Hepatic ratio of phosphatidylcholine to phosphatidylethanolamine predicts survival after partial hepatectomy in mice


  • Ji Ling,

    1. Group on the Molecular and Cell Biology of Lipids, University of Alberta, Edmonton, AB, Canada
    2. Department of Biochemistry, University of Alberta, Edmonton, AB, Canada
    Search for more papers by this author
  • Todd Chaba,

    1. Department of Laboratory Medicine and Pathology, University of Alberta, Edmonton, AB, Canada
    Search for more papers by this author
  • Lin-Fu Zhu,

    1. Department of Surgery, University of Alberta, Edmonton, AB, Canada
    Search for more papers by this author
  • René L. Jacobs,

    1. Group on the Molecular and Cell Biology of Lipids, University of Alberta, Edmonton, AB, Canada
    2. Group on the Molecular and Cell Biology of Lipids and Department of Agricultural, Food, and Nutritional Science, University of Alberta, Edmonton, AB, Canada
    Search for more papers by this author
  • Dennis E. Vance

    Corresponding author
    1. Group on the Molecular and Cell Biology of Lipids, University of Alberta, Edmonton, AB, Canada
    2. Department of Biochemistry, University of Alberta, Edmonton, AB, Canada
    • Department of Biochemistry, University of Alberta, Edmonton, AB T6G 2S2 Canada===

    Search for more papers by this author
    • fax: 780-492-3383

  • Potential conflict of interest: Nothing to report.


A major predictor of failed liver resection and transplantation is nonalcoholic fatty liver disease (NAFLD). NAFLD is linked to a wide spectrum of diseases including obesity and diabetes that are increasingly prevalent in Western populations. Thus, it is important to develop therapies aimed at improving posthepatectomy outcomes in patients with NAFLD, as well as to improve the evaluation of patients slated for hepatic surgery. Decreased hepatic phosphatidylcholine (PC) content and decreased ratio of hepatic PC to phosphatidylethanolamine (PE) have previously been linked to NAFLD. To determine if decreased hepatic PC/PE could predict survival after hepatectomy, we used mouse models lacking key enzymes in PC biosynthesis, namely, phosphatidylethanolamine N-methyltransferase and hepatic-specific CTP:phosphocholine cytidylyltransferase α. These mice were fed a high-fat diet to induce NAFLD. We then performed a 70% partial hepatectomy and monitored postoperative survival. We identified hepatic PC/PE to be inversely correlated with the development of steatosis and inflammation in the progression of NAFLD. Decreased hepatic PC/PE before surgery was also strongly associated with decreased rates of survival after partial hepatectomy. Choline supplementation to the diet increased hepatic PC/PE in Pemt−/− mice with NAFLD, decreased inflammation, and increased the survival rate after partial hepatectomy. Conclusion: Decreased hepatic PC/PE is a predictor of NAFLD and survival following partial hepatectomy. Choline supplementation may serve as a potential therapy to prevent the progression of NAFLD and to improve postoperative outcome after liver surgery. (HEPATOLOGY 2012)

Nonalcoholic fatty liver disease (NAFLD) consists of a wide spectrum of hepatic pathologies.1 The disease can progress from steatosis (triacylglycerol [TG] accumulation) to nonalcoholic steatohepatitis (NASH; steatosis with inflammation), and eventually to liver failure. It is unclear why some patients progress from steatosis to steatohepatitis, whereas others do not. However, the current consensus is that hepatic steatosis sensitizes the liver to a variety of metabolic injuries such as oxidative stress and cytokines.2 Interestingly, recent studies have linked NAFLD to aberrant hepatic phospholipid levels.3, 4

Hepatic phosphatidylcholine (PC) biosynthesis occurs by way of two pathways.5 Under normal conditions, the CDP-choline pathway is the major hepatic pathway, making 70% of hepatic PC.6 In this pathway, the rate-limiting reaction is catalyzed by CTP:phosphocholine cytidylyltransferase (CT).5 CT is encoded by two genes in mice, Pcyt1a and Pcyt1b.7, 8 In the liver, CTα (the product of the Pcyt1a gene) is the predominant isoform.8 The liver can also synthesize the remaining 30% of PC by the methylation of phosphatidylethanolamine (PE), which is catalyzed by phosphatidylethanolamine N-methyltransferase (PEMT).5, 6

