A chaperone-assisted mechanism of outer mitochondrial membrane TOM70 insertion protective against diet-induced obesity

doi:10.21203/rs.3.rs-4831443/v1

Download PDF

Research Article

A chaperone-assisted mechanism of outer mitochondrial membrane TOM70 insertion protective against diet-induced obesity

https://doi.org/10.21203/rs.3.rs-4831443/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Outer mitochondrial membrane (OMM) proteins communicate with the cytosol and other organelles including the endoplasmic reticulum (ER)^1,2. This communication is important in thermogenic adipocytes to increase energy expenditure that controls body temperature and weight^3,4. However, the regulatory mechanisms of OMM protein insertion are poorly understood. Herein, stress-induced cytosolic chaperone PPID (peptidyl-prolyl isomerase D/cyclophilin 40/Cyp40) drives OMM insertion of the mitochondrial import receptor TOM70 regulating body temperature and weight in obese mice, and respiratory/thermogenic function in brown adipocytes. PPID PPIase activity and C-terminal tetratricopeptide repeats (TPR), which show specificity towards TOM70 core and C-tail domains, facilitate OMM insertion. Our results provide an unprecedented role for ER-stress-activated chaperones in controlling energy metabolism through a selective OMM protein insertion mechanism with implications in adaptation to cold temperatures and high-calorie diets.

General Cell Biology & Physiology

Mitochondria

Protein Import

Obesity

TOM70

PERK

Obesity is a highly prevalent preclinical condition⁵. Around 40% of individuals are already overweight and 13% are obese, including adults and children^5,6. Obesity is a risk factor for several serious diseases such as type 2 diabetes (T2D)⁷, cardiovascular disease⁸, cancer⁹, or emerging viral infections^10,11. Exercise and diet^12,13 and/or bariatric surgery¹⁴-based therapies do not always succeed due to genetic components, non-compliance, or excessive cost. Activation of thermogenic brown or brown-like adipose tissue is an attractive approach to counteract obesity and associated diseases^15,16. Upon cold exposure or exercise, activated BATs initiate a thermogenic program that oxidizes fatty acids, branched-chain amino acids, and glucose at high rates¹⁷ to maintain body temperature¹⁸, improve insulin sensitivity, and balance energy metabolism¹⁹.

BAT metabolic functions rely on mitochondrial-ER communications that create a unique landscape of protein-protein interactions, signaling and trafficking pathways to maintain adaptive cellular functions during diet-induced stress^3,4. Declines or failures in mitochondrial-ER communication can lead to the development of age-related diseases including obesity and T2D ^4,20–22. At the Outer Mitochondrial Membrane (OMM), protein complexes, including the mitochondrial protein and lipid import machineries are fundamental to maintaining the mitochondrial-ER communication, but the mechanisms of OMM insertion are not completely understood. At the ER, the PKR-like endoplasmic reticulum kinase (PERK) controls mitochondrial lipid^23–25 and protein import³ resulting in increased active cristae formation and oxygen consumption³ adjusting ER signaling to energy and metabolism^2,3,26. High-fat diets (HFD) induce ER stress in BATs^27,28 and activate thermogenesis²⁰ to preserve whole-body energy balance and insulin sensitivity¹⁸. However, the molecular mechanisms of mitochondrial-ER communication, especially the components that signal to and regulate the OMM, that sustain cellular energy metabolism during HFD stress in obesity and T2D are largely unknown.

Brown adipocytes represent a relevant and robust cellular model to investigate mitochondrial-ER communication because they rely on an extensive network of mitochondria with densely packed cristae that support thermogenic function, glucose homeostasis, and body weight^18,20,29 in response to nutrient or fatty acid overloads^20,27,30,31. To identify the molecular mechanisms that control mitochondrial-ER communication and respiratory and thermogenic activities, we initially utilized either Cre-driven adipocyte-specific (AdipoQ-Cre) or brown adipocyte-specific (UCP1-Cre) PERK^−/− mice, as a model of mitochondrial-ER miscommunication³, exposed to high-fat diet (HFD) stress²⁰. Both AdipoQ-Cre and UCP1-cre PERK^−/− mice showed increased body weight gain compared to thermoneutral-housed controls (28°C, TN) (Fig. 1a and Extended Data Figs. 1a-b). We also observed lower oxygen consumption rates in HFD-fed PERK^−/− mice chronically housed at thermoneutrality (Extended Data Fig. 1c) with no differences in food intake (Extended Data Fig. 1d), suggestive of an obese phenotype caused by defective brown adipocyte thermogenic function ^20,27,30,31. Accordingly, HFD-fed PERK^−/− mice displayed cold sensitivity compared to controls (Extended Data Fig. 1e). HFD-fed PERK^−/− mice were hyperglycemic and showed glucose intolerance (Fig. 1b and Extended Data Fig. 1f-g), elevated fasting blood glucose levels (Fig. 1c and Extended Data Fig. 1h), and insulin resistance (Fig. 1c and Extended Data Fig. 1i). Together, loss of PERK-dependent mitochondrial-ER communication^3,26,32 in thermogenic adipocytes leads to whole-body energy metabolism imbalances that result in diet-induced obesity, cold sensitivity, glucose intolerance, and insulin resistance.

Electron microscopy analyses were consistent with the loss of mitochondrial-ER communication³ and showed decreases in cristae density and evident cristae widening in HFD-fed PERK^−/− mice (Fig. 1e) reminiscent of reduced MIC19 and OPA1, respectively (Extended Data Fig. 2a)^3,33–36. Accordingly, reduced respiratory capacity was observed in HFD-fed PERK^−/− mitochondria (Fig. 1f and Extended Data Fig. 2b) with imbalanced NAD⁺/NADH ratios (Extended Data Fig. 2c and Extended Data Table 1). Therefore, PERK deficiency impaired mitochondrial-ER communication and portrayed a landscape of architectural and functional respiratory/thermogenic defects concomitant with obesity and glucose dysregulation.

PERK is a positive regulator of mitochondrial cristae architecture and cellular respiration^3,26,37 by controlling the mitochondrial protein import, especially the TOM70 pathway³. Mitochondria from HFD-fed PERK^−/− iBAT showed defective in vitro import of the TOM70 substrate ³⁵S-ANT2 (Adenine Nucleotide Transporter 2) but not the TOM20 substrate ³⁵S-OTC (Ornithine Transcarbamylase) (Figs. 1g-h)³ suggesting a TOM70-dependent defect. Paradoxically, TOM70 levels in isolated mitochondria were slightly higher in HFD-fed PERK^−/− mice. However, extraction experiments in carbonate buffer, pH 11.5, to remove loosely bound or improperly inserted proteins at the OMM, showed a decrease in the steady-state levels of TOM70 and carrier substrates such as ANT2 and SLC25A44 in HFD-fed PERK^−/− mice (Fig. 1i and Extended Data Fig. 2d), indicating that the TOM70 insertion at the OMM was compromised. We also corroborated that in vitro import of ³⁵S-TOM70 into isolated HFD-fed PERK^−/− mouse mitochondria was reduced compared to controls (Fig. 1j). The TOM70 pathway maintains mitochondrial bioenergetics by transporting mitochondrial carriers and proteins with internal mitochondrial protein sequences (MTS) in an unfolded state ^38–40. Defective import of ANT2 at the IMM (Fig. 1i) was also evidenced by BN-PAGE in mitochondria from HFD-fed PERK^−/− mice (Extended Data Fig. 2e) and is consistent with the cold-sensitive phenotype observed in vivo (Extended Data Fig. 1e)^3,41. Further functional and experimental evidence implicated SLC25A44, an IMM carrier that was recently characterized as a mitochondrial branched-chain amino acid (BCAA) transporter in iBAT sustaining the high oxidative rates of BCAA, body weight, and glucose balance²². Consistent with lower insertion of SLC25A44 (Fig. 1i), leucine-isoleucine levels were elevated²² in whole iBAT tissues from HFD-fed PERK^−/− mice (Extended Data Figs. 2f-g, and Extended Data Table 1). Additionally, in vitro ¹⁴C-amino acid uptake assays showed decreases in leucine, but not glutamine, incorporation (Figs. 1k and Extended Data Fig. 2h) in mitochondria from HFD-fed PERK^−/− mice. These results indicate that mitochondrial-ER communication defects in HFD-fed PERK^−/− mice hampers the insertion of TOM70 at the OMM causing defective import of essential solute protein carriers that sustain iBAT mitochondrial thermogenic and energy metabolism functions^21,22.

The regulatory mechanisms of OMM protein insertion are still an unresolved question in cell biology^42,43. In yeast, substantial progress has identified pathways responsible for OMM protein biogenesis^44–46. Yet, in mammals, these mechanisms do not have orthologs and/or are largely unexplored in a pathophysiological context^42,44. Given the central role of TOM70 in regulating the import of essential mitochondrial solute carrier precursors and therefore mitochondrial energy metabolism⁴⁷, we investigated the mechanisms that sustain ER resident PERK-dependent TOM70 membrane insertion in the context of diet-induced obesity.

To identify regulatory proteins that control mitochondrial-ER communication driving TOM70 insertion in obese mice, we analyzed two proteomics datasets using whole iBAT and compared i) normal chow diet (NCD) vs. HFD and ii) HFD PERK^+/+ vs. HFD PERK^−/− mice. A total of 5036 proteins were commonly identified across both datasets and quantified annotating PERK-dependent proteomic changes opposing HFD (Extended Data Fig. 3a). Proteins sorted as PERK and not diet-dependent were determined by ratiometrically comparing HFD-specific and PERK-dependent protein abundance changes after multiple t-testing and FDR correction (q < 0.05) (Figs. 2a-b and Extended Data Fig. 3a). Protein network and Gene Ontology (GO) analyses suggested a very distinct mitochondrial signature in PERK^−/− iBAT with 57% (119/206) of total upregulated proteins pertaining to this organelle and decreases in cytosolic proteins particularly, chaperones with folding capacity (CHORDC1, FKBP5, PDCL3, PFDN6, and PPID) (Fig. 2c). Therefore, loss of PERK results in decreases of cytosolic chaperones that correlate with lower TOM70 membrane insertion, likely leading to upregulation of non-imported mitochondrial proteins in an attempt to sustain mitochondrial function during HFD feeding.

