Cardiolipin deficiency disrupts electron transport chain to drive steatohepatitis

60 Metabolic-dysfunction associated steatotic liver disease (MASLD) is a growing global health 61 concern with an increasing prevalence that parallels the rise in obesity.1 In the United States, 62 annual medical costs related to MASLD exceed $103 billion.2 A large portion of patients with 63 MASLD only exhibit steatosis, a silent and relatively benign early stage characterized by lipid 64 accumulation in hepatocytes without hepatocellular inflammation.3 Steatosis can then progress 65 to metabolic-dysfunction associated steatohepatitis (MASH), determined by hepatocyte injury 66 and tissue fibrosis.4 MASH is the last stage of MASLD that may be reversible, making 67 intervention at this stage particularly important.3,5 Although extensive clinical and basic research 68 have been conducted in this field, the underlying mechanisms by which fatty liver transitions to 69 MASH remain poorly understood.6-8 70 71 A defect in mitochondrial function is considered one of the hallmarks of MASLD progression in 72 both mice and humans.9-12 MASLD is initially associated with an increase in mitochondrial 73 respiratory capacity, followed by a subsequent impairment in oxidative phosphorylation 74 (OXPHOS), and increased production of mitochondrial reactive oxygen species (ROS).11,13 75 Mitochondrial ROS is thought to be caused by an inefficient electron transport chain (ETC) that 76 increases the propensity for electron leak. However, the mechanisms by which mitochondrial 77 electron leak promotes MASLD are unknown. 78 79 Cardiolipin (CL) is a phospholipid with four acyl chains conjugated to two phosphatidylglycerol 80 moieties linked by another glycerol molecule.14 CL resides almost exclusively in the inner 81 mitochondrial membrane (IMM), constituting approximately 15–20% of the mitochondrial 82 phospholipids.15 CL is synthesized by the condensation of phosphatidylglycerol (PG) and 83 cytidine diphosphate-diacylglycerol (CDP-DAG) at the IMM via the enzyme cardiolipin synthase 84 (CLS).16,17 Structural studies indicate that CL is essential for the activities of OXPHOS 85 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint enzymes.18-22 In non-hepatocytes, decreased CL leads to compromised oxidative capacity,23,24 86 impaired membrane potential,25 and altered cristae morphology.26 In particular, low CL is 87 associated with increased H2O2 production.27,28 88 89 In this manuscript, we set out to examine the changes in liver mitochondrial lipidome induced by 90 MASH. Mitochondrial CL was downregulated in four mouse models of MASLD. We then 91 performed a targeted deletion of CLS in hepatocytes and studied its effects on liver, 92 mitochondrial bioenergetics, and potential mechanisms that drive these changes. 93 94

95 Mitochondrial cardiolipin levels are decreased in mouse models of MASLD/MASH 96 Previous research from our lab in non-hepatocytes indicated that mitochondrial phospholipid 97 composition affects OXPHOS electron transfer efficiency to alter electron leak.15,29,30 MASLD 98 has been shown to alter the total cellular lipidome in liver.31 However, MASLD may also 99 influence mitochondrial content in the hepatocytes, making it difficult to discern whether these 100 are changes in the lipid composition of mitochondrial membranes and/or changes in cellular 101 mitochondrial density. Thus, we performed liquid chromatography-tandem mass spectrometry 102 (LC-MS/MS) lipidomics specifically on mitochondria isolated from four models of MASLD/MASH 103 (Figure 1). These included: 1) mice given a Western high-fat diet (HFD, Envigo TD.88137) or 104 standard chow diet for 16 weeks (Figure 1A), 2) ob/ob mice or their wildtype littermates at 20 105 weeks of age (Figure 1B), 3) mice given the Gubra Amylin NASH diet for 30 weeks (GAN, 106 Research Diets D09100310) or standard chow (Figure 1C), 4) mice injected with carbon 107 tetrachloride (CCI4) or vehicle (corn oil) for 6 weeks (Figure 1D). Importantly, none of these 108 interventions appear to alter the protein abundances of OXPHOS subunits or citrate synthase 109 (Figures 1E, 1F, 1G, and 1H), suggesting that these interventions did not alter mitochondrial 110 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint density in hepatocytes. Nevertheless, we performed the mitochondrial lipidomic analyses by 111 quantifying lipids per mg of mitochondrial proteins. 112 113 Each intervention appeared to alter different subsets of mitochondrial lipid classes (Figures 1I-L, 114 S1, and S2), as seen with our previous studies in skeletal muscle and brown adipose 115 tissues.29,30 We take these observations to mean that most physiological interventions induce 116 multiple systemic and local responses that are not mechanistically directly related to the 117 phenotype of interest (e.g., cold exposure or exercise can increase food intake, obesity could 118 affect locomotion and insulation, etc.). Although several phospholipid classes were altered 119 among the four models, strikingly, mitochondrial CL was reduced in all four MASLD/MASH 120 models (Figure 1I-M). Furthermore, PG, an essential substrate for CL synthesis, was 121 significantly increased in all MASLD/MASH models (Figure 1I-M). These changes coincided with 122 decreased transcript levels for CLS (Figure 1N, 1O, 1P, and 1Q). These observations suggest 123 that an insult in CL synthesis may be a key factor to disrupting mitochondrial function in 124 MASLD/MASH. 125 126 Hepatocyte-specific deletion of cardiolipin synthase promotes MASH 127 CL is thought to be exclusively synthesized in the IMM where CLS is localized. To study the role 128 of CL in hepatocytes, we generated mice with hepatocyte-specific knockout of CLS (CLS-LKO 129 for CLS liver knockout, driven by albumin-Cre) (Figures 2A and 2B), which successfully 130 decreased mitochondrial CL levels (Figure 2C and S3). Consistent with our previous studies in 131 non-hepatocytes, CLS deletion does not completely reduce CL levels to zero, suggesting that 132 CL generated in other tissues may be imported. Our results showed that decreased levels of CL 133 did not significantly impact body weight or composition (Figures 2D and 2E) but resulted in 134 significantly less liver mass (Figure 2F). 135 136 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint We sought to further characterize livers from control and CLS-LKO mice. Histological analyses 137 revealed that CLS deletion was sufficient to promote steatosis (Figure 2G) and fibrosis (Figure 138 2H) in standard chow-fed and high-fat fed conditions (Figures S4A and S4B). To more 139 comprehensively describe the effects of loss of hepatic CLS on gene expression, we performed 140 RNA sequencing on these livers. CLS deletion increased the expression of 713 genes and 141 decreased 1026 genes (Figure S4C). Pathway analyses revealed that many of the signature 142 changes that occur with MASLD/MASH also occurred with CLS deletion (Figures 2I and S4D). 143 This MASLD/MASH phenotype in our CLS knockout model was further confirmed with an 144 elevation of the liver enzymes AST and ALT (Figures 2J and 2K) as well as increased mRNA 145 levels of inflammatory markers (Figure 2L). We then proceeded to confirm these data by further 146 phenotyping liver tissues from control and CLS-LKO mice. 147 148 In steatohepatitis, immune cell populations in the liver become altered to activate pathological 149 immune response.32 Flow cytometry on livers from control and CLS-LKO mice indicated that the 150 loss of CL promotes a robust classic immune response found in MASH (Figure 2M). cDC2 cells 151 are a broad subset of dendritic cells with specific surface markers (e.g., CD11b, CD172a) that 152 allow them to be distinguished from other dendritic cell populations.33 This broad population of 153 dendritic cells was not different between control and CLS-LKO mice (Figure 2N). Notably, there 154 was a marked reduction in the Kupffer cell population (Figure 2O) - traditionally involved in 155 maintaining liver homeostasis whose dysfunction can lead to dysregulated immune response. 34 156 This reduction appears to be counterbalanced by a concomitant increase in Ly6Chi population, 157 which are known to typically go on to become inflammatory monocytes (Figures 2M and 2P). 158 The replacement of Kupffer cells with other inflammatory cell populations suggests a shift 159 towards a more pro-inflammatory environment, which may exacerbate liver injury and promote 160 fibrosis. Nonetheless, the MHC-II cell population and neutrophils were not increased (Figures 161 2Q and 2R) with neutrophils actually decreased (Figure 2R). The cDC1 cell population was not 162 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint different, which is traditionally elevated in response to cytotoxic T cells and might not be directly 163 related to liver fibrosis.35 Together, these findings suggest that even on a chow diet, CLS-164 deficient livers exhibit inflammatory cell infiltration, a hallmark often associated with early signs 165 of MASH. 166 167 CLS deletion promotes fatty liver but increases mitochondrial respiratory capacity 168 Hepatocyte lipid accumulation may suggest defects in substrate handling, which is often 169 manifested in systemic substrate handling. Indeed, CLS deletion modestly reduced glucose or 170 pyruvate handling, even in chow-fed conditions (Figures 3A-D). Lipid accumulation in 171 hepatocytes can occur due to an increase in lipogenesis, a decrease in VLDL secretion, or a 172 decrease in β-oxidation. However, mRNA levels for lipogenesis genes trended lower (not 173 higher), and mostly unchanged for VLDL secretion or β-oxidation (Figure 3E, Figure S5A). 174 Circulating triglycerides were not lower in CLS-LKO mice compared to control mice (Figure 175 S5B). 176 177 MASLD is known to be associated with reduced mitochondrial oxidative capacity, and such an 178 effect may also occur with CL deficiency to induce lipid accumulation. Indeed, mRNA levels of 179 several genes in the ETC were downregulated with CLS deletion, particularly those associated 180 with structural components of the ETC complexes and the electron carrier CoQ (Figure 3F). 181 Given that CL is located in the IMM where it binds to enzymes involved in OXPHOS,36-39 we 182 reasoned that the loss of CL could reduce mitochondrial oxidative capacity to promote steatosis. 183 Consistent with subcellular localization of CL, CLS deletion resulted in mitochondria with 184 disorganized membrane structures and poorly developed cristae (Figure 3G). However, 185 mitochondrial density quantified with western blots for respiratory complex subunits and citrate 186 synthase (Figure 3H), as well as mtDNA/nucDNA (Figure 3I), showed no differences in livers 187 from control and CLS-LKO mice. We thus speculated that CL lowers respiratory capacity not by 188 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint reducing the total number of mitochondria or OXPHOS respirasomes, but by reducing the 189 activity of respiratory enzymes. To our surprise, CLS deletion increased, rather than decreased, 190 mitochondrial respiration (JO2), as measured by high-resolution Oroboros respirometry (Figure 191 3J), using both with Krebs cycle substrates (Figure 3K) as well as fatty acyl substrates (Figure 192 3L). In fact, the increase in respiration induced by CLS deletion was more pronounced with fatty 193 acyl substrates than with Krebs Cycle substrates. Importantly, these changes occurred in the 194 absence of OXPHOS subunit abundance per unit of mitochondria (Figure 3M), ruling out the 195 possibility that changes in the abundance of respiratory enzymes to contribute to change in 196 respiration. A caveat to these findings is that CLS deletion promotes reduction in respiratory 197 capacity after HFD-feeding (Figures S5C and S5D). However, CLS-LKO mice are steatotic in 198 standard chow-fed condition, indicating that reduced mitochondrial fatty acid oxidation cannot 199 be the cause of steatosis at baseline. The transient increase in respiration followed by its 200 subsequent decrease is reminiscent of what is thought to occur with liver’s mitochondrial 201 respiration over the course of MASLD progression.40 202 203 High-resolution respirometry experiments were performed in isolated mitochondria from 204 hepatocytes by providing exogenous supraphysiological concentrations of substrates. While 205 these assays provide robust measurements of respiratory capacity (the potential of 206 mitochondria), they do not necessarily reflect their endogenous activity. To address this point, 207 we performed stable isotope tracing experiments using uniformly labeled 13C-palmitate in 208 murine hepa1-6 cells with or without CLS knockdown (Figure 4A).41 Surprisingly, but consistent 209 with the JO2 data, CLS deletion increased, not decreased, the incorporation of palmitate into 210 TCA intermediates (Figures 4B-D). We also performed a similar tracing experiment using 211 uniformally labeled 13C-glucose (Figure 4E-J, S5E-I) and observed increased labeling towards 212 pyruvate (Figure 4E and 4F), reduced labeling towards lactate and alanine (Figures 4G and 213 S5H), and normal labeling towards TCA intermediates except for reduced labeling towards 214 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint succinate (Figure 4H-J, S5I-M). Overall, despite the altered substrate incorporation, a decrease 215 in TCA flux does not appear to account for the steatotic phenotype observed with CLS deletion. 216 217 Low hepatic CL induces mitochondrial electron leak at IIF and IIIQ0 sites 218 Oxidative stress is thought to play a critical role in the transition from MASLD to MASH, wherein 219 sustained metabolic insult leads to hepatocellular injury and collagen deposition resulting in 220 fibrosis.7 CLS deletion promotes liver fibrosis in standard chow-fed condition (Figures 5A and 221 2H) and in HFD-fed condition (Figure S4B) that coincided with increased mRNA levels for 222 fibrosis (Figure 5B and 2I). Tissue fibrosis is often triggered by apoptosis, and CLS deletion 223 appeared to activate the caspase pathway (Figures 5C and 5D). How does deletion of CLS, a 224 mitochondrial enzyme that produces lipids for IMM, activate apoptosis? Cytochrome c is an 225 electron carrier that resides in IMM, which shuttles electrons between complexes III and IV.42 226 Under normal physiological conditions, cytochrome c is anchored to the IMM by its binding to 227 cardiolipin.39 During the initiation of intrinsic apoptosis, CL can undergo oxidation and 228 redistribution from the IMM to the outer membrane space (OMM). CL oxidation weakens its 229 binding affinity for cytochrome c, releasing it from the IMM and into the OMM where it signals 230 apoptosis.13 However, neither mitochondrial nor cytosolic cytochrome c abundance appeared to 231 be influenced by CLS deletion (Figures 5E, 5F, S6A, and S6B). 