Fatty Acid Synthase Blockade Protects Steatotic Livers from Warm Ischemia Reperfusion Injury and Transplantation


*Corresponding author: Kenneth D. Chavin chavinkd@musc.edu


Cerulenin has been shown to reduce body weight and hepatic steatosis in murine models of obesity by inhibiting fatty acid synthase (FAS). We have shown that attenuating intrahepatocyte lipid content diminished the sensitivity of ob/ob mice to ischemia/reperfusion injury and improved survival after liver transplantation. The mechanism of action is by inhibition of fatty acid metabolism by downregulating PPARα, as well as mitochondrial uncoupling protein 2 (UCP2), with a concomitant increase in ATP. A short treatment course of cerulenin prior to I/R injury is ideal for protection of steatotic livers. Cerulenin opens the potential for expanding the use of steatotic livers in transplantation.


fatty Acid synthase




The shortage of quality donor organs as compared with demand contributes to longer waiting lists and increased mortality of patients in need of a transplant. This growing disparity is forcing transplant surgeons to consider livers from ‘marginal donors.’ Donors of any age are now considered, including donors with co-morbidities such as hypertension, diabetes, and obesity. This is an increasing problem in the United States as the available donor population continues to become obese (1). Use of marginal donors may induce higher graft nonfunction rates, increase retransplantation rates, and increase recipient morbidity and mortality (2–7). Liver steatosis (hepatocyte lipid accumulation) is a primary factor for determining the usability of potential organs (3,5). Nearly one-third of all donated livers are discarded due to steatosis alone, despite the organ being biochemically normal. These discarded donor livers are considered to have a prohibitive rate of primary nonfunction (PNF) and a higher susceptibility to ischemia/reperfusion (I/R) injury, although the underlying molecular and cellular mechanisms of fatty liver transplant failure are not well understood (4,6,7).

Current pharmacologic interventions for obesity and steatosis require long-term treatment and have had mixed results. Consequently, use of pharmacologic agents to affect weight loss, along with steatosis, take weeks to months to achieve results (1,8). In the setting of a steatotic donor liver, such an approach is considered useless due to the logistical necessity for the intervention to affect the organ quickly prior to the insult of I/R. Herein lies the need for fast-acting pharmacologic agents that will facilitate stability and protect the steatotic liver from the usual insults that potentially render these organs more susceptible to PNF. Multiple investigations have explored the possibility of perturbing substrate excess-induced pathways in order to affect obesity and steatosis (9,10). Two compounds, cerulenin, and its more potent derivative C75, have been shown, with prolonged administration, to cause sustained weight loss in ob/ob mice through a number of mechanisms, along with phenotypic reversal of liver steatosis (11–13). Given the profound effect that fatty acid synthase blockade has in weight reduction and altering steatotic livers in mice, we investigated the role of a short course of administration (2 d) of cerulenin in altering the susceptibility of steatotic livers to I/R injury and transplantation.



Chemicals were purchased from Sigma Co. (St. Louis, MO, USA) with the following exceptions. RNA-Bee reagent was purchased from Tel.Test Inc. (Friendswood, TX, USA). Chloroform and isopropyl alcohol were obtained from Baker (Phillipsburg, NJ, USA). Agarose was obtained from Invitrogen (Carlesbad, CA, USA) and nylon membranes were purchased from Schleicher and Schuell (Keene, NH, USA). Quick-Hyb hybridization solution and a Random Priming Kit were purchased from Stratagene (Cedar Creek, TX, USA) and 32P- from Amersham Pharmacia Biotech (Piscataway, NJ, USA).

Animal studies

Adult 4-week-old male ob/ob mice and their lean littermates were purchased from Jackson Laboratories (Bar Harbor, ME, USA). Mice were housed in temperature- and light-controlled chambers on a 12:12 h light/dark cycle and provided with rodent chow and water ad libitum. Anesthesia was administered to 6- to 8-week-old mice at 50 mg/kg body weight intraperitoneal (i.p.) Nembutal. Mice were opened with a mid-line laparotomy, the portal triad was identified. Ischemia was initiated via a pediatric vessel loop ligature for 15 min and reperfusion was subsequently allowed for immediate, 1.5, 24 or 48 h depending on the experimental group (2). Mice were sacrificed by exsanguination and occipital-cervical subluxation according to approved IACUC protocols.