Next, using two previous independent mass spectrometry datasets analyzed for PERK-dependent signatures on purified mitochondria³, we observed that none of the cytosolic chaperones were detected except for FKBP5 and PPID, yet only PPID showed PERK dependency (Extended Data Figs. 3b-c)³. PPID (also known as CYP40) is a member of the peptidyl-prolyl isomerase family that consists of eight members named PPIA through PPIH⁴⁸. Cis/trans proline isomerization is a rate-limiting process⁴⁹ that shapes the proteome by conferring structural and functional properties to proteins, including chaperones⁵⁰. PPID was originally isolated in a complex with HSP90⁵¹ and is the only member of the family that contains tetratricopeptide repeats (TPRs) that recognize substrates and have chaperone functions⁴⁸. Despite the deep biochemical knowledge as an HSP90 partner^52–56, little is known about the cellular functions of PPID and no connections with mitochondria have been established to date. Western blot analyses confirmed the downregulation of PPID protein, but not its mRNA levels, in HFD-fed PERK^−/− mice (Fig. 2d and Extended Data Fig. 3d). Similarly, in differentiated brown adipocytes, PPID showed decreased steady-state levels in response to palmitic acid (PA) treatment upon either genetic or chemical PERK suppression, without changes in Ppid mRNA levels (Extended Data Figs. 3e-g). PPID downregulation correlated with lower TOM70 levels (Fig. 2e and Extended Data Fig. 3e). Conversely, ectopic expression of PPID restored TOM70 levels (Fig. 2e) and rescued the respiratory function from the declines caused by PA treatment and /or PERK suppression (Fig. 2f). These results suggest that in cells PPID regulates TOM70 and mitochondrial function.

We next interrogated whether PERK exerts control over PPID through the canonical eIF2α-ATF4 pathway that controls protein translation. Chemical inhibition of the PERK-eIF2α pathway with ISRIB⁵⁷ robustly inhibited its target ATF4 but did not modulate the levels of PPID even in the presence of PA (Extended Data Fig. 3h), suggesting that PERK effects are independent of the canonical eIF2α-ATF4 pathway. Given the absence of apparent Ppid mRNA levels in vivo and in cells across conditions (Extended Data Figs. 3d and 3g), we set up a series of protein analyses and cycloheximide chase experiments to test for PERK-dependent regulation. PERK suppression resulted in reduced PPID levels and higher protein ubiquitination that were more exacerbated when combined with PA treatment (Extended Data Figs. 3i-k), indicating a posttranslational control of PPID. Collectively, suppression of the ER-resident PERK downregulates the cytosolic chaperone PPID which correlates with lower TOM70 levels and decreases respiratory capacities. In contrast, increased PPID levels compensate for PERK suppression and restore TOM70 levels and cellular respiration, supporting the role of PPID in the control of mitochondrial function.

To evaluate PPID capabilities to regulate TOM70 insertion, in vitro insertion assays were performed using recombinant PPID or mitochondria from siPpid cells and ³⁵S-labeled TOM70. Exogenous addition of recombinant PPID resulted in higher in vitro mitochondrial incorporation of ³⁵S-TOM70 compared to BSA-supplemented controls (Fig. 2g). Consistently, in vitro insertion of ³⁵S-TOM70 was reduced in mitochondria from siPpid cells (Figs. 2h-i and Extended Data Fig. 3l) which also displayed defective in vitro import of the TOM70 substrates MIC19 and ANT2, but not the TOM20 substrate OTC (Extended Data Fig. 4a). In full agreement, Ppid silencing reduced the steady-state levels of TOM70, ANT2, and MIC19 in differentiated brown adipocytes, an effect that was more abruptly observed upon PA treatment (Fig. 2h, Extended Data Fig. 3l). Similar to siPpid cells, reductions in the TOM70-dependent import pathway were also observed in PERK-deficient cells (Extended Data Fig. 4a) and in vivo (Fig. 1j). We also analyzed the protein assembly of TOM40, TIM22 and TIM23 that was apparently intact under these experimental conditions while TOM70 assembly was diminished in the PPID and PERK knocked-down cells (Extended Data Fig. 4b) without changes in mitochondrial membrane potential (Extended Data Fig. 4c). High-resolution respirometry evidenced mitochondrial bioenergetic defects caused by PPID suppression, including during PA treatment (Fig. 2j). These results show that PPID assists TOM70 receptor insertion in the OMM promoting mitochondrial protein import and respiration.

To further understand the molecular basis of PPID and TOM70 interactions, we first determined whether direct physical interactions would allow PPID-assisted TOM70 protein membrane insertion. PPID contains two main ordered domains, the peptidyl-prolyl isomerase (PPIase) domain at the N-terminal and tetratricopeptide repeats (TPRs) at the C-terminal that recognize substrates^53,54. While the PPIase domain is very conserved across the PPIase family, other domains diverge across them, being PPID the only member containing TPRs⁴⁸. Structure-sequence analysis of PPID across the mouse proteome revealed domain similarities with other chaperones such as FKBP4, TOM34, or DNAJC3 and also TOM70 (Extended Data Figs. 5a-b) with a common sequence motif CX₁₄(F/Y)R (Extended Data Figs. 5b-c) highly conserved across HSP90 partners with TPRs⁵⁶. PPID binding to HSP90 C-terminal domain (CTD) was tested in wild-type, C296S, R312A TPR substitutions and H141A PPIase defective mutant (Figs. 3a and Extended Data Fig. 5d)⁵⁸. Pull-down affinity assays showed that PPID^R312A mutant completely suppressed PPID binding to HSP90 CTD (Extended Data Figs. 5e-f)⁵⁶, but the C296S substitution decreased binding to HSP90 CTD by 50%, suggesting that both mutations at the TPR domain result in different levels of dysfunctional binding. Similar pull-down experiments using ³⁵S-TOM70 and recombinant his-tagged PPID revealed that only the PPID^C296S substitution suppressed binding to TOM70 by 50% (Fig. 3a), indicating that TPR modulates substrate specificity. In isolated mitochondria, TOM70 insertion was enhanced by exogenous PPID^WT and PPID^{R312A 56}, with defective binding to HSP90, but unchanged in the presence of PPID^H141A and PPID^C296S (Fig. 3b). Likewise, the presence of PPID^WT robustly augmented the insertion of ³⁵S-TOM70 in isolated liposomes while PPIase activity inhibition using cyclosporin A and PPID^H141A and TPR mutant PPID^C296S arrested membrane incorporation (Fig. 3c and Extended Data Figs. 5g-h) revealing a mechanism that recruits both PPID PPIase and TPR activities. Next, we asked how specific PPID is towards TOM70 and its insertion. Liposome incorporation assays indicated that PPID was able to robustly insert TOM70 compared to recombinant chaperones HSP90 and HSC70 (Figs. 3d and Extended Data Fig. 5i) involved in other mitochondrial protein import pathways^59,60. Therefore, PPID binding to TOM70 is specific and independent from HSP90 interaction.

To further understand PPID-TOM70 interactions, complexes were assembled from recombinant proteins (see Methods), mildly cross-linked with succinimidyl 4-[N-maleimidomethyl]cyclohexane-1- carboxylate (SMCC), which stabilizes TOM70 dimers⁶¹, and subsequently separated by either size exclusion chromatography or glycerol gradient ultracentrifugation (Extended Data Figs. 6a-c). Immunoblot assays suggested a comigration pattern for PPID and TOM70 across fractions (Extended Data Figs. 6c-d) at ~ 100, ~150, and ~ 200 kDa. Particle masses were confirmed by light scattering mass photometry (114±8.2, 157±28, and 232±55 kDa) (Extended Data Figs. 6e). Importantly, PPID^C296S, with impaired TOM70 binding and membrane insertion (Figs. 3a-c), displayed reduced stabilization of PPID-TOM70 assemblies compared to PPID^WT and TPR mutant PPID^R312A (Extended Data Fig. 6f). This was especially evident from the fact that ~ 100 and ~ 150kDa assemblies and free PPID sedimented at earlier fractions, and are relatively more abundant, with distinct complex distributions (Extended Data Fig. 6f). For instance, fraction 2 shows clear enrichment in ~ 150 and ~ 200 kDa complexes in PPID^WT and PPID^R312A with almost no ~ 100 kDa assembly. However, PPID^C296S shows a higher representation of the ~ 100kDa complex in the same fraction 2 (Extended Data Fig. 6f) further corroborating that C296 participates in the stabilization of PPID-TOM70 complexes (Figs. 3a-c). To confirm the presence of PPID-TOM70 complexes, cross-linking mass spectrometry (XL-MS) experiments were performed. XL-MS unequivocally showed that the C-terminal TPR domain of PPID, which binds to HSP90⁵², interacts with the C-terminal (core and C-tail) domains of TOM70, where TOM70 substrates bind⁶², with minor contacts with the N-terminal chaperone clamp domain that binds to HSP90 (Figs. 3e). PPID-TOM70 physical interactions were also observed in differentiated brown adipocytes using immunoprecipitation assays (Fig. 3f). Therefore, biochemical and biophysical approaches confirmed that multiple PPID molecules can simultaneously bind to the C-terminal domain of TOM70. These interactions suggest a stabilizing TPR-TPR mechanism at the substrate binding sites in TOM70 while leaving the membrane anchoring N-terminal domain available to target its destination at the OMM (Figs. 3e). Collectively, PPID-mediated TOM70 insertion is selective and requires both TPR contacts to stabilize the OMM receptor and PPIase activity, a dual mechanism which has been observed in bacterial proteins targeted to the periplasmic membranes⁶³.

In vivo loss of PPID causes obesity, glucose intolerance, and mitochondrial dysfunction

PPID mediates TOM70 OMM insertion by direct physical interaction regulating mitochondrial function. Yet, the in vivo implications of PPID in HFD-fed mice remained elusive. To address this, adipocyte-specific PPID^−/− mice were generated using the EasiCRISPR approach⁶⁴ (Extended Data Figs. 7a-c). Crossing PPID^fl/fl mice with AdipoQ-Cre mice (Extended Data Fig. 7a) downregulated PPID (PPID^−/−) in adipose tissues (Extended Data Fig. 7c). Similar to PERK^−/− mice, HFD-fed PPID^−/− mice showed metabolic disturbances (Figs. 4a-c and Extended Data Figs. 7d-i) associated with a diabetic phenotype; a significant body weight gain (Fig. 4a and Extended Data Fig. 7d) with increased adiposity (Extended Data Fig. 7e), glucose intolerance (Fig. 4b), fasting hyperglycemia (Extended Data Fig. 7f), hyperinsulinemia (Fig. 4c) and increased circulating free fatty acids (Extended Data Fig. 7g). Energetic and metabolic defects were further supported by decreased whole-body oxygen consumption rates in HFD-fed PPID^−/− mice, without changes in food intake (Extended Data Figs. 7j-l), and the limited capacity to activate whole-body oxygen consumption through the β₃-adrenergic agonist CL316,243 under thermoneutral conditions (Extended Data Fig. 7m). In an acute thermal cold challenge, a cold-sensitive phenotype was observed in PPID^−/− lean mice (Extended Data Fig. 7n). These results suggest that similar to PERK deficiency (Fig. 1 and Extended Data Figs. 1–2), in vivo adipose-specific ablation of PPID alters body temperature and weight as well as glucose and lipid homeostasis upon HFD.

