The mitochondrial HSP90 paralog TRAP1 forms an OXPHOS-regulated tetramer and is involved in maintaining mitochondrial metabolic homeostasis

Background The molecular chaperone TRAP1, the mitochondrial isoform of cytosolic HSP90, remains poorly understood with respect to its pivotal role in the regulation of mitochondrial metabolism. Most studies have found it to be an inhibitor of mitochondrial oxidative phosphorylation (OXPHOS) and an inducer of the Warburg phenotype of cancer cells. However, others have reported the opposite and there is no consensus on the relevant TRAP1 interactors. This calls for a more comprehensive analysis of the TRAP1 interactome and of how TRAP1 and mitochondrial metabolism mutually affect each other. Results We show that the disruption of the gene for TRAP1 in a panel of cell lines dysregulates OXPHOS by a metabolic rewiring that induces the anaplerotic utilization of glutamine metabolism to replenish TCA cycle intermediates. Restoration of wild-type levels of OXPHOS requires full-length TRAP1. Whereas the TRAP1 ATPase activity is dispensable for this function, it modulates the interactions of TRAP1 with various mitochondrial proteins. Quantitatively by far the major interactors of TRAP1 are the mitochondrial chaperones mtHSP70 and HSP60. However, we find that the most stable stoichiometric TRAP1 complex is a TRAP1 tetramer, whose levels change in response to both a decline or an increase in OXPHOS. Conclusions Our work provides a roadmap for further investigations of how TRAP1 and its interactors such as the ATP synthase regulate cellular energy metabolism. Our results highlight that TRAP1 function in metabolism and cancer cannot be understood without a focus on TRAP1 tetramers as potentially the most relevant functional entity.

and rate of extracellular acidification (ECAR), caused by lactate secretion, a 145 measure of the glycolytic flux (Fig. 1k).

147
To maintain a minimal glycolytic rate and to promote pyruvate oxidation in 148 mitochondria, WT and KO cells were grown overnight in a medium containing 149 galactose and pyruvate (Gal + Pyr) as the only carbon sources [21]. Under these 150 conditions, the ECAR profile tends to mimic the OCR profile (compare Fig. 1l with  To confirm the increased Gln uptake and utilization by KO cells, indicated by the 168 OCR experiments, a quantitative flux tracing experiment was performed. For this, isotopically labelled Gln ( 13 C-Gln) was added in addition to unlabelled Glc and Pyr as 170 carbon sources (Additional file 2: Figure S2a-c and Additional file 3: Table S1 for 171 absolute quantitation of metabolites; for 13 C tracing in metabolites, see the NEI area 172 tab in Additional file 4: Table S2). Quantitation of metabolites with increased 13 C 173 abundance in KO cells are shown in Fig. 2. Both HEK293T and A549 KO cells 174 exhibited a significant increase in total Gln and glutamate concentrations (Fig. 2a), 175 further confirming that KO cells prefer Gln even in the presence of the other two 176 major carbon sources (Glc and Pyr). This is also associated with an increase in the  Table S2) and also observed a notable increase in 13 C-182 traced reduced glutathione (GSH) in both HEK293T and A549 KO cells (Fig. 2c). 183 This may indicate an adjustment to cope with increased reactive oxygen species 184 (ROS), which are often associated with increased OXPHOS [3,23].

186
Full-length TRAP1 but not its ATPase activity is essential to regulate OXPHOS 187 We next investigated which parts and functions of TRAP1 are necessary to rescue 188 the metabolic phenotype of KO cells. We designed a custom construct to express 189 TRAP1 variants with a C-terminal HA tag and an N-terminal TRAP1-MTS to ensure 190 that proteins are directed into the mitochondrial matrix (Additional file 5: Figure S3a).

191
A mitochondrially targeted EGFP construct (mito-EGFP) was used as a control 192 (Additional file 5: Figure S3b). As mentioned previously, this construct was used to 193 test whether overexpression of an unrelated protein in mitochondria might non-specifically disrupt OXPHOS function (Fig. 1h,. All TRAP1 truncation 195 mutants as well as the full-length protein were expressed with some exhibiting bands 196 corresponding to precursor proteins with uncleaved MTS and to shorter ones due to 197 N-terminal cleavage (Additional file 5: Figure S3c). The TRAP1 truncation mutants 198 were then overexpressed in the HEK293T KO cells to determine OCR profiles in the 199 presence of all three carbon sources (Fig. 3 a, c). Once again, the OCR data with the 200 mitoEGFP controls confirm a slight reduction in mitochondrial respiration due to 201 transient transfection toxicity (Figs. 1h,i,and 3a,c). However, the slightly lower OCR 202 of cells transfected with the control plasmid expressing mitoEGFP was still 203 significantly higher when compared to the OCR of cells transfected with the WT 204 TRAP1 expression plasmid ( Fig. 3 b, d). None of the TRAP1 truncation mutants 205 were able to suppress the KO OXPHOS phenotype to WT levels ( Fig. 3 b, d). This 206 indicates that a full-length TRAP1 protein is essential for normal OXPHOS 207 regulation.

