Abstract

   Recipients often suffer from hyperlactatemia during liver
   transplantation (LT), but whether hyperlactatemia exacerbates hepatic
   ischemia‐reperfusion injury (IRI) after donor liver implantation
   remains unclear. Here, the role of hyperlactatemia in hepatic IRI is
   explored. In this work, hyperlactatemia is found to exacerbate
   ferroptosis during hepatic IRI. Lactate‐primed lysine acetyltransferase
   8 (KAT8) is determined to directly lactylate mitochondrial
   phosphoenolpyruvate carboxykinase 2 (PCK2) at Lys100 and augments PCK2
   kinase activity. By using gene‐edited mice, evidence indicating that
   PCK2 exacerbates hepatic ferroptosis during IRI is generated.
   Mechanistically, PCK2 lactylate at Lys100 acts as a critical inducer of
   ferroptosis during IRI by competitively inhibiting the Parkin‐mediated
   polyubiquitination of 3‐oxoacyl‐ACP synthase (OXSM), thereby leading to
   metabolic remodeling of mitochondrial fatty acid synthesis (mtFAS) and
   the potentiation of oxidative phosphorylation and the tricarboxylic
   acid cycle. More importantly, targeting PCK2 is demonstrated to
   markedly ameliorate hyperlactatemia‐mediated ferroptosis during hepatic
   IRI. Collectively, the findings support the use of therapeutics
   targeting PCK2 to suppress hepatic ferroptosis and IRI in patients with
   hyperlactatemia during LT.

   Keywords: ferroptosis, hepatic ischemia‒reperfusion injury, lactate,
   lactylation, metabolic reprogramming, mitochondrial fatty acid
   synthesis, phosphoenolpyruvate carboxykinase 2
     __________________________________________________________________

   Recipients often suffer from hyperlactatemia during liver
   transplantation (LT). This study discovers that lactate‐primed KAT8
   directly lactylates mitochondrial PCK2 at Lys100 (PCK2^K100lac).
   PCK2^K100lac acts as a critical inducer of ferroptosis during
   ischemia‐reperfusion injury (IRI) by competitively inhibiting the
   Parkin‐mediated polyubiquitination of OXSM, thereby leading to
   metabolic remodeling of mitochondrial fatty acid synthesis. This
   process interprets hyperlactatemia‐mediated ferroptosis during hepatic
   IRI.

   graphic file with name ADVS-12-2414141-g005.jpg

1. Introduction

   Liver transplantation (LT) remains an indispensable treatment option
   for patients with end‐stage liver disease and liver malignancies.^[
   [46]^1 ^] Despite the proposal of various novel perfusion strategies,^[
   [47]^2 ^] hepatic ischemia‐reperfusion injury (IRI), a risk factor for
   primary graft dysfunction or nonfunction as well as acute and chronic
   rejection,^[ [48]^3 ^] remains inevitable. Identifying novel strategies
   to mitigate hepatic IRI would not only improve clinical outcomes but
   also enable the utilization of marginal liver grafts and thus expand
   the donor pool for life‐saving LT. Notably, due to increased anaerobic
   metabolism and reduced compensatory capacity, recipients often suffer
   from hyperlactatemia during LT surgery.^[ [49]^4 ^] Although clinical
   data indicate that hyperlactatemia is associated with primary graft
   dysfunction and mortality following LT,^[ [50]^4 , [51]^5 ^] whether
   hyperlactatemia exacerbates hepatic IRI after donor liver implantation
   remains unclear.

   Recent studies have revealed that hyperlactatemia is strongly
   associated with disturbances in several metabolic pathways, such as the
   tricarboxylic acid (TCA) cycle and oxidative phosphorylation
   (OXPHOS),^[ [52]^6 , [53]^7 ^] both of which generate excess reactive
   oxygen species and may facilitate lipid peroxidation.^[ [54]^8 ^]
   Notably, lipid peroxidation is a direct trigger of ferroptosis.
   Ferroptosis, a newly discovered form of programmed cell death driven by
   iron‐dependent phospholipid peroxidation, has been shown to play an
   essential role in IRI to various organs.^[ [55]^9 ^] Several studies
   have demonstrated that the inhibition of ferroptosis can attenuate
   apoptosis, necrosis, and inflammatory infiltration during hepatic
   IRI.^[ [56]^10 , [57]^11 ^] Moreover, as the main organ for iron
   storage, the liver is more susceptible to ferroptosis than the heart
   and kidney are. Thus, we speculate that hyperlactatemia exacerbates
   ferroptosis during hepatic IRI. However, the role of ferroptosis in
   hyperlactatemia‐mediated IRI remains unclear, and revealing the
   potential regulatory mechanisms is crucial for developing effective
   therapies to mitigate hyperlactatemia‐mediated hepatic IRI.

   Given the crucial involvement of mitochondria in various cellular
   processes, such as reactive oxygen species (ROS) production, energy
   metabolism, redox status, and iron metabolism, increasing evidence
   suggests a vital role for mitochondria in the regulation and execution
   of ferroptosis.^[ [58]^12 ^] Mitochondria are often thought of as
   predominantly catabolic organelles, breaking down carbon sources for
   use as fuel through fatty acid oxidation and the TCA cycle. In addition
   to these well‐known energy‐producing metabolic processes, a plethora of
   anabolic biochemical pathways, including those that generate
   nonessential amino acids, cholesterol, and steroid hormones, also occur
   in mitochondria. The mitochondrial fatty acid synthesis (mtFAS) pathway
   was first described in the yeasts Saccharomyces cerevisiae and
   Neurospora crassa, in which it was discovered to be required for the
   production of lipoic acid, a co‐factor required for several
   mitochondrial metabolic enzymes involved in OXPHOS and the TCA cycle,
   such as NADH:ubiquinone oxidoreductase subunit A6 (NDUFA6) and
   α‐ketoglutarate dehydrogenase (OGDH).^[ [59]^13 , [60]^14 ^] However,
   the potential role and related functional mechanisms of the mtFAS
   pathway in hepatic ferroptosis and IRI are largely unknown.

   Here, we investigated the role of hyperlactatemia in ferroptosis during
   hepatic IRI. We found a positive correlation between hyperlactatemia
   and hepatic ferroptosis during LT surgery. We then determined that
   lactate‐primed lysine acetyltransferase 8 (KAT8) directly lactylates
   phosphoenolpyruvate carboxykinase 2 (PCK2) at Lys100 and increases PCK2
   kinase activity. By using gene‐edited mice, we generated evidence
   indicating that PCK2 exacerbates hepatic ferroptosis during IRI.
   Mechanistically, lactylation of PCK2 at Lys100 suppresses
   Parkin‐mediated polyubiquitination of 3‐oxoacyl‐ACP synthase (OXSM),
   thereby leading to metabolic reprogramming to mtFAS, which in turn
   potentiates OXPHOS and TCA cycle activity. Importantly, we demonstrated
   that targeting PCK2 markedly ameliorates hyperlactatemia‐mediated
   ferroptosis during hepatic IRI. Collectively, our findings provide a
   novel PCK2‐targeted therapeutic strategy to suppress ferroptosis in
   patients with hyperlactatemia during hepatic IRI.

2. Results

2.1. Lactate Levels are Associated with Ferroptosis During Hepatic IRI

   To investigate the relationship between hyperlactatemia and hepatic
   IRI, we first assessed the correlation between serum lactate levels and
   liver injury in a cohort of adult patients undergoing LT (n = 30). The
   patients were divided into a hyperlactatemia (HLA) group (>
   = 2.1 mmol L^−1, n = 18) and a control (CTL) group (< 2.1 mmol L^−1,
   n = 12) according to whether they experienced hyperlactatemia during LT
   surgery. Compared with the CTL group, the HLA group presented higher
   serum alanine aminotransferase (ALT) and aspartate aminotransferase
   (AST) levels at postoperative day (POD)1 (Figure [61] 1A) and slower
   recovery of serum ALT and AST levels from POD1 to POD7 (Figure [62]1B).
   Moreover, liver biopsies from the HLA group presented more obvious
   alterations in liver histology and increased hepatocellular death, as
   shown by hematoxylin‒eosin (HE) staining (Figure [63]1C) and TUNEL
   staining (Figure [64]1D).

Figure 1.

   Figure 1
   [65]Open in a new tab

   Hyperlactatemia aggravates ferroptosis during hepatic IRI. A) Serum AST
   and ALT levels in LT recipients in cohort 1 on POD1 (HLA group, n = 18;
   CTL group, n = 12). B) Curve of the serum ALT and AST levels in LT
   recipients in cohort 1 from POD1 to POD7 (HLA group, n = 18; CTL group,
   n = 12). C) Representative images of HE staining of LT biopsies from
   cohort 1 (scale bar = 200 µm). D) Representative images and relative
   quantification of TUNEL staining in LT biopsies from cohort 1 (scale
   bar = 200 µm). E) ROS, MDA, GSH, and SOD levels in LT biopsies from
   cohort 1 (HLA group, n = 18; CTL group, n = 12). F) Representative
   images of Prussian blue staining of LT biopsies from cohort 1 (scale
   bar = 50 µm). G) Iron content in LT biopsies from cohort 1 (HLA group,
   n = 18; CTL group, n = 12). H) Western blot analysis of
   ferroptosis‐associated protein levels in LT biopsies before and after
   reperfusion. I) Schematic diagram of the drug treatment experiment
   utilizing hepatic IRI model mice. J) Serum AST and ALT levels in mice
   (n = 7/group) treated with/without lactate following IRI. K)
   Representative images of HE‐stained liver tissues from mice treated
   with/without lactate following IRI (scale bar = 100 µm). L) Western
   blot analysis of ferroptosis‐associated protein levels in liver tissues
   from mice treated with/without lactate following IRI. M) Representative
   TEM images of liver tissues from mice treated with/without lactate
   following IRI (scale bar = 2 µm). N) Representative images and relative
   quantification of TUNEL staining in MPH and THLE2 cells treated
   with/without lactate following H/R (scale bar = 200 µm). O) ROS, MDA,
   GSH, and SOD levels in MPH cells treated with/without lactate following
   H/R. For all the above experiments, the data are presented as the means
   ± SDs. For N‒O, 3 independent experiments (n = 3) with similar results
   were performed in triplicate. In A, D, E, G, J, N, and O, the
   statistical analyses were performed via two‐tailed unpaired Student's
   t‐tests. p < 0.05 was considered to indicate statistical significance.

   Next, we investigated whether hyperlactatemia exacerbates ferroptosis
   during LT surgery. Increased lipid peroxidation is an essential feature
   of ferroptosis.^[ [66]^15 ^] We measured the levels of ROS and
   malondialdehyde (MDA), indicators of oxidative stress, as well as the
   contents of glutathione (GSH) and superoxide dismutase (SOD), two
   crucial antioxidant factors, and found greater lipid peroxidation in
   the HLA group than in the CTL group (Figure [67]1E). The accumulation
   of iron, another important factor that triggers ferroptosis,^[ [68]^15
   ^] was significantly greater after reperfusion in the HLA group than in
   the CTL group (Figure [69]1F,G). In addition, multiple hallmarks
   associated with ferroptosis were analyzed, and the protein levels of
   nuclear factor erythroid‐derived 2‐like 2 (NRF2), solute carrier family
   7 member 11 (SLC7A11), and glutathione peroxidase 4 (GPX4), which can
   inhibit ferroptosis, were dramatically lower after reperfusion in the
   HLA group than in the CTL group; however, the expression levels of the
   ferroptosis‐promoting proteins cyclooxygenase 2 (COX2), acyl‐CoA
   synthetase long‐chain family member 4 (ACSL4) and transferrin receptor
   (TFRC) were greater in the HLA group than in the CTL group
   (Figure [70]1H). We thus speculate that hyperlactatemia may aggravate
   hepatic IRI through ferroptosis.