In mice, the inhibition of either the PEMT pathway or the hepatic CDP-choline pathway (via the deletion of hepatic CTα) results in the accumulation of hepatic TG.9, 10 It is thought that an inadequate level of hepatic PC impairs the secretion of very low density lipoprotein particles from the liver. When mice lacking PEMT (Pemt−/− mice) were fed for 10 weeks a normal choline, high-fat (HF) diet, the animals developed severe steatosis and increased plasma alanine aminotransferase.11 In addition, hepatic PC was decreased, resulting in a markedly reduced PC/PE. When Pemt−/− mice were fed for 10 weeks the HF diet supplemented with additional choline, hepatic PC levels and PC/PE were normalized, which prevented liver damage but not steatosis.12

Reduced hepatic PC/PE (1.2 versus 2.5 in normal controls) was found in patients clinically diagnosed with NASH.3 In addition, a single nucleotide polymorphism of Pemt, which results in partial loss of activity, occurred more frequently in humans with NAFLD than in healthy individuals.12 Thus, decreased PC/PE may be a good predictor of NAFLD.

In clinical and animal studies, NAFLD is associated with delayed liver regeneration, increased liver damage, and mortality in patients after hepatic resection and liver transplantation.13, 14 Because the pathology of NAFLD is correlated with obesity and insulin resistance, therapies to improve the postoperative recovery of patients with NAFLD are becoming increasingly important. Indeed, the percentage of patients who have undergone liver transplantation due to NASH has increased dramatically in the last decade (1.2% in 2001 to 9.7% in 2009).15 Furthermore, nearly half of potential donors for liver transplantation presented with some form of NAFLD.14

In this study we provide evidence that decreased hepatic PC/PE is a major contributor to the pathogenesis of NAFLD. We also identify hepatic PC/PE as a predictor of posthepatectomy outcome. Finally, we provide evidence that choline supplementation is a viable therapy to improve survival after hepatectomy by increasing hepatic PC/PE, which decreases inflammation and improves energy utilization.


ATP, adenosine triphosphate; CT, CTP:phosphocholine cytidylyltransferase; HF, high-fat diet; HF-CS, high-fat diet with choline supplementation; IL-6, interleukin 6; MRC, mitochondrial respiratory chain; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; PE, phosphatidylethanolamine; PEMT, phosphatidylethanolamine N-methyltransferase; PEPCK, phosphoenolpyruvate carboxykinase; PC, phosphatidylcholine; PC/PE, phosphatidylcholine/phosphatidylethanolamine ratio; PH, 70% partial hepatectomy; ROS, reactive oxygen species; TG, triacylglycerol; TNFα, tumor necrosis factor α; UCP2, uncoupling protein 2.

Materials and Methods


Primers for real-time quantitative polymerase chain reaction (PCR) were purchased from the Institute of Biomolecular Design at the University of Alberta and are listed in Supporting Table 1. All other reagents were from standard commercial sources.

Animal Diets and Surgery.

All procedures were approved by the University of Alberta's Institutional Animal Care Committee in accordance with guidelines of the Canadian Council on Animal Care. The liver-specific CTα knockout (LCTα−/−) and control mice (LCTα-floxed) were previously generated.10 C57Bl/6 Pemt+/+ and Pemt−/− mice were backcrossed >7 generations. All animals were exposed to a 12-hour light-dark cycle, fed ad libitum a rodent diet, and had free access to water. In some experiments animals were fed an HF diet (Bio Serv) or an HF diet supplemented with 2.7 g choline chloride per kg diet for the indicated lengths of time.

All experiments were performed on male mice between the ages of 12-15 weeks. Mice were subjected to 70% partial hepatectomy (PH) using a modified method of Higgins and Anderson.16 All surgeries were performed between 9 am and 12 pm under isoflurane anesthesia.


A portion of the liver was fixed in 10% buffered formalin and stained with hematoxylin and eosin. Each slide was scored for steatosis, hepatocellular ballooning, portal inflammation, lobular inflammation, and fibrosis.17 A modified NAFLD activity score, which was used to assess the progression of NAFLD, is the sum of steatosis, ballooning, portal inflammation, and lobular inflammation score.18, 19

Quantification of Hepatic Lipid Levels.

