tek Instruments, Inc.). For quantification of LD–\nmitochondria contact sites, we used Image J software NIH Ver 1.53. \n \nStatistical analysis  \nFor statistical analysis we used GraphPad Prism version 10.0.1 for MacOS, GraphPad \nSoftware, La Joya California USA, www.graphpad.com. \nFor all the experiments, data are representative of at least three independent \nexperiments and all attempts to reproduce were successful. P values are indicated in \nfigures or figure legends. Statistical analyses were performed using Student’s t test if \ntwo groups were analyzed or ANOVA followed by Tukey posttest if more than two \ngroups were analyzed. Statistical significance is defined as p<0.05. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nResults \nCreation and validation of an inducible mouse model of BAT-specific PLIN5 \nknockout. \nTo create this doxycycline-inducible knockout mouse line, we crossed our previously \ndescribed Plin5loxp/loxp mouse strain, in which we had introduced LoxP sites flanking \nexons 3 through 8 of the Plin5  gene by homologous recombination in C57BL/6 \nembryonic stem (ES) cells [8], with transgenic mice expressing the “tet-on” transcription \nfactor rtTA under the control of the Ucp1 gene promoter (Ucp1-rtTA), which was \ngenerously provided by Phillip Scherer [12]. Finally, we crossed Plin5 loxp/loxp; Ucp1-rtTA \nmice with mice carrying a tetracycline-responsive Cre recombinase (TRE-Cre) [13] to \nproduce BiKOPLIN5 mice. Littermate control (Control) mice lacked the TRE-Cre allele \n(Fig. 1a). \nWe evaluated different doxycycline doses, durations of doxycycline treatment, and \nhousing temperatures to determine optimal conditions to achieve significant knockout of \nPLIN5 expression specifically in BAT without activating compensatory beiging in iWAT. \nAs expected, based on our previously published data that PLIN5 is induced in BAT of \nC57BL/6J mice during housing at 6 °C [8], PLIN5 was increased 4-fold in the BAT of \nControl mice after 7 days of doxycycline diet and housing at 6 °C (Fig. 1b). In contrast, \nunder the same conditions, PLIN5 in the BAT of BiKOPLIN5 mice was ~80% less than \nin that of Control mice. No significant differences in PLIN5 protein in iWAT, liver, or \nheart were observed between BiKOPLIN5 and Control mice (Fig. 1c, d). Seven days of \ndoxycycline treatment of mice housed at 23 °C was not associated with reduced PLIN5 \nin BAT or iWAT of either BiKOPLIN5 or Control mice. \nInducible Plin5 BAT knockout causes reduction on thermogenic gene expression \nin BAT \nTo study the effects of acute BAT Plin5 knockout on thermogenic gene expression in \nBAT and iWAT, we performed real-time quantitative polymerase chain reaction (qPCR) \non RNA isolated from BAT and iWAT after 7 or 21 days of Dox administration and \nhousing at 23 °C or 6 °C. Whereas Plin5 mRNA increased in Control mice after 7 days at \n6 °C compared with those housed at 23 °C, Plin5 mRNA in BiKOPLIN5 mice did not \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nchange significantly (Figure 2a). Similarly, in the BAT of BiKOPLIN5 mice, expression of \nmost thermogenic genes (Ucp1, Dio, and Ppargc) after 7 days at 6 °C was significantly \nlower than in Control mice, but there were no differences between these genotypes in \niWAT (Fig. 2b). After 21 days of Dox diet, we observed much reduced Plin5 expression \nin the BAT of BiKOPLIN5 at both 23 °C and 6 °C compared with Control mice, as well as \nin iWAT at 6 °C. At the three-week time point, expression levels of Ucp1, Dio, Elov3 and \nPpargc in BAT were reduced in iWAT, but only that of Ppargc was statistically \nsignificant (Fig. 2c, d). \n \nTo study the effects of acute Plin5 gene disruption specifically in BAT but sparing iWAT, \nsubsequent experiments were conducted on day 8 following 7 days of housing the \nexperimental mice at 6 °C and of Dox-chow diet. \n \nAcute BAT PLIN5 knockout causes cold intolerance but not glucose intolerance. \nFirst, we evaluated body weight and daily food intake when mice were housed at 23 °C \nwithout Dox diet and then during 7 days of cold exposure with Dox diet. Food intake \nwas calculated as the average of the 3 days before day 0, when mice were maintained \nat 23 °C and not yet on the Dox diet, and again as the average of days 5, 6, and 7 of \ncold exposure and Dox diet. Body weight was measured at day 0 and again at day 7. \nWe found no differences in food intake or body weight between Control and \nBATiKOPLIN5 mice either before or after cold exposure and Dox diet (Figure 3 a, b). To \ndetermine the effect of acute PLIN5 deficiency in BAT on glucose tolerance in the \ncontext of high-fat diet (HFD) induced obesity, we fed male or female mice with 60% \nHFD without Dox for 8 weeks and housed the mice at 23 °C. After 8 weeks we shifted \nthe mice to housing at 6 °C and initiated Dox diet with 60% HFD (HFD-Dox) for 7 days. \nOn day 8-day we performed OGTT as described in Methods. Glucose tolerance was \nsimilar between control and BiKOPLIN5 mice in both male (Figure 3c) and female mice \n(Figure 3d).  \n \nGiven the reduction in thermogenic gene expression (Figure 2a) that we observed in \nBiKOPLIN5 mice, we performed a cold tolerance test to assess whether acute PLIN5 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\ndeficiency in BAT results in cold intolerance. First, we implanted a temperature \ntransponder subcutaneously and allowed the mice to recover for 3 days. Then we \ninitiated Chow-Dox diet and housed the mice at 6 °C for 7 days. On day 8 we fasted the \nmice starting at 8 a.m. and measured body temperature every two hours for 8 hours. \nMale BiKOPLIN5 mice showed consistently lower body temperatures than Control mice \nat every time point with statistically significant changes at 8 hours cold exposure (Figure \n3e). Female BiKOPLIN5 mice had statistically significant lower body temperatures than \nfemale Control mice at hours 4, 6, and 8 of the test.  (Figure 3f). We also measured \nbody weight before and after fasting during the cold tolerance test. Both male and \nfemale BiKOPLIN5 mice lost less weight during the test than Control mice (Figure 3g-h).  \n \nAcute BAT Plin5 knockout decreases BAT tissue fatty acid uptake \nBAT augments triglyceride and fatty acid clearance especially during cold exposure [15] \nand previously, we found that BAT PLIN5 overexpression increases BAT fatty acid \nuptake but has no effect on glucose uptake by that tissue [8]. To evaluate if acute \ninduction of PLIN5 deficiency in BAT changes fatty acid or glucose uptake, we \nperformed a combined oral triacylglycerol and glucose tolerance test using H3labeled \ntriolein and C14 labeled glucose, as previously described [15], in BiKOPLIN5 and Control \nmice housed for 7 days at 6 °C and fed Chow-Dox diet One hour after administration of \nthe labeled substrates via oral gavage, we harvested the indicated tissues and \nassessed uptake of the labeled substrates by scintigraphy, as described in Methods. \nFatty acid uptake was decreased in the BAT of BiKOPLIN5 mice compared with Control \nmice, but there were no statistically significant differences in iWAT, gonadal WAT \n(gWAT), heart, or liver (Figure 4a). There also were no differences in glucose uptake \nbetween BiKOPLIN5 and Control mice in any of these tissues, including BAT (Figure \n4b).  \n \nIn an independent cohort of BiKOPLIN5 and the Control mice, we studied the histology \nof the BAT, iWAT, and livers. By H&E staining we found no differences in BAT (Figure \n4c and d), iWAT (Figure 4e) or liver (Figure 4f) between the Control and BiKOPLIN5 \nmice. Both genotypes showed extensive browning of iWAT likely due to the 7 days of \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\ncold exposure we used to promote expression of Cre-recombinase in our mouse model \n(Figure 4e). The observed browning was consistent with the induction of thermogenic \ngene expression we measured in iWAT (Figure 2b). However, these changes in gene \nexpression and histology in iWAT were not sufficient to overcome PLIN5 deficiency in \nBAT for the purpose of cold tolerance. \n \nBAT Plin5 knockout impairs mitochondrial morphology and respiration \nGiven the cold intolerance we observed in mice with acute PLIN5 deficiency, we next \nevaluated mitochondria morphology and function in the BiKOPLIN5 mice by \ntransmission electron microscopy of BAT tissue and respirometry of isolated \nmitochondria. Similar to constitutive PLIN5 KO mice [8], brown adipocytes in cold-\nstressed BiKOPLIN5 mice had dysmorphic mitochondria with loss of cristae and \nswelling of mitochondria across the whole tissue (Figure 5a-top panels, lower \nmagnification). These differences were more clearly observed at higher magnification \n(Figure 5a, bottom panels). Similar to the constitutive PLIN5 knockout mice from our \nprevious work [8], BiKOPLIN5 did not show changes in mitochondria/lipid droplet \ncontact sites (Figure 5b) in terms of the percentage of mitochondria in contact with lipid \ndroplets, the ratio of mitochondria-lipid droplet contact surface to mitochondrial \nperimeter, and the ratio of mitochondria-lipid droplet contact surface to lipid droplet \nperimeter. BiKOPLIN5 mice also showed a trend to reduced