   To test this hypothesis, we established a lactate‐pretreated mouse
   hepatic IRI model (Figure [71]1I, top). The in vivo results confirmed
   that the livers of the mice in the HLA group were more susceptible to
   IRI than those of mice in the CTL group, as evidenced by higher serum
   ALT and AST levels (Figure [72]1J), poorly preserved hepatic
   architecture (Figure [73]1K), and increased cell death (Figure [74]S1A,
   Supporting Information). Consistent with the clinical observations,
   Prussian blue staining (Figure [75]S1B, Supporting Information), iron
   content analysis (Figure [76]S1C, Supporting Information), lipid
   peroxidation analysis (Figure [77]S1D), and Western blotting
   (Figure [78]1L) further demonstrated that ferroptosis was more severe
   in the livers of mice in the HLA group than in those of mice in the CTL
   group. Transmission electron microscopy (TEM) is frequently used to
   detect alterations in the morphology of mitochondria associated with
   ferroptosis.^[ [79]^15 ^] Importantly, TEM revealed that, compared with
   those of mice in the CTL group, the livers of mice in the HLA group
   exhibited ferroptosis during IRI, with smaller mitochondria and fewer
   or no mitochondrial cristae (Figure [80]1M).

   We further used mouse primary hepatocytes (MPHs) and THLE2 cells to
   validate the above results. Figure [81]S1E,F (Supporting Information)
   illustrates the protocol of MPH extraction as well as the method used
   to construct a cellular hypoxia‒reoxygenation (H/R) injury model. TUNEL
   staining (Figure [82]1N) and calcein‐acetoxymethyl
   (Calcein‐AM)/propidium iodide (PI) double‐staining (Figure [83]S1G,
   Supporting Information) confirmed that hepatocytes cultured under
   high‐lactate conditions exhibited increased cell death during H/R.
   Lipid peroxidation analysis (Figure [84]1O; Figure [85]S1H, Supporting
   Information) and Western blotting (Figure [86]S1I, Supporting
   Information) further confirmed that hepatocytes cultured under
   high‐lactate conditions exhibited increased ferroptosis during H/R.
   Consistent with these findings, TEM revealed an increased mitochondrial
   membrane density, fragmented mitochondrial outer membranes, and
   disrupted mitochondrial ridges in hepatocytes cultured with high
   lactate levels compared with hepatocytes cultured under normal
   conditions (Figure [87]S1J, Supporting Information).

   Furthermore, we investigated the role of ferroptosis in
   hyperlactatemia‐mediated hepatic IRI. Mice with hyperlactatemia were
   treated with ferrostatin‐1 (Fer‐1), a selective inhibitor of
   ferroptosis (Figure [88]1I, bottom). The results indicated that Fer‐1
   markedly attenuated hepatic IRI in mice with hyperlactatemia, as shown
   by reduced serum transaminase levels (Figure [89]S2A, Supporting
   Information), relatively well‐preserved hepatic architecture (Figure
   [90]S2B, Supporting Information), and decreased cell death (Figure
   [91]S2C, Supporting Information). Moreover, Fer‐1 markedly reduced the
   accumulation of iron (Figure [92]S2D, Supporting Information) and lipid
   peroxidation (Figure [93]S2E, Supporting Information) in the livers of
   mice with hyperlactatemia during IRI. Data from in vitro studies also
   validated the effect of Fer‐1 in attenuating hyperlactatemia‐mediated
   hepatic IRI. TUNEL staining (Figure [94]S2F, Supporting Information),
   calcein‐AM/PI double staining (Figure [95]S2G, Supporting Information),
   lipid peroxidation analysis (Figure [96]S2H, Supporting Information),
   and Western blotting (Figure [97]S2I, Supporting Information) revealed
   that treating hepatocytes with Fer‐1 to inhibit ferroptosis
   dramatically ameliorated H/R‐induced damage to hepatocytes cultured
   under high‐lactate conditions. Taken together, these data demonstrate
   that lactate exacerbates hepatic IRI through ferroptosis.

2.2. Lactate Induces the Lactylation of Mitochondrial PCK2 at K100 During
Hepatic IRI

   Lactate can alter protein expression, localization, or function via
   lysine lactylation (Kla).^[ [98]^16 ^] Using an anti‐pan‐Kla antibody,
   we observed that the global levels of lactylated lysine were
   substantially greater in the liver tissues subjected to hyperlactatemia
   during LT surgery (Figure [99] 2A), which was validated in
   lactate‐pretreated hepatic IRI models in vitro (Figure [100]2B; Figure
   [101]S3A, Supporting Information) and in vivo (Figure [102]S3B,
   Supporting Information). To gain comprehensive insight into
   hyperlactatemia‐mediated ferroptosis‐related lactylation, we used
   liquid chromatography‒tandem mass spectrometry (LC‒MS/MS) analysis to
   investigate lysine‐lactylated substrates in THLE2 cells treated
   with/without lactate following H/R. Notably, we identified 112 lysine
   lactylation sites across 75 proteins (Table [103]S1, Supplementary
   Table), with significantly increased lactylation levels at 107 lysine
   lactylation sites on 72 proteins (fold change ≥ 2, p < 0.05)
   (Figure [104]2C). We subsequently performed Kyoto Encyclopedia of Genes
   and Genomes (KEGG) functional enrichment analysis on the proteins with
   increased Kla levels. Compared with hepatocytes exposed to H/R alone,
   the proteins that exhibited increased Kla in hepatocytes cultured under
   high‐lactate conditions and exposed to H/R were enriched in
   ferroptosis‐related metabolic pathways, such as glycolysis, the
   mevalonate pathway and lipid synthesis (Figure [105]2D). We further
   analyzed the ferroptosis network associated with the substrates that
   exhibited increased Kla levels via the Search Tool for the Retrieval of
   Interacting Genes/Proteins (STRING) database. Among the
   ferroptosis‐related proteins identified, mitochondrial PCK2 is the
   rate‐limiting enzyme that catalyzes the production of glucose from
   lactate (Figure [106]2E); however, its role in ferroptosis during
   hepatic IRI remains unknown. Importantly, H/R injury in THLE2 cells
   cultured under high‐lactate conditions was markedly attenuated when
   PCK2 was depleted (Figure [107]S2C,D, Supporting Information). Thus, we
   focused on PCK2 lactylation.

Figure 2.

   Figure 2
   [108]Open in a new tab

   Lactate facilitates the lactylation of PCK2 at K100 via KAT8. A)
   Representative images of IHC staining with an anti‐pan‐Kla antibody in
   LT biopsies (scale bar = 100 µm). B) Western blot analysis of pan‐Kla
   expression and silver staining of lysates from THLE2 cells treated
   with/without lactate following H/R. C) Scatter plot of global lysine
   lactylation changes in THLE2 cells treated with/without lactate
   following H/R. D) KEGG pathway enrichment analysis of lysine‐lactylated
   proteins in THLE2 cells treated with/without lactate following H/R. E)
   Protein‒protein interaction network analysis of ferroptosis‐related
   lysine‐lactylated proteins via the STRING database. F) LT biopsies from
   patients with/without hyperlactatemia were collected for IP with an
   anti‐pan‐Kla antibody, followed by Western blotting. G) MPHs and THLE2
   cells treated with/without lactate following H/R were collected for IP
   with an anti‐pan‐Kla antibody, followed by Western blotting. H)
   Potential lactylation sites of PCK2. I) After H/R exposure, lysates of
   THLE2 cells expressing Flag‐CTL, Flag‐PCK2, Flag‐PCK2 K100R, Flag‐PCK2
   K308R, or Flag‐PCK2 K522R and cultured under high‐lactate conditions
   were incubated with an anti‐Flag antibody, and interacting proteins
   were detected with an anti‐pan‐Kla antibody via Western blotting as
   indicated. J) LC–MS/MS analysis of modified (Kla)DVARVE is shown. K)
   The Lys100 residue of PCK2 is highly conserved in different species,
   ranging from M. musculus to H. sapiens. L) Lysates of HEK293T cells
   expressing Flag‐CTL, Flag‐PCK2, Flag‐PCK2 K100R, and Flag‐PCK2 K100E
   were incubated with an anti‐Flag antibody, and interacting proteins
   were detected with an anti‐PCK2^K100lac antibody via Western blotting
   as indicated. M) After H/R exposure, lysates of MPHs and THLE2 cells
   cultured with different concentrations of lactate were incubated with
   an anti‐PCK2^K100lac antibody for Western blotting. N) The cofactor
   pocket of KAT8 is bound to acetyl‐CoA (left) and to lactyl‐CoA (right).
   The transfer groups in acetyl‐CoA and lactyl‐CoA are indicated with red
   circles. O) Before and after reperfusion, MPH lysates were incubated
   with an anti‐PCK2 antibody, and interacting proteins were detected with
   an anti‐KAT8 antibody via Western blotting as indicated. WCL (Whole
   Cell Lysate) P) After H/R exposure, lysates of THLE2 cells cultured
   with/without lactate were incubated with an anti‐PCK2 antibody, and
   interacting proteins were detected with an anti‐KAT8 antibody via
   Western blotting as indicated. Q) Western blot analysis of KAT8,
   PCK2^K100lac, and PCK2 expression in THLE2 cells expressing HA‐CTL or
   HA‐KAT8. R) Western blot analysis of KAT8, PCK2^K100lac, and PCK2
   expression in THLE2 cells transfected with siCTL or siKAT8. S) Western
   blot analysis of KAT8, PCK2^K100lac, and PCK2 expression in THLE2 cells
   cultured with/without lactate and transfected with/without siKAT8, as
   indicated. T) PEPCK activity in THLE2 cells treated with/without
   lactate following H/R. U) PEPCK activity in THLE2 cells transfected
   with PCK2 siRNA or PCK1 siRNA and treated with lactate following H/R.
   V) PEPCK activity in THLE2 cells transfected with HA‐KAT8 following
   H/R. W) PEPCK activity in THLE2 cells transfected with HA‐KAT8 with or
   without PCK1/2 siRNA following H/R. X) PEPCK activity in
   KAT8‐overexpressing PCK2^KD THLE2 cells transfected with Flag‐PCK2,
   Flag‐PCK2 K100R or Flag‐PCK2 K100E following H/R. For all the above
   experiments, the data are presented as the means ± SDs. Three
   independent experiments (n = 3) with similar results were performed in
   triplicate. In A and T‐X, the statistical analyses were performed via
   two‐tailed unpaired Student's t‐tests. p < 0.05 was considered to
   indicate statistical significance.

   We immunoprecipitated total protein extracts from liver tissues or
   hepatocytes with an anti‐PCK2 antibody, followed by Western blotting
   with an anti‐pan‐Kla antibody. Indeed, PCK2 was lactylated in
   hepatocytes, and PCK2 lactylation was markedly increased in the livers
   of lactate‐pretreated mice (Figure [109]2F) and hepatocytes cultured
   under high‐lactate conditions (Figure [110]2G). LC–MS/MS analysis
   revealed four lactylation sites in the PCK2 protein sequence, but there
   was a significant increase in lactylation at only three lactylation
   (red) sites after lactate treatment (Figure [111]2H), implying that the
   lactylation of these three sites could be regulated in hepatocytes. To
   identify the critical lactylation site of PCK2 in
   hyperlactatemia‐mediated ferroptosis during hepatic IRI,
   lactylation‐defective mutants were generated for these three
   lactylation sites by substituting K for R. We found that only the PCK2
   K100R mutant resulted in reduced lactylation in THLE2 cells cultured
   under high‐lactate conditions (Figure [112]2I). The b‐y ion matching
   diagram for PCK2 K100 is shown in Figure [113]2J. Notably, the K100
   residue of PCK2 is highly conserved in different species, ranging from
   Homo sapiens to Mus musculus (Figure [114]2K). To further confirm that
   PCK2 is lactylated at K100 during hepatic IRI, an antibody that
   specifically recognizes PCK2 lactylated at K100 (PCK2^K100lac) was
   generated, and its specificity was verified via immunohistochemistry
   (IHC) assays (Figure [115]S3E, Supporting Information) and in cells
   expressing PCK2‐WT, ‐K100R, or ‐K100E (Figure [116]2L). Indeed, lactate
   treatment strongly induced PCK2 lactylation at K100 (Figure [117]2M),
   indicating that the K100 residue of PCK2 is an essential lactylation
   site in hyperlactatemia‐mediated ferroptosis during hepatic IRI. These
   results indicate that lactate induces mitochondrial PCK2 lactylation at
   K100 during hepatic IRI.

