Decreased pyruvate dehydrogenase activity in Tafazzin-deficient cells is caused by dysregulation of pyruvate dehydrogenase phosphatase 1 (PDP1)

Cardiolipin (CL), the signature lipid of the mitochondrial inner membrane, is critical for maintaining optimal mitochondrial function and bioenergetics. Disruption of CL metabolism, caused by mutations in the CL remodeling enzyme TAFAZZIN, results in the life-threatening disorder Barth syndrome (BTHS). While the clinical manifestations of BTHS, such as dilated cardiomyopathy and skeletal myopathy, point to defects in mitochondrial bioenergetics, the disorder is also characterized by broad metabolic dysregulation, including abnormal levels of metabolites associated with the tricarboxylic acid (TCA) cycle. Recent studies have identified the inhibition of pyruvate dehydrogenase (PDH), the gatekeeper enzyme for TCA cycle carbon influx, as a key deficiency in various BTHS model systems. However, the molecular mechanisms linking aberrant CL remodeling, particularly the primary, direct consequence of reduced tetralinoleoyl-CL (TLCL) levels, to PDH activity deficiency are not yet understood. In the current study, we found that remodeled TLCL promotes PDH function by directly binding to and enhancing the activity of PDH phosphatase 1 (PDP1). This is supported by our findings that TLCL uniquely activates PDH in a dose-dependent manner, TLCL binds to PDP1 in vitro, TLCL-mediated PDH activation is attenuated in the presence of phosphatase inhibitor, and PDP1 activity is decreased in Tafazzin-knockout (TAZ-KO) C2C12 myoblasts. Additionally, we observed decreased mitochondrial calcium levels in TAZ-KO cells and treating TAZ-KO cells with calcium lactate (CaLac) increases mitochondrial calcium and restores PDH activity and mitochondrial oxygen consumption rate. Based on our findings, we conclude that reduced mitochondrial calcium levels and decreased binding of PDP1 to TLCL contribute to decreased PDP1 activity in TAZ-KO cells.

