Main

Differences exist in the response to toxic hepatic injury in neonates compared with adults. Previous studies have indicated that immature liver is partially protected from carbon tetrachloride (CCl4)-induced injury (1–3); proposed mechanisms include an underdeveloped cytochrome P-450 system and increased stimulation of hepatocellular regeneration and tissue repair mechanisms (1,4). Newer evidence shows that acute CCl4 toxicity increases nitric oxide synthase (NOS) activity resulting in increased hepatic blood flow, both in the portal vein and hepatic artery (5,6). Co-administration of CCl4 and NOS inhibitors dampens increased hepatic artery blood flow, worsening of hepatocellular damage (5,6), suggesting a protective role for NOS in this model via increased hepatic artery blood flow.

Hepatic blood flow in neonates differs from adults by nature of a ductus venosus, which shunts blood away from hepatocytes. There are also maturational differences in hepatic blood flow regulation; the hepatic artery contributes more to maintenance of sinusoidal blood flow in the immature liver than in the adult liver (7). Normally, hepatic blood flow is under the control of eNOS present in sinusoidal endothelial cells (8). Moreover, there is developmental expression of NOS in the liver. eNOS is minimally expressed in the prenatal liver, increases rapidly after birth, and reaches a maximum at 20 d of life, which is maintained in adults (9). Hepatic iNOS activity is not detectable during normal development (9).

Nitric oxide (NO) plays important roles in the liver with potential for being both cytotoxic and cytoprotective. Factors in determining beneficial versus harmful effects include location, amount, and duration of NO generation. eNOS is constitutively expressed in the endothelium of most tissues, playing an important role in regulating blood flow in response to hemodynamic shear stress and endothelial-dependent vasodilators. NO generated by eNOS is allosterically regulated and in small amounts leads to vasodilation. Unlike eNOS, iNOS is inducible and present in most liver cells. iNOS lacks fine regulation and tends to produce large quantities of NO once induced. Excessive NO leads to cell injury through nitrosylation of cellular enzymes and potentiation of oxidative stress by forming peroxynitrite (reviewed in Ref. 10). Normally only eNOS is present in the liver, whereas hepatic iNOS is upregulated by a number of conditions (endotoxemia, hemorrhagic shock, sepsis, ischemia-reperfusion, and liver regeneration). NO has been shown to protect against oxidative damage and peroxidation induced by experimental models such as alcohol (11), H2O2-mediated oxidative stress (12), and CCl4 (13).

The regulation and roles of eNOS and iNOS in CCl4-induced hepatotoxicity have not been elucidated in the developing liver. We performed this study to determine what role age-related differences in NOS expression plays in acute CCl4-induced hepatocellular injury.

METHODS

Chemicals.

Unless otherwise stated, all chemicals in this study were purchased from Sigma Chemical Co. (St. Louis, MO). Antibodies for Western blot were purchased as follows: anti-eNOS and anti-iNOS from BD Biosciences (San Diego, CA), antiactin from Chemicon (Temecula, CA), antiphospho-eNOS, antiserine/threonine protein kinase (Akt), antiphospho-Akt, and wortmannin (a phosphatidylinositol 3-kinase, PI3 K, inhibitor) were purchased from Cell Signaling (Danvers, MA).

Animal treatment.