232 233 Mitochondrial ROS has been implicated in apoptosis and fibrosis with MASLD.43-45 Using high-234 resolution fluorometry in combination with high-resolution respirometry, we quantified electron 235 leak from liver mitochondria with the assumption that almost all electrons that leak react with 236 molecular O2 to produce O2-. Using recombinant superoxide dismutase, we ensure that all O2- 237 produced is converted into H2O2, which was quantified with the AmplexRed fluorophore.46 There 238 were striking increases in mitochondrial electron leak in CLS-LKO mice compared to control 239 mice on both standard chow (Figure 5G) and high-fat diet (Figure S6C). It is noteworthy that 240 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint endogenous antioxidant pathways were insufficient to completely suppress oxidative stress 241 induced by CLS deletion (H2O2 emission shown in the 1st and 2nd bars in Figure 5G and S6C). 242 We also confirmed that JH2O2/JO2 was elevated with CLS knockdown in mitochondria from 243 murine hepa1-6 cell line (Figure S6D) suggesting that low CL induces oxidative stress in a cell-244 autonomous manner. 245 246 While unknown, CLS may possess an enzymatic activity independent of CL synthesis that may 247 contribute to electron leak. To more conclusively show that the loss of mitochondrial CL 248 contributes to oxidative stress, we supplied exogenous CL to isolated mitochondria by fusing 249 them with small unilamellar vesicles (SUVs) (Figure 5H).47 Isolated mitochondria from control 250 and CLS-LKO mice were fused with SUVs containing either CL or phosphatidylcholine (PC) 251 (Figures 5I and S6E). Remarkably, reintroducing CL to mitochondria from CLS-LKO mice 252 reduced H2O2 production back to baseline, whereas PC had no effect. Thus, loss of CL drives 253 the increased mitochondrial leak observed with CLS deletion. 254 255 How does low CL promote mitochondrial electron leak? CL is likely ubiquitous in IMM and can 256 bind to all four respiratory complexes of the ETC.20,21,36,48 There are four known sites of electron 257 leak in the IMM: 1) quinone-binding site in complex I (IQ), 2) flavin-containing site in complex I 258 (IF), 3) succinate-dehydrogenase-associated site in complex II (IIF), and the ubiquinol oxidation 259 site in complex (IIIQo) Figure 6). Electron leak at each of these sites can be quantified separately 260 using substrates and inhibitors that restrict electron flow specific to these sites. All of these sites 261 are localized to IMM, suggesting that CL has the potential to increase electron leak in any of 262 these sites. Indeed, quantification of site-specific electron leak demonstrated that CLS deletion 263 essentially increased electron leak in all these sites (Figures 6A, 6B, 6C, and 6D). 264 265 Loss of CL promotes inefficiency in coenzyme Q-dependent electron transfer 266 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint How does the loss of CL promote electron leak at these sites? We initially addressed whether 267 CL influences the formations of respiratory supercomplexes. Respiratory supercomplexes exist 268 in several combinations of multimers of Complex I, III, IV, and V and are thought to form either 269 transiently or stably to improve electron transfer efficiency.48,49 CL may play an essential role in 270 the stability of ETC supercomplexes.50,51 Using blue native polyacrylamide gel electrophoresis 271 followed by subunit-specific western blotting, we investigated supercomplex assembly in 272 isolated hepatic mitochondria from control and CLS-LKO mice (Figures 6E-P). Abundances of 273 supercomplexes associated with CIII (Figure 6L) as well as CV (Figure 6P) were reduced, while 274 singlets for CII (Figure 6J), CIII (Figure 6L), and CIV (Figure 6N) were increased in mitochondria 275 from CLS-LKO mice compared to control mice. Nevertheless, in our opinion, these changes 276 were somewhat underwhelming in that: 1) among ETC, only one of the supercomplexes, one 277 associated with CIII (I + III2 + IV1), was reduced out of nine total, and 2) the magnitude of the 278 change in supercomplex formation appeared so trivial compared to the magnitude of electron 279 leak that was observed in sites IF, IIF, and IIIQ0. Thus, while loss of some CIII supercomplexes 280 may be a contributor, we did not find these data robust enough and reasoned that there was 281 another mechanism by which CL influenced electron transfer efficiency. 282 283 Upon re-examining our site-specific electron leak data (Figure 6A-D), we noted that increases in 284 electron leak were greater at sites IIF and IIIQ0, and that these sites were proximal to coenzyme 285 Q (CoQ). CoQ, like CL, is a lipid molecule (Figure 7A), and we thought it was possible that CL 286 somehow interacts with CoQ to influence its electron transfer efficiency. Using redox mass 287 spectrometry, we measured CoQ levels in whole liver tissues from control and CLS-LKO mice 288 and found no difference in whole liver tissue CoQ levels (Figures S7A-I). However, since CoQ 289 may also be found outside of mitochondria, we performed CoQ redox mass spectrometry in 290 isolated mitochondria fractions of livers from control or CLS-LKO mice. Indeed, oxidized CoQ 291 levels were increased (Figures 7B, S7J, S7L, and S7N) in CLS-LKO mice compared to their 292 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint controls. In contrast, reduced forms of CoQ were not lower in CLS-LKO mice compared to 293 control mice (Figures 7C, S7K, S7M, and S7O). These findings indicate how CL deficiency 294 might influence CoQ-dependent electron transfer. First, low CL increased the abundance of 295 oxidized CoQ, but these oxidized CoQ were unable to become reduced at sites IQ or IIQ. Thus, 296 loss of CL appears to decrease the ability of CoQ to accept electrons, promoting electron leak at 297 IF, IQ, and IIF sites. Second, there must be a second defect, as there was a substantial increase 298 in electron leak from site IIIQ0 (Figure 6D). It can be postulated that CoQ must also be less 299 capable of efficiently donating electrons to complex III. This would be consistent with the data 300 that greater oxidized CoQ was observed despite having a normal reduced CoQ level. 301 302 Electron leak from site IIF was greater than those observed in sites IF and IQ. Data from the 303 stable isotope experiments supports this notion, where CLS deletion reduced labeling of 304 succinate indicating reduced complex II/succinate dehydrogenase (SDH) activity (Figure 4H). 305 Steady-state metabolomics (Figure S7P) also revealed reduced succinate-to-fumarate ratio, 306 suggesting reduced SDH activity (Figures 7D, S7Q, and S7R). Interestingly, in an assay that 307 measures SDH activity in a detergent-containing assay that removes CL, CLS deletion had no 308 effect (Figure 7E). Thus, loss of CL likely influences multiple processes in the ETC to increase 309 mitochondrial ROS production. 310 311 CL and CoQ are co-downregulated in liver biopsy samples from MASH patients 312 We further assessed the relationship between CL and CoQ by assessing their levels in liver 313 samples from patients undergoing liver transplant or resection due to end-stage MASH and/or 314 hepatocellular carcinoma (Table S1). A portion of liver that did not have tumor was isolated and 315 analyzed. Liver samples from patients without MASH, undergoing resection for bening live 316 rumors or metastases, were classified as healthy controls (Figure 7F). Similar to our 317 experiments in mice, we isolated mitochondria from these liver tissues and performed targeted 318 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint lipid mass spectrometry to quantify CoQ and CL. These analyses revealed striking decreases in 319 both CL (Figure 7G) and CoQ (Figure 7H) induced by MASH (tissue samples were not large 320 enough to perform CoQ redox mass spectrometry on mitochondria). A Pearson correlation 321 analysis showed a highly robust correlation between the abundances of mitochondrial CL and 322 CoQ (Figure 7I, R2 = 0.64), indicating that the variability in the abundance of CL explains 64% of 323 the variability of the abundance of CoQ. Based on our findings that CL is reduced with 324 MASLD/MASH and that loss of CL influences CoQ electron transfer efficiency, we interpret 325 these findings to mean that loss of CL destabilizes CoQ to increase its turnover. 326 327

328 In hepatocytes, disruptions of mitochondrial bioenergetics lead to and exacerbate metabolic-329 associated steatohepatitis.52 CL, a key phospholipid in the inner mitochondrial membrane, plays 330 a critical role in mitochondrial energy metabolism.23 In this manuscript, we examined the role of 331 CL in the pathogenesis of MASLD. In mice and in humans, MASLD/MASH coincided with a 332 reduction in mitochondrial CL levels. Hepatocyte-specific deletion of CLS was sufficient to 333 spontaneously induce MASH pathology, including steatosis and fibrosis, along with shift in 334 immune cell populations towards a more pro-inflammatory profile, all of which occurred in mice 335 given a standard chow diet. Paradoxically, high-resolution respirometry and stable isotope 336 experiments showed that CLS deletion promotes, instead of attenuates, mitochondrial oxidative 337 capacity in a fashion reminiscent of temporal changes that occur with mitochondrial 338 bioenergetics in human MASH.40 Our principal finding on the role of hepatocyte CL in 339 bioenergetics is that its loss robustly increases electron leak, particularly at complexes II and III. 340 This was likely due to the influence of CL on mitochondrial CoQ, whereby CLS deletion lowered 341 CoQ’s ability to efficiently transfer electrons. In humans, mitochondrial CL and CoQ were co-342 downregulated in MASH patients compared to healthy controls, with a strong correlation (R² = 343 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint 0.64) between CL and CoQ. Together, these results implicate CL as a key regulator of MASH 344 progression, particularly through its effect on CoQ redox state to promote oxidative stress. 345 346 How might CL regulate CoQ? CoQ is the main electron transporter between complex I/II and III. 347 CLS deletion in hepatocytes appeared to disrupt CoQ's ability to cycle between its oxidized and 348 reduced forms. There are several ways in which low CL might directly or indirectly influence 349 CoQ’s redox state. The primary suspect is CL interacting with complex III, as eight or nine CL 350 molecules are tightly bound to complex III38 and are thought to be essential to its function.53 351 While CL has been found to bind to other respiratory complexes, our data suggest that loss of 352 CL might disproportionately influence complex III. This is also supported by our findings that the 353 loss of CL reduced the formation of complex III-dependent supercomplex, without influencing 354 other supercomplexes. Reduced capacity for complex III to efficiently accept electrons from 355 CoQ might explain the increased electron leak at site IIIQ0 and increased level of oxidized CoQ. 356 Meanwhile, loss of CL also likely influences complex I and II, as CL has also been implicated to 357 bind these complexes.20,21,36 Complex III dysfunction is unlikely to entirely explain electron leaks 358 at sites IF, IQ, and IIF, though it is conceivable that the reduced ability of complex III to accept 359 electrons creates a bottleneck that produces electron leak at other sites, including reverse 360 electron transport at complex I.54 Conversely, complex I and II are unlikely to be the only 361 primary sites of defect as such defects probably will not promote electron leak at site IIIQO. 362 Another potential mechanism by which CL influences CoQ redox state is by CL directly 363 interacting with CoQ. As they are both lipid molecules in the IMM, low CL may reduce the lateral 364 diffusability of CoQ between respiratory complexes. Low CL might also indirectly influence CoQ 365 by contributing to changes in membrane properties, distribution of ETC in the cristae, and the 366 cristae architecture.55 Finally, increased electron leak, regardless of their origin, could have a 367 feed-forward effect by which oxidative stress disrupts redox homeostasis in other components 368 of ETC. 369 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint 370 MASH is a progressive liver disease characterized by lipid accumulation, inflammation, and 371 fibrosis in the liver.56 The progression to MASH involves a complex interplay of metabolic stress, 372 mitochondrial defects, and immune responses that collectively promote hepatocellular injury.57 373 Our findings suggest that the low mitochondrial CL level directly induces key pathological 374 features of MASH, including steatosis, fibrosis, and immune cell infiltration, even in the absence 375 of dietary or environmental stressors, such as high-fat diet. When mice were fed a standard 376 chow diet, CLS-LKO mice were more prone to lipid droplet accumulation than control mice. This 377 phenotype was exacerbated when the mice were challenged with a high-fat diet. We primarily 378 interrogated the mitochondrial bioenergetics of standard chow-fed control or CLS-LKO mice. A 379 lower respiration rate would partially explain the lipid droplet accumulation, but to our surprise, 380 CLS deletion increased JO2 regardless of substrates. Similarly, experiments using uniformly 381 labeled 13C-palmitate or 13C-glucose showed that CLS deletion promoted an overall increase in 382 the flux toward TCA intermediates, particularly for palmitate. CLS deletion did not appear to 383 increase de novo lipogenesis or reduce VLDL secretion. Thus, it is unclear what mechanisms 384 contribute to steatosis induced by CLS deletion. 