Next, tissue analyses from adipocyte-specific PPID^−/− mice revealed enlargement of iBAT and iWAT, but not liver (Extended Data Figs. 7o-p), consistent with higher adiposity detected by magnetic resonance imaging (Extended Data Fig. 7e). Hematoxylin-eosin staining evidenced whitening of iBAT tissue with larger lipid droplets (Fig. 4d) along with decreases in mitochondrial cristae density quantified by electron microscopy (Fig. 4e). PPID downregulation in iBAT showed decreases in steady-state protein levels of TOM70 and multiple carriers and reported substrates such as ANT2, SLC25A44, UCP1 and MIC19 (Fig. 4f and Extended Data Fig. 7q). Consistent with our previous results, in vitro ³⁵S-TOM70 insertion was reduced in mitochondria from PPID^−/− deficient brown adipocytes (Fig. 4g). These alterations are associated with decreases in mitochondrial respiration (Figs. 4h and Extended Data Fig. 7r) with almost no response to FCCP, suggesting that mitochondrial spare and adaptive capacity is highly limited in PPID^−/− iBAT. Together, PPID is an essential regulator of mitochondrial function by at least controlling the insertion of the OMM receptor TOM70 in vivo. Failing to stimulate PPID chaperone activities through defective PERK-dependent mitochondrial-ER communication and diet-induced obesity results in compromised mitochondrial protein import and respiratory/thermogenic function imbalances affecting body temperature, weight, and glucose homeostasis.

The prevalence of obesity and insulin resistance correlates with the onset of metabolic disturbances caused by high-calorie diets and limited energy expenditure^5,8,9. The activation of BAT thermogenesis during dietary HFD feeding stress protects against excessive body weight gain and exacerbated glucose intolerance (Figs. 1, 4, and Extended Data Fig. 1)²⁰. Therefore, identifying the components and mechanisms that increase respiratory/thermogenic functions that depend on mitochondrial-ER crosstalk in BAT is important for potential therapies to treat obesity and T2D^15,16. Our work shows unprecedented evidence of the physiological roles of cytosolic chaperone PPID as a critical molecular regulator of mitochondrial function at the mitochondrial-ER signaling axis during diet-induced obesity (Fig. 5). PPID serves as a surveillance chaperone effector that, in response to dietary stress, grants the OMM insertion of TOM70, controlling selective mitochondrial protein import (Figs. 1g-k), nutrient oxidation (Figs. 2 and 4) and mitochondrial respiration (Figs. 2f, 2j, 4h and Extended Data Fig. 7r) to maintain body weight and glucose homeostasis (Figs. 4a-c). Importantly, declines in TOM70 expression upon aging cause mitochondrial dysfunction and reduce lifespan³⁹. Therefore, the control of TOM70 insertion and function is a key checkpoint to sustain cellular and metabolic adaptive fitness. These findings support the therapeutical potential of chaperone targeting to treat metabolic diseases⁶⁵.

TOM70 protein insertion at the OMM is apparently independent from HSP90 binding (Figs. 3a-c and Extended Data Figs. 5e-f) and would be consistent with the fact that PPID activation occurs upon dissociation from the HSP90 complexes during stress⁶⁶. The insertion of OMM proteins involves three mechanisms, spontaneous insertion, chaperone-assisted⁴⁶, and/or insertase-assisted⁴². PPID TPR would stabilize TOM70 similar to cytosolic chaperones Ydj1 and Sis1 in yeast that deliver OMM precursors to Tom70⁴⁶ or cytosolic TTC1 in mammalian cells⁶⁷. Although PPID seems to be sufficient to drive the insertion of TOM70 into naked membranes (Figs. 4c-d), PPID could cooperatively work with recently reported OMM insertases MTCH1/2 to facilitate the delivery and insertion, respectively, of the receptor TOM70 from the mitochondrial-ER interspace to the OMM under conditions of cellular stress^67,68. Mechanistically, TOM70 insertion recruits in the first place the TPR domain of PPID (Figs. 3e-f) and secondly relies on its PPIase activity (Figs. 3a-c), a process that may involve multiple PPID molecules tethering TOM70 (Extended Data Fig. 6). Binding of PPID to TOM70 C-terminal domains would safeguard the receptor from substrate binding and activation^62,69 prior to its insertion while driving the transmembrane N-terminal domain to the OMM (Fig. 3e). Similarly in bacteria, SurA is a chaperone fused to two PPIase domains which folds and catalyzes the cis/trans isomerization of prolines from membrane proteins at the periplasmic space facilitating their insertion into membranes⁶³, likely representing an evolutionary conserved membrane insertion mechanism across kingdoms.

Collectively, the control of PERK-dependent mitochondrial-ER communication through the chaperone PPID addresses a fundamental question on the importance of inter-organelle signaling in preventing cellular dysfunctions and highlights the relevance of studying mitochondrial protein insertion mechanisms in pathophysiological contexts to understand cellular adaptive processes.

Cell culture

Immortalized brown preadipocytes (BAT1 cells) were obtained as described in previous work ⁷⁰. Cells were cultured in BAT1 media (DMEM media supplemented with 10% FBS and 1% penicillin/streptomycin, 1 mM pyruvate and 2 mM glutamine). Cells were plated and, 24 hours later, differentiated in differentiation media (BAT1 media with 500 µM IBMX (3-isobutyl-1-methylxanthine, Sigma I7018), 5 µM dexamethasone (Sigma D4902), 1 µM rosiglitazone. (Sigma R2408), 20 nM insulin (Sigma; I5500) and 1 nM triiodothyronine (T3, Sigma T2877). Media was replaced with maintenance media, BAT1 media supplemented with 20 nM insulin and 1 nM T3, every 2 days. Cells were fully differentiated at day 6. Cells were incubated at 37° C and 5% CO₂ in a humidified atmosphere. Gene silencing experiments were conducted on day 4 of differentiation. Cells were trypsinized and reverse-transfected using Lipofectamine RNAiMAX (ThermoFisher #13778150) following the manufacturer’s instructions and kept under maintenance media until day 6. RNAi were purchased from Integrated DNA Technologies for PPID (mm.Ri.Ppid.13.2 and mm.Ri.Ppid.13.3) and for PERK (mm.Ri.Eif2ak3.13.1 and mm.Ri.Eif2ak3.13.2). RNAi concentrations were 50 nM. For the overexpression of PPID, the mouse gene was cloned in a lentiviral plasmid (Addgene, plasmid #85133). Lentiviral particles were generated in HEK293T³ cells to transduce BAT1 cells that were selected with 5µg/mL blasticidin.

Mice

C57BL/6 mice were obtained from Jackson Laboratories. C57BL/6 PERK^loxP mice (023066 Jackson Laboratories) were crossed with adiponectin-cre (Adipoq-Cre) C57BL6 mice (010803, Jackson Laboratories) or UCP1-Cre (024670, Jackson Laboratories) to obtain adipocyte or brown adipocyte-specific PERK depleted mice (Adipoq Cre⁺/PERK^flox/flox). Adipoq Cre^-/- PERK^flox/flox littermates were used as controls. Unless indicated, mice were housed at 23° C and 12/12 h light/dark cycles with free access to food and water. Experiments started using 8-week-old mice.

C57BL/6 PPID^flox/flox mice were generated in this work using EasiCRISPR⁶⁴. Briefly, exon 4 of the mouse Ppid gene was targeted. Embryos were electroporated with gRNA molecules ((-) TGTTTGAGTACCTTTTGCTC TGG, and (+) TGTAACCCTACTCTAGTGCT AGG), recombinant Cas9 (PNA-Bio CP01) and ssDNA molecule (Synthego) that included flox sites flanking the targeted exon (TATTAGATGTCCTTAGTGTGGCATCTTCATCTTTAAGATGGAAGTGTAAGCATCTCTTGCAGATAGCCAGTTTAGAAATGGCCTGATTCTTGTTAAACTGACACTTAAAAACCTTTGTTCATTTAAGGTGGAGTGCTTAAAGTCAAACCAGAGCAAataacttcgtataatgtatgctatacgaagttatAAGGTACTCAAACATTTAGAAGTGACATTGATGTAAGAGAGTAACGTTTAGTAACCCCATCTGTTTTAAGGAAATTCATACTAGGCCTCTAACTTGTTTTTCTTTCAATGATTCTTATCTCTGCAGCATGATCGGGAGGGTTTGCTGAGCATGGCAAATGCAGGCCCCAATACGAATGGCTCTCAGTTCTTTATCACAACAGTTCCGACTCCTCATCTGGACGGGAAACATGTGGTATTTGGTCAGGTCATAAAAGGACTAGGTGTGGCAAGGACGCTTGAAAATGTAGAGGTGAATGGTGAAAAACCTGCCAAAGTAAGTAAAGTTACGGTGAAGGAAACATACAGCGTAGTTAGCATGTCGGAGACTGCAAGCCACAGGTCAGACAGTTGTTGCCTTAGACAAGTGGGAAAGAAGAAGGAGGCTGCTTTTGCTATAGGAGCCTGGACTGGCAGTAGGAGGCTGGATTGGCAACTTTTCTTAACTTTTTTTTTTTTTTTAAAGAAGGCCAAGAAGATTAGTTTAAAGGAGATATATTTATTTTTTTTCTTCAAGATCTGGCTTTCCTTGAATTGTGTAGTAGTCCTGTTGTGTAACCCTACTCataacttcgtataatgtatgctatacgaagttatTAGTGCTAGGATTAACTACCATGCCCAACAAAATGATTTTAATACTTATTTCAGCTTCAGTATGATCTAACCGTGATCTGTGTTAGATGGCCCGAGTCTTCCCATAACAGTGAGACTTTTGTGTCAGAAGTAATTAACTATGGGACACACTAGAAAG).

The resulting offspring (80 pups) were genotyped (EmeraldAmp GT PCR master mix, Takara) to detect incorporation of the transgene (270kb vs 240kb) using the following primers:

Forward: 5' GCCAGTTTAGAAATGGCCTGATTCT 3'

Reverse: 5' ATGCTCAGCAAACCCTCCCGATCAT 3'

Animals incorporating flox sites were sequenced using the following primers:

Forward: 5' AGGCCTATTTCTGATGGTGGGA 3'

Reverse: 5' TCCTGTGCACTGTCTGGTTGCT 3'

The PCR conditions were: Denaturing, 94°C, 2 min; 10 cycles: 94°C, 30s, 65°C, 20s, 68°C, 30s (decrease annealing by 0.5°C each cycle); 28 cycles: 94°C, 30s, 60°C, 20s, 72°C, 30s; 72°C for 10 minutes.

Mice were crossed until obtaining a homogeneous flox/flox population that was crossed with Adipo-Q cre mice as described above. Housing conditions were as described for PERK KO mice. All the procedures in this paper were performed in compliance with the IACUC-approved protocols at the Animal Research Facility (Beth Israel Deaconess Medical Center, Boston, MA, USA).