209
Since TRAP1 is a paralog of HSP90, a molecular chaperone that is well known to be 210 dependent on its ATPase cycle [24,25], we speculated that the ATPase activity of 211 TRAP1 might be required for OXPHOS regulation. To test this, we generated a panel 212 of point and truncation mutants that affect this enzymatic activity. Note that our 213 numbering includes the 59 amino acids of the MTS. The following ATPase activity 214 mutants were tested: the double point mutant E115A/R402A with a 10-fold reduced 215 ATPase activity relative to WT (Additional file 5: Figure S3d), the 30-fold hyperactive 216 ATPase mutant ΔStrap, and the moderately activated (2.5-fold) ATPase single point 217 mutant D158N [14]. To our surprise, all ATPase mutants are able to suppress the 218 OXPHOS phenotype of the KO cells, reducing the OCR to WT levels ( Fig. 3e-i). 219 Similar results were obtained when the OCR analysis was done with cells in medium 220 with only Gln as the carbon source (Additional file 5: Figure S3e). We further 221 confirmed the ATPase independence of the complementation by performing a 222 separate real-time OCR analysis with murine cells comparing KO MAFs stably 223 expressing either WT or the single point mutant E115A of human TRAP1 (Fig. 3j).

224
Note that the mutant E115A was designed by analogy to the yeast HSP90 E33A 225 mutant, which has been reported to be able to bind to ATP, but to be defective for 226 ATP hydrolysis [24,26]; E115A, similarly to the single mutant mentioned above, 227 binds ATP, but is defective for ATP hydrolysis [15]. Thus, the ability to hydrolyze 228 ATP, at least as well as WT TRAP1, is not essential for the regulation of OXPHOS 229 by TRAP1. While HSP90 has an exhaustive list of clients and co-chaperones [13,[27][28][29][30], the 234 interactome of its mitochondrial paralog remains poorly characterized [6]. After 235 ascertaining that a full-length TRAP1 is essential for OXPHOS regulation, we 236 wondered which proteins interact with TRAP1 and whether these might explain its 237 role in OXPHOS regulation.

239
We carried out an immunoprecipitation mass spectrometry (IP-MS) experiment with 240 WT TRAP1 and the ATPase mutants E115A/R402A and ΔStrap overexpressed in 241 HEK293T cells (Additional file 6: Figure S4a; Additional file 7: Table S3). To refine 242 this list of identified proteins, the protein interactors were first filtered for validated 243 mitochondrial proteins and then by limiting the dataset to proteins with four or more identified unique peptides. This yielded a list of 82 proteins common to WT TRAP1 245 and the two ATPase mutants; we took these to represent the most probable TRAP1 246 interactors (Additional file 8: Loss of TRAP1 has a minor impact on mitochondrial and total cellular 289 proteomes 290 We speculated that the absence of TRAP1 might destabilize some of its direct or 291 indirect interactors or lead to a compensatory upregulation of other proteins. We 292 used two separate approaches to identify such proteome changes. First, we 293 performed a quantitative SILAC MS analysis comparing WT to KO UMUC3 cells. 200 mitochondrial proteins were detected (Additional file 9: Table S5). Among this group 295 of interactors, we found little variation comparing KO to WT cells when the minimum 296 significant fold change is set to 2 (p<0.05) (Fig. 4c). Even with a cutoff of 1.5-fold, 297 only a few alterations in the mitochondrial proteome could be seen (Fig. 4c,298 Additional file 9: Table S5). With the notable exception of PHB2 (when a 1.5-fold 299 change is set as threshold), most of the mitochondrial proteins including those 300 predicted to interact with TRAP1 (especially the subunits of the ATP-synthase 301 complex highlighted by the analysis of Fig. 4b), show no significant up-or 302 downregulation in UMUC3 KO cells (Additional file 9: Table S5). Thus, the TRAP1 303 KO does not have a significant impact on the stability of the mitochondrial proteome.