Liver tissues were homogenized in 10 volumes of 50 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 1 mM dithiothreitol, and 0.1 mM phenylmethylsulfonyl fluoride, pH 7.4. Protein concentrations were determined by Bradford assay (Bio-Rad) with bovine serum albumin as standard. Total lipids were extracted from tissue homogenates (1 mg protein) using the Folch method.20 Phospholipids and neutral lipids were separated by thin-layer chromatography.9 PC and PE were quantified by a phosphorus assay.21 TG was quantified by a chemical assay.22

Real-Time Quantitative PCR.

Total RNA was isolated from snap-frozen liver tissue using TRIzol reagent (Invitrogen). RNA was treated with DNase I (Invitrogen) and reverse transcribed using Superscript II and oligo(dT)12-18 primers. Real-time quantitative PCR was performed using a Rotor-Gene 3000 instrument (Montreal Biotech). The messenger RNA (mRNA) levels were normalized to cyclophilin mRNA in the sample using a standard curve. Data were analyzed using the Rotor-Gene 6.0.19 program (Montreal Biotech).

Quantification of Hepatic Adenosine Triphosphate (ATP).

Frozen powdered liver tissue (0.01 g) was homogenized in 1 mL of 0.4 M perchloric acid. Homogenates were neutralized with 2.5 M KOH. The supernatant was diluted 1:10 with phosphate-buffered saline (PBS) (0.1 mM, pH 7.8) and assayed for ATP using the ATP assay kit (Sigma).

Quantification of Hepatic Glycogen.

Frozen powered liver tissue (0.02 g) was boiled in 30% KOH for 30 minutes and glycogen was precipitated with 3 mL of 100% EtOH. Precipitated glycogen was hydrolyzed to glucose by boiling in 3 N HCl for 3 hours. Glucose was assayed using a glucose assay kit (BioAssay Systems).

Blood Glucose.

Blood was sampled from the tail vein. Blood glucose levels were monitored using a glucometer (Accu-chek) before and 24 hours after surgery.

Statistical Analysis.

Data are represented as mean ± standard error of the mean (SEM). Correlation was determined by Pearson's correlation. Unpaired t test was used to compare two groups. One- or two-way analysis of variance (ANOVA) was used in all other comparisons. P < 0.05 was considered significant. Six to 16 animals were used per experimental group.


Decreased PC/PE Is Linked to the Progression of NAFLD Disease in LCTα−/− and Pemt−/− Mice on HF Diets.

Both the Pemt−/− mice and the LCTα−/− mice develop hepatic steatosis when challenged with an HF diet for 10 weeks.11 To study the progression of NAFLD in these animals, we placed LCTα−/− and Pemt−/− mice and their control littermates on an HF diet for 0 to 14 days to induce different degrees of NAFLD (Supporting Figs. 1, 2).

Histopathological analysis showed that Pemt+/+ mice did not exhibit NAFLD after 2 weeks on the HF diet because there was no increase in steatosis, inflammation, or hepatocellular ballooning (Fig. 1A-D). By comparison, Pemt−/− mice exhibited increasing steatosis (Fig. 1B, Supporting Fig. 3A) as well as hepatocellular ballooning (Fig. 1D) and inflammation (Fig. 1C) with the HF challenge. However, no fibrosis was detected at any stage.

Figure 1.

Progression of NAFLD in Pemt+/+ and Pemt−/− mice fed the HF diet. Liver sections were stained with hematoxylin and eosin and histopathologically graded for (A) NAFLD activity score, (B) steatosis, (C) inflammation, (D) cell ballooning (n = 3-7). *P < 0.05 versus Pemt+/+; #P < 0.05 versus t = 0 of the same genotype.

Similarly, LCTα−/− mice developed a higher grade of NAFLD when compared to LCTα-floxed mice after 2 weeks of HF diet (Supporting Fig. 4A). The LCTα−/− mice also displayed increases in steatosis and inflammation scores, but there was no significant difference in hepatocellular ballooning (Supporting Fig. 4B-D).