2.3. KAT8 Lactylates PCK2 at K100 and Promotes PCK2 Activity

   Next, we investigated the potential lysine lactyltransferase that
   lactylates PCK2. Proteins that interact with PCK2 were captured from
   THLE2 cell lysates via an anti‐PCK2 antibody and analyzed via LC‒MS/MS
   to generate a PCK2‐binding protein library (Table [118]S2, Supporting
   Information). In addition to previously reported PCK2‐binding proteins,
   including heat shock protein family D member 1 (HSPD1)^[ [119]^17 ^]
   and RAF1,^[ [120]^18 ^] KAT8 was identified as a novel binding partner
   of PCK2. KAT8, a member of the MYST protein family of histone
   acetyltransferases, was recently reported to be responsible for the Kla
   of protein substrates involved in diverse biological processes.^[
   [121]^19 ^]

   To investigate the interaction between PCK2 and KAT8, we docked lactyl
   coenzyme A (lactyl‐CoA) into the structure of KAT8. Lactyl‐CoA fits
   well in the cofactor pocket of KAT8, with the structure of the
   lactyl‐coA–KAT8 complex resembling that of the acetyl coenzyme A
   (acetyl‐CoA)–KAT8 complex (Figure [122]2N). Endogenous PCK2‐KAT8
   protein binding was validated in hepatocytes in the presence and
   absence of H/R (Figure [123]2O; Figure [124]S3F, Supporting
   Information). Coimmunoprecipitation (CoIP) and Western blot analysis of
   HEK293T cells revealed that reconstituted Flag‐PCK2 directly interacted
   with HA‐tagged KAT8 (Figure [125]S3G, Supporting Information). We then
   investigated whether lactate influences the interaction between KAT8
   and PCK2. Strikingly, lactate treatment slightly increased the binding
   of KAT8 and PCK2 after reoxygenation (Figure [126]2P). Moreover, PCK2
   lactylation at K100 increased when KAT8 was overexpressed
   (Figure [127]2Q), whereas it was suppressed by KAT8 knockdown
   (Figure [128]2R). Moreover, lactate‐induced lactylation of PCK2 at K100
   was inhibited by KAT8 knockdown (Figure [129]2S). Collectively, these
   results indicate that KAT8 interacts with and lactylates PCK2 at K100.

   We next investigated the effect of KAT8 on the regulation of PCK2
   expression and activity. Notably, exogenous overexpression or knockdown
   of KAT8 did not affect the amount of PCK2 protein expressed
   (Figure [130]2Q,R). Lactate also failed to affect PCK2 protein
   expression (Figure [131]2S), suggesting that KAT8 expression or lactate
   exposure does not directly affect PCK2 protein levels in hepatocytes.
   Interestingly, we found that lactate treatment promoted
   phosphoenolpyruvate carboxykinase (PEPCK) activity in MPH and THLE2
   cells after reoxygenation (Figure [132]2T). We next established stable
   PCK2 knockdown (PCK2^KD) THLE2 cells (Figure [133]S3H, Supporting
   Information). PCK2 knockdown affected cell activity (Figure [134]S3I,
   Supporting Information) but did not directly result in cell death
   (Figure [135]S3J, Supporting Information), which was consistent with
   previous reports.^[ [136]^20 ^] Notably, PCK2 knockdown, but not PCK1
   knockdown, markedly suppressed the activation of PEPCK by lactate in
   THLE2 cells (Figure [137]2U). Similarly, KAT8 overexpression increased
   PEPCK activity in THLE2 cells after reoxygenation (Figure [138]2V),
   whereas only PCK2 knockdown dramatically inhibited the activation of
   PEPCK caused by KAT8 overexpression in THLE2 cells (Figure [139]2W).
   Furthermore, KAT8 overexpression increased PEPCK activity in
   PCK2‐overexpressing PCK2^KD THLE2 cells but not in PCK2 K100R‐ or PCK
   K100E‐overexpressing PCK2^KD THLE2 cells (Figure [140]2X). Taken
   together, these results indicate that KAT8 promotes PCK2 activity
   through lactylation at K100 during hepatic IRI.

2.4. PCK2 Renders the Liver Susceptible to Ferroptosis and IRI

   To further investigate whether PCK2 mediates ferroptosis during hepatic
   IRI, we performed in vitro assays using THLE2 cells. The results
   demonstrated that overexpression of PCK2, but not PCK2 K100R,
   dramatically aggravated hepatocyte damage during H/R, as shown by TUNEL
   staining (Figure [141] 3A) and calcein‐AM/PI staining (Figure [142]3B).
   Moreover, quantitative analysis and Western blotting revealed that PCK2
   overexpression in hepatocytes markedly promoted the accumulation of
   intracellular iron (Figure [143]3C) and lipid peroxidation
   (Figure [144]3D), which was accompanied by an increase in the
   expression of ferroptosis‐inducing proteins and a decrease in the
   expression of ferroptosis‐inhibiting proteins (Figure [145]3E) during
   H/R. The inhibition of KAT8 acetyltransferase reversed these changes,
   including reducing hepatocyte damage (Figure [146]S4A,B, Supporting
   Information) and decreasing susceptibility to ferroptosis (Figure
   [147]S4C–E, Supporting Information) during H/R. These results suggest
   that KAT8‐mediated lactylation of PCK2 at K100 exacerbates hepatic
   ferroptosis and IRI.

Figure 3.

   Figure 3
   [148]Open in a new tab

   PCK2 renders the liver susceptible to ferroptosis during IRI. A–E)
   THLE2 cells were transfected with Flag‐CTL, Flag‐PCK2, or Flag‐PCK2
   K100R and subjected to H/R. (A) Representative images and relative
   quantification of TUNEL staining (scale bar = 200 µm). (B)
   Representative images and relative quantification of calcein‐AM/PI
   double‐staining (scale bar = 100 µm). (C) Iron content. (D) ROS, MDA,
   GSH and SOD levels. (E) Western blot analysis of ferroptosis‐associated
   protein expression. F–I) Hepatic IRI in AAV‐V or AAV‐PCK2‐treated mice
   (n = 7/group). (F) Serum AST and ALT levels at the indicated times. (G)
   Representative images of HE‐stained livers at the indicated times
   (scale bar = 1 mm). (H) Western blot analysis of ferroptosis‐associated
   protein expression. (I) Representative TEM images of liver tissues
   after reperfusion for 3 h (scale bar = 2 µm). J) Serum AST and ALT
   levels in LT recipients in cohort 2 on POD1 (low‐PCK2 group, n = 52;
   high‐PCK2 group, n = 52). K) The curve of serum ALT and AST levels in
   LT recipients in cohort 2 from POD1 to POD3 (low‐PCK2 group, n = 52;
   high‐PCK2 group, n = 52). L) Pearson correlation analysis between PCK2
   expression in donor's livers and the serum ALT or AST levels of
   recipients in cohort 2 on POD1 (n = 100). M) Representative images and
   correlation analysis of IHC staining of COX2, ACSL4, FTL, and GPX4 in
   the donor livers of cohort 2 (scale bar = 100 µm). For all the above
   experiments, the data are presented as the means ± SDs. Three
   independent experiments (n = 3) with similar results were performed in
   triplicate. The statistical analyses were performed via two‐tailed
   unpaired Student's t‐tests (A‐D, F, J) and chi‐square tests (M).
   p < 0.05 was considered to indicate statistical significance.

   To further validate the role of ferroptosis in PCK2‐mediated
   aggravation of hepatic IRI, mice were intravenously injected with a
   PCK2‐expressing adeno‐associated virus‐8 (AAV8) to generate a mouse
   strain in which the liver consistently highly expresses PCK2
   (AAV‐PCK2). An empty vector (AAV‐V) was used as a negative control. The
   quantitative real‐time polymerase chain reaction (qRT‒PCR) (Figure
   [149]S4F, Supporting Information), Western blot (Figure [150]S4G,
   Supporting Information), and IHC results (Figure [151]S4H, Supporting
   Information) collectively confirmed the ectopic expression of PCK2 in
   AAV‐PCK2 mice. A mouse hepatic IRI model was subsequently constructed.
   Elevated serum transaminase levels (Figure [152]3F) and increased
   necrotic areas (Figure [153]3G) indicated that PCK2 overexpression
   substantially aggravated hepatic IRI, most notably after 6 h of
   reperfusion. Moreover, the results of Prussian blue staining (Figure
   [154]S4I, Supporting Information), iron content analysis (Figure
   [155]S4J, Supporting Information), lipid peroxidation analysis (Figure
   [156]S4K, Supporting Information), Western blotting (Figure [157]3H),
   and TEM (Figure [158]3I) confirmed that PCK2 markedly exacerbated
   hepatic ferroptosis during IRI. Additionally, we administered Fer‐1 to
   AAV‐PCK2 mice and found that Fer‐1 significantly ameliorated hepatocyte
   IRI and ferroptosis in AAV‐PCK2 mice, as evidenced by changes in serum
   transaminase levels (Figure [159]S4L, Supporting Information), liver
   histology (Figure [160]S4M, Supporting Information), intracellular iron
   accumulation (Figure [161]S4N‐O, Supporting Information), lipid
   peroxidation (Figure [162]S4P, Supporting Information) and
   ferroptosis‐related protein expression (Figure [163]S4Q, Supporting
   Information), which further indicated that PCK2 aggravates hepatic IRI
   through ferroptosis.

   Next, we analyzed the clinical correlation between PCK2 levels in donor
   livers and hepatic IRI after LT. We utilized 1:1 propensity score
   matching (PSM) analysis to exclude cases in which clinical factors
   affected the recovery of LT recipients and enrolled 100 patients, who
   were then divided into 2 groups on the basis of the median value of
   PCK2 expression (Table[164]S3, Supporting Information). Patients who
   received a donor liver with high PCK2 expression had significantly
   higher serum ALT and AST levels on POD1 (Figure [165]3J; Table [166]S4,
   Supporting Information), as well as slower recovery of AST and ALT
   levels from POD1‐POD3 (Figure [167]3K). Further correlation analysis
   suggested that PCK2 levels in donor's livers were significantly
   positively correlated with ALT and AST levels in recipients on POD1
   (Figure [168]3L). We further assessed the clinical association between
   PCK2 expression and ferroptosis susceptibility in donor's livers. Using
   qRT‒PCR analysis, we detected a significant negative correlation
   between PCK2 expression and ferritin heavy chain 1 (FTH) and ferritin
   light chain (FTL) expression (Figure [169]S4R, Supporting Information),
   indicating that the iron reserve capacity is reduced in livers with
   high PCK2 expression. Moreover, PCK2 expression tended to be
   significantly negatively correlated with glutathione peroxidase 4
   (GPX4) expression (Figure [170]S4S, Supporting Information), further
   suggesting that patients with high PCK2 expression have a reduced
   hepatic antioxidant capacity. Most importantly, the data revealed a
   significant positive correlation between the expression of PCK2 and
   that of the ferroptosis marker COX2 (Figure [171]S4T, Supporting
   Information). Consistent with these results, IHC analysis of donor's
   livers further confirmed that the PCK2 level was positively correlated
   with the expression of the ferroptosis drivers COX2 and ACSL4 and
   strongly inversely correlated with the ferroptosis suppressors GPX4 and
   FTL (Figure [172]3M). These data suggest that high PCK2 expression may
   exacerbate hepatic ferroptosis and IRI.

2.5. PCK2 Reprograms OXSM‐Mediated mtFAS

   PCK2 is involved in hepatic gluconeogenesis.^[ [173]^21 ^] To further
   investigate whether PCK2 regulates hepatic IRI and ferroptosis through
   metabolic remodeling, we performed a mass spectrometry‐based
   metabolomic analysis of livers from AAV‐PCK2 mice and AAV‐V mice.
   Principal component analysis revealed that PCK2 overexpression led to a
   global metabolic alteration (Figure [174] 4A), with the levels of 23
   metabolites being markedly increased and the levels of 7 metabolites
   being obviously decreased (fold change ≥ 1.5, p < 0.05)
   (Figure [175]4B). Enrichment analysis of the differentially abundant
   metabolites revealed an increased abundance of metabolites involved in
   the TCA cycle, OXPHOS, and amino acid and purine metabolism in the
   livers of AAV‐PCK2 mice compared with those of the AAV‐V mice
   (Figure [176]4C). Specifically, the levels of multiple metabolites
   associated with OXPHOS (Figure [177]4D) and the TCA cycle
   (Figure [178]4E) were dramatically elevated in the livers of AAV‐PCK2
   mice.