mtDNA content, but this \nwas not statistically significant (Figure 5c). Additionally, we tested mitochondrial function \nby measuring oxygen consumption rate (OCR) for mitochondrial isolated from BAT of \nControl or BiKOPLIN5 mice housed at 6 °C and fed with Dox for 7 days. The BiKOPLIN5 \nmitochondria in comparison to Control mitochondria exhibited a blunted increase in \nmitochondrial OCR after the administration of pyruvate, as well a significant reduction in \nmaximal uncoupled respiration following the addition of carbonyl cyanide \n4(trifluoromethoxy)phenylhydrazone (FCCP) (Figure 5d). After addition of guanosine \ndiphosphate (GDP), an inhibitor of UCP1mediated uncoupled respiration, the rate of \ndecline in oxygen consumption was nearly flat in BiKOPLIN5 mitochondrial compared \nwith a steep decline in Control mitochondria.  \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\n \n \nPLIN5 localizes to the outer mitochondrial membrane in brown adipocytes during \ncold exposure of mice  \nIt has been reported that PLIN5 localizes to mitochondria in skeletal muscle, as well in \nseveral cell lines [6, 7, 19]. However, whether PLIN5 localizes to mitochondria in brown \nadipocytes is not known. Based on the profound changes in mitochondrial morphology \nand function in the BAT of BiKOPLIN5 mice, we next tested whether PLIN5 localizes to \nthe mitochondria of brown adipocytes.   \n \nTo this end, we isolated pure mitochondria as described in Methods from the BAT of \nC57BL6/J male mice that had been housed at either 23 C or 6 C for 48 h. The isolated \nmitochondria were assessed by a protease protection assay with Proteinase K to \ndetermine whether PLIN5 localizes to the outer mitochondrial membrane (OMM) or the \ninner mitochondrial membrane (IMM). We used antibodies to proteins that reside \nspecifically in each of these compartments to assess localization of PLIN5 and of \nmarker proteins: mitochondrial import receptor subunit TOM20 homolog (TOM20) as a \nmarker of the OMM, mitochondrial import inner membrane translocase subunit Tim23 \n(TIM23) as a marker of the IMM. PLIN5 was only detected in the pure mitochondria \nfraction isolated from the BAT of mice following cold exposure and not from that of mice \nat room temperature (Figure 6a). Nor was PLIN5 detectable in the mitochondria isolated \nfrom BAT of BKOPLIN5 knockout mice (negative control). After adding proteinase K to \nthe pure mitochondria sample, PLIN5 was no longer detectable by immunoblotting, as \nwas the case for TOM20. On the other hand, the inner mitochondrial membrane marker \nTIM23 was resistant to proteolysis by proteinase K. These data are consistent with \nPLIN5 localization to the OMM (Figure 6a). \n \nDiscussion \nHerein, we studied the effects of acute (7 days) depletion of PLIN5 in BAT of C57BL6/J \nmice using a doxycycline-inducible knockout of a floxed Plin5 allele (UCP1rtTa; TRE-Cre; \nPlin5flox/flox). Previously, we demonstrated that constitutive, BAT-specific Plin5 knockout \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\n(UCP1-Cre; Plin5flox/flox) in mice results in mitochondrial damage during cold exposure \nand reduced mitochondrial oxygen consumption rate [8], yet this chronic model did not \nexhibit differences in cold tolerance. With the additional evidence from the BiKOPLIN5 \nmodel presented in this report, we find that acute PLIN5 loss in brown adipocytes is \nassociated with cold-induced mitochondrial damage and dysfunction, similar to the \nmodel of chronic PLIN5 deficiency in BAT, but with cold intolerance. Acute deficiency of \nPLIN5 in BAT resulted in significant BAT phenotypes, including reduced expression of \nthermogenic genes, reduced fatty acid uptake, reduced mitochondrial DNA content, \ndecreased mitochondrial cristae density, swollen mitochondria, and impaired \nmitochondrial respiration, which together were associated with cold intolerance at 7 \ndays. In this acute setting, there was insufficient time for browning of iWAT to the \ndegree that would compensate for BAT dysfunction, at least at the level of thermogenic \ngene expression. A similar phenomenon was reported by Pereira et al., who reported \nthat BAT-specific OPA1 knockout caused mitochondrial dysfunction but also \ncompensatory WAT browning, which increased cold tolerance above even that of \ncontrol mice [20]. Our two different models of BAT PLIN5 deficiency highlight the value \nof assessing genetic loss-of-function across both acute and chronic time frames, which \nis achievable through inducible gene knockout technology. \n \nThough PLIN5 localization to mitochondria has been