385 386 Liver fibrosis is characterized by the accumulation of excess extracellular matrix components, 387 including type I collagen, which disrupts liver microcirculation and leads to injury.55 Livers from 388 CLS-LKO mice exhibited more fibrosis compared to control mice, even on a standard chow diet, 389 which was worsened when fed a high-fat diet. Indeed, transcriptomic analyses revealed that 390 CLS deletion activates pathways for fibrosis and degeneration, with many of the collagen 391 isoforms upregulated. In the MASH liver, collagen deposition is accompanied by inflammatory 392 cell infiltrate promoting an overall inflammatory environment. Flow cytometry experiments 393 further confirmed that CLS deletion led to an increase in Ly6Chi cell population, suggesting that 394 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint dying resident Kupffer cells are being replaced by Ly6Chi monocytes in the livers of CLS-LKO 395 mice.33 396 397 Early in the MASLD disease progression, mitochondria adapt to the increased energy demands 398 by increasing their respiratory capacity. In the later stages of disease progression to MASH, 399 mitochondrial respiration diminishes.58 This pattern was reminiscent of our observations in the 400 CLS-LKO mice. Livers from CLS-LKO mice fed a standard chow diet exhibited greater 401 respiratory capacity compared to that of control mice. Conversely, livers from CLS-LKO mice on 402 a high-fat diet exhibited lower respiratory capacity compared to that of control mice. We interpret 403 these findings to suggest that liver mitochondria in chow-fed CLS-LKO mice are more 404 representative of early stage of MASLD, while high-fat-diet fed CLS-LKO mice resemble later 405 stages of MASLD. 406 407 In non-hepatocytes, low CL levels have been linked to electron leak in the context of a 408 deficiency of the tafazzin gene, a CL transacylase, whose mutation promotes Barth 409 syndrome.15,26,59,60 Paradoxically, we previously showed that CLS deletion in brown adipocytes 410 does not increase electron leak.30 It is important to note that CLS deletion in our current or 411 previous study does not completely eliminate CL (likely due to an extracellular source). We do 412 not believe that CL is dispensable for efficient electron transfer in adipocytes. Rather, due to 413 unclear mechanisms, different cell types likely exhibit varying tolerances to low CL influencing 414 their bioenergetics, with brown adipocytes appearing more tolerant than hepatocytes. 415 Regardless, electron leak was elevated with CLS deletion in both standard chow and high-fat 416 diet-fed conditions. These observations mirror what has been shown in MASLD progression.61 417 The effect of CLS deletion on electron leak was due to reduced CL levels, as the reintroduction 418 of cardiolipin via SUVs completely rescued the electron leak. 419 420 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint CL is reported to be essential for the formation and stability of supercomplexes.14,22,50,51 CL has 421 a distinctive conical shape with four fatty acyl chains, which allows it to create a highly curved 422 membrane environment in the IMM, promoting close packing of protein complexes that likely 423 facilitates supercomplex formation.50 CL also directly interacts with various subunits of the ETC 424 complexes through electrostatic interactions, which help stabilize the supercomplexes by 425 anchoring them together in a specific spatial orientation to optimize electron flow.49 Somewhat 426 surprisingly, liver-specific deletion of CLS only resulted in a lower abundance of one of many 427 supercomplexes associated with CIII (I+III2+IV1) as well as the CV multimer (Vn). Because CLS 428 deletion did not completely deplete CL, we interpret these findings to mean that I+III2+IV1 and Vn 429 supercomplexes are particularly sensitive to the reduced CL level in hepatocytes. 430 431 In our study, we observed a striking correlation between CL levels and CoQ in human liver 432 samples from healthy/MASH patients (R2 of 0.64). In contrast, low mitochondrial CL induced by 433 CLS deletion coincided with a greater mitochondrial CoQ content in CLS-LKO mice compared to 434 controls. These data likely suggest that acute and robust reduction in CLS or CL level might 435 trigger a compensatory CoQ production in mice. Conversely, because the samples from 436 healthy/MASH patients were from those who had HCC, MASH samples likely came from 437 subjects that had suffered from years of MASLD pathology. In those samples, where reduction 438 in CL was quantitatively modest compared to what was induced with CLS knockout, CoQ might 439 have gradually decreased coincidental to the decrease in CL. Regardless of these differences in 440 mice and humans, it is clear that there is a relationship between CL and CoQ that is worth 441 further exploration. 442 443 In conclusion, our findings identify a critical role for CL in regulating CoQ redox state to promote 444 oxidative stress. In both mice and humans, MASLD is associated with a decrease in hepatic 445 mitochondrial CL, suggesting that low CL may be the cause of the obligatory increase in 446 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint oxidative stress known to occur with MASLD progression. We further link CL deficiency to 447 increased electron leak at Complexes II and III as sites that likely interact with CoQ to promote 448 oxidative stress. We believe that these bioenergetic changes underlie the pathogenesis of 449 MASLD, as CL deletion was sufficient to cause steatosis, fibrosis, and inflammation, 450 phenocopying many changes that occur with MASLD/MASH progression. Further research will 451 be needed to fully uncover how CL regulates CoQ, and to test whether rescuing the CL/CoQ 452 axis might be effective in treating patients with MASLD/MASH. 453 454 455 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint

456 This research was supported by National Institutes of Health (NIH) grants DK127979, 457 GM144613, DK107397, AG074535 (to KF); DK127603 (to AMP), DK130555 (to ADP); 458 HL170575, DK112826 (to WLH); GM151245 (to SMN), CA278803 (to KHF-W); DK128819, 459 DK115991 (to PNM); CA222570 (to KJE); CA272529, DK130296, DK131609 (to SAS); 460 DK091317 (to MJB, TST, STD), by American Heart Association grants 915674 (to PS) and 461 19PRE34380991 (to JMJ), by European Research Council (ERC) under the European Union’s 462 Horizon 2020 Research and Innovation Programme (Starting Grant aCROBAT agreement no. 463 639382 to ZGH), by Damon Runyon Cancer Research Foundation (Damon Runyon-Rachleff 464 Innovation Award DR 61-20 to KJE), and by the Huntsman Cancer Foundation. The University 465 of Utah Metabolomics Core Facility is supported by NIH S10 OD016232, S10 OD021505, and 466 U54 DK110858. Research reported in this publication utilized the Huntsman Cancer Institute 467 Biorepository and Molecular Pathology Shared Resources supported by NCI/NIH P30 468 CA042014. We thank Nikita Abraham and Diana Lim (University of Utah Molecular Medicine 469 Program) for assistance with figures. 470 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint STAR★Methods 471 Lead contact 472 Further information and requests for resources and reagents should be directed to and will be 473 fulfilled by the lead contact, Katsuhiko Funai ([email protected]). 474 475

availability 476 Plasmids utilized by this study are available from Sigma Aldrich. Mouse lines generated by this 477 study may be available at personal request from the lead contact. No new reagents were 478 created or used by this study. 479 480 Data and code availability 481 The data generated by this study including all images, figures, and datasets, is available upon 482 request to the lead contact, Dr. Katsuhiko Funai. Similarly, any additional information necessary 483 to reanalyze datasets is also available upon request. Code for RNA sequencing can be 484 retrieved upon request. 485 486 Experimental model and subject details 487 Human participants 488 De-identified liver samples were acquired from the University of Utah Biorepository and 489 Molecular Pathology Shared Resource from patients undergoing liver transplantation or 490 resection due to end-stage liver disease and/or liver tumor(s). All patients were classified to their 491 respective diagnosis by a pathologist at the time of initial collection. The diagnosis for individual 492 tissue samples was confirmed by a pathologist based on histology review of formalin-fixed, 493 paraffin-embedded sections takes from the same location as the tissue analyzed by targeted 494 lipid mass spectrometry. 495 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint 496 Mice 497 All mice (male and female) used in this study were bred onto C57BL/6J background. CLS-LKO 498 mice were generated by crossing the CLS conditional knockout (CLScKO+/+) generously 499 donated by Dr. Zachary Gerhart-Hines (University of Copenhagen) with mice heterozygous for 500 Albumin promotor Cre (Alb-Cre+/-) to produce liver-specific deletion of the CLS gene 501 (CLScKO+/+, Alb-Cre+/-) or control (CLScKO+/+, no Cre) mice. CLScKO+/+ mice harbor loxP sites 502 that flank exon 4 of the CLS gene. For high-fat diet studies, 8 wk CLS-LKO and their respective 503 controls began high-fat diet (HFD, 42% fat, Envigo TD.88137) feeding for 8 wks. Mice were 504 fasted 4 hours and given an intraperitoneal injection of 80 mg/kg ketamine and 10 mg/kg 505 xylazine prior to terminal experiments and tissue collection. All animal experiments were 506 performed with the approval of the Institutional Animal Care and Use Committees at the 507 University of Utah. 508 509 Cell lines 510 Hepa 1-6 murine hepatoma cells were grown in high-glucose DMEM (4.5 g/L glucose, with L-511 glutamine; Gibco 11965-092) supplemented with 10% FBS (heat-inactivated, certified, US 512 origin; Gibco 10082-147), and 0.1% penicillin-streptomycin (10,000 U/mL; Gibco 15140122). For 513 lentivirus-mediated knockdown of CLS, CLS expression was decreased using the pLKO.1 514 lentiviral-RNAi system. Plasmids encoding shRNA for mouse Crls1 (shCLS: TRCN0000123937) 515 was obtained from MilliporeSigma. Packaging vector psPAX2 (ID 12260), envelope vector 516 pMD2.G (ID 12259), and scrambled shRNA plasmid (SC: ID 1864) were obtained from 517 Addgene. HEK293T cells in 10 cm dishes were transfected using 50 μL 0.1% polyethylenimine, 518 200 μL, 0.15 M sodium chloride, and 500 μL Opti-MEM (with HEPES, 2.4 g/L sodium 519 bicarbonate, and l-glutamine; Gibco 31985) with 2.66 μg of psPAX2, 0.75 μg of pMD2.G, and 3 520 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint μg of either scrambled or Crls1 shRNA plasmid. Cells were selected with puromycin throughout 521 differentiation to ensure that only cells infected with shRNA vectors were viable. 522 523

details 524 Body composition 525 To assess body composition, mice were analyzed using a Bruker Minispec NMR (Bruker, 526 Germany) 1 week prior to terminal experiments. Body weights were measured and recorded 527 immediately prior to terminal experiments. 528 529 RNA quantification 530 For quantitative polymerase chain reaction (qPCR) experiments, mouse tissues were 531 homogenized in TRIzol reagent (Thermo Fisher Scientific) and RNA was isolated using 532 standard techniques. The iScript cDNA Synthesis Kit was used to reverse transcribe total RNA, 533 and qPCR was performed with SYBR Green reagents (Thermo Fisher Scientific). Pre-validated 534 primer sequences were obtained from mouse primer depot 535 (https://mouseprimerdepot.nci.nih.gov/). All mRNA levels were normalized to RPL32. For RNA 536 sequencing, liver RNA was isolated with the Direct-zol RNA Miniprep Plus kit (Zymo Cat#: 537 R2070). RNA library construction and sequencing were performed by the High-Throughput 538 Genomics Core at the Huntsman Cancer Institute, University of Utah. RNA libraries were 539 constructed using the NEBNext Ultra II Directional RNA Library Prep with rRNA Depletion Kit 540 (human, mouse rat). Sequencing was performed using the NovaSeq S4 Reagent Kit v1.5 541 150x150 bp Sequencing with >25 million reads per sample using adapter read 1: 542 AGATCGGAAGAGCACACGTCTGAACTCCAGTCA and adapter read 2: 543 AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGT. Pathway analyses were performed by the 544 Bioinformatics Core at the Huntsman Cancer Institute, University of Utah using the Reactome 545 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint Pathway Database. For differentially expressed genes, only transcripts with Padj 100 are included. 547 548 DNA isolation and quantitative PCR 549 Genomic DNA for assessments of mitochondrial DNA (mtDNA) was isolated using a 550 commercially available kit according to the manufacturer’s instructions (69504, Qiagen). 551 Genomic DNA was added to a mixture of SYBR Green (Thermo Fisher Scientific) and primers. 552 Sample mixtures were pipetted onto a 3840well plate and analyzed with QuantStudio 12K Flex 553 (Life Technologies). The following primers were used: mtDNA fwd, TTAAGA-CAC-CTT-GCC-554 TAG-CCACAC; mtDNA rev, CGG-TGG-CTG-GCA-CGA-AAT-T; nucDNA fwd, ATGACG-ATA-555 TCG-CTG-CGC-TG; nucDNA rev, TCA-CTT-ACC-TGGTGCCTA-GGG-C. 556 557 Western blot analysis 558 For whole liver lysate, frozen liver was homogenized in a glass homogenization tube using a 559 mechanical pestle grinder with homogenization buffer (50 mM Tris pH 7.6, 5 mM EDTA, 150 560 mM NaCl, 0.1% SDS, 0.1% sodium deoxycholate, 1% triton X-100, and protease inhibitor 561 cocktail). After homogenization, samples were centrifuged for 15 min at 12,000 × g. Protein 562 concentration of supernatant was then determined using a BCA protein Assay Kit (Thermo 563 Scientific). Equal protein was mixed with Laemmeli sample buffer and boiled for 5 mins at 95°C 564 for all antibodies except for OXPHOS cocktail antibody (at room temp for 5 mins), and loaded 565 onto 4–15% gradient gels (Bio-Rad). Transfer of proteins occurred on a nitrocellulose 566 membrane and then blocked for 1 hr. at room temperature with 5% bovine serum albumin in 567 Tris-buffered saline with 0.1% Tween 20 (TBST). The membranes were then incubated with 568 primary antibody (see Key Resource table), washed in TBST, incubated in appropriate 569 secondary antibodies, and washed in TBST. Membranes were imaged utilizing Western 570 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint Lightning Plus-ECL (PerkinElmer) and a FluorChem E Imager (Protein Simple). For isolated 571 mitochondria, identical procedures were taken with equal protein of mitochondrial preps. 