Figure 4.

   Figure 4
   [179]Open in a new tab

   Metabolomic and transcriptomic analyses revealed altered mtFAS and OXSM
   expression in the livers of AAV‐PCK2‐treated mice. A–G) Metabolite
   levels in the livers of AAV‐PCK2‐ and AAV‐V‐treated mice were analyzed
   by LC‐MS (n = 6 samples/group). (A) Dimensionality reduction analysis,
   the ellipse lines indicate 95% CI ranges with means. (B) Analysis of
   differentially abundant metabolites in the livers of AAV‐PCK2‐ and
   AAV‐V‐ treated mice. (C) Pathway analysis of the differentially
   abundant metabolites in the livers of AAV‐PCK2‐ and AAV‐V‐treated mice.
   (D) Left, schematic diagram of OXPHOS. Right, relative quantification
   of differentially abundant metabolites associated with OXPHOS in the
   livers of AAV‐PCK2‐ and AAV‐V‐treated mice. (E) Left, schematic diagram
   of the TCA cycle. Right, relative quantification of differentially
   abundant metabolites associated with the TCA cycle in the livers of
   AAV‐PCK2‐ and AAV‐V‐treated mice. (F‐I) Bulk RNA‐seq analysis of the
   livers of AAV‐PCK2‐ and AAV‐V‐treated mice. (F) Metabolism‐related
   genes with a change in expression >50% were subjected to KEGG pathway
   enrichment analysis, and plots of the metabolism‐related KEGG modules
   in which these genes were enriched were generated. (G) Schematic
   diagram of mtFAS to OXPHOS and the TCA cycle. H) Heatmap of the
   expression of the top upregulated genes involved in mtFAS. I) GSEA of
   metabolism‐related genes with a change in expression >50% in the livers
   of AAV‐PCK2‐ and AAV‐V‐treated mice. J) Western blot analysis of OXSM
   expression in the livers of AAV‐PCK2‐ and AAV‐V‐treated mice. K)
   Western blot analysis of OXSM expression in THLE2 cells transfected
   with Flag‐CTL, Flag‐PCK2, or Flag‐PCK2 K100R and subjected to H/R. For
   D‐E, the data are presented as the means ± SDs, and statistical
   analyses were performed via two‐tailed unpaired Student's t‐tests. For
   J‐K, 3 independent experiments (n = 3) with similar results were
   performed in triplicate. p < 0.05 was considered to indicate
   statistical significance.

   To further determine the metabolic gene expression signature of the
   livers of AAV‐PCK2 mice, we performed RNA sequencing (RNA‐seq) and then
   assessed the involvement of the differentially expressed
   metabolism‐related genes in 87 metabolism‐related KEGG pathways.
   Interestingly, transcriptome analysis revealed that the expression of a
   substantial portion of metabolism‐related genes was altered in the
   livers of AAV‐PCK2 mice compared with those of AAV‐V mice during
   hepatic IRI (Figure [180]4F, left). Indeed, among the 495
   differentially expressed metabolism‐related genes detected, genes
   related to OXPHOS, the TCA cycle, and mtFAS exhibited the greatest
   changes in expression (Figure [181]4F, right). However, the
   differentially expressed genes were not involved in other pathways in
   which PCK2 was previously been reported to be involved, specifically
   gluconeogenesis,^[ [182]^21 ^] suggesting that PCK2 may act through a
   nonclassical mechanism in AAV‐PCK2 mice during hepatic IRI. The mtFAS
   pathway is essential for the production of lipoic acid, a cofactor
   required for the maturation of mitochondrial dehydrogenases involved in
   OXPHOS and the TCA cycle (Figure [183]4G).^[ [184]^13 , [185]^14 ^]
   Thus, we speculated that PCK2 mediates multiple pathway alterations via
   mtFAS and exacerbates ferroptosis during hepatic IRI.

   Coincidentally, specific analysis of genes related to the mtFAS pathway
   revealed that OXSM and carbonyl reductase 4 (CBR4) were the 2 most
   robustly upregulated metabolism‐related genes after reperfusion
   (Figure [186]4H). Moreover, OXSM is a potential PCK2‐interacting
   protein (Table [187]S2, Supporting Information). Considering that OXSM
   is a rate‐limiting enzyme that catalyzes the committed step in de novo
   mtFAS,^[ [188]^13 , [189]^14 ^] we focused subsequent analyses on OXSM.
   Additionally, gene set enrichment analysis (GSEA) further confirmed
   that OXPHOS, the TCA cycle, and mtFAS were the top mitochondrial
   metabolic pathways in which the differentially expressed genes were
   enriched (Figure [190]4I), with OXSM expression being significantly
   upregulated in AAV‐PCK2 mice during hepatic IRI (Figure [191]4J); this
   finding that was verified by an increase in OXSM expression at the
   protein level in vitro (Figure [192]4K). Collectively, the results of
   the metabolomic and transcriptomic analyses revealed that OXSM‐mediated
   mtFAS is a major activated metabolic pathway in PCK2‐induced
   aggravation of hepatic ferroptosis and IRI.

2.6. Lactate Induces PCK2‐Dependent Metabolic Reprogramming During IRI

   Previous work revealed that mtFAS repletes lipoic acid and facilitates
   OXPHOS and the TCA cycle (Figure [193]4G),^[ [194]^13 , [195]^14 ^]
   whereas either OXPHOS or the TCA cycle increases cellular
   susceptibility to ferroptosis under physiological and pathological
   conditions.^[ [196]^8 ^] We found that lactate failed to dramatically
   increase the basal levels of OXSM as well as the levels of NDUFA6 and
   OGDH (Figure [197] 5A), molecular regulators that control OXPHOS and
   the TCA cycle (Figure [198]4G), respectively. However, lactate
   stimulated the expression of OXSM, OGDH, and NDUFA6 (Figure [199]5A) in
   the livers of patients after reperfusion. Consistently, lactate
   pretreatment increased OXSM, OGDH, and NDUFA6 protein levels (Figure
   [200]S5A, Supporting Information) in the livers of the mice in the IRI
   group but not in those of mice in the sham group. This finding was also
   confirmed in hepatocytes, where OXSM, OGDH, and NDUFA6 expression was
   dramatically increased by lactate stimulation but not under basal
   conditions after reoxygenation (Figure [201]5B,C). Consistent with
   these findings, qRT‒PCR assays further confirmed the significant
   upregulation of genes (Figure [202]5D,E) involved in OXPHOS and the TCA
   cycle (Figure [203]4G) in lactate‐pretreated hepatocytes. Moreover,
   real‐time oxygen consumption rate (OCR) and glycolytic rate
   measurements revealed that, after exposure to H/R, THLE2 cells cultured
   under high‐lactate conditions presented greater basal oxygen
   consumption, mitochondrial ATP respiration, and maximal oxygen
   consumption than did those cultured without lactate (Figure [204]5F).
   Moreover, basal glycolysis and glycolytic capability were markedly
   increased in lactate‐pretreated hepatocytes (Figure [205]5G),
   suggesting that lactate markedly potentiated OXPHOS and TCA cycle
   activity during hepatic IRI.

Figure 5.

   Figure 5
   [206]Open in a new tab

   Hyperlactatemia facilitates PCK2‐dependent mtFAS metabolic
   reprogramming. A) Western blot analysis of OXSM, OGDH, and NDUFA6
   expression in LT biopsies before and after reperfusion. B, C) Western
   blot analysis of OXSM, OGDH, and NDUFA6 expression in (B) MPHs and (C)
   THLE2 cells treated with/without lactate for 24 h and subjected to H/R.
   (D‐E) Real‐time qPCR validation of the top significantly downregulated
   D) OXPHOS‐related genes and E) TCA cycle‐related genes in THLE2 cells
   treated with/without lactate for 24 h and subjected to H/R. F, G)
   Real‐time Seahorse OCR and glycolytic rate measurements to evaluate the
   OXPHOS and TCA cycle rate in THLE2 cells treated with/without lactate
   for 24 h and subjected to H/R. H) Western blot analysis of OXSM, OGDH,
   and NDUFA6 expression in THLE2 cells transfected with shCTL or shPCK2,
   subjected to H/R, and cultured under high‐lactate conditions. I, J)
   Real‐time qPCR validation of the top significantly downregulated (I)
   OXPHOS‐related genes and (J) TCA cycle‐related genes in THLE2 cells
   transfected with shCTL or shPCK2, subjected to H/R, and cultured under
   high‐lactate conditions. K, L) Real‐time Seahorse OCR and glycolytic
   rate measurements to evaluate the OXPHOS and TCA cycle rate in THLE2
   cells transfected with shCTL or shPCK2, subjected to H/R, and cultured
   under high‐lactate conditions. M, N) Real‐time Seahorse OCR and
   glycolytic rate measurements to evaluate the OXPHOS and TCA cycle rate
   in THLE2 cells transfected with Flag‐CTL, Flag‐PCK2 or Flag‐PCK2 K100R,
   subjected to H/R, and cultured under high‐lactate conditions. O, P)
   Real‐time Seahorse OCR and glycolytic rate measurements to evaluate the
   OXPHOS and TCA cycle rate in THLE2 cells transfected with siCTL or
   siKAT8, subjected to H/R, and cultured under high‐lactate conditions.
   For all the above experiments, the data are presented as the means ±
   SDs. Three independent experiments (n = 3) with similar results were
   performed in triplicate. In D‐G and I‐P, the statistical analyses were
   performed via two‐tailed unpaired Student's t‐tests. p < 0.05 was
   considered to indicate statistical significance.

   Next, we investigated whether PCK2 controls mtFAS‐mediated metabolic
   reprogramming induced by lactate during hepatic IRI. In contrast to
   those in lactate‐pretreated THLE2 cells expressing shCTL, the protein
   levels of OXSM, OGDH, and NDUFA6 in PCK2^KD THLE2 cells were
   dramatically lower (Figure [207]5H). At the mRNA level, the expression
   of several factors pivotal for OXPHOS and the TCA cycle was
   consistently downregulated in PCK2^KD THLE2 cells (Figure [208]5I,J).
   Real‐time OCR and glycolytic rate measurements revealed that even in
   the presence of lactate, basal oxygen consumption, mitochondrial ATP
   respiration, maximal oxygen consumption (Figure [209]5K), basal
   glycolysis and glycolytic capability (Figure [210]5L) were decreased in
   PCK2^KD THLE2 cells following H/R, supporting our hypothesis that
   lactate promotes metabolic reprogramming in a PCK2‐dependent manner.

   We further investigated the effect of PCK2 lactylation at K100 on
   PCK2‐mediated mtFAS metabolic reprogramming. We constructed PCK2‐ or
   PCK2 K100R‐overexpressing PCK2^KD THLE2 cells. As anticipated, after
   H/R, PCK2‐overexpressing, but not PCK2 K100R‐overexpressing PCK2^KD
   THLE2 cells, presented increased OXSM, OGDH, and NDUFA6 protein levels
   (Figure [211]S5B, Supporting Information), increased mRNA levels of
   OXPHOS‐ and TCA cycle‐related genes (Figure [212]S5C,D, Supporting
   Information), and increased OCRs and glycolytic rates
   (Figure [213]5M,N, Supporting Information). Moreover, in THLE2 cells
   transfected with KAT8 siRNAs (Figure [214]S5E, Supporting Information),
   downregulated of OXPHOS‐ and TCA cycle‐related genes were also detected
   (Figure [215]5O,P; Figure [216]S5E–G, Supporting Information).
   Collectively, these results reveal an underexplored mechanism by which
   PCK2 mediates hyperlactatemia‐induced mtFAS metabolic reprogramming.