Antibodies

TOM70 (14528-1-AP), PPID (12716-1-AP), TIM23 (11123-1-AP), and MIC60 (10179-1-AP) were from Proteintech. TOM40 (sc-365467) was from Santa Cruz Biotechnology. ANT2 (14671/E2B9D), PERK (3192/C33E10), HSPA8 (D12F2), Actin (4967), ubiquitin (3936), ATF4 (11815) GAPDH (2118), Tubulin (2146) and HSP90 (C45G5) were from Cell Signaling Technology. Pan-OXPHOS (ab110413) and UCP1 (ab234430) antibodies were from Abcam. MIC19 (Chchd3) (PA531578) was from Life Technologies. TIM29 (HPA041858) was from Sigma. SLC25A44 was from a previous work²². Antibodies were used at a concentration of 1:200-1:1000.

High-fat diet in mice

8-week-old mice were fed with HFD with 60% kcal from fat (Research Diets, D12492i) for the times indicated across experiments (8–28 weeks). Mice were kept under thermoneutral conditions (30° C) or as indicated in the text and figure legends.

Cold exposure experiments

Acute cold exposure experiments using PERK^-/- mice were performed using 6-week-old lean mice or 16–28 week-old HFD-fed mice. Mice were acclimated at 30° C for 2 weeks, if not chronically housed at thermoneutrality, and then transferred to individual cages at 4° C. Rectal temperature was determined using a BAT-12 microprobe thermometer (Physitemp Bat-12 microprobe) at different time points.

Metabolic phenotypic recording

Phenotypic recording of physiological parameters was performed using a Comprehensive Laboratory Animal Monitoring System (Promethion, Sable Systems), with continuous readings. HFD and water were provided ad libitum. When indicated, the β₃-adrenergic agonist CL316,243 (Sigma, C5976) was intraperitoneally injected at a concentration of 1mg/Kg^3,21. Data was extracted using Promethion software and analyzed and curated using CalRapp⁷¹.

Glucose and insulin tolerance tests

Mice were fasted overnight (GTT) or 6 hours (ITT) before the experiment. Blood glucose levels were determined by glucometer in blood from the tail vein. Before snips, the tail ends were dipped into Bupivacaine (0.25%) for local anesthesia to reduce pain. Glucose (Sigma, G7021, 2g/Kg) or insulin (Sigma, I1882, 0.75U/Kg) were intraperitoneally injected and glucose levels were measured every 20 minutes using a glucometer.

Blood serum insulin, free fatty acids, adiponectin and leptin determination

Hormone levels in sera were determined in mice under a fed state using a commercial ELISA-based kits for insulin (Crystal Chem, 90080), leptin (Crystal Chem, 90030) and adiponectin (Crystal Chem, 80569) following the manufacturer’s instructions. Free fatty acids were determined from sera using a colorimetric assay (Abcam ab65341) following manufacturer’s instructions.

Determination of protein levels by western blot

Cells were harvested by scraping in cold PBS and, after centrifugation at 500 g, cells were resuspended in lysis buffer (50 mM HEPES, pH 7.5, 150 mM NaCl, 1% Triton X-100, 0.2% SDS, 1 mM β-mercaptoethanol and 1 mM EDTA). Lysis was performed on ice for 30 minutes. Protein was quantified using Bradford assays and denaturing Laemmli buffer was added accordingly. Samples were run in 4–12% NuPAGE gradient gels and proteins were transferred to 0.22 µm Immobilon P^SQ (Millipore) PVDF membranes. Membranes were incubated with the appropriate antibodies overnight at 4° C. Tissues were disrupted in lysis buffer using a Dounce homogenizer in the cold.

Palmitate treatment in BAT1 cells

Cells were differentiated as described above. When indicated, siRNAs were used. 100 mM palmitic acid (Sigma P9767) was prepared in 100 mM NaOH solution and heated at 70°C until dissolved. 100 mM NaOH was used as a control. DMEM/FBS medium was supplemented with 10% fatty acid-free BSA and incubated at 37°C for 30 minutes. Palmitic acid was quickly added to the DMEM/FBS/BSA medium and incubated at 37°C for 30 minutes. Solutions were stored at -20°C for 2 weeks and filter-sterilized before use. Palmitic acid was used at a concentration of 500 µM for 24 hours. When indicated other compounds were used: PERK inhibitor (GSK2606414, Millipore) at 500 nM, and eIF2⍺ inhibitor (ISRIB, Sigma, SML0843) at 250 nM.

Gene expression analysis

Total RNA was isolated using Trizol reagent (ThermoFisher Scientific, 15596026) following the manufacturer’s instructions. cDNA was generated using 2 µg of RNA and the Multiscribe Reverse Transcriptase (ThermoFisher Scientific, 4311235). For gene expression analysis, qPCR reactions were conducted using Sybr Green qPCR master mix (Applied Biosystems) and analyzed by a CFX 384 Real-Time system (Bio-Rad).

Electron microscopy

iBAT 1 mm thick slices were fixed for 1hr (0.5% Glutaraldehyde 2% Paraformaldehyde in 0.1 M sodium cacodylate buffer (pH 7.4), postfixed in 1% Osmium tetroxide (OsO4)/1.5% Potassium ferrocyanide (KFeCN6) at room temperature for 30 min. Samples were then washed in water 3x and incubated in 1% aqueous uranyl acetate for 30 min followed by 2 washes in water and subsequent dehydration in grades of alcohol (5 min each; 50%, 70%, 95%, 2x 100%). Cells were embedded in plastic by inverting a gelatin capsule filled with Epon/Araldite on top of the coverslip and polymerized at 60°C for 24 hours. Ultrathin sections (about 60 nm) were cut on a Reichert Ultracut-S microtome, picked up onto copper grids stained with lead citrate, and examined in a JEOL 1200EX Transmission electron microscope. Images were recorded with an AMT 2k CCD camera. Mitochondria images from 15 cells were analyzed using Fiji and the Trainable Weka Segmentation (TWS) plugin. Briefly, the model was trained using cristae and non-cristae cellular components as background. Gaussian blur, sobel filter and membrane projections were enabled. Images were analyzed using the generated model and cristae determined by applying a threshold to each image. Profiling of cristae across each mitochondrial section was used to define cristae abundance as peaks over distance. Controls were fitted to a normal distribution (mean±2σ) and compared to KO mouse cristae.

Mitochondria isolation

iBAT or differentiated cells were resuspended in 25 mM HEPES, pH 7.5, 75 mM sucrose, and 225 mM mannitol supplemented with 0.5 mM EGTA and phosphatase inhibitors (PhosStop, Roche). Cells were homogenized using a Dounce homogenizer performing 20 strokes. Lysates were centrifuged at 700 g for 5 minutes and supernatants containing mitochondria were transferred to a new tube. This step was repeated until no nuclear pellet was found after centrifugation. Supernatants were centrifuged at 7,500 g for 10 minutes and resuspended in the same isolation buffer three times to wash out contaminants. 1% BSA supplementation was used to remove contaminants in the first wash. Mitochondria were quantified using Bradford (BioRad, 5000201) and/or BCA (ThermoFisher, 23225) assays.

Mitochondria respiration assays

The protocol was adapted for Seahorse assays from previous works^3,72. Isolated mitochondria were prepared in MAS buffer (70mM Sucrose, 220mM Mannitol, 10mM KH₂PO4, 5mM MgCl₂, 2mM HEPES, 1mM EGTA, 0.2% BSA), pH 7.4, at a concentration of 0.4 mg/mL. 2 µg of mitochondria in 50 µL buffer was added per well in the Seahorse XFe24 plate system and spun down in the cold at 2,000 g for 20 min. Carefully, 450 µL of MAS buffer was added on top. Chemicals were added in the following order: Port A (pyruvate + malate at final concentrations of 10/6mM, respectively); Port B (GDP, 3 mM final concentration); Port C (Oligomycin at a concentration of 5 µM); and Port D (FCCP 5 µM). Each cycle consisted of 20-second mixing and 2-minute readings, with 2 readings. The maximal rate was defined as the maximal respiratory capacity upon pyruvate/malate injection. UCP1 activity was calculated as O₂^UCP1 = ΔO₂^pyr/mal-GDP=O₂^pyr/mal-O₂^GDP. Maximal FCCP respiration was calculated as O₂^FCCP = ΔO₂^{FCCP-oligomycin}=O₂^FCCP-O₂^oligomycin.

Proteomics analysis from whole iBAT tissue

Proteomics was performed as previously described^3,37. iBAT was homogenized in lysis buffer (4.0 % SS w/v, 250 mM NaCl, PhosStop (Roche) phosphatase inhibitors, EDTA free protease inhibitor cocktail (Promega), and 50 mM HEPES, pH 8.5), reduced with 5 mM TCEP at 60°C for 30 minutes, alkylated with 14 mM iodoacetamide for 45 minutes, and precipitated using trichloroacetic acid at a final concentration of 25%. Samples were centrifuged and pellets were washed three times with cold methanol. Proteins were resuspended in EPPS buffer and digested with Lys-C (1:100) for 4 hours at 37° C following trypsin digestion (1:100) at 37° C overnight. Peptides were quantified by microBCA reagent (Thermo Fisher Scientific, 23235) and TMT labeled (6-8M excess). Formic acid was added to a final concentration of 1% and peptides were cleaned up using a 50 mg Seppak column. Eluted peptides were dried down and resuspended in 5% ACN/5% formic acid buffer. Ratios were checked by HPLC and samples fractionated. Fractions were resuspended in 1% formic acid and cleaned up using C18-membrane stage tips. Samples were dried down and MS analysis was performed as previously described^3,37. The two independent datasets using whole iBAT were analyzed. One dataset compared iBAT from normal-chow diet and high-fat diet-fed mice and the other compared iBAT from PERK^+/+ and PERK^-/- HFD-fed mice. Datasets were independently processed and normalized to the control condition (NCD or WT mice). Normalized datasets were merged with 5036 common proteins. PERK-specific effects were defined by fold-change comparison between HFD-specific effects (HFD/NCD) and PERK effects (PERK^-/-/PERK^+/+) after multiple t-testing and FDR correction (q < 0.05).

High-resolution respirometry

BAT1 cells were differentiated as explained above on 6-well plates. Oxygen consumption was determined in an Oroboros O2k FluoRespirometer with a clark-type electrode. Cells were trypsinized, centrifuged at 500 g for 5 minutes, and resuspended in phenol red-free DMEM supplemented with 0.1% BSA. incubated at 37° C inside the chamber until the baseline was stable. Compounds were added to the following final concentrations: 5 µM oligomycin, 2 µM FCCP, and 3 µM rotenone/antimycin each.