572 573 Single cell preparation of liver tissue for flow cytometry 574 After mice were euthanized using isoflurane, blood was collected by cardiac puncture, the 575 abdomen was exposed and the liver collected, rinse with PBS and weighed. Liver was 576 subsequently transferred in approximately 3ml of serum-free RPMI-1640 containing 577 Collagenase D (10mg/ml; Sigma) and DNase (1mg/ml; Sigma) and incubated in a rocking 578 platform for 45 min at 37°C. The liver extract was mashed through a 70µm filter, the cell re-579 suspended in RPMI-1640 containing 10% FBS and centrifuged at 1600 rpm for 5 min. The 580 supernatant was discarded and the pellet re-suspended in approximately 4 ml of 70% Percoll, 581 then transferred in 15 ml conical tube, carefully overlay with 4 ml of 30% Percoll and centrifuged 582 1600 rpm for 25 min with the brake turned off. The non-parenchymal cells suspension from the 583 Percoll interface were removed and mixed with 10 mL of RPMI-1640 containing 10% FBS and 584 the cells were centrifuged at 1600 rpm for 5 min. Red blood cells (RBC) were removed from the 585 pelleted single cell suspensions of livers non-parenchymal cells by incubation in an ammonium 586 chloride -based 1x RBC lysis buffer (Thermofisher, eBioscience). The cells were again pelleted 587 and mixed with FACS buffer (2% BSA, 2mM EDTA in PBS), then stained with Zombie-NIR 588 viability dye (BioLegend) per manufacturer’s instructions to discriminate live vs dead cells. To 589 prevent non-specific Fc binding, the cells were incubated with Fc Block (anti-mouse CD16/32, 590 clone 93, Biolegend) for 15 min followed by the indicated antibodies cocktail for 60 min in the 591 dark on ice: CD45 (FITC, clone S18009F, Biolegend), CD11b (BVC421, clone M1/70, 592 Biolegend), F4/80 (APC, clone BM8, Biolegend), TIM4 (PerCP/Cy5.5, clone RMT4-54, 593 Biolegend), Ly6C (PE, clone HK1.4, Biolegend), MHCII (BV605, clone M5/114.15.2, Biolegend), 594 CD11c (BV785, clone N418, Biolegend) and Ly6G (PE/Cy7, clone 1A8, Biolegend). After 595 surface staining, cells were fixed with a paraformaldehyde-based fixation buffer (BioLegend). 596 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint Flow cytometric acquisition was performed on a BD Fortessa X20 flow cytometer (BD 597 Biosciebces) and data analyzed using FlowJo software (Version 10.8.1; Tree Star Inc). 598 599 Glucose tolerance test 600 Intraperitoneal glucose tolerance tests were performed by injection of 1 mg glucose per gram 601 body mass at least 6 days prior to sacrifice. Mice were fasted for 4 hours prior glucose injection. 602 Blood glucose was measured 30 minutes before glucose injection and at 0, 15, 30, 60, 90, and 603 120 minutes after injection via tail bleed with a handheld glucometer (Bayer Contour 7151H). 604 605 Pyruvate tolerance test 606 Pyruvate tolerance tests were performed by injection of 2 mg pyruvate per gram of body mass 607 in PBS adjusted to pH 7.3-7.5 at least 6 days prior to sacrifice. Blood glucose was measured 30 608 minutes before pyruvate injection and at 0, 15, 30, 45, 60, 75, 90, 105, and 120 minutes after 609 injection via tail bleed with a handheld glucometer (Bayer Contour 7151H). 610 611 Electron microscopy 612 To examine mitochondrial ultrastructure and microstructures, freshly dissected liver tissues from 613 CLS-LKO and their controls were sectioned into ≈2 mm pieces and processed by the Electron 614 Microscopy Core at University of Utah. To maintain the ultrastructure of the tissue via 615 irreversible cross-link formation, each section was submerged in fixative solution (1% 616 glutaraldehyde, 2.5% paraformaldehyde, 100 mM cacodylate buffer pH 7.4, 6 mM CaCl2, 4.8% 617 sucrose) and stored at 4°C for 48 hours. Samples then underwent 3 × 10-minute washes in 100 618 mM cacodylate buffer (pH 7.4) prior to secondary fixation (2% osmium tetroxide) for 1 hour at 619 room temperature. Osmium tetroxide as a secondary fixative has the advantage of preserving 620 membrane lipids, which are not preserved using aldehyde, alone. After secondary fixation, 621 samples were subsequently rinsed for 5 minutes in cacodylate buffer and distilled H2O, followed 622 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint by prestaining with saturated uranyl acetate for 1 hour at room temperature. After prestaining, 623 each sample was dehydrated with a graded ethanol series (2 × 15 minutes each: 30%, 50%, 624 70%, 95%; then 3 × 20 minutes each: 100%) and acetone (3 × 15 minutes) and were infiltrated 625 with EPON epoxy resin (5 hours 30%, overnight 70%, 3 × 2-hour 40 minute 100%, 100% fresh 626 for embed). Samples were then polymerized for 48 hours at 60°C. Ultracut was performed using 627 Leica UC 6 ultratome with sections at 70 nm thickness and mounted on 200 mesh copper grids. 628 The grids with the sections were stained for 20 minutes with saturated uranyl acetate and 629 subsequently stained for 10 minutes with lead citrate. Sections were examined using a JEOL 630 1200EX transmission electron microscope with a Soft Imaging Systems MegaView III CCD 631 camera. 632 633 Histochemistry 634 A fresh liver tissue was taken from each mouse and immediately submerged in 4% 635 paraformaldehyde for 12 hours and 70% ethanol for 48 hours. Tissues were sectioned at 10-μm 636 thickness, embedded in paraffin, and stained for Masson’s Trichrome to assess fibrosis or 637 hematoxylin and eosin (H&E) to determine fat droplet accumulation. Samples were imaged on 638 Axio Scan Z.1 (Zeiss). 639 640 Native PAGE 641 Isolated mitochondria (100 µg) suspended in MIM were pelleted at 12,000 x g for 15 min and 642 subsequently solubilized in 20 µL sample buffer (5 µL of 4x Native Page Sample Buffer, 8 µL 643 10% digitonin, 7 µL ddH2O per sample) for 20 min on ice and then centrifuged at 20,000 x g for 644 30 mins at 4°C. 15 µL of the supernatant (75 µg) was collected and placed into a new tube and 645 mixed with 2 µL of G-250 sample buffer additive. Dark blue cathode buffer (50 mLs 20X Native 646 Page running buffer, 50 mLs 20x cathode additive, 900 mLs ddH2O) was carefully added to the 647 front of gel box (Invitrogen Mini Gel Tank A25977) and anode buffer (50 mLs 20x Native Page 648 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint running buffer to 950 mL ddH2O) was carefully added to the back of the gel box making sure to 649 not mix. The samples were then loaded onto a native PAGE 3-12% Bis-Tris Gel (BN1001BOX, 650 Thermo Fisher Scientific), and electrophoresis was performed at 150 V for 1 hour on ice. The 651 dark blue cathode buffer was carefully replaced with light blue cathode buffer (50 mLs 20X 652 Native Page running buffer, 5 mL 20X cathode additive to 945 mLs ddH2O) and run at 30 V 653 overnight at 4°C. Gels were subsequently transferred to PVDF at 100 V, fixed with 8% acetic 654 acid for 5 min, washed with methanol, and blotted with the following primary antibodies Anti-655 GRIM19 (mouse monoclonal; ab110240), Anti-SDHA (mouse monoclonal; ab14715), Anti-656 UQCRFS1 (mouse monoclonal; ab14746), Anti-MTCO1 (mouse monoclonal; ab14705), Anti-657 ATP5a (mouse monoclonal; ab14748), Anti-NDUFA9 (mouse monoclonal; ab14713) in 5% non-658 fat milk in TBST. Secondary anti-mouse HRP antibody listed in the key resources table and 659 Western Lightning Plus-ECL (PerkinElmer NEL105001) was used to visualize bands. 660 661 Mitochondrial isolation 662 Liver tissues were minced in ice-cold mitochondrial isolation medium (MIM) buffer [300 mM 663 sucrose, 10 mM Hepes, 1 mM EGTA, and bovine serum albumin (BSA; 1 mg/ml) (pH 7.4)] and 664 gently homogenized with a Teflon pestle. To remove excess fat in the samples, an initial high-665 speed spin was performed on all samples: homogenates were centrifuged at 12,000g for 10 666 mins at 4°C, fat emulsion layers were removed and discarded, and resulting pellets were 667 resuspended in MIM + BSA. Samples were then centrifuged at 800 x g for 10 min at 4°C. The 668 supernatants were then transferred to fresh tubes and centrifuged again at 1,300 x g for 10 min 669 at 4°C. To achieve the mitochondrial fraction (pellet), the supernatants were again transferred to 670 new tubes and centrifuged at 12,000 x g for 10 min at 4°C. The resulting crude mitochondrial 671 pellets were washed three times with 0.15 M KCl to remove catalase, and then spun a final time 672 in MIM. The final mitochondrial pellets were resuspended in MIM buffer for experimental use. 673 674 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint Mitochondrial Respiration Measurements 675 Mitochondrial O2 utilization was measured using Oroboros O2K Oxygraphs. Isolated 676 mitochondria (50 µg for TCA substrate respiration and 100 µg for fatty acid respiration) were 677 added to the oxygraph chambers containing assay buffer Z (MES potassium salt 105 mM, KCl 678 30 mM, KH2PO4 10 mM, MgCl2 5 mM, BSA 1 mg/ml). Respiration was measured in response to 679 the following substrates: 0.5mM malate, 5mM pyruvate, 2.5mM ADP, 10mM succinate, 1.5 μM 680 FCCP, 0.02mM palmitoyl-carnitine, 5mM L-carnitine. 681 682 Mitochondrial JH2O2 683 Mitochondrial H2O2 production was determined in isolated mitochondria from liver tissue using 684 the Horiba Fluoromax-4/The Amplex UltraRed (10 μM)/horseradish peroxidase (3 U/ml) 685 detection system (excitation/emission, 565:600, HORIBA Jobin Yvon Fluorolog) at 37°C. 686 Mitochondrial protein was placed into a glass cuvette with buffer Z supplemented with 10 mM 687 Amplex UltraRed (Invitrogen), 20 U/mL CuZn SOD). Since liver tissue is capable of producing 688 resorufin from amplex red (AR), without the involvement of horseradish peroxidase (HRP) or 689 H2O2, phenylmethylsulfonyl fluoride (PMSF) was included to the experimental medium due to its 690 ability to inhibit HRP-independent conversion of AR to resorufin. PMSF was added to the 691 cuvette immediately prior to measurements and at a concentration that does not interfere with 692 biological measurements (100 µM). A 5-min background rate was obtained before adding 10 693 mM succinate to the cuvette to induce H2O2 production. After 4 min, 100 µM 1,3-bis(2-694 chloroethyl)-1-nitrosourea (BCNU) was added to the cuvette with 1 µM auranofin to inhibit 695 glutathione reductase and thioredoxin reductase, respectively. After an additional 4 min, the 696 assay was stopped, and the appearance of the fluorescent product was measured. 697 698 Site-specific electron leak was measured by systematically stimulating each site while inhibiting 699 the other three. Site IF was investigated in the presence of 4 mM malate, 2.5 mM ATP, 5 mM 700 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint aspartate, and 4 µM rotenone; site IQ was measured as a 4 µM rotenone-sensitive rate in the 701 presence of 5 mM succinate; site IIIQO was measured as a 2 µM myxothiazol-sensitive rate in 702 the presence of 5 mM succinate, 5 mM malonate, 4 µM rotenone, and 2 µM antimycin A; and 703 site IIF was measured as the 1 mM malonate-sensitive rate in the presence of 0.2 mM succinate 704 and 2 µM myxothiazol. As previously mentioned, electron leak is quantified using Amplex Red in 705 the presence of excess superoxide dismutase, such that both superoxide and hydrogen 706 peroxide production are accounted for by a change in fluorescence intensity (JH2O2) using high-707 resolution fluorometry (Horiba Fluoromax4®). 708 709 Phospholipidomic analysis 710 Liver tissue was homogenized in ice cold STEB (250 mM sucrose, 5 mM Tris-HCl, 1 mM EGTA, 711 0.1% fatty acid free BSA, pH 7.4, 4°C) using a tissuelyser. Mitochondria were then isolated via 712 differential centrifugation (800 x g for 10 min, 1300 x g for 10 min, 12,000 x g for 10 min at 4°C), 713 flushing each step under a stream nitrogen to prevent oxidation. Protein content was 714 determined by bicinchoninic acid assay using the Pierce BCA protein assay with bovine serum 715 albumin as a standard. To extract CoQ from mitochondria, incubations of 100 µg mitochondrial 716 protein in 250 µL ice-cold acidified methanol, 250 µL hexane, and 1146 pmol per sample of 717 CoQ standard (Cambridge Isotope Laboratories, CIL DLM-10279) were vortexed. The CoQ-718 containing hexane layer was separated by centrifugation (10 min, 17,000 x g, 4°C) and then 719 dried down under a stream of nitrogen. Dried samples were then resuspended in methanol 720 containing 2 mM ammonium formate and transferred to 1.5 mL glass mass spectrometry vials. 721 Liquid chromatography-mass spectrometry (LC-MS/MS) was then performed on the 722 reconstituted lipids using an Agilent 6530 UPLC-QTOF mass spectrometer. 723 724 Metabolomic extraction and mass spectrometry analysis 725 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint For metabolite extraction from the tissue, each sample was transferred to 2.0 mL ceramic bead 726 mill tubes (Qiagen Catalog Number 13116-50). To each sample was added 450 µL of cold 90% 727 methanol (MeOH) solution containing the internal standard d4-succinic acid (Sigma 293075) for 728 every 25 mg of tissue. The samples were then homogenized in an OMNI Bead Ruptor 24. 729 Homogenized samples were then incubated at -20 ˚C for 1 hr. After incubation the samples 730 were centrifuged at 20,000 x g for 10 minutes at 4 ˚C. 400 µL of supernatant was then 731 transferred from each bead mill tube into a labeled, fresh microcentrifuge tubes. Another internal 732 standard, d27-myristic acid, was then added to each sample. Pooled quality control samples 733 were made by removing a fraction of collected supernatant from each sample. Process blanks 734 were made using only extraction solvent and went through the same process steps as actual 735 samples. Everything was then dried en vacuo. 736 737 All GC-MS analysis was performed with an Agilent 5977b GC-MS MSD-HES and an Agilent 738 7693A automatic liquid sampler. Dried samples were suspended in 40 µL of a 40 mg/mL O-739 methoxylamine hydrochloride (MOX) (MP Bio #155405) in dry pyridine (EMD Millipore 740 #PX2012-7) and incubated for one hour at 37 °C in a sand bath. 25 µL of this solution was 741 added to auto sampler vials. 60 µL of N-methyl-N-trimethylsilyltrifluoracetamide (MSTFA with 742 1% TMCS, Thermo #TS48913) was added automatically via the auto sampler and incubated for 743 30 minutes at 37 °C. After incubation, samples were vortexed and 1 µL of the prepared sample 744 was injected into the gas chromatograph inlet in the split mode with the inlet temperature held at 745 250 °C. A 10:1 split ratio was used for analysis for most metabolites. Any metabolites that 746 saturated the instrument at the 10:1 split was analyzed at a 100:1 split ratio. The gas 747 chromatograph had an initial temperature of 60 °C for one minute followed by a 10 °C/min ramp 748 to 325 °C and a hold time of 10 minutes. A 30-meter Agilent Zorbax DB-5MS with 10 m 749 Duraguard capillary column was employed for chromatographic separation. Helium was used as 750 the carrier gas at a rate of 1 mL/min. Below is a description of the two-step derivatization 751 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint process used to convert non-volatile metabolites to a volatile form amenable to GC-MS. Pyruvic 752 acid is used here as an example. Data were collected using MassHunter software (Agilent). 