Cardiolipin (CL), the signature lipid of the mitochondrial inner membrane, is critical for maintaining optimal mitochondrial function and bioenergetics.Disruption of CL metabolism, caused by mutations in the CL remodeling enzyme TAFAZZIN, results in the life-threatening disorder Barth syndrome (BTHS).While the clinical manifestations of BTHS, such as dilated cardiomyopathy and skeletal myopathy, point to defects in mitochondrial bioenergetics, the disorder is also characterized by broad metabolic dysregulation, including abnormal levels of metabolites associated with the tricarboxylic acid (TCA) cycle.Recent studies have identified the inhibition of pyruvate dehydrogenase (PDH), the gatekeeper enzyme for TCA cycle carbon influx, as a key deficiency in various BTHS model systems.However, the molecular mechanisms linking aberrant CL remodeling, particularly the primary, direct consequence of reduced tetralinoleoyl-CL (TLCL) levels, to PDH activity deficiency are not yet understood.In the current study, we found that remodeled TLCL promotes PDH function by directly binding to and enhancing the activity of PDH phosphatase 1 (PDP1).This is supported by our findings that TLCL uniquely activates PDH in a dose-dependent manner, TLCL binds to PDP1 in vitro, TLCL-mediated PDH activation is attenuated in the presence of phosphatase inhibitor, and PDP1 activity is decreased in Tafazzin-knockout (TAZ-KO) C2C12 myoblasts.Additionally, we observed decreased mitochondrial calcium levels in TAZ-KO cells and treating TAZ-KO cells with calcium lactate (CaLac) increases mitochondrial calcium and restores PDH activity and mitochondrial oxygen consumption rate.Based on our findings, we conclude that reduced mitochondrial calcium levels and decreased binding of PDP1 to TLCL contribute to decreased PDP1 activity in TAZ-KO cells.
Cardiolipin (CL) is a unique phospholipid with four fatty acid chains and two phosphate head groups.Following synthesis in the inner mitochondrial membrane (IMM), CL undergoes a specialized remodeling process by which its predominantly saturated fatty acid chains are progressively replaced by polyunsaturated fatty acids.The importance of the CL remodeling pathway is underscored by the life-threatening disease Barth syndrome (BTHS), in which patients bear mutations in TAFAZZIN, the transacylase responsible for adding polyunsaturated fatty acid chains to monolyso-CL (MLCL) (1)(2)(3).As a result, the biochemical hallmarks of BTHS include a marked decrease in unsaturated CL species, an overall decrease in CL, and a concomitant accumulation of MLCL.The functional benefit of CL remodeling is not well understood, with some studies proposing that remodeled CL does not confer a fitness advantage compared to unremodeled CL in yeast (4,5).
In mammals, the process of CL remodeling predominantly involves the replacement of oleic acid (18:1) fatty acid chains with linoleic acid (18:2), resulting in the formation of (18:2) 4 tetralinoeleoyl-CL (TLCL) (6,7).TLCL is the most abundant CL species in various mammalian tissues, including the human heart, lymphoblasts, and skeletal muscle (7)(8)(9)(10)(11)(12).Reduced levels of TLCL have been linked to mitochondrial dysfunction, whereas a linoleate-rich high-fat diet has been shown to improve mitochondrial function and reduce mortality in a hypertensive heart failure rat model (13).While the functional distinction between TLCL and other CL species remains enigmatic, one possibility is that the acyl chain composition of CL may differentially influence lipid-protein interactions.Studies utilizing 31 P-NMR spectroscopy have revealed that CL and MLCL exhibit distinct lipid-protein associations (14,15).This disparity has also been observed in an in vitro proteinlipid binding assay, wherein bovine heart-derived CL, but not MLCL, was shown to interact with pyruvate dehydrogenase (PDH) (16).CL profiles also differ between organisms and show a general taxonomic trend towards increased incorporation of unsaturated fatty acid chains in CL from higher eukaryotes (17).CL homeostasis can be disturbed by disease conditions, including cancer (18), nonalcoholic fatty liver disease (19), Alzheimer's disease (18,20) and heart failure (21).It is poorly understood why cells maintain such a broad spectrum of CL species and what the functional difference is between specific species.
We have previously reported that PDH activity and synthesis of acetyl-CoA are decreased in CL-deficient yeast and mammalian cells (16,22), suggesting a potential regulatory role for CL in modulating PDH activity.PDH is a crucial metabolic hub that facilitates the oxidative decarboxylation of pyruvate, resulting in the formation of acetyl-CoA, CO 2 , and NADH for utilization in the tricarboxylic acid (TCA) cycle and oxidative phosphorylation (23).The mammalian PDH complex consists of multiple copies of the three enzymatic subunits: pyruvate dehydrogenase (E1), dihydrolipoamide acetyltransferase (E2), and dihydrolipoamide dehydrogenase (E3).The activity of the PDH complex is regulated by PDH kinases (PDK) and PDH phosphatases (PDP) (24)(25)(26).Inactivation of PDH occurs when any of the three serine residues on the E1 subunit is phosphorylated by PDK (27).Conversely, enzyme activation occurs upon binding of PDP to the lipoyl domain of E2, followed by dephosphorylation of E1 (28,29).In mammalian cells, two isoforms of PDP exist, including PDP1, primarily expressed in muscle tissue, and PDP2, which is most abundant in the liver (30).The presence of calcium facilitates binding between PDP1 and the lipoyl domain of E2, enhancing the dephosphorylation activity of PDP1 by over 10-fold (31).
Utilizing Tafazzin-knockout C2C12 mouse myoblasts (TAZ-KO), the current study identifies a direct and specific function of remodeled TLCL in binding to PDP1 and maintaining PDH activity.Our data suggest that impaired PDP1 activity in TAZ-KO cells is likely a dual consequence of reduced availability of TLCL to interact with PDP1 and decreased mitochondrial calcium levels.Intriguingly, treating TAZ-KO cells with calcium lactate (CaLac) restores PDH activity and rescues mitochondrial oxygen consumption rate, suggesting a potential novel therapeutic approach for treating BTHS.