reported previously, our data in \nthis study are the first to demonstrate in mouse BAT that endogenous PLIN5 enriches \non the mitochondrial outer membrane in pure mitochondrial fractions during cold \nexposure. Carole Sztalryd’s lab first reported PLIN5 mitochondrial localization in 2011 \nby demonstrating that endogenous PLIN5, but not PLIN2 or PLIN3, co-fractionates with \ncrude mitochondria isolated from rat left heart ventricle and that fluorescently tagged \nPLIN5 localizes to mitochondria in AML12 hepatocytes and HL1 cardiomyocytes [6]. \nThe authors proposed that PLIN5 promotes lipid droplet–mitochondria contacts to \nfacilitate fatty acid transfer for oxidation in mitochondria. More recently, Sarah Cohen’s \ngroup reported that interactions between the C-terminal 38 amino acids of PLIN5 and \nmitochondrial FATP4 promote lipid droplet-mitochondrial tethering for augmentation of \nfatty acid transport from lipid droplets to mitochondria for fatty acid oxidation [21]. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nGuenter Haemmerle’s group further reported that, in AML12 hepatocytes expressing \nfluorescently tagged PLIN5, mitochondria are tightly associated with PLIN5-coated lipid \ndroplets, and that this interaction persists following PKA activation [19]. Moreover, his \ngroup demonstrated that expression of PLIN5 lacking its last three amino acids did not \npromote lipid droplet-mitochondria contacts in HEK-293T cells [19]. \n \nIn contrast to these previous reports, we found no differences in lipid droplet–\nmitochondria contacts in BAT from constitutive PLIN5-KO mice, which suggests that \nPLIN5 is not required for these associations in BAT [8]. Our current study of acute \nPLIN5 deficiency confirms that PLIN5 is dispensable for lipid droplet contacts with \nmitochondria in BAT.  \n \nTwo other groups have examined the role of PLIN5 in mitochondria–lipid droplet contact \nsites in thermogenic cells. The Orian Shirihai group reported that, compared with \nthermoneutrality, cold exposure of mice reduces LD–mitochondria interactions in brown \nadipocytes. They found that cytoplasmic mitochondria not associated with lipid droplets \nexhibit a greater capacity for fatty acid oxidation than peri-droplet mitochondria. \nConsistent with this interpretation, they showed that overexpression of full-length PLIN5 \nin brown adipocytes promotes LD–mitochondria association, whereas expression of a \ntruncated PLIN5 lacking the last 20 amino acids fails to recruit mitochondria to lipid \ndroplets [16]. In line with these observations, we previously found that cold exposure \n(6 °C vs. 23 °C) decreased LD–mitochondria contacts in BAT in but not in BATiPLIN5 \nmice (PLIN5 overexpression in BAT). Notably, in mice housed at 23 °C, BATiPLIN5 \nmice exhibited a reduction in LD–mitochondria contact number and contact area \ncompared with Control mice [8]. Taken together, and in the context of the Shirihai \ngroup’s finding that cytoplasmic mitochondria in BAT are more oxidative than peri-\ndroplet mitochondria, our previous published data suggest that PLIN5 overexpression in \nBAT results in a higher proportion of mitochondria with an oxidative phenotype relative \nto Control mice. \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nIn contrast to these PLIN5 overexpression models, we did not observe differences in \nmitochondria–lipid droplet contact sites in BAT by electron microscopy quantification in \neither our constitutive BAT-specific PLIN5 knockout mice (Supplementary Figure 19 [8]) \nor the inducible BATiKOPLIN5 mice described in the current study. These findings \nsuggest that, in BAT, PLIN5 is not required for the formation or maintenance of lipid \ndroplet–mitochondria contact sites. The group of Carles Cantó has identified Perilipin 1 \n(PLIN1)—through its interaction with Mitofusin 2 (Mfn2)— as a mediator of lipid droplet–\nmitochondria tethering in BAT [22],  Whereas PLIN5 may promote mitochondrial \ntethering to lipid droplets in many cell types, it is possible that this function is assumed \nby PLIN1 or a different protein in brown adipocytes. However, the roles of PLIN5 and \nPLIN1 in mitochondria contacts with lipid droplets is far from settled as the Pingsheng \nLiu group reported that mitochondria are tightly associated with lipid droplets in \noxidative tissues but that their association is not dependent on either PLIN5 or PLIN1. \nThis conclusion was based on the proteomic identification of equivalent levels of \nmitochondrial proteins in lipid droplet fractions isolated from the BAT of wildtype, PLIN5 \ndeficient, and PLIN1 knockout mice [23]. Our