We next determined if the progression of NAFLD in either of our mouse models was associated with a decline in hepatic PC/PE. Indeed, hepatic PC and PE content of the Pemt−/− mice was altered during HF feeding compared to Pemt+/+ mice (Fig. 2A,B). Consequently, hepatic PC/PE decreased in Pemt−/− mice with time on the HF diet (Fig. 2C). Similarly, LCTα−/− mice showed a reduced PC/PE after the HF diet for 2 weeks compared to controls (Supporting Fig. 5A), although the decrease in PC/PE was not as pronounced as that of the Pemt−/− mice (Fig. 2C). Using correlation curves, we further corroborated a significant association between reduced hepatic PC/PE and the development of NAFLD (Fig. 2F). Decreased hepatic PC/PE also correlated with increasing severity of steatosis (Fig. 2D) and inflammation (Fig. 2E). However, the correlation between hepatic PC/PE and hepatocellular ballooning (r2 = 0.249, P < 0.01) was not as strong.

Figure 2.

Correlation of hepatic PC/PE with the progression of NAFLD in Pemt+/+ and Pemt−/− mice. (A) Hepatic PC, (B) PE, and (C) PC/PE were measured after Pemt+/+ and Pemt−/− mice were fed the HF diet for 0-14 days (n = 4-11). Decreased hepatic PC/PE was associated with (D) steatosis, (E) inflammation, and (F) NAFLD activity score (n = 42). *P < 0.05 versus Pemt+/+; #P < 0.05 versus t = 0 of the same genotype.

Decreased Preoperative Hepatic PC/PE Is Negatively Associated with Survival Rates Following PH.

NAFLD increases the risk of morbidity and mortality after liver surgery.13 Because hepatic PC/PE is associated with the intensity of NAFLD, we determined whether preoperative hepatic PC/PE can be used to predict survival after partial hepatectomy. Pemt−/− or LCTα−/− mice and their control littermates were fed the HF diet for 0 to 14 days followed by PH. Outcome was assessed as percent survival 24 hours after surgery. Correlation curves indicate that increased severity of NAFLD prior to surgery decreased survival after hepatectomy (Fig. 3C). Both the degree of hepatic steatosis (Fig. 3A, Supporting Fig. 3B) and inflammation (Fig. 3B) before PH are also negatively correlated with survival. Hepatocellular ballooning was very weakly correlative (r2 = 0.52). Interestingly, preoperative hepatic PC and PE levels alone have no or weak correlation with survival, respectively (Fig. 3D,E). However, decreased preoperative hepatic PC/PE was strongly predictive of mortality after PH (Fig. 3F).

Figure 3.

Correlation of survival rate after 70% partial hepatectomy with preoperative hepatic phospholipid content and histological analysis of NAFLD. Pemt−/−, LCTα−/−, and control littermates were fed the HF diet for 0-14 days and then subjected to liver resection (n = 3-7). Survival rate 24 hours after surgery was associated with preoperative (A) steatosis, (B) inflammation, and (C) NAFLD activity score as well as hepatic (D) PC, (E) PE, and (F) PC/PE.

Choline Supplementation Increases Hepatic PC/PE and Improves Survival After PH in Pemt−/− Mice.

Next we determined whether increasing hepatic PC/PE prior to surgery could improve survival after PH in mice with NAFLD. Because Pemt−/− mice exhibited a greater degree of NAFLD compared to LCTα−/− mice, we used Pemt−/− and Pemt+/+ mice. Pemt+/+ and Pemt−/− mice were fed the HF diet for 1 week (1.3 g choline/kg diet) to trigger NAFLD and then fed an HF diet supplemented with choline chloride (HF-CS: 4 g choline/kg diet) for an additional week. We expected that choline supplementation would stimulate CT-dependent PC synthesis, thereby improving the hepatic PC/PE.23 PH was performed and survival was monitored 24 hours after surgery.

The HF-CS diet normalized hepatic PC (Fig. 4A) and increased hepatic PE (Fig. 4B) in Pemt−/− mice in comparison to the HF diet. As such, hepatic PC/PE was significantly increased in Pemt−/− mice fed the HF-CS diet when compared to Pemt−/− mice fed the HF diet (Fig. 4C). Correspondingly, survival after PH increased by almost 2-fold in Pemt−/− mice fed the HF-CS diet when compared to the HF-fed Pemt−/− mice (57% versus 32%) (Fig. 4D).