Membrane potential determination

Differentiated cells were treated with either BSA or BSA-palmitate as described before on 12-well plates (Cellvis P12-1.5H-N). Prior to confocal imaging, differentiated adipocytes were incubated in the presence of TMRM (ThermoFisher, I34361) and MitoTracker Green (ThermoFisher, M7514) at a concentration of 25 and 250 nM, respectively, for 20 minutes at 37˚C and 5% CO2 to allow dyes to equilibrate. 10 µM carbonyl cyanide m-chlorophenyl hydrazone (CCCP) was added as a control to one replicate. Single Z-plane images were acquired using a Yokogawa CSU-X1 spinning disk confocal on an inverted Nikon Ti fluorescence microscope. MitoTracker Green was excited using a 488 nm laser at 20% power with a 150 ms exposure time; TMRM was excited using a 561 nm laser at 30% power for 105 ms. All images were acquired with the same settings.

Immunoprecipitation

Differentiated cells were treated with either BSA or BSA-palmitate as described before on 150 mm plates (3 per condition). Cells were washed three times with 20 mL of PBS and scraped. After centrifugation (2,000 g for 5 min), cells were lysed in IP wash buffer (50 mM HEPES, pH 8.0, 150 mM NaCl, and 1% triton) supplemented with 1% BSA, 5 mM MgCl₂, 1 mM ADP, protease inhibitors (cOmplete, Roche, 11697498001), and phosphatase inhibitors (PhosSTOP, Roche, 4906845001). Lysates were centrifuged at 20,000 g for 10 minutes and the supernatant collected, quantified by BCA assays and subjected to immunoprecipitation using anti-myc magnetic beads (Cell Signaling Technology, #5698).

Purification of recombinant ULP1, TOM70, PPID alleles, HSC70 and HSP90

All SUMO-tagged proteins were expressed in E. coli BL21 DE3 Star cells. bacteria were grown in TB medium containing 200 µg/mL ampicillin overnight at 37°C. 10 mL of this culture was poured into 1L of TB medium containing ampicillin. Cells were grown at 37°C until OD₆₀₀ reached 0.6–0.8. Cells were induced with IPTG to a final concentration of 1 mM. Cells were incubated at 37°C for 3 hours. HSP90-expressing bacteria were incubated at 16°C overnight. Bacteria were centrifuged at 6000 g for 10 minutes. Pellet was resuspended in Buffer A (25 mM HEPES, pH 7.5, 500 mM KCl, 20 mM imidazole, and 3 mM 2-mercaptoethanol) supplemented with cOmplete protease inhibitor, lysozyme, and DNase I. Cells were sonicated and lysates centrifuged at 40,000 g for 15 minutes. Supernatants were collected, filtered through a 0.45 µm syringe filter, and loaded into Co-NTA gravity columns pre-equilibrated in buffer A. Columns were washed with 20 column volumes of buffer A, and protein was eluted in Buffer B (same as A but containing 300 mM imidazole and 150 mM NaCl). Protein was then concentrated in an Amicon Ultra centrifugal filter and desalted in Buffer C (25 mM HEPES, pH 7.5, 100 KCl, and 3 mM 2-ME). Protein concentration was determined by Bradford assays. At this point, ULP1 was aliquoted in 50% glycerol buffer and preserved at -80°C. SUMO-tagged proteins were cleaved using ULP1 protease at a 1:20 (mg Protease: mg SUMO-protein) ratio overnight in the cold. ULP1 and SUMO tags were removed by loading the digested protein into a Ni-NTA column equilibrated in buffer A. Protein was collected in the flow-through. HSP90 was further purified by size exclusion chromatography (Superdex 200) in buffer C. Proteins were desalted in buffer D (25 mM HEPES, pH 7.5, and 3 mM 3-ME), concentrated, and kept at -80°C. Protein was quantified by Bradford assays.

Purification of recombinant HSP90 C-terminal domain (CTD)

HSP90 CTD was from Addgene (Plasmid #22483). E. coli BL21 Star cells were grown, harvested, and lysed under the conditions described above in 50 mM HEPES, pH 8.0, 150 mM NaCl. After clarification, supernatants were incubated with Sepharose 4b GST purification resin (Cytiva, 17075601) for 4h equilibrated in 50 mM HEPES, pH 8.0, 150 mM NaCl in the cold, 1 mM DTT and column washed 20 CV. CTD was then eluted in the same buffer but supplemented with 10 mM reduced glutathione. CTD was concentrated, desalted and cleaved with thrombin (Sigma, T6634). The cleaved protein was then incubated with GST beads in 50 mM HEPES, 150 mM NaCl for 4h in the cold and collected in the flow through. CTP was then desalted and in 50 mM HEPES, pH 8.0, quantified by Bradford assays.

In vitro PPIase assay

PPID isomerase activity was determined by the enzymatic hydrolysis of the peptide substrate N-succinyl-Ala-Leu-Pro-Phe-p-nitroanilide (Sigma, S7388). sALPF-pNA was prepared in 470 mM LiCl 2,2,2-trifluoroethanol (TFE) at 200 mM and diluted to 4 mM in the same buffer before use. Reactions were conducted in 50 mM HEPES, pH 8.0, 100 mM KCl, and 1 mM DTT with 15 nM enzyme. After 30 minutes in the cold, chymotrypsin was added to 50 µM and absorbance was monitored in continuous mode at 400 nm. The trans, but not the cis, conformation of sALPF-pNA is cleaved by chymotrypsin. Samples without PPID enzyme were used as a blank control.

Protein import of ³⁵S-labeled proteins into mitochondria

Freshly isolated mitochondria were resuspended in pre-import buffer (20 mM HEPES-KOH, pH 7.4, 250 mM sucrose, 80 mM KOAc, and 5 mM MgOAc). Protein was quantified at this point using the BCA reagent (ThermoFisher, 23225) following the manufacturer’s instructions. ³⁵S-labeled proteins were synthesized using the T7- based TnT rabbit reticulocyte system (Promega, L4610) following manufacturer’s instructions. Purified mitochondria (20–100 µg/time point) were pelleted and resuspended in 50–100 µL of import buffer (20 mM HEPES-KOH, pH 7.4, 250 mM sucrose, 80 mM KOAc, and 5 mM MgOAc, 0.1% BSA 1 mM ATP, 2 mM sodium succinate and 1 mM methionine). When indicated, membrane potential was dissipated using an FCCP (10 µM). Once mitochondria were resuspended, the TnT reticulocyte product was added to a final volume of 10% of the total reaction volume. Recombinant PPID or control BSA were added to a final concentration of 5 µM as indicated. Proteinase K was added at a final concentration of 50 µg/mL to remove non-imported precursors and reactions were incubated for 10 minutes on ice. Mitochondria were diluted in wash buffer supplemented with freshly added 0.1 mM PMSF to neutralize proteinase K and centrifuged at 12,000 g for 3 minutes. This wash step was repeated twice. For the insertion of TOM70, no proteinase K treatment was performed and instead, mitochondria were pelleted and incubated in 100 mM sodium carbonate buffer, pH 11.5, followed by ultracentrifugation (125,000, 30 minutes) and SDS or BN-PAGE. Samples transferred to PVDF membranes and autoradiographs were obtained using Typhon FLA 9500 phosphorimager (GE Healthcare).

Liposome insertion assay

Liposomes were prepared for protein insertion in a similar manner to previously published work⁶⁸ Phosphatidyl-choline (PC) and phosphatidyl-ethanolamine (PE) from egg, and 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-lissamine rhodamine B (Rh-PE) were all purchased from Avanti Polar Lipids. Lipids were mixed well as chloroform stocks to a PC:PE:Rh-PE mass ratio of 8:1.9:0.1 and then brought up to 10mM DTT. Rh-PE was added to aid in the quantification of liposome recovery. The lipid mixture was then dried by vacuum centrifugation overnight. The dried lipid film was rehydrated to a final concentration of 20 mg/mL in 50 mM HEPES, pH 7.5, 250 mM sucrose, 5 mM MgAc, and 80 mM KAc and mixed for 8 hours at 25°C. The lipid mixture was then diluted to 10mg/mL with the same buffer used for the insertion assay. TOM70 was synthesized using the TnT rabbit reticulocyte system. Reactions were incubated for 1h at 30°C. BSA, PPID, HSP90, or HSC70 were prepared to 50 µM in insertion buffer (50 mM HEPES, pH 7.5, 250 mM sucrose, 5 mM MgAc and 80 mM KAc, 1 mM DTT). When indicated cyclosporin A (CsA) was used at a final concentration of 100µM. Before the incubations, PPID was pretreated with the compound. Mixtures contained 2 mg/mL liposomes, 5 µM recombinant protein, and 10% TnT product in a total volume of 100 µL. Reactions were incubated at 37°C at the indicated time points. ³⁵S-TOM70 insertion was determined by sodium carbonate incubations (100 mM at pH 11.5) followed by centrifugation (230,000 g for 30 minutes) and SDS-PAGE.

Amino acid uptake assay

Freshly isolated mitochondria (0.25 mg/mL) were incubated mitochondria in Krebs–Ringer modified buffer (KRB)– HEPES buffer²², containing 2% BSA, 15 mM glucose, 200 nM adenosine, 1 mM MgOAc, and 2.5 µCi of ¹⁴C-Leu or ¹⁴C-Gln. No sodium salts were used. Mitochondria were washed three times in isotonic 250 mM sucrose buffer pH 7.4 and lysed in PBS supplemented with 0.2% SDS. Samples (100 µL) were transferred to a scintillation vial containing 3 mL of scintillation liquid and counts were determined.

Blue-Native PAGE electrophoresis

Mitochondria (100 µg) were solubilized in 1% digitonin in BN buffer for 30 minutes on ice. Samples were centrifuged at 18,000 g for 20 minutes and supernatant collected. Samples were run using 4–16% gradient gels according to the manufacturer’s instructions. 25 µg of protein was loaded per lane. After electrophoresis, complexes were transferred into PVDF membranes and tested against specific antibodies.

PPID pull-down affinity assays

BAT1 cells ectopically expressing PPID were differentiated on 10 cm dishes and treated with PA with or without PERK silencing as indicated. MG132 was added 4 hours before harvesting to a final concentration of 20 µM. PPID was pulled down using Co-NTA resin in 50 mM HEPES, pH 7.5, 500 mM NaCl, 1% Triton NP-40, 3 mM β-mercaptoethanol, 20 mM imidazole, protease inhibitors (cOmplete), phosphatase inhibitors (phostop). The resin was washed with 20 column volumes and protein eluted in Laemmli buffer. Samples were analyzed by SDS-PAGE and immunoblotting. PPID-CTD interactions were determined similarly using 10 µM PPID and 30 µM CTD.