findings support the notion that while \nPLIN5 overexpression may modulate the distribution or metabolic phenotype of \nmitochondrial subpopulations, PLIN5 itself is likely dispensable for basal droplet–\nmitochondria tethering in BAT. It may be that the focus of investigators on PLIN5 \nplaying a role in lipid droplet-mitochondria contacts reflects the original discovery of \nPLIN5 as a member of the Perilipin family of lipid droplet proteins [1-3]. It is important to \nrecall that PLIN5 also exists in a cytoplasmic pool and can move on and off the lipid \ndroplet. We cannot rule out the possibility that cytoplasmic PLIN5, not lipid droplet-\nbound PLIN5, is the active player on the mitochondrial outer membrane [5, 24, 25]. \n \nIf PLIN5 does not localize to mitochondria to promote the physical coupling of lipid \ndroplets with mitochondria in BAT, then what is the purpose of this localization that is \npromoted by cold exposure of mice? We observed profound disruption of mitochondrial \ncristae structure and impaired mitochondrial respiration in the BAT of BiKOPLIN5 mice \nhoused at 6°C, which was fully consistent with our findings in our previous report of \nconstitutive PLIN5 knockout in BAT. While the precise mitochondrial function of PLIN5 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nin BAT mitochondria remains to be defined, we propose that during cold exposure \nPLIN5 may interact with outer mitochondrial membrane proteins to help preserve \nmitochondrial membrane integrity during the metabolic stress of increased fatty acid flux \nand high beta-adrenergic signaling. In this regard, Guenter Haemmerle’s group reported \nthat full-length PLIN5 interacts with mitochondrial protein complexes involved in \noxidative phosphorylation and mitochondrial dynamics in AC16 cardiomyocytes, \nwhereas cardiomyocytes expressing a truncated PLIN5 variant lacking the final three \namino acids fails to exhibit these interactions and does not promote lipid droplet-\nmitochondrial contacts. [19]. While these findings shed light on potential mechanisms \nfor PLIN5 role in cardiomyocyte mitochondria, they may not elucidate its role in BAT, in \nwhich mitochondria are unique and function differently. Mitochondria in cardiomyocytes \nare optimized for efficient ATP production [26]; in contrast, mitochondria in BAT are \nspecialized for thermogenesis at the expense of ATP production [27]. In BAT, UCP1-\nmediated uncoupled respiration and additional UCP1-independent thermogenic \npathways play a central role in thermogenesis, which highlights the need to define \nPLIN5’s specific mitochondrial function in this tissue context [27]. Future work will focus \non defining this mitochondrial role of PLIN5 in brown adipocytes. \n \nIn summary, together with our previous discovery that, upon its phosphorylation by \nPKA, PLIN5 translocate to the nucleus to activate the SIRT1–PGC-1α pathway [5], our \ndata establish PLIN5 as a key integrator of lipid droplet function with thermogenic gene \nprograms and mitochondrial function. While studies in heart, liver, and muscle have \nshown that PLIN5 coordinates fatty acid handling with oxidative metabolism, here we \nextend this concept to BAT by demonstrating that PLIN5 is enriched in mitochondria \nduring adrenergic stimulation and is required to preserve mitochondrial structure and \nrespiration during cold exposure. This mitochondrial role in BAT is distinct from the \npreviously proposed function of PLIN5 in promoting lipid droplet–mitochondria contacts \nin other tissues. Instead, our findings support a model in which PLIN5 couples lipid \nmobilization with nuclear transcriptional responses and mitochondrial performance to \nmeet the energetic demands of thermogenesis. \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nData availability statement \nAll data supporting the conclusions of this study are provided in the manuscript or may \nbe requested from the corresponding author. \n \nAcknowledgments \nWe thank Philipp Scherer and UT Southwestern Touchstone Diabetes Center for the \nUCP1rtTA mouse. We thank the UT Southwestern Metabolic Phenotyping Core (Ruth \nGordillo and Syann Lee), Electron Microscopy Core (Kate Luby-Phelps), Histo \nPathology Core (Bret M. Evers and John M. Shelton), and Transgenic Core (Robert E. \nHammer) \n \nReferences \n1. Yamaguchi, T., et al., MLDP, a novel PAT family protein localized to lipid droplets and \nenriched in the heart, is regulated by peroxisome proliferator-activated receptor alpha. J \nBiol Chem, 2006. 281(20): p. 14232-40. \n2. Wolins, N.E., et al., OXPAT/PAT-1 is a PPAR-induced lipid droplet protein that promotes \nfatty acid utilization. Diabetes, 2006. 