Orthotopic liver transplant (OLTx)

An orthotopic liver transplant (OLTx) was performed as previously described. Male lean mice and male ob/ob size-matched mice were used in all surgical procedures. Vessels were exposed and a 24-gauge catheter was inserted to perfuse the liver with 100 units of heparin in 1 mL of chilled Lactated ringers solution (LR). The liver was then perfused with another 1 mL of LR and then 2 mL of University of Wisconsin (UW) solution (Viaspan, Barr Laboratory) at 4 °C. The liver was again flushed with another 1 mL of cold LR followed by 2 mL of chilled UW. The liver was then carefully dissected from the peritoneal cavity, and placed in cold LR for cuff preparation. The two-cuff technique was used as described by Kamada (3) to the donor liver. Anesthesia, laparotomy, and exposure were performed to the recipient as in the donor operation. The donor liver was then placed in the recipient in the orthotopic position. The abdomen was irrigated and viscera were replaced and the abdomen closed with a continuous 6.0 proline suture. Animals were allowed to awaken or were sacrificed after emergence from anesthesia if they appeared sick.

Serum alanine aminotransferase (ALT)

Whole blood was collected and centrifuged to obtain serum, which was evaluated for serum alanine aminotransferase (ALT) using a clinical laboratory Synchron LY2D System (Beckman Coulter Inc., Fullerton, CA, USA) and expressed as IU/L.

Serum endotoxin concentration

Murine blood serum endotoxin analysis was performed according to manufacturer's instructions using the chromogenic endpoint assay (Associates of Cape Cod, Inc. (Falmouth, MA, USA) Cat. no. C0180).


Hematoxylin and eosin (H&E) and Oil-Red O staining were performed as previously described (14,15).

Northern blot analysis

Total cellular RNA preparations and Northern blot were performed as described previously (16). Murine UCP2 cDNA was a gift from Dr Paul Dowell (Department of Biological Chemistry, Johns Hopkins School of Medicine). Mouse PPARα cDNAs were made by RT-PCR according to standard protocols, and the primer sequences are available upon request. For normalization, the Northern blots were stripped and re-hybridized with an 18S RNA cDNA probe.

ATP content assay

ATP concentration was measured by luminometer according to the manufacturer's (Cylex, Columbia, MD, USA) instructions.

Western blot analysis

Western blot analysis was performed as previously described (17) in collaboration with Dr. Daniel Ricquier. The blot was incubated with the first antibody for 4 h with UCP2 (605) antibody (Daniel Ricquier) at a dilution of 1:1000 followed by three washes, each for 10 min in Tris buffered saline with Tween 20 (TBST). The blot was placed in a secondary antibody (Pierce Biotechnology, Rockford, IL, USA) goat anti-rabbit IgG at a dilution of 1:7500 for 1 h. Detection was done by treatment with substrate (Cell Signaling) Lumi-Glo Chemiluminescent and exposure to film for 15 min. A minimum of six animals/treatment/time point were analyzed. Western blot samples were standardized against Cox 1 expression.