2.7. PCK2 Restricts the Parkin‐Mediated Polyubiquitination of OXSM

   Next, we investigated the underlying mechanisms by which PCK2 regulates
   ferroptosis during hepatic IRI. By treating cells with
   phosphoenolpyruvate, a direct metabolite of PCK2, we excluded the
   possibility that PCK2 regulates OXSM, OGDH, and NDUFA6 expression in a
   metabolism‐dependent manner (Figure [217]S6A, Supporting Information).
   As previously described, OXSM is a potential PCK2‐binding protein
   (Table [218]S2, Supporting Information); thus, we analyzed the
   interaction between PCK2 and OXSM. Indeed, CoIP analysis indicated that
   PCK2 interacted with endogenous OXSM in hepatocytes (Figure [219] 6A).
   We also confirmed the interaction between PCK2 and OXSM by coexpressing
   Flag‐PCK2 and MYC‐OXSM in HEK293T cells (Figure [220]6B,C). The results
   of the glutathione‐S‐transferase (GST) pull‐down assay further
   confirmed the direct interaction between PCK2 and OXSM
   (Figure [221]6D). Additionally, cell fractionation analysis revealed
   that PCK2 and OXSM colocalized in mitochondria (Figure [222]6E). Taken
   together, these data indicate that PCK2 physically interacts with OXSM.

Figure 6.

   Figure 6
   [223]Open in a new tab

   PCK2 abrogates the Parkin‐induced degradation of OXSM. A) MPH and THLE2
   cell lysates were incubated with an anti‐PCK2 antibody, and interacting
   proteins were detected with an anti‐OXSM antibody via Western blotting
   as indicated. B, C) Lysates of HEK293T cells expressing Flag‐PCK2 and
   MYC‐OXSM were incubated with an anti‐Flag or anti‐MYC antibody, and
   interacting proteins were detected with an anti‐MYC or anti‐Flag
   antibody via Western blotting as indicated. D) GST‐PCK2 or GST was
   incubated with total lysates of THLE2 cells expressing MYC‐OXSM for 2
   hours, and then Western blotting with anti‐MYC or anti‐GST antibodies
   was performed as indicated. E) The localization of PCK2 and OXSM in
   MPHs and THLE2 cells was analyzed by immunoblotting. F) Western blot
   analysis of OXSM expression in THLE2 cells transfected with Flag‐PCK2
   or Flag‐PCK2 K100R. G) Western blot analysis of OXSM expression in
   THLE2 cells transfected with siCTL or siPCK2 and cultured under normal
   or high‐lactate conditions. H) Western blot analysis of OXSM expression
   in MPHs and THLE2 cells treated with the indicated concentration of
   lactate (Lac). I) Western blot analysis of OXSM expression in THLE2
   cells transfected with Flag‐CTL, Flag‐PCK2, or Flag‐PCK2 K100R and
   treated with CHX for the indicated times. J) Western blot analysis of
   OXSM expression in THLE2 cells transfected with siCTL or siPCK2 and
   treated with CHX for the indicated times. K) Western blot analysis of
   OXSM protein expression in the indicated Flag‐PCK2‐overexpressing THLE2
   cells transfected with siCTL or siPCK2 and treated with/without MG132
   or lactacystin as indicated. L, M) THLE2 cells were transfected as
   indicated, and the cell lysates were then immunoprecipitated with an
   anti‐OXSM antibody, and detection was performed with an anti‐His
   antibody. N) THLE2 cell lysates were incubated with an anti‐Parkin
   antibody, and interacting proteins were detected with an anti‐OXSM
   antibody via Western blotting as indicated. O) Western blot analysis of
   OXSM protein expression in THLE2 cells transfected with HA‐CTL or
   HA‐Parkin and treated with/without MG132. P) THLE2 cells were
   transfected as indicated, the cell lysates were then immunoprecipitated
   with an anti‐OXSM antibody, and detection was performed with an
   anti‐His antibody. Q) Western blot analysis of OXSM expression in THLE2
   cells transfected with HA‐CTL or HA‐Parkin and treated with CHX for the
   indicated times. R) The interaction between Parkin and OXSM in THLE2
   cells transfected with different concentrations of Flag‐PCK2 was
   evaluated via CoIP assays followed by Western blot analysis. S and U)
   THLE2 cells were transfected as indicated, and the cell lysates were
   then immunoprecipitated with an anti‐MYC antibody, and detection was
   performed with an anti‐His antibody. T and V) Western blot analysis of
   OXSM protein expression in THLE2 cells transfected with the indicated
   vectors. For all the above experiments, three independent experiments
   (n = 3) with similar results were performed in triplicate.

   Since PCK2 can affect OXSM expression (Figure [224]5H; Figure [225]S5B,
   Supporting Information), we investigated how PCK2 modulates OXSM
   protein levels. Strikingly, PCK2 overexpression barely affected OXSM
   mRNA levels in vitro and in vivo (Figure [226]S6B,C, Supporting
   Information). However, overexpression of PCK2 but not PCK2 K100R
   increased OXSM protein levels in a dose‐dependent manner
   (Figure [227]6F). In contrast, the silencing of PCK2 markedly decreased
   the protein levels (Figure [228]6G, top) but not the mRNA levels of
   OXSM (Figure [229]S6D, Supporting Information). Additionally, the
   protein level (Figure [230]6H), but not the mRNA level (Figure
   [231]S6E, Supporting Information), of OXSM was markedly increased in
   lactate‐pretreated hepatocytes, and silencing of PCK2 reversed the
   lactate‐induced increase in OXSM protein levels (Figure [232]6G,
   bottom). Moreover, overexpression of PCK2 but not PCK2 K100R increased
   the half‐life of the OXSM protein (Figure [233]6I), whereas silencing
   of PCK2 significantly facilitated the degradation of the OXSM protein
   (Figure [234]6J) and reversed the impact of lactate on OXSM stability
   (Figure [235]S6F, Supporting Information).

   We then investigated the potential mechanisms by which PCK2 inhibits
   the degradation of the OXSM protein. Interestingly, the addition of the
   proteasome inhibitor MG132 effectively abrogated the effects of
   silencing PCK2 on OXSM protein degradation (Figure [236]6K, top), but
   the addition of the lysosome inhibitor chloroquine did not prevent the
   degradation of the OXSM protein (Figure [237]S6G, Supporting
   Information). Consistently, the addition of the proteasome inhibitor
   lactacystin also blocked OXSM protein degradation (Figure [238]6K,
   bottom), indicating that PCK2 silencing promoted OXSM degradation via
   the proteasome pathway.

   We next sought to determine whether PCK2 inhibits OXSM degradation by
   regulating the ubiquitination of OXSM. As anticipated, ectopic
   expression of PCK2 (Figure [239]6L), not PCK2 K100R (Figure [240]S6H,
   Supporting Information), substantially decreased the level of
   polyubiquitylated OXSM. Conversely, PCK2 knockdown markedly increased
   OXSM polyubiquitination (Figure [241]6M). Furthermore, the
   polyubiquitination of the OXSM protein was markedly inhibited in
   lactate‐pretreated hepatocytes (Figure [242]S6I, Supporting
   Information), and silencing of PCK2 robustly reversed the inhibitory
   effect of lactate on OXSM polyubiquitination in hepatocytes (Figure
   [243]S6J, Supporting Information). These data illustrate that PCK2
   regulates OXSM protein abundance through the ubiquitin‐proteasome
   pathway.

   We then aimed to identify the E3 ligase that mediates the
   ubiquitination‐mediated degradation of the OXSM protein. We studied the
   effects of several potential E3 ligases that were previously reported
   to affect mitochondrial metabolism^[ [244]^22 ^] and found that Parkin
   induced drastic OXSM degradation (Figure [245]S6K, Supporting
   Information). We then investigated whether Parkin induces the
   degradation of OXSM and whether PCK2 affects this process. As expected,
   we found that Parkin interacted with OXSM (Figure [246]6N) to decrease
   OXSM protein levels through proteasome‐mediated degradation
   (Figure [247]6O), which caused OXSM polyubiquitination (Figure [248]6P)
   and thus shorten the half‐life of OXSM (Figure [249]6Q) in THLE2 cells.
   These findings imply that Parkin is a critical E3 ligase of OXSM.
   Importantly, PCK2 overexpression decreased the binding of Parkin to
   OXSM in a dose‐dependent manner (Figure [250]6R). Moreover, PCK2
   overexpression largely abolished the Parkin‐induced polyubiquitination
   of the OXSM protein (Figure [251]6S) and thereby increased the OXSM
   protein level (Figure [252]6T). Furthermore, when Parkin expression was
   presilenced in lactate‐cultured THLE2 cells, silencing of PCK2 failed
   to promote OXSM polyubiquitination (Figure [253]6U) and reduce OXSM
   levels (Figure [254]6V). These data suggest that PCK2 competitively
   binds to OXSM and thereby abolishes the interaction of OXSM with the E3
   ligase Parkin, which leads to the disruption of the Parkin‐dependent
   degradation of the OXSM protein.

2.8. PCK2‐Mediated Ferroptosis is Dependent on OXSM During Hepatic IRI

   To determine whether the ferroptosis‐promoting effects of PCK2 are
   dependent on OXSM during IRI, we inhibited OXSM in PCK2‐overexpressing
   THLE2 cells and analyzed whether lactate could induce mtFAS. A 90%
   reduction in OXSM expression induced a significant decrease in OGDH and
   NDUFA6 expression (Figure [255] 7A) as well as the expression of
   critical genes involved in OXPHOS (Figure [256]7B) and the TCA cycle
   (Figure [257]7C) in PCK2‐overexpressing THLE2 cells. Moreover,
   real‐time OCR and glycolytic rate measurements indicated that silencing
   OXSM in PCK2‐overexpressing THLE2 cells repressed the lactate‐induced
   increase in basal oxygen consumption, mitochondrial ATP respiration,
   and maximal oxygen consumption (Figure [258]7D) as well as basal
   glycolysis and glycolytic capability (Figure [259]7E), confirming OXSM
   dependence. Moreover, TUNEL staining (Figure [260]7F), calcein‐AM/PI
   staining (Figure [261]7G), iron content analysis (Figure [262]7H),
   lipid peroxidation analysis (Figure [263]7I), Western blotting
   (Figure [264]7J), and TEM (Figure [265]7K) further revealed that
   silencing OXSM markedly inhibited hyperlactatemia‐mediated hepatic IRI
   and ferroptosis in PCK2‐overexpressing THLE2 cells.

Figure 7.

   Figure 7
   [266]Open in a new tab

   PCK2‐mediated OXSM‐dependent ferroptosis during hepatic IRI. A) Western
   blot analysis of PCK2, OXSM, OGDH, and NDUFA6 expression in THLE2 cells
   cotransfected with Flag‐CTL or Flag‐PCK2 and siCTL or siOXSM as
   indicated. (B‐K) PCK2‐overexpressing THLE2 cells cultured under
   high‐lactate conditions were transfected with siCTL or siOXSM and
   subjected to H/R. B, C) Real‐time qPCR analysis of the top
   significantly downregulated (B) OXPHOS‐related genes and (C) TCA
   cycle‐related genes. D, E) Real‐time Seahorse OCR and glycolytic rate
   measurements to evaluate the OXPHOS and TCA cycle rate. F)
   Representative images and relative quantification of TUNEL staining
   (scale bar = 200 µm). G) Representative images and relative
   quantification of calcein‐AM/PI double‐staining (scale bar = 100 µm).
   H) Iron content. I) ROS, MDA, GSH and SOD levels. J) Western blot
   analysis of ferroptosis‐associated protein expression. K)
   Representative TEM images (scale bar = 1 µm). (L‐R) AAV‐PCK2‐treated
   mice (n = 7/group) infected with AAV‐shCTL or AAV‐shOXSM were treated
   with lactate and subjected to hepatic IRI. L) Western blot analysis of
   OXSM, OGDH, and NDUFA6 expression in the liver. M) Serum AST and ALT
   levels. N) Representative images of HE‐stained livers (scale bar =
   200 µm). O) Representative images of Prussian blue‐stained livers
   (scale bar = 200 µm). P) Iron content in the liver. Q) ROS, MDA, GSH,
   and SOD levels in the liver. R) Western blot analysis of
   ferroptosis‐associated protein expression in the liver. For all the
   above experiments, the data are presented as the means ± SDs. Three
   independent experiments (n = 3) with similar results were performed in
   triplicate. In B‐I, M, P, and Q, the statistical analyses were
   performed via two‐tailed unpaired Student's t‐tests. p < 0.05 was
   considered to indicate statistical significance.

