Phosphatidylethanolamine made in the inner mitochondrial membrane is essential for yeast cytochrome bc1 complex function

Of the four separate PE biosynthetic pathways in eukaryotes, one occurs in the mitochondrial inner membrane (IM) and is executed by phosphatidylserine decarboxylase (Psd1p). Deletion of Psd1, which is lethal in mice, compromises mitochondrial function. We hypothesize that this reflects inefficient import of non-mitochondrial PE into the IM. To test this, we re-wired PE metabolism in yeast by re-directing Psd1p to the outer mitochondrial membrane or the endomembrane system. Our biochemical and functional analyses identified the IMS as the greatest barrier for PE import and demonstrated that PE synthesis in the IM is critical for cytochrome bc1 complex (III) function. Importantly, mutations predicted to disrupt a conserved PE-binding site in the complex III subunit, Qcr7p, impaired complex III activity similar to PSD1 deletion. Collectively, these data demonstrate that PE made in the IM by Psd1p is critical to support the intrinsic functionality of complex III and establish one likely mechanism.


39
Conservation of the Psd pathway from bacteria to humans likely reflects the 40 endosymbiotic origin of mitochondria which in turn suggests that mitochondrial PS and PE 41 metabolism has been preserved to optimize mitochondrial performance 6 . Indeed, deletion of 42 phosphatidylserine decarboxylase (Pisd in mouse and humans and PSD1 in yeast) in eukaryotic 43 cells decreases cellular growth, impairs oxidative phosphorylation, produces aberrant 44 mitochondrial morphology, and diminishes PE levels in cells and mitochondria 7,9-12 . The Psd 6 that the supplemental ethanolamine concentration was increased to 10mM), although subtle 130 differences between the strains were noted. Overall, these findings indicate that the Kennedy 131 pathway cannot fully compensate for the absence of Psd1p function, potentially due to 132 inefficient trafficking of PE from the ER to mitochondrial OM and/or from the OM to the IM.

133
Validation of constructs that re-direct Psd1p to OM or ER membranes 134 To interrogate whether the cytosol and/or the IMS is a barrier that prevents extra-135 mitochondrially produced PE from replacing PE normally made in the IM, we generated chimeric 136 Psd1p constructs that are localized to either the ER or OM to redirect PS and PE metabolism.  was not impaired by these modifications to its NH2-terminus ( Fig 2B). All three integrated 149 constructs were similarly expressed relative to each other and over-expressed compared to 150 endogenous Psd1p. Importantly, both OM-Psd1p and ER-Psd1p rescued the ethanolamine 151 auxotrophy of psd1Δpsd2Δ yeast indicating that these constructs are fully functional and 152 capable of generating levels of PE necessary for cellular growth (Fig 2C).

7
Previously, localization of ER-Psd1p was established by virtue of its enrichment in the 154 40,000 x g pellet (P40) after subcellular fractionation by gravity centrifugation and its sensitivity 155 to endoglycosidase H, which revealed a mobility shift following SDS-PAGE post-treatment 29 . To 156 confirm the OM localization of OM-Psd1p, its protease accessibility was determined in intact 8 similar to the OM control Tom70p, OM-Psd1p was completely degraded in intact mitochondria 161 verifying that it was successfully re-localized to the OM with the bulk of the enzyme facing the 162 cytosol. As expected given the presence of inter-organelle contact sites, a proportion of ER-163 Psd1p co-fractionated with crude mitochondria (Supplementary Fig 1) and demonstrated 164 protease-sensitivity in intact mitochondria (Fig 2D), a topology that is consistent with its N-165 glycosylation status (Fig 2E). Thus, a portion of ER-Psd1p is retained in the ER and/or resides 166 in an endosomal compartment that is co-purified with mitochondria.  abundance of PE (Fig 3A and 3G). In psd1Δ and psd1Δpsd2Δ membranes, the levels of 173 phosphatidylinositol (PI) was increased ( Figure 3B and 3H) and CL decreased (Fig 3C and 3I), 174 consistent with previous reports 9,28,33 . A compensatory increase in PS was notably absent in 175 these yeast strains (Fig 3D and 3J). While mitochondrial PE levels were modestly decreased in 176 the absence of Psd2p (Fig 3G), this decrease failed to result in a respiratory growth defect (Fig   177   1). Combined, these results indicate that Psd2p contributes to the pool of PE associated with 178 mitochondria which is nonetheless unable to functionally replace PE made by Psd1p.

179
Interestingly, OM-Psd1 and ER-Psd1 yeast contained significantly higher relative 180 amounts of PE than IM-Psd1 in both cellular and mitochondrial membranes (Fig 3A and 3G

208
A limitation of the radiolabeling-based phospholipid analysis is that it utilized crude 209 mitochondria isolated after physical disruption of intact yeast with glass beads. As such,

235
To directly assess OXPHOS capacity in these strains, oxygen consumption was 236 monitored in isolated mitochondria using an O2 electrode after the addition of ADP and 237 ascorbate tetramethyl-p-phenyldiamine (TMPD) which promotes proton pumping by complex IV 11 ( Fig 4B-4E    which is capable of rescuing strains with reduced CoQ biosynthesis 39 , failed to improve 300 respiratory growth of psd1Δ or psd1Δpsd2Δ yeast ( Supplementary Fig 8). Lastly, psd1Δ, 301 psd1Δpsd2Δ, OM-Psd1, and ER-Psd1 respiratory growth was not further impaired in medium 302 lacking para-amino benzoic acid ( Supplementary Fig 9), a molecule that can be used to produce  significantly increased CL in psd1Δpsd2Δ yeast to psd1Δ levels (Fig 6C and 6G). Significant 15 changes in the abundance of other phospholipid species in psd1Δpsd2Δ yeast supplemented 323 with either choline or ethanolamine were not observed ( Supplementary Fig 10). As such, these 324 results indicate that the Kennedy Pathway for PE but not PC production, is metabolically-linked 325 to CL biosynthesis and/or stability. Moreover, these results suggest that the ability of 326 ethanolamine to improve complex IV activity in psd1Δpsd2Δ yeast is due to its unanticipated 327 capacity to increase CL levels, a phospholipid known to be important for complex IV function 42 .