Figure 4.

Choline supplementation to the HF diet improves hepatic PC/PE and survival after PH. Pemt+/+ and Pemt−/− mice were fed for 2 weeks the HF diet or for 1 week the HF diet plus 1 week the HF diet supplemented with choline (HF-CS). Preoperative hepatic (A) PC, (B) PE, and (C) PC/PE levels (n = 4-7). (D) Survival rate 24 hours after surgery (6-16). *P < 0.05 versus Pemt+/+; #P < 0.05 versus t = 0 of the same genotype.

The HF-CS diet had no effect on body weight (Supporting Fig. 6A) and steatosis was not prevented in Pemt−/− mice (Fig. 5A, Supporting Fig. 6B). However, the livers of Pemt−/− mice fed the HF-CS diet exhibited reduced inflammation and NAFLD activity in comparison to Pemt−/− mice fed the HF diet (Fig. 5B,C). Likewise, Pemt−/− mice fed the HF-CS diet showed attenuated hepatic macrophage infiltration as measured by decreased mRNA of macrophage markers cluster of differentiation 68 (CD68) and F4/80 and proinflammatory cytokines interleukin (IL)-6 and tumor necrosis factor alpha (TNFα) (Fig. 6A).

Figure 5.

Choline supplementation to the HF diet attenuates NAFLD in Pemt−/− mice. Pemt+/+ and Pemt−/− mice were fed the HF diet for 2 weeks or for 1 week the HF diet plus 1 week the HF diet supplemented with choline (HF-CS). Livers were analyzed for (A) steatosis, (B) inflammation, and (C) NAFLD activity score (n = 4-7). *P < 0.05 versus Pemt+/+; #P < 0.05 versus t = 0 of the same genotype.

Figure 6.

Choline supplementation to the HF diet reduces inflammation and increases ATP levels in Pemt−/− mice. Pemt+/+ and Pemt−/− mice were fed the HF diet for 2 weeks or the HF diet for 1 week plus 1 week of the HF diet supplemented with choline (HF-CS). (A) mRNA expressions of CD68, F4/80, TNFα, and IL-6 (n = 3-8). (B) Hepatic ATP content (n = 5). (C) mRNA expression of UCP2 (n = 3-4). *P < 0.05 versus Pemt+/+; #P < 0.05 versus t = 0 of the same genotype.

Choline supplementation of the Pemt−/− mice on the HF diet prevented the depletion of hepatic ATP levels observed in Pemt−/− mice fed the HF diet (Fig. 6B). Correspondingly, HF-fed Pemt−/− mice displayed increased hepatic expression of uncoupling protein-2 (UCP2), an inner mitochondrial membrane protein that decreases ATP production,24 which was reversed when the Pemt−/− mice were fed the HF-CS diet (Fig. 6C).

Choline Supplementation Prevented Hypoglycemia in Pemt−/− Mice, Resulting in Decreased Mortality After PH.

After PH, the remaining liver undergoes regeneration while preserving tissue-specific functions necessary for survival, which includes the maintenance of euglycemia.25 Both Pemt+/+ and Pemt−/− mice fed the HF diet experienced drops in plasma glucose levels after PH (Fig. 7A). However, the Pemt−/− mice had significantly lower glucose levels than Pemt+/+ mice at 24 hours. The Pemt−/− mice were also less ambulatory and displayed signs of hypothermia (e.g., shivering, cold to the touch). Unsurprisingly, the survival rate of the Pemt−/− mice was low, whereas all Pemt+/+ mice survived (Fig. 4D). When the mice were fed the HF-CS diet, there were no significant differences in plasma glucose between the Pemt+/+ and Pemt−/− mice 24 h after PH (Fig. 7A). Thus, more Pemt−/− mice fed the HF-CS diet recovered from the PH than Pemt−/− mice fed the HF diet (Fig. 4D).

Figure 7.