Statistical analysis

Experimental results were analyzed for their significance (Student's t-test). Significance was established at the 95% confidence level (p < 0.05)


Fatty acid synthase blockade protects steatotic livers from I/R injury

To determine if fatty acid synthase inhibition would protect steatotic animals subjected to hepatic I/R, 6- to 8-week-old male ob/ob mice were treated with cerulenin or vehicle i.p once daily for 48 h. In situ warm hepatic ischemia with bowel congestion was induced in the mice by complete occlusion of the porta-hepatis for 15 min, followed by 24 h of reperfusion (2). A protective effect was observed in steatotic animals subjected to warm ischemia after pre-treatment with cerulenin. There was 30% survival in the vehicle treatment group as compared with 95% survival with cerulenin treatment, p < 0.01 (Figure 1A). In addition to the survival benefit noted, 48 h pre-treatment with cerulenin grossly altered the phenotype of the ob/ob mouse liver, such that the livers appeared ‘normal’ and not steatotic. Histology of ob/ob livers demonstrated decreased steatosis and a change from macro to micro steatosis with cerulenin treatment (60 mg/kg/day for 2 d) (Figure 1B (H&E) and D (ORO) untreated vs. C (H&E) and E (ORO) cerulenin treated). Further histological analysis revealed decreased centrilobular necrosis and a greater percentage of viable hepatocytes with cerulenin treatment (Figure 1G) when compared with the samples from vehicle-treated animals (Figure 1F) after I/R. Consequently, cerulenin was able to protect steatotic livers from the detrimental effects of warm I/R, in part, by decreasing hepatic necrosis.

Figure 1.

Cerulenin treatment prior to ischemia/reperfusion (I/R) enhances survival. Twenty-four hour survival data of ob/ob cerulenin-treated vs. ob/ob vehicle-treated mice after 15 min of warm ischemia. (A) Average percent survival for vehicle- (n = 23) and cerulenin-treated mice (n = 21) are 30% and 95%, respectively. There is a statistically significant difference between the two groups (p < 0.01). (B–G) Histology of vehicle- and cerulenin-treated ob/ob livers prior to I/R by hematoxylin and eosin (H&E) (B vs. C), respectively, and Oil Red O staining (D vs. E), respectively. Vehicle-treated livers after 15 min of ischemia and 24 h of reperfusion demonstrated severe necrosis by H&E as compared with cerulenin pre-treatment (F vs. G).

Pronounced alteration in liver phenotype, histology and animal survival after warm I/R with cerulenin treatment suggests that fatty acid synthase blockade could be a viable approach to pharmacologically alter the state of steatotic livers to prevent PNF. Using our nonarterialized steatotic mouse liver transplant model, we investigated the ability of cerulenin pre-treatment to be protective in the setting of cold I/R and transplantation (18). Four-week-old ob/ob donor mice were pre-treated with vehicle or cerulenin for 2–7 d (60 mg/kg/d). Animals received the treatment i.p. daily for the prescribed time. Weight-matched lean C57Bl/6 mice were used as lean control and recipient animals. Non-arterialized transplants were performed as described by Birsner and others (18). These data represent only technically successful transplants. The animals that awoke from anesthesia had no signs of bile leak or evidence of bleeding on necropsy. Successfully transplanted untreated ob/ob donor livers to lean recipients had only a 36%, 2 h post-operative survival and 0% survival at 24 h (18). Histology of ob/ob-transplanted livers to lean recipients showed severe coagulative necrosis consistent with a transplanted liver that has developed PNF. Cerulenin pre-treatment altered this outcome. With increasing pre-treatment with cerulenin there were marked increases in short- (Figure 2A) and long-term survival. The maximum protective effect was observed with 7 d of cerulenin treatment (Figure 2B). These data clearly suggested a protective advantage mediated by cerulenin. It is acknowledged that a 7-d course of cerulenin treatment of a donor organ is unrealistic in the cadaveric transplant setting; however, it may be feasible in the setting of living donor donation. Of note, long-term survival in this model is measured in hours to days, not weeks. This may be due to the nonarterialized nature of the model and/or additional factors related to transplantation of a steatotic liver.

Figure 2.

Survival after murine orthotopic liver transplant (OLTx). (A) ob/ob to lean untreated control (n = 12) was compared with groups receiving different regimens of 2 (n = 12), 4 (n = 4), and 7 (n = 4) days of cerulenin treatment once per day. Survival improved significantly for the 4- and 7-d treatment groups compared with control (p < 0.05). 100% survival was observed in the lean-to-lean OLTx (n = 6). (B) Long-term survival is displayed in the Kaplan–Meier graph showing cerulenin treatment and survival of murine steatotic OLTx. Cold ischemic time in all groups was similar at 85.7 ± 11.8 (min).