   To investigate the role of OXSM in PCK2‐mediated ferroptosis in vivo,
   we silenced OXSM in the livers of AAV‐PCK2 mice via injection of an
   AAV9 vector harboring short hairpin RNAs (shRNA) sequences targeting
   the mouse OXSM gene (hereafter, AAV‐PCK2‐shOXSM mice) and observed an
   80% decrease in OXSM protein levels after 4 weeks (Figure [267]7L).
   Scramble shRNA was used as a negative control (AAV‐PCK2‐shCTL mice). As
   expected, despite lactate treatment, OXSM silencing led to reductions
   in the protein levels of OGDH and NDUFA6 (Figure [268]7L) as well as
   the expression of critical genes involved in OXPHOS (Figure [269]S7A,
   Supporting Information) and the TCA cycle (Figure [270]S7B, Supporting
   Information). Moreover, OXSM silencing ameliorated hepatic IRI
   (Figure [271]7M,N), and inhibited hepatic ferroptosis
   (Figure [272]7O,R) in AAV‐PCK2‐shOXSM mice. Overall, these data
   indicate that the ferroptosis‐promoting effects of PCK2 occur through
   OXSM during hepatic IRI.

2.9. Targeting PCK2 is a Promising Therapeutic Strategy

   Currently, numerous strategies have been proposed to attenuate hepatic
   ferroptosis and IRI,^[ [273]^23 ^] but no ferroptosis inhibitors are
   currently available for clinical application. Given our findings, we
   investigated whether PCK2 could serve as a potential therapeutic target
   for the hyperlactatemia‐induced exacerbation of hepatic ferroptosis and
   IRI. To this end, we performed gene editing or pharmacological
   interventions in vivo. We first injected an AAV9 vector expressing an
   shRNA against PCK2 into mice (hereafter, AAV‐shPCK2 mice) and
   successfully knocked down PCK2 in the liver (Figure [274]S7A–C,
   Supporting Information), then, we treated the AAV‐shPCK2 mice with
   lactate prior to hepatic IRI. Indeed, PCK2 knockdown substantially
   ameliorated hyperlactatemia‐mediated hepatic IRI, as shown by
   variations in serum transaminase levels (Figure [275] 8A), HE staining
   (Figure [276]8B), and TUNEL staining (Figure [277]8C). Moreover,
   Prussian blue staining (Figure [278]8D), iron content analysis
   (Figure [279]8E), lipid peroxidation analysis (Figure [280]8F,G),
   Western blotting (Figure [281]8H) and TEM (Figure [282]8I) further
   demonstrated that silencing PCK2 markedly inhibited
   hyperlactatemia‐mediated hepatic ferroptosis during IRI.

Figure 8.

   Figure 8
   [283]Open in a new tab

   PCK2 is a promising therapeutic target for attenuating
   hyperlactatemia‐mediated hepatic ferroptosis. A–H) Mice infected with
   AAV‐shCTL or AAV‐shPCK2 (n = 5/group) were treated with lactate and
   subjected to hepatic IRI. (A) Serum AST and ALT levels at the indicated
   times. (B) Representative images of HE‐stained livers at the indicated
   times (scale bar = 100 µm). (C) Representative images of TUNEL‐stained
   livers at the indicated times and relative quantification of the data
   (scale bar = 100 µm). (D) Representative images of Prussian
   blue‐stained livers (scale bar = 200 µm). (E) Iron content in the
   liver. (F‐G) ROS, MDA, GSH and SOD levels in the liver. (H) Western
   blot analysis of ferroptosis‐associated protein expression in the
   liver. I) Representative TEM images of livers (scale bar = 2 µm). J–P)
   Mice (n = 7/group) were treated with the indicated KAT8 inhibitors and
   lactate and subjected to hepatic IRI. (J) Serum AST and ALT levels. (K)
   Top, representative images of HE‐stained livers. Bottom, representative
   images of TUNEL‐stained livers and relative quantification of the data
   (scale bar = 100 µm). (L) Representative images of Prussian
   blue‐stained livers (scale bar = 200 µm). (M) Iron content in the
   liver. (N‐O) ROS, MDA, GSH and SOD levels in the liver. (P) Western
   blot analysis of ferroptosis‐associated protein expression in the
   liver. Q) Schematic showing the regulation of hepatic ferroptosis
   during IRI by lactate, which enhances KAT8‐mediated PCK2 lactylation at
   K100. This effect diminishes the Parkin‐dependent degradation of OXSM,
   leading to mtFAS metabolic reprogramming and the potentiation of OXPHOS
   and TCA cycle activity, thereby exacerbating ferroptosis and IRI in the
   liver. For all the above experiments, the data are presented as the
   means ± SDs. In A, C, E‐G, J‐K, and M‐O, the statistical analyses were
   performed via two‐tailed unpaired Student's t‐tests. p < 0.05 was
   considered to indicate statistical significance.

   In addition to this genetic approach, we used two commercially
   available, selective KAT8 inhibitors, KAT8‐IN‐1 and MC4033, to
   pharmacologically and systemically target KAT8. Consistent with the
   gene knockdown results, measurement of serum transaminase levels
   (Figure [284]8J), HE staining (Figure [285]8K, top), TUNEL staining
   (Figure [286]8K, bottom), Prussian blue staining (Figure [287]8L), iron
   content analysis (Figure [288]8M), lipid peroxidation analysis
   (Figure [289]8N,O) and Western blotting (Figure [290]8P) confirmed the
   ability of KAT8 inhibition to suppress hyperlactatemia‐mediated hepatic
   IRI and ferroptosis. Importantly, Western blotting revealed that
   KAT8‐IN‐1 and MC4033 significantly decreased the lactylation of PCK2 at
   K100 in the liver during IRI (Figure [291]8P). In conclusion, these
   results serve as a basis for the clinical translation of therapeutics
   targeting the KAT8‒PCK2 axis to alleviate IRI in patients with
   hyperlactatemia undergoing LT.

3. Discussion

   Owing to disrupted hepatic metabolism, hyperlactatemia is a prominent
   feature of the perioperative period in LT recipients. Here, we explored
   the direct links among hyperlactatemia, ferroptosis, and hepatic IRI.
   As shown in Figure [292]8Q, hyperlactatemia exacerbates hepatic
   ferroptosis during IRI by promoting PCK2 activity. Mechanistically, in
   the context of hyperlactatemia, KAT8 functions as a lysine lactylation
   transferase and mediates PCK2 lactylation at K100, thereby activating
   the kinase activity of PCK2 in mitochondrial metabolism. PCK2
   subsequently acts as a critical inducer of ferroptosis during IRI by
   competitively inhibiting the Parkin‐mediated ubiquitination of OXSM,
   potentiating the production of mtFAS products, and maintaining OXPHOS
   and TCA cycle activity. Importantly, IRI is a pivotal risk factor for
   primary graft dysfunction or nonfunction as well as early mortality in
   patients undergoing LT.^[ [293]^3 ^] Our data demonstrate that
   targeting PCK2 can effectively reduce the expression of OXSM,
   ultimately inhibiting ferroptosis, thus representing an avenue for
   effectively alleviating hyperlactatemia‐mediated hepatic IRI in the
   clinic.

   Although various forms of cell death, including apoptosis, necrosis,
   and pyroptosis, are involved in the development of hepatic IRI, and
   while numerous targets for these forms of cell death have been
   proposed, the clinical efficacy of available treatments is currently
   limited, indicating that new mechanisms of cellular injury remain to be
   investigated. Ferroptosis is rapidly gaining attention in the field of
   liver diseases, as excessive iron accumulation is a primary
   characteristic of most liver diseases, and generally, the liver is
   vulnerable to oxidative damage. Recent studies have revealed that
   hyperlactatemia leads to impaired hepatic iron metabolism and systemic
   iron homeostasis imbalance through increased hepatic hepcidin
   expression.^[ [294]^24 ^] Huang reported that lactate‐mediated
   upregulation of NADPH‐dependent NOX4 expression contributes to ROS
   generation and tissue damage.^[ [295]^25 ^] However, the association
   between hyperlactatemia and hepatic ferroptosis has never been
   investigated. Considering the physiological features of the liver and
   the factors that induce ferroptosis, we hypothesized that the liver is
   more susceptible to ferroptosis during IRI when stimulated with
   lactate. In our study, we comprehensively evaluated the relationships
   between lactate levels and hepatic IRI, iron metabolism, and oxidative
   damage. Moreover, treatment with Fer‐1 specifically inhibited
   ferroptosis and effectively reduced hepatic IRI, demonstrating that the
   liver is indeed more prone to IRI in a high‐lactate microenvironment.
   Thus, it is imperative to thoroughly explore the potential mechanisms
   that regulate ferroptosis to develop effective therapies for combating
   hyperlactatemia‐mediated exacerbation of IRI.

   The current understanding of the function of lactate extends from its
   origins as a byproduct of glycolysis to its essential role in cellular
   metabolism as a factor that coordinates signaling among different
   cells, organs, and tissues.^[ [296]^26 ^] Lysine lactylation is a newly
   discovered posttranslational modification that links gene regulation to
   cellular metabolism through metabolic dysregulation and epigenetic
   modifications, affecting the microenvironment and disease
   development.^[ [297]^27 ^] Although the role of lactate in hepatic IRI
   has not been reported, the involvement of lactylation in tissue injury
   has been extensively reported. Astrocytic LRP1 enables mitochondrial
   transfer to neurons and mitigates ischemic stroke by suppressing ARF1
   lactylation in the brain.^[ [298]^28 ^] Heat shock protein A12A
   (HSPA12A) maintains aerobic glycolytic homeostasis and histone 3
   lactylation in cardiomyocytes to attenuate myocardial IRI.^[ [299]^29
   ^] Similarly, Du et al. recently revealed that HSPA12A protects the
   liver from IRI by suppressing glycolysis‐mediated high mobility group
   box 1 (HMGB1) lactylation and secretion from hepatocytes to inhibit
   macrophage chemotaxis and inflammatory activation.^[ [300]^30 ^]
   Herein, considering that lactate accumulates in the liver, we first
   explored the role of lactylation in the regulation of ferroptosis and
   hepatic IRI. Through proteomics followed by functional enrichment
   analysis and protein‒protein interaction network analysis, we screened
   the mitochondrial protein PCK2. Moreover, our data demonstrated that
   lactate promoted KAT8‐mediated lactylation of PCK2 and increased PCK2
   kinase activity, which exacerbated hepatic ferroptosis and IRI. It is
   believed that PCK2 participates in the pyruvate cycle in
   glucose‐starved cells and provides carbon for gluconeogenesis and
   ultimately several downstream anabolic pathways.^[ [301]^21 ^] However,
   through metabolomics and transcriptomics analyses, we observed that, in
   the livers of AAV‐PCK2 mice subjected to IRI, the pathways altered most
   substantially were OXPHOS and the TCA cycle, suggesting that PCK2 might
   regulate hepatic IRI injury through a noncanonical pathway.