Choline supplementation to the HF diet increases survival of Pemt−/− mice after PH. (A) Blood glucose levels in Pemt+/+ and Pemt−/− mice fed either the 2 weeks HF diet or the 1 week HF diet and then 1 week HF diet with choline supplementation (HF-CS) (n = 6-7). (B) Hepatic glycogen content (n = 7-9). (C) mRNA expression of PEPCK 24 hours after PH (n = 3-6). *P < 0.05 versus Pemt+/+; #P < 0.05 versus t = 0 of the same genotype.

Hypoglycemia experienced by the HF-fed Pemt−/− mice after PH could be the result of impaired gluconeogenesis and/or glycogenolysis.25 Glycogen levels in Pemt+/+ and Pemt−/− on both diets were significantly decreased after PH compared to the levels before surgery, suggesting that glycogenolysis occurs normally in Pemt−/− mice (Fig. 7B). However, Pemt−/− mice fed the HF diet had significantly lower glycogen levels before PH in comparison to Pemt+/+ mice. No difference in preoperative glycogen levels was observed in Pemt−/− mice on HF-CS in comparison to Pemt+/+ mice (Fig. 7B). The expression of phosphoenolpyruvate carboxykinase (PEPCK), the rate-limiting enzyme in hepatic gluconeogenesis, was also much lower in the HF-fed Pemt−/− mice after PH in comparison to controls (Fig. 7C). This difference in PEPCK expression was not observed in Pemt−/− mice on the HF-CS diet in comparison to Pemt+/+ mice.


Our studies have provided evidence that decreased hepatic PC/PE is associated with the progression of NAFLD. Pemt−/− mice developed NAFLD to a larger degree than LCTα−/− mice when fed the HF diet. Coincidentally, hepatic PC/PE in Pemt−/− mice also dropped to a far greater degree than LCTα−/− mice. This was unexpected because the CT-mediated pathway is quantitatively the more important pathway for PC production than the PEMT pathway.6 However, because PEMT catalyzes the conversion of PE to PC, PEMT may play a pivotal role in regulating the balance between PC and PE. Indeed, PEMT activity in vitro is up-regulated as the consequence of elevated PE in the endoplasmic reticulum.26 Chow fed Pemt−/− mice also have increased hepatic PE levels and decreased PC/PE in comparison to control mice.27 Interestingly, PE levels in Pemt−/− mice were normalized after 2 weeks of the HF diet. This may reflect a compensatory decrease in the CDP-ethanolamine pathway as result of elevated PE and/or decreased PC/PE.28

The distribution and composition of PC and PE are regulated to maintain membrane integrity and to regulate the flow of substances across the plasma and intracellular membranes.29, 30 Decreased hepatic PC/PE can trigger inflammation and steatohepatitis by negatively affecting plasma membrane permeability, leading to leakage of hepatocellular content into the extracellular space.3 This can activate Kupffer cells and promote the infiltration of neutrophils and other macrophages, which will increase cytokine-mediated injury of hepatocytes. In particular, TNFα is known to induce oxidative damage to the mitochondrial DNA and increase the permeability of the mitochondrial outer membrane, thus impairing the mitochondrial respiratory chain (MRC).31 Dysfunction of the MRC increases reactive oxygen species (ROS) productivity, which further contributes to hepatocellular injury and the pathogenesis of NASH.31 Indeed, patients with NASH and rodent models of diet-induced steatohepatitis often present with depleted hepatic ATP content.32, 33

Pemt−/− mice on 2 weeks of the HF diet showed increased production of TNFα and depletion of hepatic ATP content, suggestive of an impaired MRC. Furthermore, we also observed increased UCP2 expression in 2-week HF-fed Pemt−/− mice, a marker of oxidative stress.24 UCP2 mediates the movement of protons across the inner mitochondrial membrane into the matrix, which decreases the formation of mitochondrial ROS, but impairs ATP synthesis.34

Choline supplementation increased hepatic PC/PE in Pemt−/− mice that have developed NAFLD, prevented cytokine production, reduced UCP2 expression, and normalized ATP content. Therefore, decreased hepatic PC/PE may facilitate the development of NASH by promoting inflammation and MRC dysfunction. Interestingly, Pemt−/− mice still accumulated hepatic TG when hepatic PC/PE was increased with choline supplementation. Therefore, the accumulation of hepatic TG may not be the major cause of the metabolic injuries that leads to the progression of NAFLD. Instead, decreased PC/PE could have a more significant role in contributing to the development of NAFLD.