753 Metabolites were identified and their peak area was recorded using MassHunter Quant. This 754 data was transferred to an Excel spread sheet (Microsoft, Redmond WA). Metabolite identity 755 was established using a combination of an in-house metabolite library developed using pure 756 purchased standards, the NIST library and the Fiehn library. There are a few reasons a specific 757 metabolite may not be observable through GC-MS. 758 759 Mitochondrial phospholipids enrichment 760 Isolated mitochondria (500 μg) from 2-month-old mice were incubated in fusion buffer [220 mM 761 mannitol, 70 mM sucrose, 2 mM Hepes, 10 mM KH2PO4, 5 mM MgCl2, 1 mM EGTA, 10 mM 762 glutamate, 2 mM malate, 10 mM pyruvate, and 2.5 mM ADP (pH 6.5)] for 20 min at 30°C under 763 constant stirring agitation in the presence of 15 nmol of small unilamellar vesicles (SUVs). After 764 fusion, mitochondria were layered on a sucrose gradient (0.6 M) and centrifuged 10 min at 765 10,000g at 4°C to remove SUV. Pellet was then washed in mitochondrial buffer [250 mM 766 sucrose, 3 mM EGTA, and 10 mM tris-HCl, (pH 7.4)]. 767 768 Succinate dehydrogenase assay 769 Liver succinate dehydrogenase activity was measured using the colorimetric SDH Detection 770 Assay Kit (ab228560). Briefly, 10 mg liver tissue was rapidly homogenized in assay buffer, 771 samples were centrifuged at 10,000 x g for 10 min, and supernatant transferred to a fresh tube. 772 20 µL of positive controls or sample was added to each well and the volume adjusted to 50 µL 773 with SDH assay buffer. A SDH reaction mix was prepared using 46 µL SDH assay buffer, 2 µL 774 SDH probe, and 2 µL SDH substrate mix per sample and added to each well for a final volume 775 of 100 µL. Absorbance was measured at 600 nm at 25°C with a microplate reader in kinetic 776 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint mode. Absorbance was followed for 30 minutes and time points 10 and 30 min were selected in 777 the linear range to calculate succinate dehydrogenase activity of the samples. 778 779 Serum AST and ALT 780 Mice were sacrificed by CO2 inhalation and blood samples collected via cardiac puncture into 20 781 mL of heparin and centrifuged for collection of plasma within 1 hour of blood collection and 782 frozen at -80oC until analysis. Plasma samples from mice were processed in a single batch for 783 determination of serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) 784 levels using a DC Element chemistry analyser (HESKA). 785 786 Quantification and statistical analyses 787 All data presented herein are expressed as mean ± SEM. The level of significance was set at p 788 < 0.05. Student’s t-tests were used to determine the significance between experimental groups 789 and two-way ANOVA analysis followed by Tukey’s HSD post hoc test was used where 790 appropriate. The sample size (n) for each experiment is shown in the figure legends and 791 corresponds to the sample derived from the individual mice or for cell culture experiments on an 792 individual batch of cells. Unless otherwise stated, statistical analyses were performed using 793 GraphPad Prism software. 794 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint Figure Legends 795 Figure 1. Hepatic mitochondrial phospholipidome in mouse models of MASLD 796 (A) H&E stains of livers from mice given standard chow or a Western HFD for 16 wks. 797 (B) H&E stains of livers from 20 wk old wildtype or ob/ob mice. 798 (C) Masson’s Trichrome stains of livers from mice given standard chow or the GAN diet for 30 799 wks. 800 (D) Masson’s Trichrome stains of livers from mice injected with vehicle or carbon tetrachloride 801 for 6wks. 802 (E) Representative western blot of OXPHOS subunits and citrate synthase in liver tissues from 803 mice given standard chow or a Western HFD for 16 wks (n=4 per group). 804 (F) Representative western blot of OXPHOS subunits and citrate synthase in liver tissues from 805 20 wk old wildtype or ob/ob mice (n=4 per group). 806 (G) Representative western blot of OXPHOS subunits and citrate synthase in liver tissues from 807 mice given standard chow or the GAN diet for 30 wks (n=4 per group). 808 (H) Representative western blot of OXPHOS subunits and citrate synthase in liver tissues from 809 mice injected with vehicle or carbon tetrachloride for 6 wks (n=4 per group). 810 (I) Mitochondrial phospholipidome from mice given standard chow or HFD. 811 (J) Mitochondrial phospholipidome from 20 wk old wildtype or ob/ob mice. 812 (K) Mitochondrial phospholipidome from mice given standard chow or the GAN diet for 30 wks. 813 (L) Mitochondrial phospholipidome from mice injected with vehicle or carbon tetrachloride for 6 814 wks. 815 (M) Venn Diagram comparing mitochondrial phospholipidome from all four models of MASLD: 816 HFD, ob/ob, GAN, or carbon tetrachloride. 817 (N) CLS message for livers of mice given standard chow or a Western HFD for 16 wks. 818 (O) CLS message for livers from 20 wk old wildtype or ob/ob mice. 819 (P) CLS message for livers from mice given standard chow or the GAN diet for 30 wks. 820 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint (Q) CLS message for livers from mice injected with vehicle or carbon tetrachloride for 6 wks. 821 Statistical significance was determined by 2-way ANOVA (I, J, K, and L) and unpaired Student's 822 T test (N, O, P, and Q). 823 824 Figure 2. Hepatocyte-specific deletion of CLS induces MASLD/MASH 825 (A) A schematic for hepatocyte-specific deletion of CLS in mice. 826 (B) CLS mRNA abundance in livers from control and CLS-LKO mice (n=4 and 7 per group). 827 (C) Abundance of mitochondrial CL species in liver from control and CLS-LKO mice (n=5 and 6 828 per group). 829 (D) Body mass (n=13 and 11 per group). 830 (E) Body composition (n=6 and 7 per group). 831 (F) Liver mass (n=10 and 13 per group). 832 (G) H&E stains for control or CLS-LKO mice fed a chow diet, mice are 8wks old. 833 (H) Masson's Trichrome stains for control or CLS-LKO mice. 834 (I) RNA sequencing data for genes associated with MASH, liver regeneration, and HCC for 835 control and CLS-LKO mice (n=7 and 5 per group). 836 (J) Serum AST from control and CLS-LKO mice (n=6 and 7 per group). 837 (K) Serum ALT from control and CLS-LKO mice (n=6 and 7 per group). 838 (L) mRNA abundance of TNFα, TGFβ, IL-12, and MCP1 in control and CLS-LKO livers (n=5 and 839 7 per group). 840 (M) Representative image of flow cell population gating for control and CLS-LKO livers (n=5 and 841 7 per group). 842 (N) Flow cytometry of cDC2 cell population in control and CLS-LKO livers (n=5 and 7 per 843 group). 844 (O) Flow cytometry of F4/80+ cell population in control and CLS-LKO livers (n=5 and 7 per 845 group). 846 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint (P) Flow cytometry of Ly6Chi cell population in control and CLS-LKO livers (n=5 and 7 per 847 group). 848 (Q) Flow cytometry of inflammatory monocyte cell population in control and CLS-LKO livers 849 (n=5 and 7 per group). 850 (R) Flow cytometry of neutrophil cell population in control and CLS-LKO livers (n=5 and 7 per 851 group). 852 (S) Flow cytometry of cDC1 cell population in control and CLS-LKO livers (n=5 and 7 per 853 group). 854 855 Figure 3. CLS deletion increases mitochondrial respiratory capacity 856 (A) Glucose tolerance test (IPGTT) performed 7 days prior to sacrifice date (n=6 and 7 per 857 group). 858 (B) Area under the curve for IPGTT. 859 (C) Pyruvate tolerance test (PTT) performed 7 days prior to sacrifice date (n=6 and 8 per 860 group). 861 (D) Area under the curve for PTT. 862 (E) RNA sequencing pathway analysis related to lipogenesis, VLDL, and beta-oxidation for 863 control and CLS-LKO mice (n=6 and 5 per group). 864 (F) mRNA levels for genes associated with components of OXPHOS. 865 (G) Transmission electron microscopy images of liver mitochondria from control and CLS-LKO 866 mice (scale bars, 1 μm). 867 (H) Representative western blot of whole liver tissue OXPHOS subunits and citrate synthase 868 between control and CLS-LKO mice (n=3 per group). 869 (I) Ratio of mitochondrial to nuclear DNA in liver tissue (n=8 per group). 870 (J) Representative tracing from high-resolution respirometry during TCA cycle intermediate 871 respiration. 872 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint (K) JO2 consumption in isolated liver mitochondria from control and CLS-LKO mice fed a chow 873 diet in response to 0.5mM malate, 5mM pyruvate, 2.5mM ADP, 10mM succinate, and 1.5 μM 874 FCCP (n=6 per group). 875 (L) JO2 consumption in isolated liver mitochondria from control and CLS-LKO mice fed a chow 876 diet in response to 0.02mM palmitoyl-carnitine, 5mM L-carnitine, and 2.5mM ADP (n=6 per 877 group). 878 (M) Representative western blot of isolated mitochondria OXPHOS subunits between control 879 and CLS-LKO mice (n=4 per group). 880 881 Figure 4. Stable isotope tracing with [U-13C] palmitate and [U-13C] glucose in hepa1-6 cells 882 with or without CLS deletion 883 (A) Schematic illustration of the labeling process during stable isotope tracing with [U-13C] 884 palmitate or [U-13C] glucose. Blue or green circles represent 13C-labeled carbons, and red 885 circles represent unlabeled 12C carbons. The pathway shows the flow from palmitate to beta-886 oxidation or glucose through glycolysis to the tricarboxylic acid (TCA) cycle, with key 887 intermediates labeled. 888 (B) Levels of labeled succinate from palmitate tracing in hepa1-6 cells (n=6 for shSC and 889 shCLS). 890 (C) Levels of labeled malate from palmitate tracing in hepa1-6 cells (n=6 for shSC and shCLS). 891 (D) Levels of labeled fumarate from palmitate tracing in hepa1-6 cells (n=6 for shSC and 892 shCLS). 893 (E) Levels of labeled pyruvate from glucose tracing in hepa1-6 cells (n=6 for shSC and shCLS). 894 (F) Levels of labeled lactic acid from glucose tracing in hepa1-6 cells (n=6 for shSC and 895 shCLS). 896 (G) Levels of labeled acetyl-CoA from glucose tracing in hepa1-6 cells (n=6 for shSC and 897 shCLS). 898 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint (H) Levels of labeled succinate from glucose tracing in hepa1-6 cells (n=6 for shSC and 899 shCLS). 900 (I) Levels of labeled fumarate from glucose tracing in hepa1-6 cells (n=6 for shSC and shCLS). 901 (J) Levels of labeled citrate from glucose tracing in hepa1-6 cells (n=6 for shSC and shCLS). 902 903 Figure 5. CL deficiency promotes mitochondrial electron leak 904 (A) Electron microscopy images for control and CLS-LKO mice depicting fibrosis via red arrows. 905 Scale bars are 2 μM. 906 (B) Quantitative PCR analysis of fibrotic markers (Col1a1 and Des) in liver tissue from control 907 and CLS-LKO mice (n=5 and 7 per group). 908 (C) Representative image for western blot analysis of cleaved caspase-3 in liver tissue from 909 control and CLS-LKO mice (n=4 per group). 910 (D) Representative image for western blot analysis of cleaved caspase-7 in liver tissue from 911 control and CLS-LKO mice (n=4 per group). 912 (E) Western blot analysis and quantification of cytochrome c levels in mitochondrial fraction from 913 liver tissue of control and CLS-LKO mice (n=7 per group). 914 (F) Western blot analysis and quantification of cytochrome c levels in cytosolic fraction from liver 915 tissue of control and CLS-LKO mice (n=7 per group). 916 (G) H2O2 emission and production in isolated liver mitochondria from control and CLS-LKO mice 917 fed a chow diet, stimulated with succinate, or succinate, auranofin, and BCNU (n=3 and 4 per 918 group). 919 (H) Schematic representation of rescue experiment. Isolated mitochondria from CLS-LKO mice 920 were enriched with small unilamellar vesicles (SUVs) containing either cardiolipin (CL) or 921 phosphatidylcholine (PC). 922 (I) Quantification of H2O2 production in liver mitochondria enriched with CL or PC SUVs in 923 control and CLS-LKO mice (n=4 per group). 924 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint 925 Figure 6. Influence of CL deficiency on site-specific electron leak and supercomplex 926 formation 927 (A) Electron leak at site IQ in mitochondria from control and CLS-LKO mice (n=7 per group). 928 (B) Electron leak at site IF in mitochondria from control and CLS-LKO mice (n=7 per group. 929 (C) Electron leak at site IIF in mitochondria from control or CLS-LKO mice (n=7 per group). 930 (D) Electron leak at site IIIQ0 in mitochondria from control or CLS-LKO mice (n=7 per group). 931 (E) Abundance of respiratory supercomplex I formation using the GRIM19 antibody in isolated 932 mitochondria from livers taken from control and CLS-LKO mice (n=4 per group). 933 (F) Quantification of E. 934 (G) Abundance of respiratory supercomplex I formation using the NDUFA9 antibody in isolated 935 mitochondria from livers taken from control and CLS-LKO livers (n=4 per group). 936 (H) Quantification of G. 937 (I) Abundance of respiratory supercomplex II formation using the SDHA2 antibody in isolated 938 mitochondria from livers taken from control and CLS LKO livers (n=4 per group). 939 (J) Quantification of I. 940 (K) Abundance of respiratory supercomplex III formation using the UQCRFS1 antibody in 941 isolated mitochondria from livers taken from control and CLS-LKO livers (n=4 per group). 942 (L) Quantification of K. 943 (M) Abundance of respiratory supercomplex IV formation using the MTCO1 antibody in isolated 944 mitochondria from livers taken from control and CLS-LKO livers (n=4 per group). 945 (N) Quantification of M. 946 (O) Abundance of respiratory supercomplex V formation using the ATP5A antibody in isolated 947 mitochondria from livers taken from control and CLS-LKO livers (n=4 per group). 948 (P) Quantification of O. 949 950 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint Figure 7. CL deficiency disrupts coenzyme Q homeostasis in mice and humans 951 (A) Chemical structure of Coenzyme Q (CoQ) in its oxidized (ubiquinone) and reduced 952 (ubiquinol) forms. 953 (B) Oxidized CoQ levels in isolated mitochondrial fractions from livers taken from control and 954 CLS-LKO mice (n=7 per group). 955 (C) Reduced CoQ levels in isolated mitochondrial fractions from livers taken from control and 956 CLS-LKO livers (n=7 per group). 957 (D) Succinate-to-fumarate ratio from untargeted metabolomics showing differential abundance 958 of TCA cycle metabolites in livers taken from CLS-LKO mice compared to controls (n=5 and 7 959 per group). 960 (E) Activity of succinate dehydrogenase (SDH) in control and CLS-LKO livers (n=6 per group). 961 (F) Representative histological images using H&E stain on human liver samples from patients 962 with advanced steatohepatitis. 963 (G) Analysis of CL in human liver samples from patients with advanced steatohepatitis (n=10 964 and 16 per group). 965 (H) Analysis of CoQ in human liver samples from patients with advanced steatohepatitis (n=10 966 and 16 per group). 967 (I) Pearson correlation analysis of CL and CoQ levels in human liver samples (R² = 0.64). 968 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint

969 970 1. Younossi, Z.M., et al. Global epidemiology of nonalcoholic fatty liver disease—Meta‐971 analytic assessment of prevalence, incidence, and outcomes. Hepatology 64, 73-84 972 (2015). 973 2. Miao, L.T., Giovanni; Byrne, Christopher D; Cao, Ying-Ying; Zheng, Ming-Hua. Current 974 status and future trends of the global burden of MASLD. Trends in Endocrinology & 975 Metabolism 29, 1043-2760 (2024). 976 3. Saito, K., et al. Characterization of hepatic lipid profiles in a mouse model with 977 nonalcoholic seatohepatitis and subsequent fibrosis. 5(2015). 978 4. John, B.S.S. Nonalcoholic Steatohepatitis (NASH). in StatPearls (StatPearls Publishing, 979 Treasure Island, 2023). 980 5. Vigano, L., Lleo, A. & Aghemo, A. Non-alcoholic fatty liver disease, non-alcoholic 981 steatohepatitis, metabolic syndrome and hepatocellular carcinoma—a composite 982 scenario. Hepatobiliary Surgery and Nutrition 7, 130-133 (2018). 983 6. Fan, J.-G. & Cao, H.-X. Role of diet and nutritional management in non-alcoholic fatty 984 liver disease. Gastroenterology and Hepatology 28, 87-87 (2013). 985 7. Gordon, D.L., et al. Biomarkers of NAFLD progression: a lipidomics approach to an 986 epidemic. Journal of Lipid Research 56, 722-736 (2015). 987 8. Fu, S., et al. Aberrant lipid metabolism disrupts calcium homeostasis causing liver 988 endoplasmic reticulum stress in obesity. Nature 473, 528-531 (2011). 989 9. Simoes, I.C.M., Fontes, A., Pinton, P., Zischka, H. & Wieckowski, M.R. Mitochondria in 990 non-alcoholic fatty liver disease. The International Journal of Biochemistry & Cell Biology 991 95, 93-99 (2018). 992 10. Leveille, M. & Estall, J.L. Mitochondrial Dysfunction in the Transition from NASH to HCC. 993 Metabolites 9, 233 (2019). 994 11. Einer, C., et al. Mitochondrial adaptation in steatotic mice. 59, 483-495 (2018). 995 12. Sunny, N.E., Bril, F. & Cusi, K. Mitochondrial Adaptation in Nonalcoholic Fatty Liver 996 Disease: Novel Mechanisms and Treatment Strategies. 28(2018). 997 13. Noureddin, M. & Sanyal, A.J. Pathogenesis of NASH: the Impact of Multiple Pathways. 998 17, 350-360 (2018). 999 14. Pfeiffer, K., et al. Cardiolipin stabilizes respiratory chain supercomplexes. 278, 52873-1000 52880 (2003). 1001 15. Johnson, J.M., et al. Targeted overexpression of catalase to mitochondria does not 1002 prevent cardioskeletal myopathy in Barth syndrome. 121, 94-102 (2018). 1003 16. Houtkooper, R.H. & Vaz, F.M. Cardiolipin, the heart of mitochondrial metabolism. Cell. 1004 Mol. Life Sci. 65, 2493-2506 (2008). 1005 17. Schlame ab, M. & Hostetler, K.Y. Cardiolipin synthase from mammalian mitochondria. 1006 Biochimica et Biophysica Acta (BBA) - Molecular and Cell Biology of Lipids 1348, 207-1007 213 (1997). 1008 18. Paradies, G., Paradies, V., Ruggiero, F.M. & Petrosillo, G. Role of Cardiolipin in 1009 Mitochondrial Function and Dynamics in Health and Disease: Molecular and 1010 Pharmacological Aspects. Cells 8, 728 (2019). 1011 19. Funai, K., Summers, S.A. & Rutter, J. Reign in the membrane: How common lipids 1012 govern mitochondrial function. Current opinion in cell biology 63, 162-173 (2020). 1013 20. Jussupow, A., Luca, A.D. & Kaila, V.R. How cardiolipin modulates the dynamics of 1014 respiratory complex I. Science Advances 5(2019). 1015 21. Schwall, C.T., Greenwood, V.L. & Alder, N.N. The stability and activity of respiratory 1016 Complex II is cardiolipin-dependent. Biochimica et Biophysica Acta (BBA) - Molecular 1017 and Cell Biology of Lipids 1817, 1588-1596 (2012). 1018 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint 22. Bazan, S., et al. Cardiolipin-dependent Reconstitution of Respiratory Supercomplexes 1019 from Purified Saccharomyces cerevisiae Complexes III and IV*. 288, 401-411 (2013). 1020 23. Dudek, J., et al. Cardiolipin deficiency affects respiratory chain function and organization 1021 in an induced pluripotent stem cell model of Barth syndrome. Stem Cell Research 11, 1022 806-819 (2013). 1023 24. Sustarsic, E.G., et al. Cardiolipin Synthesis in Brown and Beige Fat Mitochondria Is 1024 Essential for Systemic Energy Homeostasis. Cell Metabolism 28, 159-174 (2018). 1025 25. Jiang, F., et al. Absence of cardiolipin in the crd1 null mutant results in decreased 1026 mitochondrial membrane potential and reduced mitochondrial function. 275, 22387-1027 22394 (2000). 1028 26. Xu, Y., Sutachan, J.J., Plesken, H., Kelley, R.I. & Schlame, M. Characterization of 1029 lymphoblast mitochondria from patients with Barth syndrome. 85, 823-830 (2005). 1030 27. Hsu, P., et al. Cardiolipin remodeling by TAZ/tafazzin is selectively required for the 1031 initiation of mitophagy. Autophagy 11, 643-652 (2015). 1032 28. Le, C.H., et al. Tafazzin deficiency impairs CoA-dependent oxidative metabolism in 1033 cardiac mitochondria. Journal of Biological Chemistry 295, 12485-12497 (2020). 1034 29. Heden, T.D., et al. Mitochondrial PE potentiates respiratory enzymes to amplify skeletal 1035 muscle aerobic capacity. Science Advances 5, 8352 (2019). 1036 30. Johnson, J.M.P., Alek D; Balderas, Enrique; Sustarsic, Elahu G; Maschek, J Alan; 1037 Brothwell, Marisa J; Alejandro, Jara-Ramos; Panic, Vanja; Morgan, Jeffrey T; Villanueva, 1038 Claudio J; Sanchez, Alejandro; Rutter, Jared; Lodhi, Irfan J; Cox, James E; Fisher-1039 Wellman, Kelsey H; Chaudhuri, Dipayan; Gerhart-Hines, Zachary; Funai, Katsuhiko. 1040 Mitochondrial phosphatidylethanolamine modulates UCP1 to promote brown adipose 1041 thermogenesis. Science Advances 9, 7864 (2023). 1042 31. Kartsoli, S., Kostara, C.E., Tsimihodimos, V., Bbairaktari, E.T. & Christodoulou, D.K. 1043 Lipidomics in non-alcoholic fatty liver disease. World Journal of Hepatology 12, 436-450 1044 (2020). 1045 32. Ehud Zigmond, C.V. Two Roads Diverge in the Sick Liver, Monocytes Travel Both. Cell 1046 Press 10, 1016 (2020). 1047 33. Zahao, X.-A.C., Guang-Mei; Liu, Yong; Chen, Yu-Xin; Wu, Hong-Yan; Chen, Jin; Xiong, 1048 Ya-Li; Tian, Chen; Wang, Gui-Yang; Jia, Bei; Xia, Juan; Wang, Jian; Yan, Xiao-Min; 1049 Zhang, Zhao-Ping; Huang, Rui; Wu, Chao. Inhibitory effect of silymarin on CCl4-induced 1050 liver fibrosis by reducing Ly6Chi monocytes infiltration. International Journal of Clinical & 1051 Experimental Pathology 10, 11941-11951 (2017). 1052 34. Daemen, S.G., Anastasiia; Kalugotla, Gowri; He, Li; Chan, Mandy M.; Beals, Joseph W.: 1053 Liss, Kim H.; Klein, Samuel; Feldstein, Ariel E.; Finck, Brian N.; Artyomov, Maxim N.; 1054 Schilling, Joel D. Dynamic Shifts in the Composition of Resident and Recruited 1055 Macrophages Influence Tissue Remodeling in NASH. Cell Reports 34, 108626 (2021). 1056 35. Liyanage, S.E.G., Peter J.; Ribeiro, Joana; Cristante, Enrico; Sampson, Robert D.; 1057 Luhmann, Ulrich F.O.; Ali, Robin R.; Bainbridge, James W. Flow cytometric analysis of 1058 inflammatory and resident myeloid populations in mouse ocular inflammatory models. 1059 Experimental Eye Research 151, 160-170 (2016). 1060 36. Fry, M. & Green, D.E. Cardiolipin requirement for electron transfer in complex I and III of 1061 the mitochondrial respiratory chain. J. Biol Chem 256, 1874-1880 (1981). 1062 37. Paradies, G., Petrosillo, G., Pistolese, M. & Ruggiero, F.M. Reactive oxygen species 1063 affect mitochondrial electron transport complex I activity through oxidative cardiolipin 1064 damage. Gene 286, 135-141 (2002). 1065 38. Lange, C., Nett, J.H., Trumpower, B.L. & Hunte, C. Specific roles of protein–1066 phospholipid interactions in the yeast cytochrome bc1 complex structure. EMBO J 20, 1067 6591-6600 (2001). 1068 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint 39. Robinson, N.C. Functional binding of cardiolipin to cytochrome c oxidase. J Bioenerg 1069 Biomembr 25, 153-163 (1993). 1070 40. Koliaki, C., et al. Adaptation of hepatic mitochondrial function in humans with non-1071 alcoholic fatty liver is lost in steatohepatitis. Cell Metabolism 21, 739-746 (2015). 1072 41. Siripoksup, P., et al. Sedentary behavior in mice induces metabolic inflexibility by 1073 suppressing skeletal muscle pyruvate metabolism. J Clin Invest 134(2024). 1074 42. Drose, S. & Brandt, U. The Mechanism of Mitochondrial Superoxide Production by the 1075 Cytochrome bc1 Complex. Journal of Biological Chemistry 283, 21649-21654 (2008). 1076 43. Marra, F. & Svegliati-Baroni, G. Lipotoxicity and the gut-liver axis in NASH pathogenesis. 1077 Hepatology 68, 280-295 (2018). 1078 44. Buzzetti, E., Pinzani, M. & Tsochatzis, E.A. The multiple-hit pathogenesis of non-1079 alcoholic fatty liver disease (NAFLD). 65, 1038-1048 (2016). 1080 45. Turkseven, S., et al. Mitochondria-targeted antioxidant mitoquinone attenuates liver 1081 inflammation and fibrosis in cirrhotic rats. American Physiological Society 318, 298-304 1082 (2020). 1083 46. Karakuzu, O.C., Melissa R.; Liu, Yi; Garsin, Danielle A. Amplex Red Assay for 1084 Measuring Hydrogen Peroxide Production from Caenorhabditis elegans. Bio Protec. 9, 1085 3409 (2019). 1086 47. Prola, A.B., Jordan; Vandestienne, Aymeline; Piquereau, Jerome; Denis, Raphael GP; 1087 Guyot, Stephane; Chauvin, Hadrien; Mourier, Arnaud; Maurer, Marie; Henry, Celine; 1088 Khadhraoui, Nahed; Gallerne, Cindy; Molinie, Thibaut; Courtin, Guillaume; Gressette, 1089 Melanie; Solgadi, Audrey; Dumont, Florent; Castel, Julien; Ternacle, Julien; Demarquoy, 1090 Jean; Malgoyre, Alexandra; Koulmann, Nathalie; Derumeaux, Genevieve; Giraud, Marie-1091 France; Joubert, Frederic; Veksler, Vladimir; Luquet, Serve; Relaix, Frederic; Tiret, 1092 Laurent; Pilot-Storck, Fanny. Cardiolipin content controls mitochondrial coupling and 1093 energetic efficiency in muscle. Science Advances 7(2021). 1094 48. Pennington, E.R., Funai, K., Brown, D.A. & Shaikh, S.R. The role of cardiolipin 1095 concentration and acyl chain composition on mitochondrial inner membrane molecular 1096 organization and function. Biochimica et Biophysica Acta (BBA) - Molecular and Cell 1097 Biology of Lipids 1864, 1039-1052 (2019). 1098 49. Mileykovskaya, E. & Dowhan, W. Cardiolipin-Dependent Formation of Mitochondrial 1099 Respiratory Supercomplexes. Chemistry and Physics of Lipids 179, 42-48 (2014). 1100 50. Arnarez, C., Marrink, S.J. & Periole, X. Molecular mechanism of cardiolipin-mediated 1101 assembly of respiratory chain supercomplexes. Chemical Science (2016). 1102 51. Bottinger, L., et al. Phosphatidylethanolamine and Cardiolipin Differentially Affect the 1103 Stability of Mitochondrial Respiratory Chain Supercomplexes. 5, 677-686 (2012). 1104 52. Dornas, W. & Schuppan, D. Mitochondrial oxidative injury: a key player in nonalcoholic 1105 fatty liver disease. Gastrointestinal and Liver Physiology 319, G400-G411 (2020). 1106 53. Gomez, B. & Robinson, N.C. Quantitative Determination of Cardiolipin in Mitochondrial 1107 Electron Transferring Complexes by Silicic Acid High-Performance Liquid 1108 Chromatography. Analytical Biochemistry 267, 212-216 (1999). 1109 54. Brand, M.D. Mitochondrial generation of superoxide and hydrogen peroxide as the 1110 source of mitochondrial redox signaling. Free Radical Biology and Medicine 100, 14-31 1111 (2016). 1112 55. Decker, S.T.F., Katsuhiko. Mitochondrial membrane lipids in the regulation of 1113 bioenergetic flux. Cell Metabolism 36, 1963-1978 (2024). 1114 56. Bessone, F., Razori, M.V. & Roma, M.G. Molecular pathways of nonalcoholic fatty liver 1115 disease development and progression. 76, 99-128 (2019). 1116 57. Ines C. M. Simoes, R.A., Jose Teixeira, Agnieszka Karkucinska-Wieckowska, Adriana 1117 Carvalho, Susana P. Pereira, Rui F. Simoes, Sylwia Szymanska, Michal Dabrowski, 1118 Justyna Janikiewicz, Agnieszka Dobrzyn, Paulo J. Oliveira, Yaiza Potes, Mariusz R. 1119 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint Wieckowski. The Alterations of Mitochondrial Function during NAFLD Progression-- An 1120 Independent Effect of Mitochondrial ROS Production. International Journal of Molecular 1121 Sciences 22, 6848 (2021). 1122 58. Simoes, I.C., Fontes, A., Pinton, P., Zischka, H. & Wieckowski, M.R. Mitochondria in 1123 non-alcoholic fatty liver disease. 95, 93-99 (2017). 1124 59. Goncalves, R.L.S., Schlame, M., Bartelt, A.L., Brand, M.D. & Hotamisligil, G.S. 1125 Cardiolipin deficiency in Barth syndrome is not associated with increased 1126 superoxide/H2O2 production in heart and skeletal muscle mitochondria. 595, 415-432 1127 (2021). 1128 60. Xu, Y., et al. Loss of protein association causes cardiolipin degradation in Barth 1129 syndrome. Nature Chemical Biology 12, 641-647 (2016). 1130 61. Simoes, I.C.A., Ricardo; Teixeira, Jose; Karkucinska-Wieckowska, Agnieszka; Carvalho, 1131 Adriana; Pereira, Susana P.; Simoes, Rui F.; Szymanska Sylwia; Dabrowski, Michal; 1132 Janikiewicz, Justyna; Dobrzyn, Agnieszka; Oliveira, Paulo J.; Potes, Yaiza; Wieckowski, 1133 Mariusz R. . The Alterations of Mitochondrial Function during NAFLD Progression-- An 1134 Independent Effect of Mitochondrial ROS Production. Int J Mol Sci 22, 6848 (2021). 1135 1136 1137 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint Graphical Abstract .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint 100 µm A E I B F J C G D H L N Veh CCl4Veh CCl4 WT ob/ob PC PE CL PI PS LPC PG LPE PC PE CL PI PS LPC PG LPE CIII CIV CII CI CS 55 48 40 30 20 45 55 48 40 30 20 45 55 48 40 30 20 45 55 48 40 30 20 45 (kDa)(kDa)(kDa) K PC CV PE CIII CL CIV PI PS CII LPC PG CI LPE (kDa) CV CIII CIV CII CI CS CS CV CIII CIV CII CI CS 1.0 0.5 0.0 1.5 * 1.0 0.5 0.0 P 1.5 0.5 0.0 O 1.5 1.0 ** ** ** 1.0 0.5 0.0 Q 1.5 Chow HFD H&E WT Ob/ob H&E Chow GAN Masson’s TrichromeMasson’s Trichrome 100 µm 100 µm 100 µm 100 µm 100 µm 100 µm CV Z-score -2 -1 0 Z-score -0.2 0 0.2 0.4 1 2 -2 0 2 Z-score Relative LevelsRelative LevelsRelative LevelsRelative Levels Crls1 mRNA * * ** ** **** ****** **** * **** ***** **** Figure 1 HFD GAN ob/ob CCl4 PC PE CL PI PS LPC PG LPE Chow HFD *** *** * **** ***** -1 0 1 Z-score M Chow HFD WT ob/ob Chow GAN PS LPE PG Veh CCl4 PI PE CL LPC PC 100 µm Chow GAN .