305
Second, we did a label free quantitation (LFQ) MS analysis of the total cellular 306 proteome with WT and KO HEK293T and HCT116 cells cultured with the three 307 different cocktails of carbon sources (Glc + Pyr + Gln, Gal + Pyr only, Gln only; 308 Additional file 10: Table S6). We reduced the initial list of 4578 proteins to 2660 309 proteins by using as criterion the identification of at least seven unique peptides per 310 protein (Additional file 11: Table S7). The comparison of the LFQ KO /LFQ WT ratios for 311 these proteins from cells cultured in medium with all three carbon sources did not 312 reveal any significant changes (Additional file 6: Figure S4d, e). Although a few 313 proteins were observed outside the 2-fold limit, they were not consistent across 314 HEK293T and HCT116 cells to warrant a correlation with the loss of TRAP1. The

315
LFQ ratio profiles turned out to be similar for media with other combinations of 316 carbon sources (Additional file 11: Table S7).
In toto, all three MS experiments indicated that while TRAP1 interacts with multiple 319 mitochondrial proteins, its loss does not have much of an impact on the 320 mitochondrial or cellular proteomes. with a protein such as TRAP1 with a pI of 6.40 in a separating gel at pH 8.8 to be 334 reasonably well correlated with molecular weight and size. When blotted for 335 endogenous TRAP1, a single molecular complex of ~300 kDa could be seen, which 336 is absent from KO cells (Fig. 5). However, the molecular weight of the detected 337 complex was not exactly what was expected if a TRAP1 dimer was in a complex with 338 mtHSP70, HSP60 or even both proteins. Moreover, looking at overexpressed WT or 339 ATPase mutant TRAP1 side by side, we found that the E115A/R402A mutant forms 340 a complex of the same size as WT TRAP1 whereas the hyperactive ATPase mutant 341 (ΔStrap) seems to form a slightly larger or conformationally different, more slowly 342 migrating complex (Fig. 5).

344
To determine what the 300 kDa TRAP1 complex contains, we expressed a TRAP1-345 GST fusion protein and GST alone as a negative control, and applied the workflow 346 described in Additional file 12: Figure S5A for a GST-pulldown MS analysis. Upon 347 setting the cutoff for an interactor at a minimum of eleven unique peptides, no 348 mitochondrial chaperone could be detected in the excised gel piece. Apart from 349 TRAP1, only proteins that were also co-purified with GST alone could be identified 350 (Additional file 12: Figure S5b, Additional file 13: Table S8). Hence, the high  The TRAP1 complex is induced in response to OXPHOS perturbations 358 Based on the hypothesis that an oligomerized TRAP1 complex might be the 359 functional entity of TRAP1, we checked its levels when OXPHOS is inhibited with a 360 prolonged exposure of HEK293T cells to hypoxia in various media (Fig. 6a).

361
Although the baseline levels of the TRAP1 complex vary in cells adapted to different 362 carbon sources in normoxia (left part of Fig. 6a), we saw a consistent increase in the 363 levels of the TRAP1 complex when cells were placed in hypoxia. It is notable that the 364 maximum increase in the levels of the TRAP1 complex was observed with cells 365 grown in Gal + Pyr medium when they were exposed to hypoxia (Fig. 6a). Cells with 366 this carbon source combination exclusively rely on OXPHOS for respiration 367 (Additional file 1: Figure S1, compare panels d and e). Considering that the ATP synthase is one of the major OXPHOS complexes that is inhibited by prolonged 369 hypoxia [39] and that we had found ATP-synthase components to be amongst the 370 main TRAP1 interactors (see Fig. 4b), we asked whether inhibition of the ATP-371 synthase complex would affect TRAP1 oligomerization (Fig. 6b). To this end, we 372 compared the levels of the TRAP1 complex from HEK293T cells exposed to hypoxia 373 or to the ATP-synthase inhibitor oligomycin under normoxic conditions. Under 374 hypoxic conditions, the induction of the TRAP1 complex is slow and only seems to 375 initiate around 6 hrs. (Fig. 6b). The slow time course may reflect the slow depletion 376 of oxygen from the medium and cells rather than a characteristic of mitochondria or 377 the TRAP1 complex. There is also an overall increase in the levels of TRAP1 378 protomers in cells exposed to hypoxia (Fig. 6b, middle panel with SDS-PAGE), but 379 this induction does not appear to be HIF1α-mediated (Additional file 14: Figure S6a).

380
In contrast, oligomycin induces a more rapid accumulation of the TRAP1 complex 381 above basal level without a noticeable concomitant increase in total TRAP1 protein 382 levels ( Fig. 6b).  Figure S7). Therefore, we tested whether 411 inhibition of ATP synthase could override the effects of complex I and III inactivation 412 (Fig. 7b). This was examined at the 3 and 6 hr time points with a combination of 413 rotenone + antimycin and oligomycin + rotenone + antimycin in parallel. Indeed, 414 inhibition of ATP synthase was able to override the suppressive effect of the 415 combined inhibition of complexes I and III on the TRAP1 complex in HEK293T cells, 416 as can be most clearly seen at the 6 hr time point (Fig. 7b).