TLCL decreases PDH phosphorylation and rescues PDH activity in TAZ-KO cells
BTHS is caused by mutations in the CL remodeling enzyme TAFAZZIN.As a result of this remodeling defect, BTHS cells produce substantially less polyunsaturated CL and accumulate the remodeling intermediate MLCL.In mammalian cells, the predominant CL species in the heart and muscle is polyunsaturated TLCL, and several studies have suggested that BTHS pathology results specifically from a lack of TLCL enrichment in the heart (32)(33)(34).However, a mechanistic link between the lack of TLCL and BTHS pathology has not been demonstrated.
We have previously reported that acetyl-CoA synthesis is decreased in TAZ-KO cells (16).Consistent with this, the activity of PDH, which generates acetyl-CoA, is reduced in these cells (16).PDH activity deficiency has also been observed in a Tafazzin-knockdown mouse model (35).However, the mechanism whereby CL/TAFAZZIN regulates PDH is unknown.Insight into this mechanism came from our observation of a consistent increase in phosphorylated (inactive) PDH in TAZ-KO cells, which was rescued by supplementation of bovine heart CL (16).These data suggest that CL regulates PDH activity by decreasing PDH phosphorylation.However, how specific CL species (particularly TLCL) and in vivo remodeling of CL acyl chain composition modulate PDH activity remains unclear.
To determine which CL species can rescue PDH activity, we supplemented mitochondrial extracts from wild-type (WT) and TAZ-KO cells with either (18:0) 4 CL, (18:1) 4 CL, (18:2) 4 CL (TLCL), or (18:2) 3 MLCL and then assayed PDH phosphorylation and activity.Although total PDH levels were unchanged between treatment conditions (data not shown), TLCL significantly reduced the inhibitory phosphorylation of PDH (Fig. 1A) and restored PDH activity in TAZ-KO mitochondria in a dose-dependent manner (Fig. 1, B and C).Supplementation of saturated or (18:2) 3 MLCL did not significantly reduce PDH phosphorylation, suggesting that remodeling of CL, and thereby enrichment of TLCL, is critical for maintaining optimal PDH activity.Despite the presence of three polyunsaturated linoleic acid acyl chains (18:2) 3 on MLCL (identical to the four acyl chains of TLCL), it did not rescue PDH activity, indicating that the number of acyl chains per se is also an important determinant in the ability of CL to restore PDH function.

TLCL-mediated rescue of PDH activity is dependent on PDP1
CL physically interacts with a myriad of different proteins in addition to the PDH complex (16,36).Because the acyl chain structure of CL is predicted to influence its binding dynamics, we hypothesized that the unique ability of TLCL to restore PDH activity may result from its binding affinity for components of the PDH complex and its associated regulatory enzymes PDK and PDP.To this end, we performed a proteinlipid overlay assay incorporating different PDH/PDK/PDP components to test their interaction with TLCL.PDP1 and PDK4 are the predominant isoforms of PDP and PDK expressed in C2C12 myoblasts (37) (Fig. 2A) and were therefore the isoforms tested in this analysis.Intriguingly, only PDP1 was shown to bind with TLCL (Fig. 2B), suggesting that restoration of PDH activity may result from the direct interaction between TLCL and PDP1.Although PA is also bound to PDP1 (and to E1), we have previously demonstrated that PA does not restore PDH activity in TAZ-KO mitochondrial extracts (16).
To further investigate whether restoration of PDH activity by TLCL is dependent on PDP1, we tested the ability of TLCL to activate PDH in the presence of the serine/threonine phosphatase inhibitor PhosSTOP (38).Our results showed that PDP1 inhibition attenuated TLCL-mediated activation of PDH in TAZ-KO mitochondria (Fig. 3, A and B).As expected, TLCL supplementation alone resulted in decreased inhibitory phosphorylation of PDH (p-PDH) (Fig. 3A, lane 1 vs. 3).However, pretreating cells with phosphatase inhibitor blocked the ability of TLCL to reduce p-PDH levels (Fig. 3A, lane 3 vs. 4).The observations regarding phosphorylation of PDH were mirrored in PDH activity, where pretreatment with phosphatase inhibitor substantially reduced PDH activity in TLCLsupplemented cells (Fig. 3B).Taken together, these findings suggest that remodeled TLCL enhances PDH activity by Tafazzin deficiency dysregulates PDP1, reducing PDH activity binding to PDP1, thereby facilitating the dephosphorylation of PDH-E1.