Determination of PPID binding to CTD and ³⁵S-TOM70

Binding assays to CTD were performed by incubating PPID (5µM) with CTD (10µM) at 37°C for 1 hour in 50 mM HEPES, pH 8.0, 200 mM KCl, 5 mM MgCl_2, and 1 mM DTT. Co-NTA resin, equilibrated in the same buffer, was added and incubated at RT under rotation for 30 minutes. The resin was washed three times and proteins were eluted in a buffer supplemented with 500 mM imidazole. For the binding to ³⁵S-TOM70, the same protocol was followed but adding 10% of the reticulocyte ³⁵S-labeled protein.

Cycloheximide chase experiments

BAT1 cells (7.5x10⁵) were differentiated on 6-well plates as indicated above. On the day of the experiment, cycloheximide was freshly prepared in DMSO and added to a final concentration of 100µg/mL. Samples were taken at different time points, quantified by Bradford assays (BioRad, 5000201), and analyzed by SDS-PAGE.

Polar metabolomics

iBAT was powdered in liquid nitrogen and homogenized. Metabolites were extracted using 80:20% methanol:water solution at -80° C for 15 minutes. Extracts were centrifuged in the cold for 10 minutes at 18,000 g and supernatants were dried down in the cold. Samples were submitted to the Metabolomics Core at the Beth Israel Deaconess Medical Center for further polar metabolite detection by QTRAP® 5500 System and Reverse-Phase Ion-Pairing Chromatography.

Purification of PPID-TOM70 complexes

Purified TOM70 (40µM) and PPID (120µM) were incubated at 37°C for 1 hour in 50 mM HEPES, pH 8.0, 100 mM KCl, 1 mM MgCl₂, 0.5 mM TCEP. Complexes were stabilized by crosslinking with 1 mM SMCC at room temperature for 20 minutes. Reactions were quenched with 10 mM TRIS-HCl, pH 8.0, and 10 mM DTT. Complexes were separated in a Superdex 200 increase size exclusion chromatography column equilibrated in the same reaction buffer or by 10–30% glycerol gradients in a Beckman Coulter XL-Optima with rotor SW41Ti at 300,000 g for 18h. Fractions were analyzed by SDS-PAGE.

Light scattering mass photometry

Light scattering mass photometry analyses were performed on different samples containing glycerol-gradient separated complexes. Samples were diluted to 10 nM in 50 mM HEPES, pH 8.0, 100 mM KCl, 1 mM MgCl₂, 0.5 mM TCEP, and 1-minute movies with ~ 1,000–3,000 counts each were recorded using a Refeyn TwoMP mass photometer. Results were automatically fitted to a Gaussian distribution using the DiscoverMP 2.3 software.

Thermal shift assays

Purified PPID WT and H141A mutant (10µM) were mixed with either vehicle or 100µM cyclosporin A in the presence of SYPRO orange (ThermoFisher, S6650) at a 20x concentration. A thermal ramp (0.5°C/min) was generated using a BioRad CFX384 thermal cycler. Melting curves were fitted to the Boltzmann equation to determine T_m values.

Cross-linking mass spectrometry analyses

Complexes crosslinked with SMCC were denatured with 8 M urea. Cysteine residues were reduced with TCEP and alkylated with iodoacetamide. Urea was diluted to 2 M prior to a tryptic digest (1:50 ratio of trypsin:protein) overnight at 32°C. Digestions were quenched with the addition of 1% formic acid (FA) and peptides were dried by vacuum centrifugation and desalted over a C18 column. Digests were resuspended in 1% FA and analyzed on an Ultimate 3000 RSLCnano system coupled to a Fusion Lumos mass spectrometer (Thermo Scientific). Peptides were separated across a 60-min gradient of 6–30% acetonitrile in 1% FA over a 50-cm C18 column (ES903; Thermo Scientific) and electrosprayed (1.9 kV, 300°C) into the mass spectrometer with an EasySpray ion source (Thermo Scientific). Precursor ion scans (375–1325 m/z) were obtained in the orbitrap at 120,000 resolution in profile (RF lens % = 30). Fragment ion scans were acquired in the orbitrap at 30,000 resolution (0.5 m/z isolation window, HCD at 30% NCE).

Raw data were searched with pLink2 (http://pfind.org/software/pLink/) ⁷³ against a target-decoy database containing TOM70 and PPID construct sequences and pLink2 default contaminants with the following search parameters: mass accuracy of ±10 ppm for precursors and ±20 ppm for fragment ions, two missed cleavages by trypsin, oxidation of Met, carbamidomethylation of Cys, phosphorylation of Ser/Thr/Tyr, and crosslinking of Cys/Lys with SMCC. Search results were filtered to 1% FDR at the PSM level. Representative spectra from each crosslinked species in the resulting list were manually inspected. Any crosslinked spectra that did not contain multiple abundant fragment ions from each linked peptide or did contain multiple unexplained peaks were removed from the list.

Statistics

Graphs indicate mean ± SD. Two-sample t-tests were performed when comparing two samples. In case of repeated measurements, multiple testing correction or FDR was applied. One-way or two-way ANOVA tests were performed to determine differences between groups or calculate regression models, respectively, followed by multiple comparison-adjusted testing. For mouse metabolic experiments, data were analyzed using CalR⁷¹. Significances were considered if p ≤ 0.05. Statistics depict comparisons to controls unless indicated.

Deposited data

The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE⁷⁴ partner repository with the dataset identifiers (forthcoming for publication) (mass spectrometry data from mice) and PXD053607 (cross-linking mass spectrometry).

Competing interest

The authors declare no competing interests.

Author contributions

Conceptualization: PLM, PP. Methodology: PLM, TV, MJ, SPG, SK, HJ, MJE, PP. Investigation: PLM, TV, MR, KZ, JMP, CFB, ALP, JHS, TDJ, MJ, AS. Visualization: PLM. Funding acquisition: PLM, PP, CFB. Project administration: PLM, PP. Supervision: PP, PLM. Writing – original draft: PLM. Writing – review & editing: PLM, TV, CFB, SK, AS, HJ, PP.

Acknowledgments

NIH NIDDK grants R01DK081418, R01DK089883 and R01DK136640 (PP). NIDDK NIH award K99DK133502 (PLM). NIH grant 1T32GM145407-01 (TV). UMDF Accelerator Award (JMP). F32 GM125243 (NIGMS) and the Charles A. King Trust Postdoctoral Fellowship Program (CFB). American Heart Association Postdoctoral Fellowship (24POST1198357) (JHS). Cancer Research Institute Irvington Postdoctoral Fellowship (CRI5298) (TDJ).

Data availability

All the data and materials used in this work will be available upon reasonable request to the corresponding author.