   The mtFAS pathway, an evolutionarily conserved mitochondrial
   biosynthetic pathway, generates lipoic acid, which is a required
   cofactor for several mitochondrial metabolic enzymes involved in OXPHOS
   and the TCA cycle.^[ [302]^13 , [303]^14 ^] On the basis of the
   transcriptomic analysis results, we speculated that PCK2 acts on the
   mtFAS pathway to alter OXPHOS and the TCA cycle, thereby exacerbating
   ferroptosis during hepatic IRI. Using in vivo and in vitro IRI models,
   we confirmed that PCK2 alters the expression of OGDH and NDUFA6, which
   are pivotal metabolic proteins, the OCRs, and the glycolytic rate.
   Moreover, we confirmed that PCK2 plays a core role in the
   hyperlactatemia‐induced increase in mtFAS pathway activity and the
   subsequent increases in OXPHOS and TCA cycle activity, thereby
   exacerbating hepatic ferroptosis during IRI. Unlike cytosolic fatty
   acid synthase (FASN), mtFAS is a type II FAS pathway that takes place
   in the mitochondrial matrix. mtFAS and FASN generate de novo fatty
   acids via the same chemical process, with the major difference being
   that in cytosolic FAS, each enzyme is a domain of a single FASN
   polypeptide rather than a discrete enzyme.^[ [304]^13 , [305]^14 ^] The
   mitochondrial ketoacyl synthase OXSM is responsible for catalyzing the
   condensation reaction in the first step of mtFAS and generating
   ketoacyl intermediates for a series of reduction and dehydration
   reactions by subsequent enzymes in the mtFAS cycle.^[ [306]^13 ^] Our
   data suggest that PCK2 restricts Parkin‐linked ubiquitination‐mediated
   degradation of OXSM and that OXSM silencing markedly impairs the
   activation of PCK2 during mtFAS and downstream activation of the TCA
   cycle and OXPHOS. Furthermore, our data strongly indicate that mtFAS
   has the potential as a therapeutic target for hepatic IRI, but no
   inhibitors are available for clinical application. Here, we attempted
   to target PCK2 and found that, after inhibiting PCK2 expression or
   activity, OXSM expression was effectively suppressed, resulting in the
   inhibition of ferroptosis and markedly alleviating
   hyperlactatemia‐mediated hepatic IRI.

   There are several limitations to this study that should be addressed in
   future research. PCK2 is a mitochondrial gene involved in glycogen
   synthesis and is specifically expressed at high levels in
   hepatocytes,^[ [307]^31 ^] we thus focused on the role of PCK2 mainly
   in hepatocytes. It has been extensively reported that multiple factors
   in the immune microenvironment, such as macrophages, induce
   ferroptosis,^[ [308]^32 ^] but whether lactate exacerbates hepatic
   ferroptosis and IRI via immune cells requires further investigation. In
   summary, our findings not only reveal the mechanism by which
   hyperlactatemia mediates hepatic IRI but also identify PCK2 as a
   therapeutic target. Further clinical trials are necessary to
   investigate the beneficial effects of the rational utilization of PCK2
   inhibitors on hepatic IRI injury in patients with altered serum lactate
   levels.

4. Experimental Section

Human Samples

   A retrospective analysis of all adults undergoing LT at West China
   Hospital of Sichuan University between May 2017 and August 2024 was
   performed. The patients were divided into two cohorts. Cohort 1
   included 30 patients between May and August 2024, of whom 18
   experienced hyperlactatemia during LT surgery and 12 did not. Lactate
   levels in venous blood were measured after donor liver implantation but
   before portal vein opening during LT surgery. The correlation between
   the serum ALT and AST levels of these 30 patients from POD1 to POD7 and
   lactate levels were assessed. Cohort 2 included 247 patients between
   May 2017 and April 2024. The clinical data of 100 patients from POD1 to
   POD3 were used to analyze the relationship between PCK2 expression
   levels and hepatic IRI through 1:1 PSM. Liver tissues were harvested
   before reperfusion (ischemia group) and after 1.5 h of reperfusion
   (reperfusion group). A portion of each liver sample was fixed in 4%
   formaldehyde and then embedded in paraffin, and the other portion was
   frozen directly in liquid nitrogen. All clinical data were obtained
   from records from the Chinese Liver Transplant Registry (CLTR,
   [309]http://www.cltr.org/). Organs were donated voluntarily, and
   informed consent forms were signed by all involved donors or their
   families and by recipients. All samples were used only for scientific
   research, and the studies were performed according to the principles
   outlined in the Declaration of Helsinki. This study was approved by the
   Ethics Committee on Biomedical Research, West China Hospital of Sichuan
   University (2020, No 385).

Experimental Animals

   Eight‐week‐old male C57BL/6 mice were obtained from Byrness Weil
   Biotech Ltd. (Chengdu, China). All the mice were housed in a specific
   pathogen‐free (SPF) environment on a 12‐h light/dark cycle under
   controlled light, temperature, and humidity conditions. Food and water
   were available ad libitum. All operations on experimental animals were
   performed in accordance with the National Institutes of Health's Guide
   for the Care and Use of Laboratory Animals. All operations were
   approved by the Animal Care and Use Committee of West China Hospital of
   Sichuan University (2020338A).

Animal Adeno‐Associated Virus (AAV) Production and Infection

   To overexpress PCK2 in the liver, a hepatocyte‐specific AAV vector,
   AAV8‐TBG‐PCK2 (AAV‐PCK2) was constructed by GeneChem Co., Ltd.
   (Shanghai, China). An empty AAV8 vector (AAV‐V) was used as a negative
   control. Additionally, to knock down PCK2 or OXSM in the liver,
   AAV9‐TBG vectors harboring shRNAs against PCK2 or OXSM (AAV‐shPCK2 or
   shOXSM, designed and synthesized by GeneChem) were injected into the
   mice. Scramble shRNA (AAV‐sh) was used as a negative control. The TBG
   promoter was used to achieve hepatocyte‐specific overexpression or
   knockdown. For AAV transduction, viruses (1.5 × 10^11 viral
   particles/mouse) were delivered via tail vein injection. After viral
   infection for 4 weeks, the mice were used for subsequent experiments.
   The AAV9‐TBG vectors harboring shRNAs were provided by Hanbio
   Biotechnology Co., Ltd. (Shanghai, China).

Mouse Hepatic IRI Model

   A mouse model of hepatic IRI was established as previously reported.^[
   [310]^33 ^] In brief, after the mice were anesthetized with isoflurane,
   an upper abdominal midline laparotomy was performed to expose the
   liver. The blood supply to the left and median lobes was then occluded
   via an atraumatic clamp to induce ischemia for 90 min. Whitening of the
   ischemic liver lobe indicated successful ischemia induction. After the
   clamp was removed unless the indicated time points, liver samples, and
   serum were collected after 6 h of reperfusion for subsequent
   examination after reperfusion. Mice subjected to the same operation
   without vascular occlusion served as sham controls.

   To investigate the roles of hyperlactatemia, ferroptosis and KAT8
   inhibition in hepatic IRI, C57BL/6 mice was treated with lactate
   (120 mg k^−1g, [311]M10168, Abmole, Shanghai, China) in normal saline
   and Fer‐1 (20 mg k^−1g, S7243, Selleck, Houston, TX), KAT8‐IN‐1
   (10 mg k^−1g, HY‐W015239, MCE, NJ, USA), or MC4033 (10 mg k^−1g,
   HY‐149302, MCE) dissolved in 2% dimethyl sulfoxide (DMSO, D5879‐500ML,
   Sigma‒Aldrich). In accordance with the manufacturers’ instructions,
   lactate was intraperitoneally administered 1 h prior to ischemia
   induction, and Fer‐1, KAT8‐IN‐1, and MC4033 were intraperitoneally
   administered 2 h prior to ischemia induction. Mice in the control group
   were treated with the same volume of normal saline or DMSO.

Mouse Liver Function Analysis

   Whole blood samples were collected and centrifuged for 8 min at
   4000 rpm to obtain serum. The serum levels of ALT and AST in the mice
   were measured via a biochemical analyzer (Sysmex) at the Clinical Test
   Analysis Department of Haiqi Pharmaceutical Technology Co., West China
   Hospital (Chengdu, China).

Primary Hepatocyte Isolation and Culture

   MPHs were isolated from the indicated male C57BL/6 mice via the type IV
   collagenase perfusion method as previously reported.^[ [312]^33 ^] In
   brief, the liver was digested by sequential perfusion of 50 mL of warm
   (37 °C) liver perfusion medium (PYG0081, Boster, Wuhan, China) and a
   solution containing 0.5% type IV collagenase ([313]LS004188,
   Worthington, NJ, USA) via the portal vein. The liver was then excised,
   minced, and soaked in 0.5% type IV collagenase for further digestion;
   the digested material was filtered through a 70 µm cell strainer and
   washed with sterile PBS. MPHs in DMEM (Gibco, 11 885 084, NY, USA)
   supplemented with 10% fetal bovine serum (FBS, Gibco, 16 140 063, NY,
   USA) were seeded in plates coated with type I rat tail collagen (Merck,
   C3867, Shanghai, China) and cultured in an incubator containing 5%
   CO[2]at 37 °C.

Cell Culture

   The THLE2 cell line was purchased from the Meisen Chinese Tissue
   Culture Collection (MeisenCTCC, Zhejiang, China). The HEK293T cell line
   was purchased from the Cell Bank of the Chinese Academy of Sciences
   (Shanghai, China). THLE2 cells were cultured in BEGM (MeisenCTCC), and
   HEK293T cells were cultured in high‐glucose DMEM (Gibco). The culture
   media were supplemented with 10% FBS and 1% penicillin‐streptomycin.
   The cells were cultured in a humidified incubator at 5% CO[2] and
   37 °C. All the cell lines were validated via STR DNA profiling and
   tested negative for mycoplasma via PCR.

In Vitro H/R Model

   To simulate hepatic IRI in vitro, H/R injury to MPH and THLE2 cells was
   induced. In brief, the cell medium was replaced with serum‐free medium,
   and the cells were cultured under hypoxic conditions (1% O[2], 5%
   CO[2], and 94% N[2]) for 4 h; then, the medium was replaced with
   complete medium, and the cells were transferred to a normoxic incubator
   for 2 h to allow reoxygenation. To investigate the role of lactate in
   H/R injury, MPHs, and THLE2 cells were treated with lactate (20 mM) for
   0.5 h prior to hypoxia. Control cells were treated with normal saline.

RNA Interference

   Lipofectamine 3000 reagents (Invitrogen, Carlsbad, MA, USA) were used
   for siRNA transfection as described previously.^[ [314]^18 , [315]^34
   ^] siRNAs (generated by Tsingke, Tianjin, China) were used at a final
   concentration of 50 nM. After transfection, the cells were cultured for
   an additional 48 h and then used in subsequent assays. Information
   about the siRNAs used in this study is provided in Table [316]S5
   (Supporting Information).

Plasmid Construction and Transfection

   PCR‐amplified human wild‐type PCK2, KAT8, and OXSM transcripts were
   cloned and inserted into the GV141‐Flag, GV141‐HA, and GV141‐Myc
   vectors, respectively. PCR‐amplified human wild‐type Parkin, MUL1,
   MARCH5, and RNF185 transcripts were cloned and inserted into the
   GV141‐HA vector. The ubiquitin (Ub) sequence was subsequently cloned
   and inserted into the GV141‐His vector. The site‐specific mutants used
   in this study were designed and synthesized by GeneChem (Shanghai,
   China). Additionally, to knock down PCK2 in THLE2 cells, shRNAs against
   PCK2 were designed and synthesized by GeneChem. Scramble shRNA was used
   as a negative control. All reagents used for the experiments described
   in this section were purchased from GeneChem. Information about the
   shRNAs used in this study is provided in Table [317]S6 (Supporting
   Information). Transfection was performed with Lipofectamine 3000
   reagent as described previously.^[ [318]^18 , [319]^34 ^]

CoIP

   CoIP was performed via a Co‐IP kit (Abs955, Absin, Shanghai, China) as
   described previously.^[ [320]^18 , [321]^34 ^] In brief, hepatocytes
   were homogenized in IP lysis buffer (20 mM Tris–HCl (pH 7.5), 0.5%
   NP‐40, 250 mM NaCl, 3 mM EDTA, 3 mM EGTA, 1 mM DTT, 1 mM cocktail, 1 mM
   phosphoSTOP, 1 mM NEM, and 1 mM NAM). A total of 500 µg of the extract
   was incubated first with the indicated primary antibody or IgG as a
   negative control for 4 h and then with protein A/G‐Sepharose beads for
   2 h at 4 °C. After extensive washes with PBS, the immunoprecipitates
   were used in subsequent assays. All the antibodies used were purchased
   from commercial manufacturers, as listed in Table [322]S7 (Supporting
   Information).