Binding of PDP1 to PDH is reduced in TAZ-KO cells
Next, we examined the reasons for decreased PDH phosphorylation in TAZ-KO cells upon treatment with TLCL.The previous experiment suggests that restoration of PDH activity by TLCL is dependent on PDP1 (Fig. 3).Therefore, we hypothesized that TLCL enhances the activity of PDP1.PDP1 activates PDH by first binding to the lipoyl domain of PDH-E2 and then subsequently binding to the PDH-E1 subunit to dephosphorylate the key regulatory serine residues (28,29,39,40).To test the hypothesis that TLCL facilitates binding of PDP1 to PDH-E1, we employed two approaches.First, we used recombinant PDP1 protein as bait and performed a pull-down assay using WT and TAZ-KO cell lysates.Our results revealed decreased interaction between PDP1 and both the E1 and E2 subunits of PDH in TLCL-deficient TAZ-KO cells (Fig. 4A), despite no significant difference in steady-state levels of PDP1 protein (Fig. 4B).As a complementary approach, we implemented co-immunoprecipitation (Co-IP) and consistently observed a decrease in the interaction between endogenous PDP1 and both E1 and E2 in TAZ-KO lysates (Fig. 4C).These findings indicate that the binding of PDP1 to PDH subunits E1 and E2 is decreased in TAZ-KO cells, supporting a mechanism whereby diminished TLCL levels result in a lower affinity of PDP1 for its PDH substrate.

PDP1 activity is reduced in TAZ-KO cells
Based on the previously reported findings that TLCL levels and PDH activity are substantially reduced in TAZ-KO cells (16,37) and our data demonstrating that TLCL enhances PDH activity in a PDP1-dependent manner (Fig. 3), we predicted that PDP1 activity is decreased in TAZ-KO cells.To directly test PDP1 activity, we employed an in vitro PDP1 activity assay modified from Shan et al., (41).In this assay, we incubated increasing amounts of recombinant PDP1 protein with commercial PDH complex (Sigma, P7032) following incubation with WT or TAZ-KO lysates.To validate the system, we first confirmed that the addition of recombinant PDP1 to PDH effectively resulted in reduced levels of phosphorylated PDH relative to PDP1-free controls.We then tested our hypothesis by assaying PDH phosphorylation following the addition of recombinant PDP1 protein that had been previously incubated with either WT or TAZ-KO lysates.We observed decreased PDP1-mediated PDH dephosphorylation following incubation of PDP1 with TAZ-KO cell lysate.Specifically, incubation of PDP1 with WT lysate resulted in a 70% reduction in PDH phosphorylation vs.only a 10% reduction following TAZ-KO lysate incubation (Fig. 5).To achieve a similar degree of PDH dephosphorylation, 30% more PDP1 protein was required in the reaction mixture when it had been previously incubated with lysate from TAZ-KO cells vs. WT cells.This suggests that PDP1 is modified in the TAZ-KO intracellular environment, resulting in diminished in vitro enzyme activity.Notably, the PDH phosphorylation measured in this assay reflected only that of the exogenous PDH complex, as levels of endogenous PDH-E1 pulled down from cell lysates during the PDP1 incubation were negligible (data not shown).Taken together, these findings suggest that PDP1 activity is reduced in TAZ-KO cells, as evidenced by decreased binding affinity between PDP1 and its PDH substrate, decreased in vitro dephosphorylation activity, and abrogated rescue of PDH activity and phosphorylation in TLCL-supplemented mitochondria pretreated with phosphatase inhibitor.