Clinical and animal studies have shown NAFLD impairs recovery after liver resections.13 In liver transplantations, steatotic grafts also increase the chance of primary nonfunction and recipient morbidity and mortality.35 Furthermore, the degree of NAFLD also predisposes the liver to injury due to ischemia/reperfusion during hepatectomy, further contributing to liver dysfunction and mortality after surgery.36 Therefore, the presence of NAFLD is often assessed before hepatic surgery.

Currently, NAFLD is diagnosed by the histopathological evaluation of liver biopsies stained with hematoxylin and eosin.13 For the first time, we show hepatic PC/PE can predict survival after PH as well as histopathological analysis. In practice, assaying hepatic PC/PE may be safer than histological analysis in determining risk of liver surgery because the measure of hepatic PC/PE would require one versus multiple biopsies.13, 37 Quantification of hepatic PC/PE would also avoid personal bias in the staining and the histopathological evaluation of the liver samples. Therefore, measuring hepatic PC/PE should be studied in humans to determine postoperative risk in addition to the standard histopathological analysis.

The majority of short-term patient morbidity and mortality after liver resection is caused by inadequate function of the remnant liver.13 To support the demands of rapid cellular proliferation and essential liver function after PH, sufficient ATP is required.25 Our data suggest that decreased hepatic PC/PE may lead to impaired MRC function. Therefore, low hepatic PC/PE may contribute to a negative outcome after PH due to a failure of the remnant liver to meet energy requirements. Postoperative mortality experienced by Pemt−/− mice fed the HF diet coincided with the onset of hypoglycemia. This was attributed to depleted hepatic glycogen stores presurgery and an impaired induction of gluconeogenesis after PH. Of note, IL-6 inhibits insulin-stimulated glycogen synthesis in rat hepatocytes.38 Therefore, the inflammatory response induced by low hepatic PC/PE in HF-fed Pemt−/− mice may also contribute to mortality after PH by impairing glycogen storage.

Increasing the preoperative hepatic PC/PE by dietary choline supplementation doubled the survival rate in Pemt−/− mice with severe NAFLD by preventing inflammation and MRC dysfunction, thus avoiding the detrimental drop in blood glucose after surgery. Preventing inflammation and MRC dysfunction also protects hepatocytes from cytokine/ROS-mediated necrosis during ischemia/reperfusion and liver regeneration.36 Therefore, a high dietary intake of choline before PH could improve the postoperative outcome of hepatectomy or liver transplantation in individuals presenting with inflammation (e.g., patients with NASH, hepatocellular carcinoma, or hepatitis). The HF-CS diet did not improve survival in Pemt−/− mice by 100%. However, it is possible a longer duration or higher dose of choline supplementation may further improve NAFLD and survival after hepatectomy.

Reduced hepatic PC/PE has been observed in patients with NASH.3, 4 This suggests that a low hepatic PC/PE may be a common characteristic of inflammation of the liver, not just a negative effect of genetic inhibition of hepatic PC biosynthesis. Therefore, choline supplementation could prove to be a preventative therapy for individuals susceptible to NASH and/or NASH patients before the development of cirrhosis and hepatic carcinomas. Intravenous administration of choline has already proven effective in resolving hepatic steatosis, high serum aminotransferase levels, and low plasma choline levels in patients receiving long-term intravenous feeding.39 Choline supplementation was also effective in preventing steatosis and fibrosis induced by a high sucrose diet or 15% ethanol in rats.40 Recently, high dietary intake of choline beneficially reduced plasma inflammatory biomarkers such as C-reactive protein and TNFα.41

In conclusion, we provide evidence that hepatic PC/PE is inversely correlated with the degree of NAFLD. As such, quantification of hepatic PC/PE could be an excellent determinant of NAFLD and predictor of postliver surgery outcome. Choline supplementation in mice with NAFLD prevented hepatic inflammation and NASH development by increasing hepatic PC/PE. Therefore, choline supplementation could be a viable therapy to prevent NASH development and to decrease the mortality rate postliver surgery in patients with NAFLD.


We thank Susanne Lingrell, Randal Nelson, and Audric Moses for excellent technical assistance. We thank Dr Andrew Mason for helpful comments.