55(12): p. 3418-28. \n3. Dalen, K.T., et al., LSDP5 is a PAT protein specifically expressed in fatty acid oxidizing \ntissues. Biochim Biophys Acta, 2007. 1771(2): p. 210-27. \n4. Kimmel, A.R. and C. Sztalryd, Perilipin 5, a lipid droplet protein adapted to mitochondrial \nenergy utilization. Curr Opin Lipidol, 2014. 25(2): p. 110-7. \n5. Gallardo-Montejano, V.I., et al., Nuclear Perilipin 5 integrates lipid droplet lipolysis with \nPGC-1 alpha/SIRT1-dependent transcriptional regulation of mitochondrial function. \nNature Communications, 2016. 7. \n6. Wang, H., et al., Perilipin 5, a lipid droplet-associated protein, provides physical and \nmetabolic linkage to mitochondria. J Lipid Res, 2011. 52(12): p. 2159-2168. \n7. Bosma, M., et al., The lipid droplet coat protein perilipin 5 also localizes to muscle \nmitochondria. Histochem Cell Biol, 2012. 137(2): p. 205-16. \n8. Gallardo-Montejano, V.I., et al., Perilipin 5 links mitochondrial uncoupled respiration in \nbrown fat to healthy white fat remodeling and systemic glucose tolerance. Nat Commun, \n2021. 12(1): p. 3320. \n9. in Guide for the Care and Use of Laboratory Animals. 2011: Washington (DC). \n10. Fernandez-Valdivia, R., et al., A mouse model to dissect progesterone signaling in the \nfemale reproductive tract and mammary gland. Genesis, 2010. 48(2): p. 106-13. \n11. Kong, X., et al., IRF4 is a key thermogenic transcriptional partner of PGC-1alpha. Cell, \n2014. 158(1): p. 69-83. \n12. Sun, K., et al., Brown adipose tissue derived VEGF-A modulates cold tolerance and energy \nexpenditure. Mol Metab, 2014. 3(4): p. 474-83. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\n13. Perl, A.K., et al., Early restriction of peripheral and proximal cell lineages during \nformation of the lung. Proc Natl Acad Sci U S A, 2002. 99(16): p. 10482-7. \n14. Bickel, P.E., et al., Flotillin and epidermal surface antigen define a new family of \ncaveolae-associated integral membrane proteins. Journal of Biological Chemistry, 1997. \n272(21): p. 13793-13802. \n15. Bartelt, A., et al., Brown adipose tissue activity controls triglyceride clearance. Nature \nMedicine, 2011. 17(2): p. 200-U93. \n16. Benador, I.Y., et al., Mitochondria Bound to Lipid Droplets Have Unique Bioenergetics, \nComposition, and Dynamics that Support Lipid Droplet Expansion. Cell Metab, 2018. \n27(4): p. 869-885 e6. \n17. Price, N.L., et al., SIRT1 Is Required for AMPK Activation and the Beneficial Effects of \nResveratrol on Mitochondrial Function. Cell Metabolism, 2012. 15(5): p. 675-690. \n18. Wieckowski, M.R., et al., Isolation of mitochondria-associated membranes and \nmitochondria from animal tissues and cells. Nat Protoc, 2009. 4(11): p. 1582-90. \n19. Kien, B., et al., Lipid droplet-mitochondria coupling via perilipin 5 augments respiratory \ncapacity but is dispensable for FA oxidation. J Lipid Res, 2022. 63(3): p. 100172. \n20. Pereira, R.O., et al., OPA1 deletion in brown adipose tissue improves thermoregulation \nand systemic metabolism via FGF21. Elife, 2021. 10. \n21. Miner, G.E., et al., PLIN5 interacts with FATP4 at membrane contact sites to promote \nlipid droplet-to-mitochondria fatty acid transport. Dev Cell, 2023. 58(14): p. 1250-1265 \ne6. \n22. Boutant, M., et al., Mfn2 is critical for brown adipose tissue thermogenic function. EMBO \nJ, 2017. 36(11): p. 1543-1558. \n23. Cui, L., et al., Lipid droplets and mitochondria are anchored during brown adipocyte \ndifferentiation. Protein Cell, 2019. 10(12): p. 921-926. \n24. Bartholomew, S.R., et al., Distinct cellular pools of perilipin 5 point to roles in lipid \ntrafficking. Biochim Biophys Acta, 2012. 1821(2): p. 268-78. \n25. Wolins, N.E., D.L. Brasaemle, and P.E. Bickel, A proposed model of fat packaging by \nexchangeable lipid droplet proteins. FEBS Lett, 2006. 580(23): p. 5484-91. \n26. Dorn, G.W., 2nd, Mitochondrial dynamics in heart disease. Biochim Biophys Acta, 2013. \n1833(1): p. 233-41. \n27. Ikeda, K. and T. Yamada, UCP1 Dependent and Independent Thermogenesis in Brown \nand Beige Adipocytes. Front Endocrinol (Lausanne), 2020. 