GST Pull‐Down Assay

   GST‐fused PCK2 was expressed in Escherichia coli Rosetta cells and
   purified. Synthetic proteins were purified to >98% purity via
   high‐pressure liquid chromatography. HEK293T cell extracts expressing
   the indicated proteins were mixed with 5 mg of GST derivatives bound to
   glutathione‐Sepharose beads in 0.5 mL of modified binding buffer (25 mM
   Tris‐HCl at pH 7.2, 150 mM NaCl, 1 mM DTT, 0.5 mM EDTA, 0.5 mM EGTA,
   1 mM cocktail, 1 mM phosphoSTOP, 1 mM NEM, and 1 mM NAM). The binding
   reaction was performed at 4 °C overnight, and the beads were
   subsequently washed 4 times with binding buffer and subjected to
   Western blot analysis.

LC‐MS/MS Analysis

   Immunoprecipitates of MPHs and THLE2 cells were resolved by SDS‒PAGE;
   the gel bands were digested with 10 ng µL^−1 trypsin; and the proteins
   were eluted with 0.1% formic acid and 75% acetonitrile. The eluted
   proteins were then subjected to quality control and qualitatively
   analyzed via a Q ExactiveTM HF‐X mass spectrometer by Novogene Co.,
   Ltd. (Tianjin, China) to obtain raw proteome data. The raw proteome
   data files were directly imported into Proteome Discoverer 2.2 software
   for database searching, peptide spectrum matching and protein
   quantification. Target protein‐protein interaction networks were
   constructed via the STRING database ([323]https://string‐db.org/).

Metabolomic Analysis

   Metabolomic analyses were conducted by Bioprofile Technology Co., Ltd
   (Shanghai, China). In brief, after being frozen in liquid nitrogen,
   liver tissues were pulverized in 250 µL of mixed solvent
   (chloroform:methanol: water, 1:2:1). The lysate was sonicated and
   centrifuged for 10 min at 12000 rpm. The aqueous supernatant was
   transferred to a gas chromatography vial containing internal standards,
   and the pellet was resuspended in 250 µL of methanol and homogenized
   with a T10 basic homogenizer at 4 °C for 30 s. Aliquots of the
   supernatant and the mixture in the vial were vacuum‐dried after a
   second centrifugation. The samples were analyzed on an LC–MS/MS
   instrument (Thermo Fisher Ult3000‐Exploris 480 Orbitrap).

RNA Sequencing

   Total RNA was extracted from mouse livers via a TRIzol reagent kit
   (Takara, Dalian, China). Transcriptome sequencing and subsequent data
   analysis were conducted by Novogene Co., Ltd. (Tianjin, China).
   Bioinformatics analyses were performed as previously described.^[
   [324]^18 , [325]^34 ^] In brief, the R “limma” package was used to
   identify differentially expressed genes between the indicated groups on
   the basis of the following criteria: false discovery rate (FDR) <0.05
   and |log2(fold change (FC))|>1. The Database for Annotation,
   Visualization, and Integrated Discovery (DAVID) tool was used to
   analyze the enriched GO and KEGG pathways. GSEA was performed to
   analyze the GO categories in which the signature gene sets from the
   gene profiles between the indicated groups were enriched.

TEM

   Fresh liver tissues of appropriate size were obtained and immediately
   fixed in 2.5% phosphate‐glutaraldehyde for 4 h. After 2 washes in
   dimethylarsenic acid sodium buffer, the samples were dehydrated in an
   ethanol gradient and then fixed, embedded, and sectioned (70 nm). The
   sections were then viewed with a Tecnai 10 (100 kV) transmission
   electron microscope. For each sample, five fields of view were randomly
   selected, and twenty mitochondria were examined in each field of view.

Histology and IHC

   Histology and IHC were performed as previously described.^[ [326]^18 ,
   [327]^34 ^] In brief, paraffin‐embedded liver tissues were serially
   sectioned at a thickness of 5 µm. The sections were then stained with
   HE for routine histological examination via light microscopy. To
   measure the iron content, selected sections were stained with Prussian
   blue. In addition, serial tissue sections were deparaffinized,
   subjected to antigen retrieval using 0.01 m citric acid buffer (pH 6.0)
   at 95 °C for 15 min, and incubated overnight at 4 °C with primary
   antibodies prior to incubation with HRP‐conjugated AffiniPure goat
   anti‐rabbit IgG (Proteintech, SA00001‐2, Wuhan, China; 1:5000) for 1 h
   at room temperature. Information about the primary antibodies used in
   this study is available in Table [328]S7 (Supporting Information).

TUNEL Assay

   Liver tissue sections and formalin‐fixed paraffin‐embedded cells were
   deparaffinized, and TUNEL (BOSTER, MK1028, Wuhan, China) staining was
   performed according to the manufacturer's instructions.

Calcein‐AM/PI Staining

   To analyze the death of hepatocytes directly, calcein‐AM/PI staining
   was performed via a Calcein/PI Cell Viability/Cytotoxicity Assay Kit
   (Beyotime, C2015S, Wuhan, China) in accordance with the manufacturer's
   instructions. Five fields of view for each sample were randomly
   photographed (Zeiss, Oberkochen, Germany), and the average value for
   the five fields of view is reported for each sample. The cell death
   ratio was calculated as the number of PI‐positive cells/the number of
   total cells.

Lipid Peroxidation Analysis

   The levels of ROS, MDA, GSH, and SOD were measured via commercially
   available kits from Nanjing Jiancheng Bioengineering Institute
   (Nanjing, China) in accordance with the manufacturer's instructions. In
   brief, liver tissues or cells were rinsed with PBS, homogenized in RIPA
   lysis buffer on ice, and sonicated. After sonication, the samples were
   centrifuged at 4000 rpm for 10 min to remove debris. The protein
   concentration of each sample was determined with a BCA protein assay
   kit (P0009, Beyotime). Using a microplate reader (Thermo, Multiskan
   SkyHigh, A51119500C), the ROS level (E004‐1‐1) was measured on the
   basis of the content of 2,7‐dichlorofuorescin diacetate detected at
   488 nm; the MDA level (A003‐1‐2) was measured on the basis of the
   content of thiobarbituric acid‐reactive substances detected at 532 nm;
   the GSH level (A006‐2‐1) was measured on the basis of the content of
   2‐nitrobenzoic acid‐reactive substances detected at 412 nm; and the SOD
   level (A001‐3‐2) was measured on the basis of the content of
   WST1‐reactive substances detected at 450 nm. Finally, ROS, MDA, GSH,
   and SOD levels were normalized to the protein concentration for
   statistical analysis.

Measurement of Iron Content

   Total elemental iron was measured via a tissue iron assay kit
   (A039‐2‐1, Jiancheng) according to the manufacturer's instructions. The
   iron content was normalized to the protein concentration for
   statistical analysis.

Measurement of PEPCK Activity

   Intracellular PEPCK activity was measured via a phosphoenolpyruvate
   carboxylase test kit (A130‐1‐1, Jiancheng) according to the
   manufacturer's protocol. PEPCK activity was normalized to the protein
   concentration for statistical analysis.

Seahorse Assay

   The indicated THLE2 cells were plated in an XF96 plate ([329]W21715,
   Seahorse Biosciences, USA) at a density of ≈1 × 10^4 cells/well. Oxygen
   consumption was measured with a Seahorse XFe96 Analyzer (Agilent) to
   analyze the maximum OXPHOS respiration capacity. After basal readings
   were obtained, 1.5 mM oligomycin, 0.5 mM FCCP, and 0.5 mM
   rotenone/antimycin A were injected in sequence. All drugs were provided
   in the Cell Mito Stress Test Kit for XFe/XF Analyzers (Agilent,
   103015–100). The glycolytic rate was also measured with a Seahorse
   XFe96 Analyzer. After basal readings were obtained, 0.5 mM rotenone/AA
   and 50 mM 2‐deoxy‐D‐glucose (2‐DG) were injected in sequence. All drugs
   were provided in the Seahorse XF Glycolytic Rate Assay Kit (Agilent,
   103344‐100).

qRT‒PCR

   Total RNA was isolated from tissues and cells via TRIzol (Thermo Fisher
   Scientific, 15 596 018, WI, USA). The RNA concentration and purity were
   measured via a spectrophotometer. The RNA was reverse transcribed via
   the PrimeScript RT Reagent Kit (Takara), and qPCR was performed via
   SYBR Premix Ex Taq (Takara) according to the manufacturer's protocol.
   The expression levels were normalized to the endogenous expression
   level of β‐actin, which was used as the control. The thermocycling
   conditions were as follows: initial denaturation at 95 °C for 30 s,
   followed by 60 cycles of denaturation at 95 °C for 5 s, annealing at
   55 °C for 30 s, and extension at 72 °C for 30 s. Information about the
   primers used in this study is available in Table [330]S8 (Supporting
   Information).

Western Blot Analysis

   Protein was extracted from liver tissues and hepatocytes via RIPA
   buffer (Sigma‒Aldrich, Darmstadt, Germany) containing 1%
   phenylmethylsulfonyl fluoride (PMSF) (Beyotime, ST506, Shanghai, China)
   and phosphatase inhibitors (Beyotime, P1045). The protein concentration
   of each sample was determined with a BCA protein assay kit (Beyotime,
   P0009). Denatured protein samples (10 µg) were separated via 8%
   SDS‒PAGE (Zoman, Beijing, China) and transferred onto nitrocellulose
   (NC) membranes. The NC membranes were blocked with 5% nonfat dried milk
   in Tris‐buffered saline (pH 7.4) containing 0.1% Tween 20 (TBST) for
   1 h at 4 °C and subsequently reacted with specific primary antibodies
   diluted in TBST at 4 °C overnight. All primary antibodies were used at
   a dilution of 1:1000. The membranes were subsequently washed three
   times in TBST for 10 min each and incubated with an HRP‐conjugated
   AffiniPure goat anti‐rabbit IgG (H + L) (Proteintech, SA00001‐2, Wuhan,
   China; 1:4000) secondary antibody for 1 h at room temperature. The
   signals were detected via enhanced chemiluminescence (Meilunbio,
   Dalian, China). An anti‐β‐actin antibody was used for the normalization
   of protein expression. For cycloheximide (CHX) chase assays,
   hepatocytes were treated with 20 mg mL^−1 CHX for 24 h after
   transfection and collected at the indicated time points, and the
   resulting cell lysates were subjected to Western blotting. Information
   about the primary antibodies used in this study is available in Table
   [331]S7 (Supporting Information).

Molecular Docking Analysis

   The structure of KAT8 (AFDB accession: AF‐[332]Q9H7Z6‐F1‐v4) was
   downloaded from the AlphaFold Protein Structure Database. The docking
   of KAT8 with lactyl coenzyme A (lactyl‐CoA) or acetyl coenzyme A
   (acetyl‐CoA) was assessed via the GRAMM Docking Web Server
   ([333]https://gramm.compbio.ku.edu) and further analyzed via Discovery
   Studio software according to previously reported methods.^[ [334]^35 ^]

Statistical Analysis

   All the statistical analyses in this study were performed via SPSS 22.0
   (IBM Corp.), and the figures were produced via GraphPad Prism 6.0 or R
   software (version 3.6.1). All the experiments were independently
   performed at least three times, and n represents the number of
   independent samples or mice from each subgroup. The data are expressed
   as the means ± standard deviations (SDs). The normal distribution of
   the data was assessed by the Kolmogorov‐Smirnov test. For normally
   distributed data, the data were analyzed via unpaired Student's t‐test
   or one‐way ANOVA followed by Bonferroni's post hoc test, where
   appropriate. For non‐normally distributed data, Kruskal–Wallis tests
   were used to compare differences among more than two groups, and
   Wilcoxon tests to compare differences between two groups. For analysis
   of the clinical data, the correlation between two parameters was
   evaluated via Pearson correlation analysis. Statistical significance
   was evaluated on the basis of p values, and p < 0.05 was considered to
   indicate statistical significance.

Conflict of Interest

   The authors declare no conflict of interest.

Author Contributions

   Y. S., J. Y., and J. Y. designed the project. J. Y., M. Y., Z. W., J.
   W., K. Z., J. W., Q. Z., M. C., and T. L. performed the experiments and
   acquired the data. J. Y. and J. Y. analyzed the results and wrote the
   manuscript.

Supporting information

   Supporting Information
   [335]ADVS-12-2414141-s002.pdf^ (2MB, pdf)

   Supplementary Table1
   [336]ADVS-12-2414141-s001.xlsx^ (286.9KB, xlsx)

Acknowledgements