Mitochondrial calcium levels are decreased in TAZ-KO cells
Calcium enhances the activation of PDP1 by increasing its binding affinity for PDH-E2 (29,38,42).Therefore, impaired binding of PDP1 to PDH-E1/E2 suggests a potential secondary mechanism of PDP1 inhibition in TAZ-KO cells resulting from decreased mitochondrial calcium levels.In support of this, a recent study demonstrated that the expression of proteins associated with the mitochondrial calcium uniporter (MCU) is substantially reduced in BTHS patient-derived tissues and that MCU-mediated calcium uptake is impaired in TAZ-KO C2C12 myoblasts (43).
To further investigate mitochondrial calcium uptake and concentration, we titrated cells with calcium chloride to increase intracellular calcium levels while simultaneously tracking Fluo-4 AM (calcium indicator) and TMRM  (mitochondrial membrane potential (MMP) probe) fluorescence by confocal microscopy (44,45).To precisely measure mitochondrial calcium, cells were permeabilized to clear cytosolic calcium and imaged in an intracellular buffer containing 1 μM thapsigargin, which eliminates ER calcium signals by inhibiting the predominant ER Ca 2+ ATPase (46).Steady-state analysis revealed decreased mitochondrial calcium signals in TAZ-KO mitochondria compared to WT (Fig. 6A).TMRM analysis demonstrated that the MMP is also reduced in TAZ-KO cells, which is consistent with our previous observation (Fig. 6B) (47).Using this methodology, calcium uptake can be visualized by an increase in mitochondrial matrix Fluo-4 AM signal.TAZ-KO cells showed no significant difference in the rate of mitochondrial calcium uptake relative to WT through 120 s of imaging (Fig. 6C).Interestingly, after the gradual addition of calcium, TAZ-KO cells showed a drop in the TMRM signal at approximately 120 s of imaging (Fig. 6D).The concomitant drop in TAZ-KO mitochondrial Fluo-4 AM signal at this timepoint reflects the rapid loss of mitochondrial calcium following the loss of membrane potential.These data indicate that the decreased levels of mitochondrial matrix calcium in TAZ-KO cells are not due to decreased mitochondrial calcium uptake but a reduced capacity to maintain calcium levels as high as WT cells.In summary, these data suggest that decreased PDP1 activity and binding affinity for PDH-E2 are due, at least in part, to reduced mitochondrial calcium levels in TAZ-KO cells.

CaLac treatment rescues PDH activity and OCR in TAZ-KO cells
In light of our findings that TAZ-KO cells have reduced mitochondrial calcium levels, we sought to test if restoring mitochondrial calcium would rescue PDH activity.To test this, we treated cells with calcium lactate (CaLac), a compound that passively diffuses across cell membranes and increases free calcium levels, thereby circumventing deficiencies in intracellular calcium transport (48).
Treatment of TAZ-KO cells with CaLac resulted in a reduction in PDH phosphorylation (Fig. 7, A and B).Moreover, the oxygen consumption rate (OCR) was increased in TAZ-KO cells treated with CaLac.Even at a concentration of 2.5 mM CaLac, basal OCR was restored to WT levels (Fig. 7C), and a higher concentration of CaLac (12.5 mM) further enhanced the OCR beyond WT levels (Fig. 7D).These findings indicate that increasing mitochondrial calcium levels improves PDH activity and mitochondrial respiratory function in TAZ-KO cells.