Pregnant Sprague-Dawley rats were obtained at 14 d of gestation and kept in plastic cages with free access to food and water with alternating 12-h periods of light and darkness. In experiments requiring fetal tissue, the dam was killed by CO2 narcosis and cervical dislocation, followed by delivery of fetuses by cesarean section. In age-related experiments, newborn rats were kept with the mother with free access to breast-feeding before weaning, and standard rat chow and water post weaning. The age of rats studied were day of life (DOL) 0, 2, 5, 10, and 20. Adult, male Sprague-Dawley rats (200–250 g) were used as the adult time point, to avoid any potential effect of estrous on NOS expression. In the study of CCl4-induced hepatotoxicity, pups of either sex from various litters of equal postnatal age were randomized to one of the three experimental groups: 1) control group (vehicle, corn oil), 2) CCl4 [100 μL/kg body weight (BW), i.p.], or 3) L-N(G)-nitroarginine methyl ester (L-NAME, 2 mg/kg BW, i.p.) 1 h before CCl4 injection. Animals were killed 24 h following injection. For study of posttranslational regulation, DOL-20 pups were randomly divided into four experimental groups: 1) control group (vehicle, corn oil), 2) CCl4 (100 μL/kg BW, i.p.), 3) wortmannin (1.4 mg/kg BW, i.p.) 1 h before and 12 h after CCl4 injection, or 4) wortmannin alone (1.4 mg/kg BW, i.p.). Animals were killed 24 h after injection. The study was approved by the Institutional Animal Care and Use Committee at the University of Mississippi Medical Center.

Evaluation of CCl4 hepatotoxicity.

Blood samples were collected and centrifuged at room temperature for 10 min at 13,000 rpm. Serum was assayed for alanine aminotransferase (ALT) activity following the protocol of an ALT assay kit (Sigma Chemical Co. Diagnostics, St. Louis, MO). Liver samples from each aged animal were fixed in 10% formalin, paraffin-embedded, and sections obtained and stained with hemotoxylin and eosin for histologic evaluation.

Determination of lipid peroxidation.

Tissue concentration of malondialdehyde (MDA), an intermediate of lipid peroxidation, was measured as described previously (14).

Real time RT-PCR for NOS.

cDNA was synthesized using iScript cDNA Synthesis system (Bio-Rad Hercules, CA) on total cellular RNA. Controls without reverse transcriptase were performed for each sample to ensure absence of genomic DNA. Real time PCR was carried out in a real time thermal cycler (iCycler, Bio-Rad) using iQ SYBR Green Supermix (Bio-Rad). Cycling conditions were 5 min at 95°C, followed by 40 cycles of 15 s at 95°C, 15 s at 55°C, then 60 s at 72°C. PCR specificity was tested via analysis of the melting curve and agarose gel electrophoresis. To semiquantify input amounts of templates, standard curves were constructed with serial dilutions of cDNA sample from a positive control (brain tissue from lipopolysaccharide-treated rat). To standardize results, interpolated values for each sample were divided by the value of the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Primers were designed with Primer3 software (15) and checked for absence of cross-reactivity by BLAST search. The primer pairs used, product size, and positive controls are shown in Table 1.

Table 1 Real time PCR primer sequences, product size, and positive controls
Full size table

Western blotting for eNOS, iNOS, and Akt.

Liver tissue was homogenized in 50 mM HEPES, pH 7.4, 1 mM EDTA and protease inhibitor cocktail, followed by centrifugation at 15,000g for 30 min. Protein was separated by 7.5% SDS-PAGE, transferred to nitrocellulose membranes, and visualized by Enhanced Chemiluminescence System (Amersham Health). Densitometric quantification of Western blot signal intensity of autoradiograms was performed using a Molecular Dynamics P.D densitometer.

Measurement of eNOS and iNOS activity.

NOS activity was estimated by formation of nitrite/nitrate from l-arginine through the in vitro reaction of NOS, according to the method described by Kelley et al. (16). Nitrite/nitrate generated in the reaction mixture by NOS activity was measured with a Nitric Oxide Synthase Assay kit (Calbiochem, San Diego, CA). An eNOS inhibitor L-N (5)-(1-iminoethyl)-ornithine (L-NIO) at final concentration of 10 μM in the reaction mixture was used in a separate set of eNOS activity assays to confirm eNOS specific activity.

Statistical analysis.

Values are expressed as mean ± SE (SEM). Statistical analysis was performed using unpaired t test (for only two groups) and analysis of variance (for more than two groups). Statistical significance was considered at p < 0.05.

RESULTS

Age related eNOS and iNOS expression.