328
These findings further underscore that PE made within the IM is necessary for the full activity of    Psd1p accumulation or maturation 29 ( Fig 7B) and resulted in an ethanolamine auxotrophy 43 339 ( Fig 7C). As anticipated, PS and PE levels were drastically reduced in strains lacking Cho1p 340 (Fig 7D-F). In comparison to psd1Δpsd2Δ, cho1Δpsd1Δpsd2Δ yeast had a significant increase 341 in CL and PI levels (Fig 7G and 7H), the former of which may be associated with its enhanced 342 respiratory growth (Fig 7K). Deletion of cho1Δ in the psd1Δpsd2Δ background restored PC to

343
WT levels ( Fig 7J) but failed to increase the levels of the CL precursor PA (Fig 7I). Growth of 344 cho1Δ was decreased compared to WT in YPD and rich lactate medium and similar to 345 psd1Δpsd2Δ in SCEG supplemented with ethanolamine ( Fig 7K). Notably, the activities of 346 complex III ( Fig 7L) and IV (Fig 7M) were reduced in the absence of Cho1p. Since Psd1p and 347 essential subunits of complexes III and IV were expressed normally in the absence of Cho1p, singly or in combination with Psd1p and Psd2p (Fig 7N)   interaction. Importantly, the amount of Qcr7p E82R or Qcr7p E82D detected in cell and mitochondrial 362 lysates (Fig 8B and 8C) was similar to WT indicating that neither mutation compromised Qcr7p 363 stability (the Qcr7p E82R variant was consistently upshifted compared to WT following SDS-364 PAGE). Further, Qcr7p E82R and Qcr7p E82D supported respiratory growth in rich or minimal 365 medium ( Fig 8D). Despite being sufficiently functional to promote respiratory growth, complex III 366 activity was significantly decreased for Qcr7p E82R and Qcr7p E82D to a similar degree as when 367 Psd1p is missing (Fig 8E). Surprisingly, complex IV activity was also decreased in Qcr7p E82R but 368 not Qcr7p E82D (Fig 8F). The impaired respiratory complex activity for Qcr7p E82R and Qcr7p E82D 369 was independent of any changes in the steady state amount of their subunits, (Fig 8C). These

385
Given these conflicting results, we first tested the ability of ethanolamine supplementation to 386 improve the respiratory growth phenotype of yeast lacking Psd1p. While we observed that 387 growth of psd1Dpsd2D could be rescued on respiratory medium supplemented with 388 ethanolamine, growth after 3 days was delayed and impaired similarly to psd1D yeast.

389
Perplexingly, supplementation of psd1Dpsd2D with ethanolamine to generate PE through the 390 Kennedy pathway did not significantly increase mitochondrial PE levels and instead resulted in 391 a significant, albeit modest, increase in CL (Fig 6C-F

413
( Fig 5F and Fig 6B). Since complex III was impaired in yeast deficient in mitochondrial PS and 414 PE that contained a normal amount of fully processed Psd1p (cho1D, Fig 7K)

444
Intriguingly, the improved complex IV functionality provided by ethanolamine correlated 445 best with an increased abundance of CL and not PE (Fig 6C-F)

454
Depletion of PS in the psd1Dpsd2D background restored CL levels to that of psd1D and cho1D 455 which were still comparatively reduced vs WT (Fig 7G). This increase in CL also coincided with 456 improved growth for cho1Dpsd1Dpsd2D yeast (Fig 7K). It is possible that in the absence of PS, 457 metabolic pathways that either promote PA formation in the ER 56 or promote PA import to the 458 IM 57,58 are stimulated. Moving forward, it will be important to distinguish between these non-459 mutually exclusive models.

460
Another outstanding question raised by our study is exactly how PE made by Psd1p in

482
These structural observations will guide future efforts to determine the role(s) of PE as it relates 483 to complex III activity.

487
Yeast strains and growth conditions

488
All yeast strains used in this study are listed in Table 1

557
To improve growth on rich lactate, psd1Δpsd2Δ yeast were grown in the presence of 2mM 558 choline prior to harvesting mitochondria (except where indicated in Fig 6A,B and in Fig 7).

559
Purification of mitochondria by sucrose step gradient was performed as previously described 29 .

610
After state 4 rate was measured, 10µM CCCP was added to induce uncoupled respiration, and 611 the rate was followed for either 2 minutes or until oxygen level reached zero.

613
Complex III and IV activities were measured as described 66,67 . To measure complex III activity,

630
Preparation of yeast cell extracts, submitochondrial fractionation, phospholipid analysis, 1D BN-631 PAGE, and immunoblotting were performed as described previously 17,29 . Immunoblots using IR 632 800 CW secondary antibodies were imaged using an Odyssey CLx Imaging System.

633
Immunoblots and TLC plates were quantitated by Quantity One Software by Bio-Rad  Supplementary Fig 7. Quantitation of CoQ synthome subunits and additional mitochondrial proteins. Densitometry analysis of steady state protein amounts in isolated mitochondria (30 µg) from the indicated strains (representative immunoblots shown in Fig 5F). Analysis versus WT (*) or psd1Dpsd2D (#) was performed by one-way ANOVA ± S.E.M. for n=4.