Discussion
In this study, we highlight the significance of CL remodeling to maintain the function of PDH, the gatekeeping enzyme for the TCA cycle.PDP1 activity, which activates PDH, is reduced in TAZ-KO mitochondria.We hypothesize that at least two factors contribute to PDP1 inhibition in TAZ-KO cellsecreased TLCL, which binds PDP1 and likely enhances its interaction with PDH, and decreased mitochondrial calcium, which has been shown to inhibit PDP1 binding and activation (29,38,42).
Many studies aimed at understanding TAFAZZIN deficiency in the context of BTHS have focused on indirect downstream phenotypes without identifying the primary/ direct consequence of impaired CL remodeling.The ex vivo experiments conducted in our study identified a direct and  specific role for remodeled TLCL in binding to PDP1, thereby facilitating the activation of PDH through dephosphorylation.This model is supported by our finding that when phosphatase activity is inhibited, TLCL is unable to rescue PDH activity, suggesting dependency on PDP1 activity.Although PDP1 binding and activity are enhanced by calcium (42), the ability of TLCL to rescue PDH in isolated mitochondria using a calcium-free buffer suggests that this mechanism operates independently of modulating mitochondrial calcium levels.Instead, TLCL may serve as a scaffolding molecule to promote the interaction between PDP1 and PDH, thereby facilitating PDH activation.This is supported by the observed dose-dependency of TLCL rescue, wherein increasing levels of TLCL would effectively create additional sites for PDP1 to interact with PDH.
Interestingly, we did not observe decreased calcium uptake associated with Tafazzin deficiency as Ghosh et al. reported.This discrepancy may be the result of differences in how calcium treatments were administered to cells (gradual addition of calcium in the current study vs. an acute, highly concentrated single injection of calcium by Ghosh et al.).However, both studies provide evidence of decreased mitochondrial matrix calcium levels in TAZ-KO cells relative to WT (43).Given the importance of calcium homeostasis, the exact nature of how decreased mitochondrial calcium affects TAZ-KO cells remains to be determined.However, it can be excluded that calcium activates PDP1 activity by facilitating the binding of PDP1 to PDH-E1, as we did not observe a difference in the binding affinity between PDP1 and PDH-E1 (as measured by immunoprecipitation) when TAZ-KO cells were treated with either exogenous calcium or the calcium chelator EGTA (data not shown).Therefore, it is likely that decreased mitochondrial calcium acts in conjunction with the reduction in TLCL to reduce the dephosphorylation of PDH-E1 by PDP1.
Treating TAZ-KO cells with CaLac rescued both PDH phosphorylation and OCR in a dose-dependent manner, suggesting that decreased calcium levels play a role in TAZ-KOassociated PDH deficiency and mitochondrial dysfunction.Interestingly, Ghosh et al. did not observe enhanced PDH dephosphorylation or activity following treatment with isoprenaline hydrochloride, a β-adrenergic agonist known to indirectly increase mitochondrial calcium uptake (43).In light of the ability of CaLac to diffuse passively across membranes (a characteristic not shared by isoprenaline hydrochloride), these two sets of findings collectively suggest that TAZ-KO cells possess a defect in maintaining mitochondrial calcium homeostasis.However, CaLac treatment has also been shown to enhance the activity of lactate dehydrogenase B (LDHB), which converts lactate into pyruvate-the primary substrate of PDH (48).Therefore, we cannot rule out the possibility that CaLac treatment enhances PDH activity in part by increasing substrate availability.However, we view this possibility as less likely, due to previous reports of both lactate and pyruvate accumulation in TAZ-KO cells (16,37).This is also supported by our finding that cells treated with GSK2837808A, a specific lactate dehydrogenase A inhibitor that favors LDHB-mediated conversion of lactate to pyruvate, failed to activate PDH activity and rescue OCR in TAZ-KO cells (data not shown).Interestingly, treating TAZ-KO cells with 12.5 mM CaLac restored OCR to levels exceeding what was observed in untreated WT cells.Although CaLac likely enhances the activity of calcium-sensitive PDP1, the observation that phospho-PDH levels are similar between these groups suggests that CaLac may act through an additional mechanism to enhance mitochondrial function.Indeed, several enzymes in the TCA cycle and ETC are also calcium-sensitive, including isocitrate dehydrogenase, oxoglutarate dehydrogenase, cytochrome c oxidase, and ATP synthase (49).Therefore, CaLac treatment may represent an especially robust means for mitigating mitochondrial dysfunction.Overall, this study addresses a gap in our understanding of BTHS by identifying a novel and direct role for remodeled TLCL in binding to PDP1 to promote activation of PDH.While an exact understanding of the in vivo binding dynamics between TLCL, PDP1, and PDH-E1/E2 will require additional studies, the current work suggests that reduced mitochondrial calcium levels likely contribute to diminished PDP1 and PDH activity in the context of Tafazzin deficiency.

Mitochondrial extraction
Cells (1X10 7 ) were washed once with PBS and homogenized with a glass homogenizer in mitochondrial isolation buffer (280 mM sucrose, 0.25 mM EDTA, 0.02 M Tris (pH 7.4)).Cell debris was removed by centrifugation at 720g for 5 min.Mitochondria were subsequently collected by centrifugation at 14,812g for 10 min.Protein concentration was determined by Bradford assay (Bio-Rad).