11: p. 498. \n \n \n \n \n \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nFigure legends \nFigure 1. Validation and generation of BiKOPLIN5 mice \n(a) Schematic representation for the strategy for doxycycline-inducible disruption of \nPlin5 gene. (b) Western blot (WB) depicting PLIN5 in BAT (left panel) and quantification \n(right panel) and (c) iWAT (left panel) and quantification (right panel) in BiKOPLIN5 and \ncontrol mice housed at 23 oC or 6 oC for 7 days. (d) WB depicting PLIN5 from liver (left \npanel) and heart (right panel) of Control or BiKOPLIN5 mice. For WB experiments, mice \nwere housed at the indicated temperatures and administered Dox for 7 days.” \n \nFigure 2. BiKOPlin5 mice exhibit reduction in thermogenic gene expression in \nBAT \n(a) Quantitative polymerase chain reaction (qPCR) for the indicated genes from BAT or \n(b) iWAT from Control or BiKOPlin5 mice housed at 6 oC for 7 days. (c) qPCR for the \nindicated genes from BAT or (d) iWAT from control or BiKOPlin5 mice housed at 6 oC \nfor 21days. (n=4 per group). Statistical analysis was performed using two-way ANOVA \nfollowed by Tukey post-test for multiple comparisons. Significant p values (p < 0.05) are \nshown in the figure. \n \nFigure 3 BiKOPLIN5 showed reduced cold tolerance but not glucose intolerance \n(a) Body weight and (b) food intake of Control and BiKOPLIN5 mice fed a chow diet \ncontaining 600 mg/kg Dox and exposed to cold for 7 days (n = 18 per group). (c) Oral \nglucose tolerance tests (OGTT) in males and (d) females Control and BiKOPLIN5 mice. \nFor OGTT studies, mice were fed a 60% HFD and housed at 23 °C for 8 weeks, then \nswitched to a 60% HFD containing 600 mg/kg Dox and housed at 6 °C. OGTT was \nperformed on day 8 of Dox treatment and cold exposure (n = 10–14 per group). (e) Cold \ntolerance tests in male and (f) female Control and BiKOPLIN5 mice. (g) Body weight \nchanges during the cold tolerance test in males and (h) females. For cold tolerance \nassays, n = 18–19 for males per group and n = 6 for females per group. Statistical \nanalysis was performed using an unpaired Student’s t-test. Significant p-values (p < \n0.05) are indicated in the figure. \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nFigure 4 BiKOPLIN5 showed reduced fatty acid uptake but not glucose uptake in \nBAT \n(a) Fatty acid and (b) glucose uptake on the indicated tissues from BiKOPLIN5 or \nControl mice. For this assay the mice were fed 600 mg/kg diet chow Dox and housed at \n6 oC for 7 days. Assay was performed on day 8. n= 6 per group. Statistical analysis was \nperformed using unpaired Student t test. Significant p values (p < 0.05) are shown in the \nfigure. (c) Representative image of hematoxylin and eosin staining from BAT at 10x \nmagnification or (d) 40x magnification from control or BiKOPLIN5 mice.  \n(e) Representative image of hematoxylin and eosin staining from iWAT at 10x \nmagnification from control or BiKOPLIN5 mice. (f) Representative image of hematoxylin \nand eosin staining from liver at 10x magnification from control or BiKOPLIN5 mice. \nScale bar is shown in the figure. n=3 per group.  \n \nFigure 5 Plin5 deletion in BAT causes mitochondrial dysfunction \n(a) BAT electron micrograph from Control (left panel) or BIKOPLIN5 (right panel) mice \nhoused at 6 oC and fed with 600 mg/kg Dox diet for 7 days. Top panel is 800x and lower \npanel is 5000x magnification. M=mitochondria, LD=lipid droplets. (b) Electron \nmicrograph quantification of mitochondria in contact with lipid droplets (top panel), \nmitochondria-lipid droplet contact/mitochondria perimeter (middle panel) and \nmitochondria lipid droplet contact/lipid droplet perimeter (bottom panel). n=10 EM fields \nper group. Statistical analysis was performed using unpaired Student t test. Significant p \nvalues (p < 0.05) are shown in the figure. (c) BAT Mitochondrial DNA quantification from \ncontrol or BiKOPLIN5 mice housed at 6 oC and fed with 600 mg/kg Dox diet for 7 days \n(d) Oxygen consumption rate (OCR) from BAT mitochondria isolated from control or \nBiKOPLIN5 mice housed at 6 oC and fed with 600 mg/kg Dox diet for 7 days. \nMitochondria were sequentially injected with Pyruvate, GDP, ADP, Oligomycin and \nFCCP n=3 per group (d). Statistical analysis was performed using unpaired Student t \ntest. Significant p values (p < 0.05) are shown in the figure. \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nFigure 6 PLIN5 localize on the outer mitochondrial membrane in BAT \n(a) Western blot for PLIN5, Actin, TIM23 and TOM20 from whole cell lysate and pure \nmitochondria with and without treatment with Proteinase K from control or BKOPLIN5 \nmice housed at 23 or 6 oC for 16 hours.  \n \nFigure 7 Working model of Perilipin 5 acute and chronic knockout in BAT  \n(a) Both acute and chronic PLIN5 loss led to impaired mitochondrial structure and \nfunction in BAT. However, the physiological consequences differ depending on the \nduration of PLIN5 depletion. In the acute setting, mitochondrial damage occurs before \ncompensatory mechanisms can be engaged, resulting in pronounced cold intolerance. \nIn contrast, chronic PLIN5 knockout also causes mitochondrial dysfunction but allows \ntime for compensatory thermogenic activation in iWAT, which maintains thermogenesis \nand prevents overt cold intolerance. PLIN5-Control mice, cKOPLIN5-Constitutively KO \nmice (BKOPLIN5), iKOPLIN5-inducible KO mice (BiKOPLIN5) \n \n \n \n \n \n \n \n \n \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nFigure 1\n23 6\nBiKO\nPlin5\nBiKO\nPlin5\nPLIN5\nGDI\nCtl Ctl\nTemp\nO C\n50\n50\n37\n75\niWAT\nM\nPLIN5\nGDI\nTemp\nO C\n50\n50\n37\n75\nM BiKO\nPlin5\nBiKO\nPlin5\nCtl Ctl\n23 6\nPLIN5/GDI\n23 6Temp\nO C\nPLIN5/GDI\n23 6Temp\nO C\na\nb\nc d\nControl\nBiKOPLIN5\n50\n50\nPLIN5\nActin\nLiver\nControl BiKOPLIN5\n50\n50\nPLIN5\nGDI\nHeart\nControl BiKOPLIN5\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nControl\nBiKOPLIN5\n0\n1\n2\n3\n4\n5\np < 0.0001\np < 0.0001Plin5\n23 o C 6 o C\n0\n1\n2\n3\n4\n5\np < 0.0001\np < 0.0001\nUcp1\n23 o C 6 o C\n0\n1\n2\n3\n4\np < 0.0001\np = 0.0019\np < 0.0001\nDio\n23 o C 6 o C\n0\n2\n4\n6 p = 0.0001\np = 0.0076 23 o C 6 o C\nElov3\n0\n1\n2\n3\n4\np < 0.0001\np < 0.0001\n23 o C 6 o C\nPpargc\nRelative gene \nexpression \n(normalized to \n18S)\n0\n2\n4\n6\n8 p < 0.0001\np < 0.0001\n0\n5\n10\n15\np < 0.0001\np < 0.0001\n0\n5\n10\n15\np < 0.0001\np < 0.0001\n0\n5\n10\n15 p < 0.0001\np < 0.0001\n0\n2\n4\n6 p = 0.0057\np = 0.0002\nPlin5 Ucp1 Dio Elov3 Ppargc\n23 o C 6 o C 23 o C 6 o C 23 o C 6 o C 23 o C 6 o C 23 o C 6 o C\n7 days at 6 °C \n21 days at 6 °C \niWAT\niWAT\nBAT\nBAT\nPlin5 Ucp1 Dio Elov3 Ppargc\nPlin5 Ucp1 Dio Elov3 Ppargc\n23 o C 6 o C 23 o C 6 o C 23 o C 6 o C 23 o C 6 o C 23 o C 6 o C\n23 o C 6 o C 23 o C 6 o C 23 o C 6 o C 23 o C 6 o C 23 o C 6 o C\n0\n5\n10\n15 p < 0.0001\np < 0.0001\n0\n5\n10\n15\np = 0.0015\np = 0.0002\n0\n5\n10\n15 p = 0.0096\n0\n1\n2\n3\n4\n5 p = 0.0236\np < 0.0001\np < 0.0001\n0\n1\n2\n3\n4\n5 p = 0.0013\np < 0.0001\np < 0.0001\n0\n2\n4\n6\n8\n10\np < 0.0001\np < 0.0001\np = 0.0451\n0\n5\n10\n15\n20\np < 0.0001\np < 0.0001\n0\n5\n10\n15\n20 p < 0.0001\np < 0.0001\n0\n5\n10\n15\np < 0.0001\np < 0.0001\n0\n1\n2\n3\n4\np < 0.0001\np < 0.0001\np = 0.0082\na\nb\nc\nd\nFigure 2\nRelative gene \nexpression \n(normalized to \n18S)\nRelative gene \nexpression \n(normalized to \n18S)\nRelative gene \nexpression \n(normalized to \n18S)\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nBody \ntemp (o C)\nHours fasting \n0 2 4 6 8\n32\n34\n36\n38\np=0.0344\nControl\nBiKOPLIN5\nWeight\n loss \n(grams)\n-1.5\n-1.0\n-0.5\n0.0\np= 0.0003\n0\n2\n4\n6\nFood \nintake\n(grams/day)\nBody \nweight \n(grams)\nDays on cold \nand Dox diet \n0 7\n0\n10\n20\n30\nBlood \nglucose\n(mg/dl)\nTime (min)\na b c\ne f g h\n-1.5\n-1.0\n-0.5\n0.0\np=0.0009\nFigure 3\nWeight\n loss \n(grams)\nHours fasting \n0 2 4 6 8\n32\n34\n36\n38\n0.0392\n0.0032 0.0003\nBody \ntemp (o C)\n0 15 30 60 120\n0\n100\n200\n300\n400\n0 15 30 60 120\n0\n100\n200\n300\n400\nTime (min)\nBlood \nglucose\n(mg/dl)\nd\nMale\nFemaleMale\nFemale\nFemaleMale\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nDPM/\nmg of tissue\n0\n500\n1000\n1500 p=0.17\np=0.35 p=0.35\np=0.40\np=0.17\nBAT gWAT LiverHeartiWAT\nFatty acid uptake\nGlucose uptakea\n0\n40000\n80000\n120000\np=0.02\np=0.18 p=0.36\np=0.23\np=0.27\nBAT gWAT LiverHeartiWAT\nFigure 4 Control\nBiKOPLIN5\nDPM/\nmg of tissue\n200 µm\n 200 µm\n200 µm\n 200 µm\n 200 µm\n 200 µm\n50 µm\n 50 µm\nBAT BAT\niWAT Liver\ne\nc d\nf\nb\nControl BiKOPLIN5 Control BiKOPLIN5\nControl BiKOPLIN5 Control BiKOPLIN5\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nd\n0.0\n0.5\n1.0\n1.5\nPyruvate\nGDP\nADP\nOligo\nFCCP\nµmol/l/min/\nµg protein\n1 2 3 4 5 6 7 8 9 10 11 12\np<0.0001\np<0.0001\nTime (min)\nc\n0.0\n0.5\n1.0\n1.5 0.0821\nmtDNA \nRelative amount\nControl\nBiKOPLIN5\n0\n20\n40\n60\n% Mitochondria\n in \ncontact \nwith LD\n0\n5\n10\n15\n0\n1\n2\n3\n4\nMitochondria-LD\n contact/\nMito perimeter (%)\nMitochondria-\nLD contact/\nLD perimeter (%)\nFigure 5\nba\nControl BiKOPlin5\n10 µm\n 10 µm\n1 µm 1 µm\nControl BiKOPlin5\nM\nLD\nLD\nM\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nFigure 6\na\nNo specific \nPure \nmitochondria\nPure \nMitochondria\n+ Proteinase K\nWhole\ncell\nlysate\nM 6 23 6 6 23 6 6 236 6 23 6 \nControl KO Control KO KO KO\nPLIN5\nTOM20\nTIM23\nActin\nM\nControl Control\n23 6 6 \nKOControl\n50\n75\n25\n35\n15\n25\n15\n50\n35\nTemp\nO C\n( )\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint \n\nFigure 7\na\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 1, 2026. ; https://doi.org/10.64898/2025.12.31.697198doi: bioRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}