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint 0 100 200 300 400ALT (U/L) ALTK 0 600 400 200 1,000 800 AST (U/L) ASTJ * J CLS-LKOCtrl 9.54% 1.54% 88.0% 32.8% 0.43% 65.5% Gated on CD45+CD11b+F4/80+ M Control CLS-LKO Mass (g) Fat Lean 0 5 10 15 ED 0 10 20 30Body Mass (g) 20 0 80 60 40 100 cDC2 (Relative Levels) 2 0 8 6 4 10 F4/80+ (Relative Levels) 0 20 40 60 Ly6Chi (Relative Levels) cDC2 F4/80+ Ly6Chi ** 0.0 0.5 1.0 1.5 2.0 cDC1 (Relative Levels) 0 10 5 20 15 25 0.0 0.5 1.0 1.5 2.0 cDC1NeutrophilsN O P Q Inflammatory Monocytes R S *** Inflammatory Monocytes (Relative Levels) 0.0 0.5 1.0 1.5 2.0 Liver mass (g) ** F Neutrophils (Relative Levels) 20 TNFα TGFβ IL-12 MCP1 0.0 0.5 1.0 1.5 Inflammation Relative mRNA Levels ✱✱ ✱✱ CLS -LKO Ctrl H&E G H Masson’s Trichrome CtrlCLS-LKO 100 μm 100 μm 100 μm 100 μm I Ctrl CLS-LKO Col1a1 Col16a1 Col13a1 Mmp13 Il12b Cdc20 Aurkb Cdca8 Tuba8 Kifc1 Klc3 Kif18b Psat1 MASH Liver regenerationHCC -0.5 0 0.5 1.0 L A 5´ ARM FRT Neo 4 FRT loxP loxP loxP3´ ARM lacZ CLS exon 4 Alb-Cre mediated excision B 0.0 0.5 1.0 1.5 Relative mRNA Levels **** CLS mRNA Cardiolipin Species 0 10000 8000 6000 4000 2000 Cardiolipin (pmol lipid*mg protein-1) C **** ** CL Levels Body Mass Body Composition Liver Mass Z-score Figure 2 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint AUC (x103) 0 10 20 30 D G Electron Microscopy 1 µm 1 µm CtrlCLS-LKO J 3 5 7 9 11 13 15 17 19 21 23 Time (min) 30 20 10 0 -10 -20 -30 O2 slope neg. [pmol/(s * mL)] Ctrl CLS-LKO Mal/pyr ADP Succinate FCCP CV CIII CIV CII M Isolated Mitochondria kDa 55 48 40 30 A 0 10 20 30AUC (x103) B CV CIII CIV CII CS H Whole Tissue Lysate kDa 55 48 40 30 52 -30 0 15 30 60 90 120 Time After Injection (min) 0 100 200 300 400 Blood Glucose (mg/dL) * * Ctrl CLS-LKO -30 0 15 30 45 60 75 90 105120 Time After Injection (min) 0 50 100 150 200 Blood Glucose (mg/dL) Ctrl CLS-LKO C Control CLS-LKO K L Mal/Pyr ADP Succ FCCP 0 500 1000 1500 2000 JO2 (pmol * s-1 * mg mito-1) ✱✱✱✱ ✱✱✱ PLM ADP 0 100 200 300 400 500 JO2 (pmol * s-1 * mg mito-1) ✱✱✱✱ E Acly Acaca Fasn Acss2 Scd1 Dgat1 Dgat2 Apob Mttp Tm6sf2 Apoe Cpt1a Cpt2 Acadvl Hadha Acadl Acadm Acads Ctrl CLS-LKO * * ** ** * Lipogenesis VLDL Secretionβ-oxidation -2 -1 0 1 2 Ndufs1 Sdha Sdhb Sdhd Uqcrc1 Uqcrc2 Uqcrb Cox5a Pdss2 mt-Co1 Ctrl CLS-LKO ** ** * * * * * -2 -1 0 1 F Ctrl CLS-LKO 0.0 0.2 0.4 0.6 0.8 1.0 mtDNA/nucDNA I Glucose Tolerance Pyruvate Tolerance mtDNA/nucDNA TCA Respiration Fatty Acid Respiration Z-score Z-score Figure 3 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint [U-13C] Palmitate Citrate M+2 ns0.25 0.20 0.15 0.10 0.05 0.00 H * Succinate M+2 0.00 0.02 0.04 0.06 0.08 E Isotope Enriched Fraction Pyruvate M+3 0.0 0.2 0.4 0.6 I Fumarate M+2 0.10 0.08 0.06 0.04 0.02 0.00 ns B shSC shCLS.. [U-13C] Palmitate shSC shCLS [U-13C] Glucose J Lactate M+3 **** 0.0 0.2 0.4 0.6 G Isotope Enriched Fraction * Acetyl-CoA M+2 0.0 0.1 0.2 0.3 0.4 * F Isotope Enriched Fraction Isotope Enriched Fraction Isotope Enriched Fraction Isotope Enriched Fraction 0.00 0.05 0.10 0.15 0.20 Succinate M+2 Isotope Enriched Fraction ✱✱ 0.00 0.05 0.10 0.15 0.20 0.25 Malate M+2 Isotope Enriched Fraction ✱ 0.00 0.05 0.10 0.15 0.20 Fumarate M+2 Isotope Enriched Fraction ✱✱ C D * A [U-13C] Glucose β-oxidation Figure 4 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint col11 DES 0 1 2 3 4 Fibrosis Relative mRNA levels ** B 0 2 1 3 4 SUVs % Electron Leak (JH2O2 /JO 2) ** I Ctrl CLS-LKO H CL or PC SUVs Isolated mitochondria enriched with CL or PC SUVs * 0 1 2 3 4 Cleaved Caspase 3C 19 kDa 17 kDa Cleaved Caspase 3 * 0 5 10 15 G % Electron Leak (JH2O2 /JO 2) Cleaved Caspase 3 (Relative Abundance) Cleaved Caspase 7 0 1 2 3 P=0.1503 D 19 kDa 17 kDa Cleaved Caspase 7 Cleaved Caspase 7 (Relative Abundance) 0.0 0.5 1.0 1.5 2.0 E 17 kDa Mitochondrial Cytochome c Mitochondrial Cyt C (Relative Abundance) 17 kDa 2.5 2.0 1.5 1.0 0.5 0.0 Cytosolic Cytochrome c F Cytosolic Cyt C (Relative Abundance) Ctrl CLS-LKO Electron Microscopy A 2 µm 2 µm Control CLS-LKO Mitochondrial Electron Leak Figure 5 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint JH2O2 Production (pmol * s-1 * mg mito-1) JH2O2 Production (pmol * s-1 * mg mito-1) JH2O2 Production (pmol * s-1 * mg mito-1) C D IV QH2/Q ½O2 H2 O e- e- I IQ IIIQo e- e- O2 O2- O2 2O - Myxothiazol Malonate Rotenone Succinate O2 2O - e- e- e- IF e- NADH Malate e- O2- e- IIQ II e- IIF e- IIIQi e- III e- Antimycin A e- Cyt C Complex IV (MTCO1) IV1 M III2+ IV1 III2+ IV2 Complex II (SDHA2) II I Complex III (UQCRFS1) K 2: I+III2+IV1 1: I+III2 III2 SC Complex I (GRIM19) E I 2: I+III2+IV 1: I+III2+IIn 5: I2+III2 SC II 0.0 0.5 1.0 1.5 CII Singlet SDHA (Relative Abundance) J * 0.0 0.5 1.0 1.5 CIII SupercomplexesL III 2 0.0 0.5 1.0 1.5 CIII Singlet UQCRSF1 (Relative Abundance) UQCRSF1 (Relative Abundance) * * 0.5 1.0 1.5 2.0 CV Supercomplex ATP5A (Relative Abundance) ATP5A (Relative Abundance) 0.0 0.0 P Vn V 0.5 1.0 1.5 CV Singlet * 0.0 0.5 1.0 1.5 2.0 CIV Supercomplexes 0.0 0.5 1.0 1.5 CIV SingletN IV1 MTCO1 (Relative Abundance) MTCO1 (Relative Abundance) * I 0 1 2 3 CI Singlet GRIM19 (Relative Abundance) 0 IIF 1,000 2,000 3,000 IQ O2 0 1,000 2,000 A P=0.0879 III QO 0 1,000 2,000 3,000 * ** Complex V (ATP5A) O Vn V 3,000 Site IQ JH2O2 Production (pmol * s-1 * mg mito-1) B IF 0 1,000 2,000 3,000 ** Site IF Site IIF Site IIIQ0 H 0.0 0.5 1.0 1.5 2.0 GRIM19 (Relative Abundance) F CI Supercomplexes Complex I (NDUFA9) 2: I+III2+IV11: I+III2+IIn 5: I2+III2 SC I G 0.0 0.5 1.0 1.5 2.0 2.5 ✱ CI Supercomplexes NDUFA9 (Relative Abundance) I 0.0 0.5 1.0 1.5 CI Singlet NDUFA9 (Relative Abundance) Control CLS-LKO Figure 6 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint 0 5 10 15 20 * Succinate:Fumarate Succinate:Fumarate Ratio D 0.000 0.005 0.010 0.015 0.020 SDH Activity (mU/mg liver) SDH Activity E CL 70:7 CL 72:6 CL 72:7 CL 74:8 CL 74:9 CL 72:8 0 5000 10000 15000 CL (pmole lipid * mg protein-1) ✱✱✱✱ G CoQ7 CoQ8 CoQ9 CoQ10 0 20000 40000 60000 80000 100000 CoQ (pmole lipid * mg protein-1) ✱✱✱✱ H 0 5000 10000 15000 20000 0 20000 40000 60000 80000 100000 CL vs CoQ CL (pmole lipid * mg protein-1) CoQ (pmole lipid * mg protein-1) R squared = 0.64 I Healthy MASH H&E 20 μm 20 μm A B CoQ8 CoQ9 CoQ10 0 1 2 3 4 Oxidized CoQ (Relative Abundance) ✱✱✱✱ Oxidized CoQ C CoQ8H2 CoQ9H2 CoQ10H2 0 1 2 3 4 Reduced CoQ (Relative Abundance) ns Reduced CoQ F Healthy MASH CL Levels CoQ Levels Figure 7 Control CLS-LKO .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint 0 5000 10000 15000 20000 0 2000 4000 6000 0 5000 10000 15000 0 2000 4000 6000 8000 0 20000 40000 60000 LPC (pmol lipid * mg protein-1) 0 10000 20000 30000 40000 50000 0 5000 10000 15000 25000 0 1000 2000 3000 0 5000 10000 15000 20000 0 500 1000 1500 PG 0 500 1000 1500 0 500 1000 1500 2000 2500 LPE 0 500 1000 1500 2000 PI 0 500 1000 1500 2000 LPE (pmol lipid * mg protein-1) Chow HFD Chow HFD WT ob/ob WT ob/ob M 10000 A B C D E F G H I J K L N O P **** **** **** **** **** **** **** **** **** **** **** ******** **** **** **** **** * **** **** **** **** 20000 **** **** ** **** ** ** CL (pmol lipid * mg protein-1) PC (pmol lipid * mg protein-1) PE (pmol lipid * mg protein-1) (pmol lipid * m -1g protein ) PS (pmol lipid * mg protein-1) PS (pmol lipid * mg protein-1) CL (pmol lipid * mg protein-1) PC (pmol lipid * mg protein-1) PG (pmol lipid * mg protein-1) PE (pmol lipid * mg protein-1) PI (pmol lipid * mg protein-1) LPC (pmol lipid * mg protein-1) -1(pmol lipid * mg protein ) *(pmol lipid mg protein-1) 0 1000 2000 3000 4000 5000 ✱✱✱✱ ✱✱✱✱ 0 1000 2000 3000 Supplemental Figure S1 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint 0 500 1000 1500 2000 2500 0 1000 2000 3000 4000 5000 PS 0 1000 2000 3000 0 2000 4000 6000 0 2000 4000 6000 8000 10000 0 200 400 600 800 1000 0 5000 10000 0 200 400 600 800 1000 0 2000 4000 6000 8000 0 200 400 600 0 2000 6000 8000 0 100 200 300 400 500 0 500 1000 1500 2000 2500 0 100 200 300 400 0 1000 2000 3000 4000 0 100 200 300 400 **** Chow GAN Chow GAN Corn Oil CCl4 Corn Oil CCl4 A B C D E F J 15000 G H I K L M N O P CL (pmol lipid * mg protein-1) PC (pmol lipid * mg protein-1) PE (pmol lipid * mg protein-1) PI (pmol lipid * mg protein-1) CL (pmol lipid * mg protein-1) PC (pmol lipid * mg protein-1) PE (pmol lipid * mg protein-1) PI (pmol lipid * mg protein-1) PS (pmol lipid * mg protein-1) LPC (pmol lipid * mg protein-1) PG (pmol lipid * mg protein-1) LPE (pmol lipid * mg protein-1) LPC (pmol lipid * mg protein-1) PG (pmol lipid * mg protein-1) LPE (pmol lipid * mg protein-1) * -1(pmol lipid mg protein ) ******** **** **** ******** **** * **** ** **** * **** * ******** *** ****** **** *** *** **** 4000 **** **** **** **** ** ** **** **** **** **** **** **** **** *** Supplemental Figure S2 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint Control CLS-LKO Control CLS-LKO A B C D E F G H I J K L M N O Chow HFD 0 20000 40000 60000 80000 ** ** ** ** ** * **** **** ** 0 2000 4000 6000 8000 0 50000 100000 150000 200000 0 5000 10000 15000 0 20000 40000 60000 80000 100000 LPC (pmol lipid * mg protein-1) PC (pmol lipid * mg protein-1) PE (pmol lipid * mg protein-1)PC (pmol lipid * mg protein-1)LPC (pmol lipid * mg protein-1) *(pmol lipid mg protein-1) -1(pmol lipid * mg protein )-1(pmol lipid * mg protein ) 0 5000 10000 15000 20000 25000 **** PG (pmol lipid * mg protein-1) 0 50000 100000 150000 PE (pmol lipid * mg protein-1) 0 2000 4000 10000 **** PG (pmol lipid * mg protein-1) 6000 8000 0 50000 100000 150000 200000 PI (pmol lipid * mg protein-1) 0 5000 10000 15000 LPE 0 50000 100000 150000 200000 PI 0 5000 10000 15000 LPE 0 5000 10000 15000 PS (pmol lipid * mg protein-1) 0 2000 4000 6000 CL (pmol lipid * mg protein-1) 0 50000 100000 150000 PS (pmol lipid * mg protein-1) Supplemental Figure S3 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint C D Glyoxylate metabolism and glycine degradation Complex I biogenesis Branched chain amino acid catabolism Respiratory electron transport ATP synthesis Citric acid TCA cycle and respiratory electron transport Respiratory electron transport Metabolism of amino acids and derivatives Golgi to ER retrograde transport Keratinization RMTs methylate histone arginines TP53 transcription of cell cycle genes Lymphoid and non-lymphoid immunoregulatory interactions Signaling by PDGF Mitotic prometaphase Megakaryocyte differentiation and platelet function NCAM signaling for neurite out growth Megakaryocyte development and platelet production PRC2 methylates histones and DNA Rho GTPase effectors COPI-dependent golgi to ER retrograde traffic Chromosome maintenance O-glycosylation of TSR domain containing proteins Non-integrin membrane ECM interactions Collagen chain trimerization Mitotic spindle checkpoint Molecules associated with elastic fibres Deposition of new CENPA containing nucleosomes Condensation of prophase chromosomes ECM proteoglycans Amplification of signal from the kinetochores Elastic fibre formation Assembly of collagen fibrils Kinesins Resolution of sister chromatid cohesion Rho GTPases activate formins Integrin cell surface interactions Collagen biosynthesis and modifying enzymes Collagen degradation Collagen formation Extracellular matrix organization Degradation of the extracellular matrix -4 -2 0 2 4 Normalized Enrichment Score A BControl HFD CLS-LKO HFD Control HFD CLS-LKO HFD H&E Masson’s Trichome 100 m 100 m100 m100 m Supplemental Figure S4 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint A ATGL DGAT1 0.0 FASN SCD1 0.5 1.0 1.5 2.0 Relative mRNA Levels B 0 50 100 150TG (mg/dL) C FCCP 0 Mal/Pyr ADP Succ 500 1000 1500 J O 2 (pmol * s-1 * mg mito-1) ** ***C Control CLS-LKO 0 PLM ADP 100 200 300 400 **D E Isotope Enriched Fraction Glycine M+2 0.00 0.02 0.04 0.06 0.08 **** Isotope Enriched Fraction Aspartate M+2 0.00 0.05 0.10 0.15 0.20 ns F G 3-Phosphoglyceric acid M+3 0.5 0.4 0.3 0.2 0.1 0.0Isotope Enriched Fraction ns Isotope Enriched Fraction *** Alanine M+3 0.0 0.1 0.2 0.3 0.4 I Isotope Enriched Fraction Malate M+2 ns 0.10 0.08 0.06 0.04 0.02 0.00 H J O 2 (pmol * s-1 * mg mito-1) shSC shCLS [U-13C] Glucose Supplemental Figure S5 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint 0 1 2 3 4 SUVs % Electron Leak (JH2O2/JO2) Ctrl ns ns * E 0 5 10 15 20 JH2O2(pmol * s -1* mg mito-1) / JO2 (pmol * s-1* mg mito-1) **** C 17 kDa Mitochondrial cytochrome c 0 20000 40000 60000Relative Abundance A 17 kDa Cytosolic cytochromec 0 10000 5000 20000 15000Relative Abundance B Control CLS-LKO 1 0 2 3 4 5 * *** JH2O2 (pmol * s-1 * mg mito-1) / JO2 (pmol * s-1 * mg mito-1) D shSC shCLS HFD Supplemental Figure S6 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint 0 1000 2000 3000 Total CoQ 8 (pmol lipid * mg protein-1) Total CoQ9 (pmol lipid/mg protein) (pmol lipid/mg protein) Total CoQ10 (pmol lipid/mg protein) Total CoQ8 H2 (pmol lipid/mg protein)Mitochondrial CoQ8H2 (pmol lipid/mg protein) 100000 80000 60000 40000 20000 0 Mitochondrial CoQ8 (pmol lipid/mg protein) 800 600 400 200 0 Total CoQ9 H2 (pmol lipid/mg protein) Total CoQ10 H2 (pmol lipid/mg protein) 50 40 30 20 10 0 5000 0 15000 10000 20000 0 200000 400000 600000 *** **** Mitochondrial CoQ 9 0 5000 10000 15000 0 150 100 50 200 Mitochondrial CoQ 9 H2 (pmol lipid/mg protein) 500 400 300 200 100 0 0 5000 10000 15000 Mitochondrial CoQ10 (pmol lipid/mg protein) 400 300 200 100 0 0 10 20 30 40 Mitochondrial CoQ10 H 2 (pmol lipid/mg protein) * D E F G H I J K L M N O Whole Liver Tissue Total CoQ Control CLS-LKO Isolated Mitochondria CoQ P Cholesterol UracilFumaric acid L-Serine Succinic acid Fructose-6-phosphate Glucose 6-phosphate -1 0 1 2 0 1 2 3 4 Log2 Fold Change -Log10 (p-value) Down Regulated Up Regulated 0 500 1,000 1,500 2,000 TotalCoQ8 A (pmol lipid * mg protein-1) Whole Liver Tissue Oxized or Reduced CoQ B 0 5,000 10,000 15,000 20,000 TotalCoQ8(pmol lipid * mg protein-1) 1,000 800 600 400 200 0 C TotalCoQ10 (pmol lipid * mg protein-1) Q ns 0.0 2.0x10 6 1.0x10 6 4.0x10 6 3.0x10 6 Fumarate (AUC) R 0.0 5.0x106 1.0x107 1.5x107 Succinate (AUC) Supplemental Figure S7 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint Key Resource Table REAGENT or RESOURCE SOURCE IDENTIFIER Antibodies GRIM19 Abcam ab110240 SDHA Abcam ab14715 UQCRFS1 Abcam ab14746 MTCO1 Abcam ab14705 ATP5a Abcam Ab14748 NDUFA9 Abcam Ab14713 Total OxPhos Antibody cocktail