Potential mechanisms of the protective effects of FAS blockade on ischemia/reperfusion

Cerulenin, an antibiotic originally isolated from the fungus Cephalosporium caerulens,has a structure of (2S)(3R)-2,3-epoxy-4-oxo-7,10-dodecadienoyl amide (19). Cerulenin does not inhibit respiration in yeast, and its mode of action is independent of the TCA cycle and its intermediates. The mode of action is through the inhibition of fatty acid synthase (FAS) and HMG-CoEnzyme A Synthase in yeast and keto-acyl acyl-carrier protein synthase in Escherichia coli. When administered to mammals, the drug is rapidly neutralized by mammalian epoxidases, given that the epoxide ring present at the 2,3 position of the compound is essential for both the inhibition of the enzyme activities and the antimicrobial activity of the compound. There are a number of potential mechanisms explaining the protective effects of cerulenin in steatotic I/R. These include the contribution of an antibiotic effect, an alteration in fatty acid metabolic pathways, and/or an alteration in mitochondrial uncoupling energy homeostasis.

Cerulenin was not working as an antibiotic

One explanation accounting for our observation of cerulenin's protective effects is its antimicrobial properties. Previous attempts at gut sterilization with antibiotics have shown promise in ameliorating the effects of endotoxic shock; however, its clinical application has failed. To further explore this potential mechanism, we examined the effect of treatment on circulating endotoxin levels after total porta-hepatis occlusion and reperfusion. Using our warm ischemia model as previously described, 6- to 8-week-old ob/ob mice were pre-treated with cerulenin or vehicle (60 mg/kg) for 48 h prior to ischemia. Immediately upon reperfusion the animals were sacrificed, and serum was collected and measured for endotoxin. Cerulenin had a small, but statistically significant, effect on the levels of circulating endotoxin (p < 0.05) (Figure 3). Endotoxin levels measured at different time points after reperfusion were also examined. Very early after the establishment of reperfusion (min), most of the endotoxin was likely absorbed by the liver and lungs during the first-pass of the blood through the body. Consequently the endotoxin concentrations found post immediate reperfusion (time > 15 min) were equivalent to those observed prior to I/R. This result suggested that cerulenin may partially be protective against I/R through its antibiotic effects, thereby dampening the consequences of endotoxin exposure during reperfusion. This, however, cannot account for the degree of protection observed when the only animal exposed to cerulenin was the donor in our liver transplant model. Consequently, the recipient is exposed to an endotoxin insult during reperfusion which is independent of cerulenin treatment. Therefore, the formal possibility of an antibiotic effect as a component of protection in this model is eliminated.

Figure 3.

Serum endotoxin levels of cerulenin-treated ob/ob mice after I/R. Serum endotoxin was determined at baseline (n = 8) and immediately after reperfusion after 15 min of warm ischemia. Vehicle control (n = 6) or cerulenin (n = 8) was administered intraperitoneal (i.p.) once daily for 2 d prior to ischemia. There is a statistically significant difference between the vehicle- and cerulenin-treated groups (p < 0.05).

Cerulenin inhibited fatty acid metabolism by affecting ATP and its associated signaling