CL supplementation and PDH activity
Mitochondria isolated from 1X10 7 myoblasts were suspended in a 1:1 ratio of 50 mM NaCl and 50 mM imidazole (pH 7.4).Individual CL species were prepared fresh in 50% ethanol prior to each experiment.For every 100 μg of mitochondria, 1.5 μl of 10% digitonin (protein/digitonin ratio 100:150) was added, followed by the addition of 1 μl of CL (20 mg/ml) or 4 μl of CL (5 mg/ml) or 50% ethanol (control).The interaction was kept on ice for 2.5 h, and then 2 μl of 10% digitonin was added to each group to reach a final protein/ digitonin ratio of 100:350.For PDH activity, 10 to 30 μg of lysate was used with a PDH activity assay kit (MAK183, Sigma) at 37 C.The mitochondrial lysate was extracted with the addition of a phosphatase inhibitor in the corresponding assay (PhosSTOP, Roche).

RNA extraction and real-time qPCR analysis
cDNA was synthesized from total RNA using SuperScript (Invitrogen).Real-time qPCR was carried out with the QuantSudio3 real-time qPCR machine (Applied Biosystems) using PowerUp SYBR Green reagents (Thermo Fisher).Relative mRNA transcript levels were quantified with the 2-ΔΔCT method (where CT is the threshold cycle), using Actb as an internal control.Primers used in this study are listed in Table 1.

Construction of bacterial expression vector for PDP1 cDNA
Methods were adapted and modified from the protocol by Huang et al. (30).Briefly, the coding region of PDP1 cDNA corresponding to the mature polypeptide was constructed by PCR using cDNA as a template and native Q5 Hot Start High Fidelity Polymerase (NEB).The resulting PCR product contained NcoI and XhoI restriction sites flanking the coding region.The forward primer (5 0 -AAA CCA TGGCTT CTA CAC CGC AGA AAT TTT AC-3 0 ), corresponding to bases 215 to 237 of PDP1 cDNA, included the NcoI restriction site.The reverse primer (5 0 -TTT CTC GAG CTG TTC CTG GTT TTG GTA TGC-3 0 ), corresponding to bases 1594 to 1664, contained the XhoI restriction site.The primers were designed based on the transcript sequenced archived on the Ensembl website (MGP_C3HHeJ_T0064368.1).The resulting cDNA was digested with NcoI and XhoI endonucleases and subcloned between the NcoI and XhoI sites of the pET-28a expression vector to produce a carboxyl-terminal fusion with His 6 tag encoded by the vector (plasmid PDP1).The plasmids were sent for sequencing to confirm the correct orientation of PDP1 (Genewiz).
Expression and purification of PDP1 protein PDP1 and pGro7 (Takara) were co-transformed into BL21 (DE3) competent cells.The resulting transformants were then selected on LB agar plates containing kanamycin (100 μg/ml) and chloramphenicol (34 μg/ml).The cells were pre-cultured the day before and diluted by 1:50 with 100 ml of fresh cell medium containing 0.1% L-arabinose.The cells were then incubated at 37 C with constant shaking at 230 rpm the next day until the OD 600 of the culture reached 0.5 to 0.7.The cells were induced with 0.4 mM β-D-thiogalactopyranoside and incubated for another 2 h at 37 C.The cells were then harvested and resuspended in lysis buffer containing 50 mM HEPES, 150 mM NaCl, 20 mM imidazole supplemented with 0.4 mM phenylmethylsulfonyl fluoride (PMSF), 0.5% Tri-tonX100, and 10 mM β-mercaptoethanol.The cells were sonicated on ice to avoid foaming or excessive heat.The extraction was clarified by removing inclusion bodies at 600g for 30 min at 4 C.The supernatant was purified and concentrated using nickel resin (HisPur Ni-NTA, Thermo Scientific) and centrifugal filters (Amicon Ultra, Millipore).