Abcam MS604-300 Citrate Synthetase Abcam Ab96600 Cytochrome c Cell Signaling 11940S Caspase-3 Cell Signaling 9661S Caspase-7 Cell Signaling 9491S Bacterial and virus strains Second-generation lentiviral-mediated knockdown system NEB Stable Competent E. Coli NEB C3040H Biological samples Chemicals, peptides, and recombinant proteins Amplex Red Reagent ThermoFisher Scientific A12222 Auranofin Sigma Aldrich A6733 Carmustine (BCNU) Sigma Aldrich C0400 SPLASH Mix Avanti Polar Lipids 330707 Cardiolipin Mix I Avanti Polar Lipids LM6003 Bovine Serum Albumin Sigma Aldrich A7030 Protease Inhibitor Cocktail Thermo Scientific 78446 Tamoxifen Sigma Aldrich T5648 Sunflower Oil Sigma Aldrich S5007 TRIzol Thermo Scientific 15596018 Mini-PROTEAN TGX Gels BioRad 4561086 ECL PerkinElmer 104001EA Malate Sigma Aldrich M7397 Pyruvate Sigma Aldrich P2256 GDP Sigma Aldrich G7127 CL 316,243 Sigma Aldrich C5976 ADP Sigma Aldrich A5285 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint ATP Sigma Aldrich A9187 Glutamate Sigma Aldrich G5889 Succinate Sigma Aldrich S3674 Carnitine Sigma Aldrich 8.40092 Palmitoyl-CoA Sigma Aldrich P9716 Palmitoyl-L-carnitine Sigma Aldrich P1645 SYBR Green Thermo Scientific A25776 4% Paraformaldehyde Thermo J19943-K2 Opti-MEM Gibco 31985 DMEM Gibco 1195-092 FBS Gibco 10082-147 Penicillin-streptomycin Gibco 15140122 Critical commercial assays Pierce BCA Protein Assay Kit Thermo Scientific 23227 iScript cDNA Synthesis Kit BioRad 1708891 Deposited data Experimental models: Cell lines HEK293T cells ATCC CTRL-3216 Hepa 1-6 murine hepatoma cells ATCC CRL-1830 Experimental models: Organisms/strains Mouse: CLS conditional knockout (CLS-cKO) Sustarsic et al. 2018 N/A Mouse: CLS-LKO This paper N/A Mouse: Alb-Cre Jackson Laboratory 003574 Oligonucleotides RT qPCR Primer ATGL F (CCACTCACATCTACGGAGCC) www.IDTDNA.com RT qPCR Primer ATGL R (TAATGTTGGCACCTGCTTCA) www.IDTDNA.com RT qPCR Primer DGAT1 F (GACGGCTACTGGGATCTGA) www.IDTDNA.com RT qPCR Primer DGAT1 R (TCACAACACACCAATTCAGG) www.IDTDNA.com RT qPCR Primer FAS F (GGATAGCTGTGTAGTGTAACCAT) www.IDTDNA.com RT qPCR Primer FAS R (GGTCATCGTGATAACCACACA) www.IDTDNA.com RT qPCR Primer SCD1 F (GCTCTACACCTGCCTCTTCG) www.IDTDNA.com .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint RT qPCR Primer SCD1 R (CAGCCGAGCCTTGTAAGTTC) www.IDTDNA.com RT qPCR Primer CLS F (TGACCTATGCAGATCTTATTCCA) Johnson et al. 2019 RT qPCR Primer CLS R (TGGCAGAGTTCGGTATCTGA) Johnson et al. 2019 RT qPCR Primer TNFa F (CCACCACGCTCTTCTGTCTAC) www.IDTDNA.com RT qPCR Primer TNFa R (AGGGTCTGGGCCATAGAACT) www.IDTDNA.com RT qPCR Primer Taz F (CCCTCCATGTGAAGTGGCCATTCC) Johnson et al. 2019 RT qPCR Primer Taz R (TGGTGGTTGGAGACGGTGATAAGG) Johnson et al. 2019 mtDNA F: (TTAAGACACCTTGCCTAGCCACAC) Mouse Primer Depot NCI/NIH mtDNA R: (CGGTGGCTGGCACGAAATT) Mouse Primer Depot NCI/NIH nucDNA F: (ATGACGATATCGCTGCGCTG) Mouse Primer Depot NCI/NIH nucDNA R: (TCACTTACCTGGTGCCTAGGGC) Mouse Primer Depot NCI/NIH Recombinant DNA Sc Addgene 1864 Crls1 Sigma Aldrich TRCN0000123937 psPAX2 Addgene 12260 pMD2.G Addgene 12259 Software and algorithms GraphPad Prism 9.0 GraphPad N/A Other .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint Supplemental Table S1. Patient demographic information Healthy MASH Healthy MASH Age at time of collection 50.3 + 9.6 yrs 62.2 + 7.1 yrs Sex Male: 1 Female: 10 Male: 9 Female: 8 Alcohol use? N/A Yes: 0 No: 17 Race White: 9 African American: 1 Asian: 1 White: 10 Unknown: 7 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint Supplemental Figure Legends Figure S1. Mitochondrial phospholipidome from Figure 1I and 1J. (A) Abundance of mitochondrial CL species in liver of control mice or mice fed a HFD for 16 weeks (n=5 per group). (B) Abundance of mitochondrial PC species in liver of control mice or mice fed a HFD for 16 weeks (n=5 per group). (C) Abundance of mitochondrial PE species in liver of control mice or mice fed a HFD for 16 weeks (n=5 per group). (D) Abundance of mitochondrial PI species in liver of control mice or mice fed a HFD for 16 weeks (n=5 per group). (E) Abundance of mitochondrial PS species in liver of control mice or mice fed a HFD for 16 weeks (n=5 per group). (F) Abundance of mitochondrial LPC species in liver of control mice or mice fed a HFD for 16 weeks (n=5 per group). (G) Abundance of mitochondrial PG species in liver of control mice or mice fed a HFD for 16 weeks (n=5 per group). (H) Abundance of mitochondrial LPE species in liver of control mice or mice fed a HFD for 16 weeks (n=5 per group). (I) Abundance of mitochondrial CL species in liver from control mice or leptin-deficient mice, 30 weeks old (n=6 per group). (J) Abundance of mitochondrial PC species in liver from control mice or leptin-deficient mice, 30 weeks old (n=6 per group). (K) Abundance of mitochondrial PE species in liver from control mice or leptin-deficient mice, 30 weeks old (n=6 per group). (L) Abundance of mitochondrial PI species in liver from control mice or leptin-deficient mice, 30 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint weeks old (n=6 per group). (M) Abundance of mitochondrial PS species in liver from control mice or leptin-deficient mice, 30 weeks old (n=6 per group). (N) Abundance of mitochondrial LPC species in liver from control mice or leptin-deficient mice, 30 weeks old (n=6 per group). (O) Abundance of mitochondrial PG species in liver from control mice or leptin-deficient mice, 30 wks old (n=6 per group). (P) Abundance of mitochondrial LPE species in liver from control mice or leptin-deficient mice, 30 wks old (n=6 per group). Figure S2. Mitochondrial phospholipidome from Figure 1K and 1L. (A) Abundance of mitochondrial CL species in livers from mice injected with corn oil or carbon tetrachloride for 10 wks (n=5 and 7 per group). (B) Abundance of mitochondrial PC species in livers from mice injected with corn oil or carbon tetrachloride for 10 wks (n=5 and 7 per group). (C) Abundance of mitochondrial PE species in livers from mice injected with corn oil or carbon tetrachloride for 10 wks (n=5 and 7 per group). (D) Abundance of mitochondrial PI species in livers from mice injected with corn oil or carbon tetrachloride for 10 wks (n=5 and 7 per group). (E) Abundance of mitochondrial PS species in livers from mice injected with corn oil or carbon tetrachloride for 10 wks (n=5 and 7 per group). (F) Abundance of mitochondrial LPC species in livers from mice injected with corn oil or carbon tetrachloride for 10 wks (n=5 and 7 per group). (G) Abundance of mitochondrial PG species in livers from mice injected with corn oil or carbon tetrachloride for 10 wks (n=5 and 7 per group). (H) Abundance of mitochondrial LPE species in livers from mice injected with corn oil or carbon .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint tetrachloride for 10 wks (n=5 and 7 per group). (I) Abundance of mitochondrial CL species in livers from mice fed the Gubra-Amylin MASH diet or chow for 30 wks (n=6 per group). (J) Abundance of mitochondrial PC species in livers from mice fed the Gubra-Amylin MASH diet or chow for 30 wks (n=6 per group). (K) Abundance of mitochondrial PE species in livers from mice fed the Gubra-Amylin MASH diet or chow for 30 wks (n=6 per group). (L) Abundance of mitochondrial PI species in livers from mice fed the Gubra-Amylin MASH diet or chow for 30 wks (n=6 per group). (M) Abundance of mitochondrial PS species in livers from mice fed the Gubra-Amylin MASH diet or chow for 30 wks (n=6 per group). (N) Abundance of mitochondrial LPC species in livers from mice fed the Gubra-Amylin MASH diet or chow for 30 wks (n=6 per group). (O) Abundance of mitochondrial PG species in livers from mice fed the Gubra-Amylin MASH diet or chow for 30 wks (n=6 per group). (P) Abundance of mitochondrial LPE species in livers from mice fed the Gubra-Amylin MASH diet or chow for 30 wks (n=6 per group). Figure S3. Mitochondrial phospholipidome from standard chow or high-fat diet fed control and CLS-LKO livers. (A) Abundance of mitochondrial PC species in liver from control or CLS-LKO mice, 8 wks old (n=5 and 6 per group). (B) Abundance of mitochondrial PE species in liver from control or CLS-LKO mice, 8 wks old (n=5 and 6 per group). (C) Abundance of mitochondrial PI species in liver from control or CLS-LKO mice, 8 wks old (n=5 and 6 per group). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint (D) Abundance of mitochondrial PS species in liver from control or CLS-LKO mice, 8 wks old (n=5 and 6 per group). (E) Abundance of mitochondrial LPC species in liver from control or CLS-LKO mice, 8 wks old (n=5 and 6 per group). (F) Abundance of mitochondrial PG species in liver from control or CLS-LKO mice, 8 wks old (n=5 and 6 per group). (G) Abundance of mitochondrial LPE species in liver from control or CLS-LKO mice, 8 wks old (n=5 and 6 per group). (H) Abundance of mitochondrial CL species in liver from control or CLS-LKO mice fed a high-fat diet for 8 wks (n=11 and 12 per group). (I) Abundance of mitochondrial PC species in liver from control or CLS-LKO mice fed a high-fat diet for 8 wks (n=11 and 12 per group). (J) Abundance of mitochondrial PE species in liver from control or CLS-LKO mice fed a high-fat diet for 8 wks (n=11 and 12 per group). (K) Abundance of mitochondrial PI species in liver from control or CLS-LKO mice fed a high-fat diet for 8 wks (n=11 and 12 per group). (L) Abundance of mitochondrial PS species in liver from control or CLS-LKO mice fed a high-fat diet for 8 wks (n=11 and 12 per group). (M) Abundance of mitochondrial LPC species in liver from control or CLS-LKO mice fed a high fat diet for 8 wks (n=11 and 12 per group). (N) Abundance of mitochondrial PG species in liver from control or CLS-LKO mice fed a high-fat diet for 8 wks (n=11 and 12 per group). (O) Abundance of mitochondrial LPE species in liver from control or CLS-LKO mice fed a high fat diet for 8 wks (n=11 and 12 per group). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint Figure S4. Additional histological and transcriptomic data from control and CLS-LKO mice. (A) H&E stains for control and CLS-LKO mice fed a HFD for 8 wks. (B) Masson’s Trichrome stains for control and CLS-LKO mice fed a HFD for 8 wks. (C) Volcano plot of genes differentially expressed in livers taken from control and CLS-LKO mice (n=5 and 7 per group). (D) Normalized enrichment scores in RNA sequencing using Reactome database for most significantly affected pathways in livers taken from control and CLS-LKO mice (n=5 and 7 per group). Figure S5. Additional metabolic, mitochondrial, and fluxomic phenotyping data with CLS deletion. (A) Relative mRNA levels of lipogenic genes in livers from control and CLS-LKO mice (n=6 and 7 per group). (B) Serum triglycerides for control and CLS-LKO mice (n=6 and 7 per group). (C) JO2 consumption in isolated liver mitochondria from control or CLS-LKO mice fed a Western HFD for 8 wks in response to 0.5 mM malate, 5 mM pyruvate, 2.5 mM ADP, 10 mM succinate, and 1.5 μM FCCP (n=11 and 12 per group). (D) JO2 consumption in isolated liver mitochondria from control or CLS-LKO mice fed a Western HFD for 8 wks in response to 0.02 mM palmitoyl-carnitine, 5 mM L-carnitine, and 2.5 mM ADP (n=7 per group). (E) Levels of labeled glycine from glucose tracing in hepa1-6 cells (n=6 for shSC and shCLS). (F) Levels of labeled aspartate from glucose tracing in hepa1-6 cells (n=6 for shSC and shCLS). (G) Levels of labeled 3-phosphoglyceric acid from glucose tracing in hepa1-6 cells (n=6 for shSC and shCLS). (H) Levels of labeled alanine from glucose tracing in hepa1-6 cells (n=6 for shSC and shCLS). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint (I) Levels of labeled malate from glucose tracing in hepa1-6 cells (n=6 for shSC and shCLS). Figure S6. Additional mitochondrial phenotyping data with CLS deletion. (A) Western blot of cytochrome c levels in isolated mitochondria from HFD-fed control and CLS-LKO mice (n=6 per group). (B) Western blot of cytochrome c levels in cytosolic fractions from HFD-fed control and CLS- LKO mice (n=6 per group). (C) H2O2 emission and production in isolated liver mitochondria from control or CLS-LKO mice fed a Western HFD, stimulated with succinate, or succinate and auranofin and BCNU (n=9 and 8 per group). (D) H2O2 emission and production in isolated liver mitochondria from hepa1-6 CLS knockdown cells stimulated with succinate, or succinate and auranofin and BCNU (n=3 per group). (E) Quantification of electron leak using SUV to enrich mitochondria in control mice (n=4 per group). Figure S7. Additional data on coenzyme Q (A) Mass spectrometric analysis of total CoQ8 levels in whole liver tissue from control and CLS-LKO mice (n=7 per group). (B) Mass spectrometric analysis of total CoQ9 levels in whole liver tissue from control and CLS-LKO mice (n=7 per group). (C) Mass spectrometric analysis of total CoQ10 levels in whole liver tissue from control and CLS-LKO mice (n=7 per group). (D) Oxidized CoQ8 levels in whole liver tissue from control and CLS-LKO livers (n=7 per group). (E) Reduced CoQ8 levels in whole liver tissue from control and CLS-LKO livers (n=7 per group). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint (F) Oxidized CoQ9 levels in whole liver tissue from control and CLS-LKO livers (n=7 per group). (G) Reduced CoQ9 levels in whole liver tissue from control and CLS-LKO livers (n=7 per group). (H) Oxidized CoQ10 levels in whole liver tissue from control and CLS-LKO livers (n=7 per group). (I) Reduced CoQ10 levels in whole liver tissue from control and CLS-LKO livers (n=7 per group). (J) Oxidized CoQ8 levels in isolated mitochondria from control and CLS-LKO livers (n=7 per group). (K) Reduced CoQ8 levels in isolated mitochondria from control and CLS-LKO livers (n=7 per group). (L) Oxidized CoQ9 levels in isolated mitochondria from control and CLS-LKO livers (n=7 per group). (M) Reduced CoQ9 levels in isolated mitochondria from control and CLS-LKO livers (n=7 per group). (N) Oxidized CoQ10 levels in isolated mitochondria from control and CLS-LKO livers (n=7 per group). (O) Reduced CoQ10 levels in isolated mitochondria from control and CLS-LKO livers (n=7 per group). (P) Volcano plot from untargeted metabolomics showing differential abundance of TCA cycle metabolites. (Q) Fumarate levels from metabolomics data (n=5 and 7 per group). (R) Succinate levels from metabolomics data (n=5 and 7 per group). metabolites in CLS-LKO livers compared to controls (n=5 and 7 per group). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.10.617517doi: bioRxiv preprint