A second mechanism of action by which cerulenin may mediate the observed increase in survival is through altering the response of secondary signals and proinflammatory cytokines. In the nonstressed steatotic hepatocyte, normal fatty acid metabolism is enhanced and, thus, certain signaling pathways are up-regulated. Such alterations include mitochondrial uncoupling protein 2 (UCP2) upregulation by stimulation through a PPAR-α response pathway, which is increased because of an excess of fatty acids. By transiently blocking the stimulus for UCP2 expression, either at the level of fatty acids (substrate) and/or at a downstream enzymatic reaction, UCP2 expression should be downregulated. This will improve the energy status of the cell by increasing ATP levels, resulting in protection of the steatotic hepatocyte from the effects of I/R (13). We have previously demonstrated that the ATP levels in ob/ob mice are markedly decreased relative to lean controls (2). Furthermore, ATP levels plummet during hepatic warm ischemia and never return to normal thus contributing to ob/ob mortality after this insult. Given an improvement in survival and decrease in hepatocellular injury with cerulenin, we sought to determine if the hepatic protection was due to alterations in ATP content. Again, using our warm ischemia model, 6- to 8-week-old ob/ob mice were treated with cerulenin or vehicle (60 mg/kg/day) for 48 h and then subjected to 15 min of ischemia followed by reperfusion. Animals were sacrificed immediately, 90 min, and 24 h after reperfusion. Tissues were freeze-clamped and then analyzed for the concentration of ATP per mg tissue by a luciferin/luciferase ATP detection assay. Immediately after reperfusion, ATP was reduced by 79% in controls treated with vehicle compared with baseline; ATP levels after cerulenin treatment were only reduced by 63%. These significant differences in the concentration of ATP between vehicle-treated and cerulenin-treated groups were lost at 90 min and 24 h. The ATP data collected 24 h after reperfusion for vehicle-treated mice represented a minority of the original group as only 20% of all animals survived at this time point (Figure 4). Consequently, the ATP levels measured for this group represented only the survivors and thus are not representative of the population studied. In addition, we examined the effects of cerulenin treatment on ATP levels prior to I/R. Figure 5A shows a significant increase in ATP prior to I/R in the cerulenin-treated mice compared with the vehicle-treated and baseline mice.

Figure 4.

Cerulenin treatment prior to I/R results in increased ATP. Whole tissue ATP level at baseline and in ob/ob cerulenin-treated vs. ob/ob vehicle-treated mice at 0, 90 min, and 24 h after 15 min of warm ischemia. All groups had n = 4 with the exception of cerulenin-treated at 24 h (n = 5). There was a statistically significant difference between the two groups at immediate reperfusion (p < 0.05).

Figure 5.

UCP2 Northern and Western blot analysis of cerulenin treatment. (A) Tissue levels of ATP in cerulenin- and vehicle-treated mice prior to I/R. All ATP values were standardized to protein. (B and C) Twenty-four hours after ischemia/reperfusion (I/R), cerulenin-treated mice (n = 7) expressed significantly less UCP2 than did vehicle-treated mice (n = 4) (p < 0.05). Baseline (n = 4) and vehicle baseline (n = 4) levels were also measured. The densitometric values are expressed as averages standardized to baseline. (D) UCP2 Western blot analysis was performed on tissue from 1 and 24 h after I/R. Cerulenin-treated mice (lanes 7–9 and 13–15) expressed significantly less UCP2 than did vehicle-treated mice (lanes 4–6 and 10–12), respectively. The greatest effect was seen at the 1-h time point. Vehicle baseline levels (lanes 1–3) were also measured. Additional controls were wild type (lane 16), UCP2 knockout mouse (lane 17), and lung positive control (lane 18). All lanes were compared and standardized to COX2. (E) Densitometric analysis of the Western blot of UCP2 (D). The ratio of UCP2:COX2 was determined for each group. * p < 0.05

To further investigate the dramatic observed differences in ATP levels after cerulenin treatment, we evaluated the influence of cerulenin on UCP2 expression. Northern blot analysis was performed on RNA isolated from 6-week-old ob/ob mice treated with cerulenin or vehicle. The level of UCP2 mRNA that was expressed in ob/ob mice at baseline did not change with vehicle treatment alone. I/R plus vehicle treatment increased UCP2 mRNA expression. This was consistent with our previous findings and was expected (2). Also, as expected, cerulenin treatment led to a decrease in UCP2 expression after I/R (Figure 5B,C). The results observed by measuring the mRNA level correlated well with those measuring UCP2 protein expression (Figure 5D). Densitometric analysis of the Western Blot was performed and was consistent in that the protein levels of UCP-2 (Figure 5E) showed to a first approximation that the trend of UCP-2 protein expression followed that of the expression of the mRNA synthesis. These data clearly suggest that one mechanism by which FAS blockade was protecting the steatotic livers in both our warm and cold ischemia models was by downregulating UCP2 expression, resulting in an increase in ATP content.