OCR measurements in intact mammalian cells
WT and TAZ-KO C2C12 myoblasts were plated in XF96well cell culture microplates (Agilent Technologies) at 30,000 cells/well in 80 μl of DMEM complete growth medium and incubated at 37 C, 5% CO 2 for 4 h.Before OCR measurement, 200 μl of pre-warmed growth medium was added to each well, and cells were incubated at 37 C for 30 min.OCR measurements were carried out in intact cells using a Seahorse XFe96 Extracellular Flux Analyzer (Agilent Technologies).Calcium lactate (CaLac, Sigma) was injected through port D at a final concentration of 5 μM and 2.5/12.5 μM individually.For the mitochondrial stress test, oligomycin, carbonyl cyanide p-trifluoro-methoxyphenyl hydrazone (FCCP), and antimycin A were sequentially injected through ports A, B and C to achieve final concentrations of 2 μM, 1 μM, and 0.5 μM, respectively.

Statistical analysis
All values are presented as mean ± SD.Statistical analyses were performed in GraphPad Prism software using a twotailed unpaired Student's t test on data obtained from at least three independent experiments performed with different biological replicates.

Figure 1 .
Figure 1.PDH activity is rescued by supplementation of C(18:2) 4 CL.A, phosphorylated PDH (p-PDH) levels were assayed in mitochondria obtained from WT and TAZ-KO myoblasts.Mitochondria were incubated on ice for 2.5 h with the indicated CL species.The signal intensities of protein bands were quantified using ImageJ software (bottom panel).The protein band was normalized to NDUFB6 and quantified relative to the WT control-treated sample.B and C, mitochondria from WT and TAZ-KO myoblasts were treated with or without C(18:2) 4 CL, and PDH activity was assayed as described in "Experimental procedures".Data points represent mean ±S.D. (error bars) for each individual biological replicate of each group.* 0.01<p < 0.05, ** 0.001<p < 0.01, *** <p < 0.001.

Figure 2 .
Figure 2. C(18:2) 4 CL binds to PDP1 in vitro.A, PDP and PDK isoform mRNA levels were evaluated in WT myoblasts by qPCR.Expression levels were normalized to Actb as an internal control.Data points represent mean ±S.D. (error bars) for each individual biological replicate of each group.*** <p < 0.001.B, the indicated CL species, phosphatidic acid (PA), phosphatidylcholine (PC), and phosphatidylserine (PS), were serially diluted and spotted onto a PVDF membrane that was incubated overnight in buffer containing 25 μg of the indicated recombinant proteins PDP1.Interactions were detected by immunoblotting with an antibody against the His tag.

Figure 3 .
Figure 3. PDP1 activity is required for TLCL-mediated rescue of PDH activity.A, phosphorylated PDH (p-PDH) levels and PDH activity (B) were evaluated in mitochondria from TAZ-KO myoblasts treated with or without phosphatase inhibitor and supplemented for 2.5 h with or without C(18:2) 4 CL.Data points represent mean ±S.D. (error bars) for each individual biological replicate of each group.* 0.01<p < 0.05, ** 0.001<p < 0.01.

Figure 4 .
Figure 4.The binding of PDP1 to PDH is reduced in TAZ-KO cells.A, WB analysis of the bait-prey pull-down assay.B, cell lysates from WT and TAZ-KO cells were subjected to WB analysis using the indicated antibodies.C, coimmunoprecipitation assay for PDP1, PDH-E1, and PDH-E2 protein interaction.The indicated myoblast lysates were immunoprecipitated using the antibodies specified on the left side of the figure, and eluted protein was subjected to WB analysis by probing for the specific proteins listed on the right side.The negative control lane represents WT lysate passed through empty beads lacking primary antibodies.

Figure 5 .
Figure 5. PDP1 activity is decreased in TAZ-KO cells.Increasing amounts of recombinant PDP1 protein were incubated with WT or TAZ-KO lysates.PDP1 eluted from this mixture was subsequently incubated with PDH complex for 1 h.The dephosphorylation activity of PDP1 was evaluated by assaying phosphorylated PDH (p-PDH) using WB analysis.p-PDH levels were quantified by dividing the amount of p-PDH signal by total PDH-E1 in each sample, and values were normalized relative to the first lane in each group (containing the lowest concentration of eluted PDP1).

Table 1 qPCR
Primers used in this paper