Ontogeny of hepatic eNOS, as determined by western blot analysis and activity (Fig. 1A and B), was similar to our previous study (9). eNOS was presented before birth, albeit at low levels. On DOL-2 eNOS protein and activity were significantly increased compared with earlier times. There was further increase until a maximum was attained at DOL-20, which persisted in the adult. NOS mRNA was determined by real time RT-PCR. Validation of the PCR and primers was performed by serial dilution of cDNA. The coefficiency of the PCR was >0.995 and the efficiency of the PCR was 100 ± 10%. eNOS mRNA showed a pattern similar to its protein expression, with the exception of a relatively lower level of expression in the adult (Fig. 1C). iNOS mRNA, protein, and activity were undetectable at all ages.

Figure 1
figure 1

Developmental expression of eNOS and iNOS. (A) Semiquantitative NOS protein assay was performed by western blot. Densitometric analysis was based on three blots and standardized to actin. Values are depicted as means ± SE expressed as a percentage of actin; n = 3 for each time point; *p < 0.05 compared with eNOS at 18 d gestation (g-18); †p < 0.05 compared with DOL-5. iNOS protein was not detectable as shown in the representative blot. (B) eNOS activity was measured with (gray) or without specific eNOS inhibitor (black) in the reaction mixture. Inhibitor was used to confirm that OD change was specific for eNOS activity. Values are means ± SE; n = 6–8 for each time point; *p < 0.05 compared with g-18; †p < 0.05 compared with DOL-5. iNOS activity was not detectable at any age. (C) Expression of eNOS mRNA as analyzed by real time PCR and standardized to GAPDH. Values are means ± SE expressed as ratio of eNOS to GAPDH; n = 3 (measured in triplicate) for each time point; *p < 0.05 compared with eNOS at 18 d. iNOS mRNA was not detectable by real time PCR at any age.

Full size image

Hepatic damage due to CCl4.

Hepatic injury, measured by serum ALT, was noted in all ages (DOL-0 to DOL-20, Fig. 2A). ALT increase was mild; about 3-fold over control values. Co-administration of CCl4 and the nonselective NOS inhibitor L-NAME increased hepatic damage compared with CCl4 alone starting on DOL-2 (Fig. 2A). Administration of L-NAME alone had no effect on ALT (16.69 ± 3.88 U/mL). Histologic examination showed typical features of CCl4 toxicity from DOL-2 to adult (Fig. 3C and D), such as ballooning degeneration with pycnotic nuclei and individual hepatocyte necrosis; fibrosis was not seen. In adults, the findings were most pronounced in zone 3 (central vein), with some involvement of zone 2. In less mature animals, histologic features were found throughout the hepatic lobule, although with a prominence in zone 3. Animals exposed to CCl4 on DOL-0 had minimal evidence of injury, limited to scattered necrotic hepatocytes (Fig. 3B).

Figure 2
figure 2

Liver injury and lipid peroxidation induced by CCl4 with or without L-NAME. Columns represent control (white), CCl4 (black), and CCl4/L-NAME (gray) groups. (A) Liver injury was measured by ALT. Values are means ± SE; n = 3 each group at each time point; *p < 0.05 compared with respective controls in each age group; †p < 0.05 compared with CCl4 treatment alone in the same age group. (B) Lipid peroxidation was measured by MDA. CCl4 treatment increased MDA generation. Co-administration of CCl4 and L-NAME had no effect on the degree of lipid peroxidation. Values are means ± SE; n = 3 for each time point; *p < 0.05 compared with respective controls.

Full size image
Figure 3
figure 3

Histologic examination. Representative photomicrographs of liver sections from control rat (A) and CCl4-treated rats: DOL-0 (B), DOL-2 (C), and adult (D). Liver section of DOL-2 represents the histology of DOL-2, 5, 10, and 20 since they have similar histopathological changes. N = necrosis; S = swollen cells; E = extramedullary hematopoiesis. Magnifications were 100× and 200×. The bar on each picture is 10 μm in length.

Full size image

Determination of lipid peroxidation.