417
Having found that the levels of the TRAP1 complex change upon inhibiting 419 OXPHOS, we wondered what would happen if OXPHOS were upregulated. This 420 question is not trivial to address experimentally as it appears that most cells in 421 culture operate OXPHOS at or close to maximal capacity. We decided to culture 422 HEK293T cells on glucose as the only carbon source and then to force them to divert  wild-type levels of OXPHOS requires full-length TRAP1. While this is not surprising, 466 it was unexpected that the ATPase activity of TRAP1 does not correlate with its 467 ability to restore OXPHOS to WT levels. This finding strongly suggests that the In view of the evidence that a TRAP1 tetramer may be the primary "functional unit" of 532 TRAP1, we reasoned that its levels might be influenced by fluctuating OXPHOS.

533
Indeed, when we inhibited OXPHOS by exposure of cells to hypoxia, we observed 534 that the levels of the TRAP1 complex increased with a corresponding increase in the 535 total mitochondrial protomer levels as observed with native and denaturing PAGE, 536 respectively. However, this increase in TRAP1 complex and total protomer levels 537 cannot be attributed to HIF1α as its overexpression does not induce TRAP1 mRNA connection with the ATP synthase is further supported by our finding that multiple 542 subunits comprising the ATP-synthase complex interact with TRAP1. Although the induction of the TRAP1 complex was consistent with the pharmacological inhibition 544 of ATP synthase across multiple cell lines, the variation in its protomer levels was 545 not. While the TRAP1 complex is induced by inhibition of ATP synthase, it is reduced 546 by inhibition of complexes I or III. Surprisingly, we found that inhibition of ATP 547 synthase overrides the latter effect. This pharmacological epistasis experiment 548 argues that ATP synthase is a primary TRAP1 interactor in the ETC. The opposite 549 "perturbation" of OXPHOS, that is its stimulation by an inhibitor of lactate 550 dehydrogenase, similarly promotes the formation of the TRAP1 tetramer. Thus, for 551 reasons that remain to be elucidated, the "functional unit" of TRAP1 is sensitive to 552 both an induction or a decline in OXPHOS.

554
In toto, although the precise molecular mechanism for how TRAP1 regulates 555 OXPHOS remains to be uncovered, we know now that the overall levels of TRAP1 556 may not be correlated or relevant to OXPHOS regulation as previously thought [6]. It 557 is really its tetrameric form that needs to be quantitated and structurally and 558 functionally dissected in more detail to understand how TRAP1 contributes to 559 regulating OXPHOS and mitochondrial homeostasis.  Table S9).  Table   618 S9). The transfection procedure was similar to the one described for HEK293T and 619 HCT116 cells, but the clonal isolation was performed with the mCherry reporter 620 using FACS sorting under aseptic conditions. The sorted clones were subcultured 621 and finally immunoblotted for TRAP1 to identify clones that were devoid of the 622 protein.         expression vectors for TRAP-GST and GST using the Jetprime transfection reagent at 70% confluency. 24 hrs after transfection, mitochondrial lysates were prepared in 818 lysis buffer (10 mM Tris-HCl pH 7.5, 50 mM NaCl, 1mM EDTA, 0.1% Triton X-100, 819 1mM DTT, 10% glycerol, 10 mM sodium molybdate, protease inhibitor cocktail 820 (A32965, Thermo Scientific)) as described before. 1 mg clarified mitochondrial 821 lysates prepared in lysis buffer was incubated overnight with 50 µl glutathione-822 conjugated magnetic agarose beads (Thermo Scientific) at 4ºC on a spinning rotor.

823
The beads were washed four times with the same buffer and the proteins were  Table   848 S9) using the Jetprime transfection reagent. On the same day, one set was exposed 849 to hypoxia (1% O2, overnight) and the third set was left in normoxia. On day 3, each Microsoft Excel. The differences between various groups was analyzed with a two 861 tailed Students t-test. Until specified, the error bars represent the standard error of 862 the mean with *p<0.05, **p<0.01, and ***p<0.001 denoting the difference between 863 the means of two compared groups considered to be statistically significant. Each      Glc as the only carbon source before and after the addition of oligomycin.

1221
(l, m) OCR traces of WT and KO HEK293T cells grown in media with Gal + Pyr (l) 1222 and Gln (m) as the only carbon sources.