Dichloroacetate improves hepatic and systemic energy metabolism during sepsis

Dichloroacetate improves hepatic and systemic energy metabolism during sepsis 1 2 Rabina Mainali1*, Manal Zabalwi2*, David Long2, Nancy Buechler1, Shalini Jain3, Ellen 3 Quillen2, Chia-Chi Key2, Xuewei Zhu2, Peter W. Stacpoole4, Jennifer Martinez5, Charles 4 E. McCall2#, Matthew A, Quinn1,2,5# 5 6 1Department of Pathology, Section on Comparative Medicine Wake Forest School of 7 Medicine, Winston-Salem, North Carolina 8


65
Sepsis represents a life-threatening condition that results from dysregulated in-66 flammatory responses to disseminated and uncontrolled infection. If not diagnosed and 67 treated early after its onset, sepsis may limit survival by inducing early shock and subse-68 quent immunometabolic paralysis. Sepsis accounts for one in three hospital deaths in the 69 U.S. and millions of deaths each year globally, highlighting its hazard to public health 1,2 .

70
The high mortality rate associated with sepsis reflects the lack of a clinically viable mo-71 lecular-based therapeutic target. Therefore, understanding the pathogenesis of sepsis at 72 both the molecular and organismal levels is major gap in public health and of the utmost 73 importance. 74 Sepsis is a biphasic phenomenon, with clearly defined acute and chronic phases 75 3,4 . The acute phase of sepsis is driven by high energy consumption and hyper-inflamma-76 tion characterized by oxidative stress, followed by cellular reprogramming to a low energy, 77 anti-inflammatory state of immunometabolic paralysis, with accompanying organ fail-78 ure. 2,5 . In addition to inflammatory responses, dysregulation of the host's metabolism dur-79 ing sepsis plays a vital role in its progression and severity 6-8 . Compared to healthy indi-80 viduals, energy homeostasis is significantly altered in septic patients 7,8 . The early phase 81 of sepsis is a hypermetabolic condition marked by increased catabolism of fats, proteins 82 and carbohydrates, associated with high rates of oxygen consumption and ATP synthesis 83 9-11 . Following the acute phase is a hypometabolic state where ATP production and mito-84 chondrial respiration decrease 9 . Immune cells and some organ cells enter an energy 85 conserving "hibernation-like" state as a protective mechanism to lower the metabolic de-86 mands of the cell and help with its recovery 6,12,13 . However, staying in this hypometabolic 87 state for a prolonged period can lead to organ dysfunction and failure 12,13 .

88
Of particular interest is that, during the hyper-inflammatory anabolic phase of sep-  accordance with the changes observed in the oxidative phosphorylation pathway, we 127 found a negative enrichment score for TCA cycle enzymes (GO:0006099) in the liver of 128 CLP mice (Fig. 1c). Thus, our transcriptome data indicate that sepsis impairs 129 mitochondrial metabolism in the liver.

130
Next we wanted to assess if the transcriptional changes elicited by sepsis would 131 manifest in altered hepatic TCA cycle metabolism. Therefore, we performed global 132 unbiased metabolomic screening in isolated hepatocytes from control and septic mice by 133 Ultrahigh Performance Liquid Chromotography-Tandem Mass Spectroscopy (UPLC-134 MS/MS). In line with altered transcriptional regulation of the TCA cycle, we found that 135 sepsis altered the relative abundance of multiple metabolites involved in the TCA cycle 136 (Fig. 1d). In particular, significant elevation of citrate and cis-aconitate was observed in 137 septic hepatocytes at 30h (Fig. 1e&f). Unlike macrophages, which shift their ratio of 138 succinate and a-ketoglutarate to favor succinate accumulation over a-ketoglutarate 18-20 , 139 we found no changes in the levels of these metabolites in hepatocytes (Fig. 1G&H).   (Fig. 2a). Given the induction of ROS metabolic genes we wanted to 157 determine if sepsis alters key metabolites involved in redox balance, specifically the 158 cysteine-glutathione transsulfuration redox regulatory cycle. Hepatocytes isolated during 159 sepsis-induced tolerance did not adjust intracellular homocysteine levels (Fig. 2b), 160 however, cystathionine significantly decreased (Fig. 2c). Furthermore, we found a trend 161 for decreased cysteine levels in the liver (p=0.0534) (Fig. 2d) and significantly decreased 162 glycine (Fig. 2e). Most dramatically, septic mice depleted hepatocyte glutathione (Fig. 2f). 163 During oxidative stress, the pool of glutathione shifts from reduced glutathione (GSH) to 164 oxidized glutathione (GSSG). Additional evidence of increased hepatic oxidative stress 165 was a significant shift in the cellular glutathione pools, favoring GSSG over GSH in CLP 166 mice compared to sham controls (Fig. 2g).  of septic mice (Fig. 3a). Sepsis also decreased transcription output that supports fatty 176 acid metabolic process components in the liver, as evidenced by a negative GSEA en-177 richment score (Fig. 3b).

178
Since hepatic lipid metabolism significantly changes during sepsis, we further in-179 vestigated the effect of sepsis on both the fatty acid oxidation and biosynthetic processes 180 pathways. Slight changes occurred in the expression profiles of genes involved in fatty 181 acid oxidation in response to CLP after 30 h. In contrast, we found a striking increase in 182 genes involved in fatty acid biosynthesis (Fig. 3c). We then asked whether the transcrip-   found that TCA metabolites that accumulate in hepatocytes during sepsis-citrate, cis-234 aconitate, and itaconate-returned almost to sham levels after DCA treatment (Fig. 5c).

235
Of particular note, we also observed that DCA treatment did not restore fumarate levels, 236 but rather decreased this TCA intermediate decreased even further than measured in 237 septic livers not exposed to the drug (Fig. 5c).

238
Next, we focused on metabolites proximal to the glutathione synthesis redox path-239 way, given the transcriptional reversal of the conjugation to glutathione pathway in DCA 240 treated septic mice and distal to methionine (Fig. 5b). We found that DCA not only caused 241 a reversal of cystathionine and taurine low levels in septic mice, but also increased accu-242 mulation of these metabolites in the liver after DCA (Fig. 5d). No significant differences   Given that DCA restores levels of sepsis-dysregulated circulating metabolites, we 267 hypothesized that PDK inhibition might also reverse the sepsis-induced anorexia re-268 sponse and starvation. To test this hypothesis, we housed septic mice treated with or 269 without DCA in metabolic cages for after sepsis onset. Consistent with previous reports          Total food intake Total food intake (g) ** ***