Determination of MDA showed an increase in lipid peroxidation following administration of CCl4 (Fig. 2B). Simultaneous administration of the CCl4 and L-NAME had no effect on the degree of lipid peroxidation compared with CCl4 treatment alone.

NOS expression during CCl4 hepatotoxicity.

eNOS and iNOS mRNA was determined by real time PCR and expressed as a ratio to GAPDH. Administration of CCl4 with or without L-NAME had no effect on eNOS expression (Fig. 4A). iNOS mRNA was not detected in any animal. eNOS and iNOS proteins were determined by Western blot analysis using actin as the constitutively expressed protein. Similar to mRNA, administration of CCl4 with or without L-NAME had no effect on eNOS protein expression (Fig. 4B). iNOS protein was not detected in any animal (Fig. 4D). CCl4 administration had no effect on eNOS activity on DOL-0; however, eNOS activity increased significantly at all ages beginning on DOL-2. Co-administration of CCl4 and L-NAME had no effect on eNOS activity on DOL-0, but decreased activity below control levels in all ages from DOL-2 (Fig. 4C). iNOS activity was not detected under any conditions.

Figure 4
figure 4

Expression of eNOS in the liver of newborn rats treated with CCl4 with or without L-NAME. Columns represent control (white), CCl4 (black), and CCl4/L-NAME (gray) groups. (A) mRNA of eNOS measured by real time RT-PCR was not significantly different in between controls and CCl4-treated rats with or without L-NAME in the same age groups. eNOS mRNA was standardized to GAPDH. Values are means ± SE expressed as ratio of eNOS to GAPDH; n = 3 (measured in triplicate). (B) Western blot showed no significant difference in eNOS protein expression between controls and CCl4-treated rats with or without L-NAME in the same age groups. Values are depicted as means ± SE expressed as a percentage of actin; n = 3 for each time point. (C) eNOS activity was significantly increased in CCl4-treated rats as compared with respective control. Co-administration of L-NAME with CCl4 decreased eNOS activity below control levels. Values are means ± SE; n = 3 for each time point (measured in triplicate). (D) iNOS was not induced in CCl4-treated rats with or without L-NAME as shown in the representative blot of DOL-20 rats. *p < 0.05 compared with respective control in each age group; †p < 0.05 compared with CCl4 treatment alone.

Full size image

Posttranslational regulation of eNOS.

There is a disconnect between eNOS expression and activity in CCl4-induced hepatotoxicity. Evidence indicates that eNOS activity can be activated by posttranslational modification mediated by phosphorylation of eNOS through the PI3k/Akt signaling pathway (17,18). We examined posttranslational regulation of eNOS in our experimental model. CCl4 treatment resulted in increased phosphorylation of both Akt and eNOS (the active forms of Akt and eNOS, respectively) as determined by western blot analysis (Fig. 5A and B). Increased phosphorylation of eNOS was associated with increased eNOS activity (Fig. 6A). Treatment with wortmannin (inhibitor of PI3 K/Akt signaling) reduced eNOS activity below control values. Similarly, increased eNOS activity seen with CCl4 administration was negated by co-administration of wortmannin and CCl4 (Fig. 6A). Western blot analysis showed that wortmannin attenuated phosphorylation of Akt and eNOS both in control and CCl4-treated rats (Fig. 5A and B). Co-administration of CCl4 and wortmannin potentiated the degree of hepatic injury compared with CCl4 alone; wortmannin alone did not cause hepatic injury (Fig. 6B).

Figure 5
figure 5

Determination of phosphorylation of Akt and eNOS by western blot. There was an increase in phosphorylation of both Akt (A) and eNOS (B) in the liver of CCl4-treated DOL-20 rats. Administration of wortmannin (Wort) significantly reduced the phosphorylation both in control (Cont) and CCl4-treated animals. Values are means ± SE expressed as a percentage of actin; n = 3; *p < 0.05 compared with respective control; †p < 0.05 compared with CCl4 treatment alone. A representative blot is shown.