Ronayne CT, Latorre-Muro P (2024) Navigating the landscape of mitochondrial-ER communication in health and disease. Front Mol Biosci 11:1356500
Bennett CF, Latorre-Muro P, Puigserver P (2022) Mechanisms of mitochondrial respiratory adaptation. Nat Rev Mol Cell Biol 23:817–835
Latorre-Muro P et al (2021) A cold-stress-inducible PERK/OGT axis controls TOM70-assisted mitochondrial protein import and cristae formation. Cell Metab 33:598–614e7
López-Crisosto C et al (2015) ER-to-mitochondria miscommunication and metabolic diseases. Biochim Biophys Acta Mol Basis Dis 1852:2096–2105
Barnes AS (2011) The epidemic of obesity and diabetes: Trends and treatments. Tex Heart Inst J 38:142–144
Wang Y et al (2021) Has the prevalence of overweight, obesity and central obesity levelled off in the United States? Trends, patterns, disparities, and future projections for the obesity epidemic. Int J Epidemiol 49:810–823
Kalyani RR, Golden SH, Cefalu WT (2017) Diabetes and aging: Unique considerations and goals of care. Diabetes Care 40:440–443
Chia CW, Egan JM, Ferrucci L (2018) Age-related changes in glucose metabolism, hyperglycemia, and cardiovascular risk. Circ Res 123:886–904
Luo C et al (2020) Obesity/Type 2 diabetes-associated liver tumors are sensitive to cyclin D1 deficiency. Cancer Res 80:3215–3221
Nakeshbandi M et al (2020) The impact of obesity on COVID-19 complications: a retrospective cohort study. Int J Obes 44:1832–1837
Al-Sabah S, Al-Haddad M, Al-Youha S, Jamal M, Almazeedi S (2020) COVID-19: Impact of obesity and diabetes on disease severity. Clin Obes 10:e12414
Fock KM, Khoo J (2013) Diet and exercise in management of obesity and overweight. J Gastroenterol Hepatol (Australia) 28:59–63
Magkos F, Hjorth MF, Astrup A (2020) Diet and exercise in the prevention and treatment of type 2 diabetes mellitus. Nat Rev Endocrinol 16:545–555
Wolfe BM, Kvach E, Eckel RH (2016) Treatment of Obesity: Weight Loss and Bariatric Surgery. Circ Res 118:1844–1855
Cypess AM, Kahn CR (2010) Brown fat as a therapy for obesity and diabetes. Curr Opin Endocrinol Diabetes Obes 17:143–149
Peirce V, Vidal-Puig A (2013) Regulation of glucose homoeostasis by brown adipose tissue. Lancet Diabetes Endocrinol 1:353–360
Orava J et al (2011) Different Metabolic Responses of Human Brown Adipose Tissue to Activation by Cold and Insulin. Cell Metab 14:272–279
Cannon B, Nedergaard J (2004) Brown adipose tissue: function and physiological significance. Physiol Rev 84:277–359
Mendez-Gutierrez A, Osuna-Prieto FJ, Aguilera CM, Ruiz JR, Sanchez-Delgado G (2020) Endocrine Mechanisms Connecting Exercise to Brown Adipose Tissue Metabolism: a Human Perspective. Curr Diab Rep 20:40
Feldmann HM, Golozoubova V, Cannon B, Nedergaard J (2009) UCP1 Ablation Induces Obesity and Abolishes Diet-Induced Thermogenesis in Mice Exempt from Thermal Stress by Living at Thermoneutrality. Cell Metab 9:203–209
Kazak L et al (2015) A Creatine-Driven Substrate Cycle Enhances Energy Expenditure and Thermogenesis in Beige Fat. Cell 163:643–655
Yoneshiro T et al (2019) BCAA catabolism in brown fat controls energy homeostasis through SLC25A44. Nature 572:614–619
Sassano ML et al (2023) PERK recruits E-Syt1 at ER–mitochondria contacts for mitochondrial lipid transport and respiration. J Cell Biol 222:e202206008
Bobrovnikova-Marjon E et al (2012) PERK utilizes intrinsic lipid kinase activity to generate phosphatidic acid, mediate Akt activation, and promote adipocyte differentiation. Mol Cell Biol 32:2268–2278
Perea V et al (2023) PERK signaling promotes mitochondrial elongation by remodeling membrane phosphatidic acid. EMBO J 42:e113908
Kato H et al (2020) ER-resident sensor PERK is essential for mitochondrial thermogenesis in brown adipose tissue. Life Sci Alliance 3:e201900576
Liu Z et al (2017) Reducing Smad3/ATF4 was essential for Sirt1 inhibiting ER stress-induced apoptosis in mice brown adipose tissue. Oncotarget 8:9267–9279
Alcalá M et al (2017) Increased inflammation, oxidative stress and mitochondrial respiration in brown adipose tissue from obese mice. Sci Rep 7:16082
Nedergaard J, Cannon B (2018) Brown adipose tissue as a heat-producing thermoeffector. Handb Clin Neurol 156:137–152
Fromme T, Klingenspor M (2011) Uncoupling protein 1 expression and high-fat diets. American Journal of Physiology - Regulatory Integrative and Comparative Physiology. 300. https://doi.org/10.1152/ajpregu.00411.2010. R1-8 Preprint at
Glick Z, Teague RJ, Bray GA (1981) Brown adipose tissue: Thermic response increased by a single low protein, high carbohydrate meal. Sci (1979) 213:1125–1127
Lebeau J et al (2018) The PERK Arm of the Unfolded Protein Response Regulates Mitochondrial Morphology during Acute Endoplasmic Reticulum Stress. Cell Rep 22:2809–2817
Civiletto G et al (2015) Opa1 overexpression ameliorates the phenotype of two mitochondrial disease mouse models. Cell Metab 21:845–854
Frezza C et al (2006) OPA1 Controls Apoptotic Cristae Remodeling Independently from Mitochondrial Fusion. Cell 126:177–189
Friedman JR, Mourier A, Yamada J, McCaffery M, J., Nunnari J (2015) MICOS coordinates with respiratory complexes and lipids to establish mitochondrial inner membrane architecture. Elife 1–61 (2015)
Stephan T et al (2020) MICOS assembly controls mitochondrial inner membrane remodeling and crista junction redistribution to mediate cristae formation. EMBO J 39:e104105
Balsa E et al (2019) ER and Nutrient Stress Promote Assembly of Respiratory Chain Supercomplexes through the PERK-eIF2α Axis. Mol Cell 74:877–890e6
Backes S et al (2018) Tom70 enhances mitochondrial preprotein import efficiency by binding to internal targeting sequences. J Cell Biol 217:1369–1382
Liu Q et al (2022) Tom70-based transcriptional regulation of mitochondrial biogenesis and aging. Elife 11:e75658
Wei X et al (2020) Mutations in TOMM70 lead to multi-OXPHOS deficiencies and cause severe anemia, lactic acidosis, and developmental delay. J Hum Genet 65:231–240
Kazak L, Spiegelman BM (2020) Mechanism of futile creatine cycling in thermogenesis. Am J Physiol Endocrinol Metab 319:E947–E949
Ellenrieder L, Mårtensson CU, Becker T (2015) Biogenesis of mitochondrial outer membrane proteins, problems and diseases. Biol Chem 396:1199–1213
Gupta A, Becker T (2021) Mechanisms and pathways of mitochondrial outer membrane protein biogenesis. Biochim Biophys Acta Bioenerg 1862:148323
Waizenegger T, Schmitt S, Zivkovic J, Neupert W, Rapaport D (2005) Mim1, a protein required for the assembly of the TOM complex of mitochondria. EMBO Rep 6:57–62
Jores T et al (2018) Cytosolic Hsp70 and Hsp40 chaperones enable the biogenesis of mitochondrial β-barrel proteins. J Cell Biol 217:3091–3108
Drwesh L et al (2022) A network of cytosolic (co)chaperones promotes the biogenesis of mitochondrial signal-anchored outer membrane proteins. Elife 11
Pfanner N, Warscheid B, Wiedemann N (2019) Mitochondrial proteins: from biogenesis to functional networks. Nat Rev Mol Cell Biol 20:267–284
Davis TL et al (2010) Structural and biochemical characterization of the human cyclophilin family of peptidyl-prolyl isomerases. PLoS Biol 8:e1000439
Dunyak BM, Gestwicki JE (2016) Peptidyl-Proline Isomerases (PPIases): Targets for Natural Products and Natural Product-Inspired Compounds. J Med Chem 59:9622–9644
Wedemeyer WJ, Welker E, Scheraga H (2002) A. Proline cis-trans isomerization and protein folding. Biochemistry 41:14637–14644
Hoffmann K, Handschumacher RE (1995) Cyclophilin-40: Evidence for a dimeric complex with hsp90. Biochem J 307:5–8
Ratajczak T, Carrello A (1996) Cyclophilin 40 (CyP-40), mapping of its hsp90 binding domain and evidence that FKBP52 competes with CyP-40 for hsp90 binding. J Biol Chem 271:2961–2965
Dornan J et al (1999) Purification, characterization and crystallization in two crystal forms of bovine cyclophilin 40. Acta Crystallogr D Biol Crystallogr 55:1079–1082
Taylor P et al (2001) Two structures of cyclophilin 40: Folding and fidelity in the TPR domains. Structure 9:431–438
Mark PJ et al (2001) Human cyclophilin 40 is a heat shock protein that exhibits altered intracellular localization following heat shock. Cell Stress Chaperones 6:59–70
Ward BK et al (2002) A structure-based mutational analysis of cyclophilin 40 identifies key residues in the core tetratricopeptide repeat domain that mediate binding to Hsp90. J Biol Chem 277:40799–40809
Sidrauski C, McGeachy AM, Ingolia NT, Walter P (2015) The small molecule ISRIB reverses the effects of eIF2α phosphorylation on translation and stress granule assembly. Elife e05033 (2015)
Baker JD et al (2017) Human cyclophilin 40 unravels neurotoxic amyloids. PLoS Biol 15:e2001336
Young JC, Hoogenraad NJ, Hartl FU (2003) Molecular chaperones Hsp90 and Hsp70 deliver preproteins to the mitochondrial import receptor Tom70. Cell 112:41–50
Bhangoo MK et al (2007) Multiple 40-kDa heat-shock protein chaperones function in Tom70-dependent mitochondrial import. Mol Biol Cell 18:3414–3418
Fan ACY, Gava LM, Ramos CHI, Young JC (2010) Human mitochondrial import receptor Tom70 functions as a monomer. Biochem J 429:553–563
Gordon DE et al (2020) Comparative host-coronavirus protein interaction networks reveal pan-viral disease mechanisms. Science 370:eabe9403
Hagan CL, Silhavy TJ (2011) Kahne, D. β-barrel membrane protein assembly by the bam complex. Annu Rev Biochem 80:189–210
Miura H, Quadros RM, Gurumurthy CB, Ohtsuka M (2018) Easi-CRISPR for creating knock-in and conditional knockout mouse models using long ssDNA donors. Nat Protoc 13:195–215
Henstridge DC, Whitham M, Febbraio MA (2014) Chaperoning to the metabolic party: The emerging therapeutic role of heat-shock proteins in obesity and type 2 diabetes. Mol Metab 3:781–793
Blackburn EA et al (2015) Cyclophilin40 isomerase activity is regulated by a temperature-dependent allosteric interaction with Hsp90. Biosci Rep 35:e00258
Muthukumar G et al (2024) Triaging of ⍺-helical proteins to the mitochondrial outer membrane by distinct chaperone machinery based on substrate topology. Mol Cell 84:1–19
Guna A et al (2022) MTCH2 is a mitochondrial outer membrane protein insertase. Science 378:317–322
Backes S et al (2021) The chaperone-binding activity of the mitochondrial surface receptor Tom70 protects the cytosol against mitoprotein-induced stress. Cell Rep 35:108936
Uldry M et al (2006) Complementary action of the PGC-1 coactivators in mitochondrial biogenesis and brown fat differentiation. Cell Metab 3:333–341
Mina AI et al (2018) CalR: A Web-Based Analysis Tool for Indirect Calorimetry Experiments. Cell Metab 28:656–666e1
Shabalina IG et al (2010) Uncoupling protein-1 is not leaky. Biochim Biophys Acta Bioenerg 1797:773–784
Chen ZL et al (2019) A high-speed search engine pLink 2 with systematic evaluation for proteome-scale identification of cross-linked peptides. Nat Commun 10:3404
Perez-Riverol Y et al (2022) The PRIDE database resources in 2022: A hub for mass spectrometry-based proteomics evidences. Nucleic Acids Res 50:D543–D552

The authors declare no competing interests.