Northern blot analysis demonstrated that I/R did not interfere with PPARα expression, but pre-treatment with cerulenin led to downregulation of PPARα in liver samples subjected to I/R from ob/ob mice [(60 mg/kg/day) for 2 d or vehicle alone before I/R]. There was also an increased mRNA level of UCP2 in ob/ob liver samples after I/R, but that increase was blocked by pre-treatment with cerulenin (Figure 6).

Figure 6.

Northern blot analysis of PPARα and UCP2. Mouse livers from ob/ob animals receiving 2-d treatments of cerulenin or vehicle were subjected to 15 min of ischemia and 24 h reperfusion as indicated. 18S RNA was used for normalization.


Causes of obesity are multifactorial. One common finding associated with obesity is that cellular metabolism is altered as a function of substrate excess. Substrate excess is associated with an increase in circulating metabolized fatty acids. In the liver, the biochemical breakdown of fatty acids is through a pathway of acetyl-CoA to malonyl-CoA. The downstream impact of increased fatty acids is an alteration in the normal biochemical pathways which results in increases of PPARα,γ,δ, and UCP2 mRNA expression. Efforts to alter this metabolism and affect weight gain have focused on all stages of the pathway (20). Interference with these intermediates alters energy metabolism, ATP production and Krebs cycle metabolism, resulting in weight loss. Additional efforts have focused on altering the neurochemical axis, which promotes safety and reduces food consumption. Each intervention, however, requires prolonged treatment to have its desired effect. In cadaveric liver transplantation, there is only a short, finite amount of time that an intervention to alter the fat composition of the donor organ would be considered practical. Long-term treatments may be effective for living donation where an extended treatment course, greater than 1 week, is possible.

Cerulenin, a potent fatty acid synthase blocker, has been shown to sustain weight loss in ob/ob mice with phenotypic reversal of liver steatosis (11,12). The mechanism of action of this inhibitor is by directly blocking the metabolism of fatty acids and through an as yet uncharacterized neuropeptide that inhibits metabolic rate. In this report, we have described the use of cerulenin to successfully alter steatotic livers and render them protected from warm and cold ischemia reperfusion injury. This work is significant in that it demonstrates an alteration in phenotypically and histologically steatotic livers to ‘lean-like’ livers, which is an essential factor for survival after I/R. Furthermore, using this approach, a short-term treatment course resulted in a reproducible outcome with an increase in survival of the graft and animal which was consistently associated with amelioration of the steatotic condition.

Cerulenin works both by a direct inhibitory affect on FAS and by suppressing the appetite of the animals receiving the drug. Using the short, 2-d single dose treatment, we accomplished alterations which improved survival. From a mechanistic perspective, this appears to be at the level of ATP homeostasis. Protection resulted from an increase in available ATP prior to I/R and by altering the downstream signaling of fatty acid metabolism through a decrease in UCP2 expression. Altering each component was key to survival because upon reperfusion, more ATP was available. The levels of UCP2 continue to be a good indicator of whether or not a steatotic liver will survive I/R-based injury. Levels of UCP2 were consistent with previous work with ob/ob mice and hepatocyte cultures, which demonstrated that the concentration of ATP is lower and was inversely proportional to the degree of UCP-2 expression (2). Given that cerulenin pre-treatment results in both an increase in ATP and a decrease in UCP2 expression, this compound appears to be an ideal pharmacologic intervention for protecting steatotic livers with a short treatment course prior to I/R injury.


We would like to thank Dr Daniel Ricquier from the Centre National De La Recherche Scientifique (Paris France), for his work on this project.