ExtendedDataTable1MetabolomicsdataPERKvsPERKiBAT.xlsx
Extended Data Table 1. Metabolomics data PERK^+/+ vs PERK^-/- iBAT.xlsx
extendedDataresearchSquare.pdf
Extended Data Figure Legends Extended Data Fig. 1. Determination of physiological parameters in HFD-fed PERK^-/- mice. a, Body weight over 16 weeks under HFD in either AdipoQ-Cre (left panel) or UCP-Cre (right panel) expressing PERK^flox/flox mice. WT = PERK^+/+, KO = PERK^-/-. AdipoQ-Cre n_WT=5, n_KO=6; UCP1-Cre n_WT=6, n_KO=10. Mean±SD and two-way ANOVA statistics are presented. b, Body weight gain (%) over 16 weeks under HFD in UCP-Cre expressing PERK^flox/flox mice. WT = PERK^+/+ (n=6) and KO = PERK^-/- (n=10). Mean±SD and two-way ANOVA statistics are presented. c, Oxygen consumption in adipocyte-specific HFD-fed PERK^+/+(WT, n=5) and PERK^-/- (KO, n=5) mice at 30°C. ANCOVA statistics are presented. d, Food intake in adipocyte-specific HFD-fed PERK^+/+(WT, n=5) and PERK^-/-(KO, n=5) mice at 30°C. Mean±SD is presented. Two-way ANOVA showed no differences. e, Rectal temperature at different time points upon cold exposure in adipocyte-specific 20-week HFD-fed PERK KO mice (PERK^-/-, n=5) and controls (PERK^+/+, n=6). Mean±SD and two-way ANOVA statistics are presented. f, Glucose tolerance test. Blood glucose levels at different time points in brown adipocyte-specific 20-week HFD-fed PERK KO mice (PERK^-/-, n=7) and controls (PERK^+/+, n=5). Mice were fasted for 16 hours, and glucose tolerance was measured after intraperitoneal injection of glucose (2g/Kg). Mean± SD and two-way ANOVA statistics are presented. g, Blood glucose levels after 16-hour fasting, from panel f, in brown adipocyte (UCP1-Cre)-specific PERK^-/- (KO, n=7) and PERK^+/+ (WT, n=5) controls. Mean± SD and two-sample t-test are presented. h, Insulin tolerance test. Blood glucose levels at different time points in adipocyte-specific 20-week HFD-fed PERK KO mice (PERK^-/-, n=10) and controls (PERK^+/+, n=6). Mice were fasted for 6 hours, and insulin tolerance was measured after intraperitoneal injection of insulin (0.75U/Kg). Mean± SD and two-way ANOVA statistics are presented. Extended Data Fig. 2. Impact of PERK-dependent mitochondrial-ER communication suppression on mitochondrial proteins and function upon HFD. a, Mitochondrial protein levels determined by SDS-PAGE in isolated mitochondria from PERK^+/+ and PERK^-/- HFD-fed mice, one animal per lane. b, Oxygen consumption rates in isolated iBAT mitochondria from HFD-fed PERK-/- and PERK+/+ mice. n=4, Mean±SD is presented. c, Relative NAD⁺/NADH ratios in PERK^+/+ and PERK^-/- HFD-fed mice. Mean±SD and two-sample t-test are presented. d, Densitometry protein quantitation from the experiments shown in Fig. 1i, n=4, mean±SD and two-way ANOVA with Dunnett’s multiple test correction are presented. Mitochondria from three mice were pooled used per replicate. e, Assembly of ANT2 into the ANT protein complex by blue-Native PAGE immunoblotting. f, Volcano plot representing metabolomics data from whole iBAT tissues from PERK^+/+ (WT) and PERK^-/- (KO) mice. g, Relative leucine-isoleucine levels in PERK^+/+ and PERK^-/- HFD-fed mice. Mean±SD and two-sample t-test are presented. h, In vitro incorporation of ¹⁴C-leucine after 10 minutes into isolated mitochondria from PERK^-/- (KO, n=3) and PERK^+/+ (WT, n=3) HFD-fed mouse iBAT. Mean± SD and two-sample t-test are presented. Extended Data Fig. 3. Regulation of PPID levels during high fat dietary stress. a, Proteomics data analysis volcano plots and workflow. b, PPID protein levels in mitochondria from differentiated brown adipocytes exposed to norepinephrine (NE) or NE supplemented with a PERK inhibitor (NE+PERKi). MIC19 was used as a positive control c, PPID levels in mitochondria from differentiated brown adipocytes determined by western blot analysis under conditions described in b. d, Relative Ppid mRNA levels in PERK^+/+ (n=3) and PERK^-/- (n=4) HFD-fed mice. One animal per data point. No statistical differences were found. e, Protein levels in differentiated brown adipocytes upon PERK silencing (siPERK) and/or palmitate stimulation. f, PPID levels in differentiated brown adipocytes determined by western blot analysis after palmitic acid (PA) stimulation or PA supplemented with PERK inhibitor (PA+PERKi). g, Relative Ppid mRNA levels in differentiated brown adipocytes under conditions explained in f,. One-way ANOVA statistics are presented. h, Levels of PPID in cells upon inhibition of the PERK-eIF2a pathway with ISRIB in the presence or absence of palmitic acid (PA). ATF4 was used as a positive control. i, Determination of PPID half-life by cycloheximide chase experiments. Differentiated brown adipocytes were treated with palmitic acid (PA) and PERK was knock-down (siPERK) as indicated. j, Graphical representation of PPID levels from the experiments described in i,. n=4, mean±SD and two-way ANOVA with Dunnett’s multiple test correction are presented. k, Determination of PPID ubiquitination by pull-down assays from differentiated brown adipocytes under conditions described in j. MG132 was added at 20 mM 4 hours before harvesting the cells. l, Densitometry protein quantitation from the experiments shown in Fig. 2h, n=4, mean±SD and two-way ANOVA with Dunnett’s multiple test correction are presented. Extended Data Fig. 4. PPID regulates mitochondrial protein import. a, In vitro import of either ³⁵S-ANT2, ³⁵S-MIC19 and ³⁵S-OTC into isolated mitochondria from brown adipocytes upon PERK or PPID knock-down and palmitate stimulation. B=blank tube containing only precursor, DY=membrane potential. Below, graphical representation from the experiment represented above. Mean± SD and two-way ANOVA with Dunnett’s multiple test correction are presented (n³3). b, Blue-Native PAGE immunoblots assessing the assembly of TOM40, TOM70, TIM22 and TIM23 complexes under the conditions described in a. c, Determination of mitochondrial membrane potential by TMRM and MitoTracker Green staining in differentiated brown adipocytes under the conditions described in a. Extended Data Fig. 5. PPID conservation and characterization of residues with key biological activities. a, ProteinBlast analysis using mouse PPID protein sequence (NP_001343255.1) and top hits. E-value indicates the expect parameter for a specific hit, the lower, the more significant. b, Alignment of the top hit domains indicated in a. c, Superimposition of aligned domains between PPID (PDB: 1iHG) and TOM70 (PDB: 7KDT) and detail of the CX₁₄(F/Y)R motif indicated in b. d, Kinetic PPIase assay of different PPID mutants. PPIase activity is normalized to controls (WT). Mean± SD and one-way ANOVA with Dunnett’s multiple test correction are presented (n=3). e, Determination of PPID TPR activity by its ability to bind to HSP90 C-terminal domain (CTD) using pull-down affinity assays with Co-NTA resin. f, Quantification of protein binding from e. Mean± SD and one-way ANOVA with Tukey’s multiple test correction are presented (n=3). g, ³⁵S-TOM70 insertion into naked liposomes in the presence of PPID WT and H141A mutant with or without 100mM cyclosporin A (CsA). One-way ANOVA statistics are presented for individual comparisons (n=4). h, Thermal shift assay experiments evaluating the binding of CsA to PPID WT and H141A mutant. Thermal shifts are annotated for three independent assays. i, Coomassie stained gel showing isolated PPID, HSC70 and HSP90 used in Fig. 3d. Extended Data Fig. 6. In vitro reconstitution of PPID-TOM70 complexes. a and b, Separation of PPID-TOM70 complexes by Size Exclusion Chromatography (SEC) after incubation at 37°C. c, Separation of PPID-TOM70 complexes by glycerol gradient ultracentrifugation after incubation at 37°C. d, Immunodetection of PPID or TOM70 on selected fractions from c. e, Light scattering mass photometry on the indicated fractions from c. f, Comparative separation of PPID-TOM70 complexes by glycerol gradient ultracentrifugation after incubation at 37°C upon incubation with either PPID WT or C296S or R312 alleles. Bar plot on the right represents relative abundances across fractions 1 through 4 determined by densitometry. Extended Data Fig. 7. Effects of PPID suppression in adipocytes in HFD-fed mice. a, Design and generation of PPID^flox/flox mice using EasiCRISPR. b, Example of genotyping during the generation of adipocyte-specific PPID^-/- mice. c, PPID levels across different tissues in adipocyte-specific PPID^-/- lean mice. d, Body weight over 8 weeks under HFD in adipocyte-specific PPID^+/+ (n=13) and PPID^-/- (n=12). Mean±SD is presented. e, Magnetic Resonance Imaging to determine lean and fat composition in PPID WT (n=8, PPID^+/+) and KO (n=6, PPID^-/-) mice. Mean±SD and multiple t-test comparisons are presented. f, Blood glucose levels after 16-hour fasting in adipocyte-specific HFD-fed PPID^-/- (KO, n=6) and PPID^+/+ (WT, n=8) controls. Mean± SD and two-sample t-test are presented. g-i, Determination of free fatty acids, adiponectin and leptin levels in mouse sera. Mean± SD, n³4, and One-way ANOVA with Tukey’s multiple comparisons are presented. j, Oxygen consumption in adipocyte-specific HFD-fed PPID^+/+(WT, n=7) and PPID^-/- (KO, n=6) mice at multiple environmental temperatures. ANCOVA statistics are presented. k, Body weight in adipocyte-specific PPID^-/- (KO, n=6) and PPID^+/+ (WT, n=8) controls after 12 weeks under HFD. Mean± SD and two-sample t-test are presented. l, Food intake in adipocyte-specific HFD-fed PPID^+/+(WT, n=8) and PPID^-/- (KO, n=6) mice at 30°C. Mean±SD is presented. No significant differences were detected. m, Increment in oxygen consumption (DO2) upon CL316,243 stimulation in mice housed at 30°C (WT n = 4, PERK^-/-n = 5, and PPID^-/- n=5, mean ± SD, *p < 0.05; BMR, basal metabolic rate). Two-way ANOVA statistics are presented. n, Rectal temperature over time in PPID WT (n=8, PPID^+/+) and KO (n=6, PPID^-/-) male mice. Mean±SD and multiple two-way ANOVA comparisons are presented. o, Tissue weights in adipocyte-specific HFD-fed PPID^+/+(WT, n=5) and PPID^-/- (KO, n=5) mice. Mean±Max/min and multiple t-test comparisons are presented. p, Detail of iBAT enlargement in HFD-fed PPID^-/- compared to PPID^+/+ mice. q, Quantitation of protein levels from the experiment presented in Fig. 4f. Mean± SD and multiple t-tests with Holm-Šídák correction are presented. r, Oxygen consumption rates in isolated iBAT mitochondria from PPID^-/- and PPID^+/+ mice. n=4, Mean±SD is presented.

Download PDF

Version 1

posted

You are reading this latest preprint version

A chaperone-assisted mechanism of outer mitochondrial membrane TOM70 insertion protective against diet-induced obesity

Status:

Version 1

Abstract

Figures

Main

Disruption of PERK-dependent mitochondrial-ER communication results in defective outer mitochondrial membrane TOM70 insertion

PPID is an ER-stress activated cytosolic chaperone that controls TOM70 insertion and mitochondrial respiration

PPID and TOM70 interact through their tetratricopeptide repeat (TPR) domains

Discussion

Methods

Cell culture

Mice

Antibodies

High-fat diet in mice

Cold exposure experiments

Metabolic phenotypic recording

Glucose and insulin tolerance tests

Blood serum insulin, free fatty acids, adiponectin and leptin determination

Determination of protein levels by western blot

Palmitate treatment in BAT1 cells

Gene expression analysis

Electron microscopy

Mitochondria isolation

Mitochondria respiration assays

Proteomics analysis from whole iBAT tissue

High-resolution respirometry

Membrane potential determination

Immunoprecipitation

Purification of recombinant ULP1, TOM70, PPID alleles, HSC70 and HSP90

Purification of recombinant HSP90 C-terminal domain (CTD)

Protein import of 35S-labeled proteins into mitochondria

Liposome insertion assay

Amino acid uptake assay

Blue-Native PAGE electrophoresis

PPID pull-down affinity assays

Determination of PPID binding to CTD and 35S-TOM70

Cycloheximide chase experiments

Polar metabolomics

Purification of PPID-TOM70 complexes

Light scattering mass photometry

Thermal shift assays

Cross-linking mass spectrometry analyses

Statistics

Deposited data

Declarations

Competing interest

Author contributions

Acknowledgments

Data availability

References

Additional Declarations

Supplementary Files

Status:

Version 1

Protein import of ³⁵S-labeled proteins into mitochondria

Determination of PPID binding to CTD and ³⁵S-TOM70