Full size image
Figure 6
figure 6

eNOS activity and liver function of newborn rats treated with CCl4 with or without wortmannin. (A) eNOS activity was significantly increased in CCl4-treated rats compared with controls (Cont). Administration of wortmannin (Wort) reduced eNOS activity both in control and CCl4-treated animals. Values are means ± SE; n = 3 for each time point (measured in triplicate). (B) Liver function as measured by ALT showed that wortmannin (Wort) potentiated the degree of hepatic injury compared with CCl4 alone. Administration wortmannin alone did not cause hepatic injury. Values are means ± SE; n = 3 (measured in triplicate). *p < 0.05 compared with control (Cont); †p < 0.05 compared with CCl4 treatment alone.

Full size image

DISCUSSION

The role of NOS in hepatic disease is complicated. The effect of NOS, mediated through production of NO, has variable, and sometimes, contradictory effects. NO can protect the liver by scavenging lipid radicals and inhibiting the lipid peroxidation chain reaction (13). Alternatively, hydroxyl radicals formed by the reaction of NO and superoxide anion via peroxynitrate may result in hepatotoxicity (10). NO is also important in regulating hepatic blood flow, with an indirect effect on hepatocellular injury (6). In this study, we sought to characterize the role of NOS in CCl4-induced hepatotoxicity in the developing liver.

Our data shows that administration of CCl4 leads to mild increase in serum ALT in the newborn rat consistent with previous studies (1–3). Administration of CCl4 led to an increase in hepatic eNOS activity with no detectable iNOS activity. These findings are in contrast to some studies showing induction of iNOS in adult rats following exposure to CCl4 (6,13,19), but in agreement with other studies showing exposure to CCl4 results in eNOS activation (20) without upregulation of iNOS (21,22). In this study, co-administration of CCl4 and L-NAME led to a worsening of hepatic injury, suggesting a protective role for eNOS in CCl4-induced toxicity.

Studies have shown that newborn rat is “resilient” to CCl4 toxicity (1,3,4) based on serum ALT levels and lethality compared with adult rats. Resiliency has been explained by undeveloped cytochrome P450, increased hepatocellular growth, and tissue repair mechanisms in young rats (1,4). Yet, in our study, histopathological changes were present in all animals. These data are consistent with a previous study, where livers from DOL-1 rats were minimally damaged by CCl4 compared with rats at DOL-4, 10, 14, 21, and adults (23). CCl4 is bioactivated by cytochrome P450 to free radicals. The major isozyme for the process is cytochrome P450 2E1 (24). Ontogeny studies have demonstrated that cytochrome P450 2E1 protein is not detectable in the liver of DOL-1 rat; starts to increase rapidly on DOL-3; and reaches approximately 50% of adult levels on DOL-28 (25). The ontogeny of cytochrome P450 2E1 is mirrored by hepatic damage in our and previous studies (23), where no significant histologic alterations are seen at DOL-0 and DOL-1 but similar histologic damage is seen regardless of age after DOL-2. Lower serum ALT in CCl4-treated newborn rats does not indicate less hepatic injury compared with adult rats. Rather, the lower serum ALT in newborns could be simply because of less availability of the hepatic enzyme to leak into the blood stream during CCl4 toxicity. Supporting evidence comes from a previous study showing that hepatocellular ALT content is significantly lower in newborn than in adult rats (26).

Previous studies report mixed results when assessing inhibition of NOS in CCl4-induced hepatotoxicity, particularly the effect on peroxidation. Administration of a nonselective NOS-inhibitor (L-NAME) increases lipid peroxidation following exposure to CCl4, suggesting NO plays a protective role in CCl4 intoxication through interaction with reactive oxygen species (ROS) (13). Conversely, a selective iNOS inhibitor (L-NIL) inhibited peroxidation following CCl4 administration suggesting that NO plays a deleterious role in CCl4 intoxication in adult rats (19). We measured hepatic tissue MDA following CCl4 exposure showing that, while lipid peroxidation occurs, it is not altered by co-administration of CCl4 and L-NAME despite increased hepatic injury. This data would suggest that in this model NO does not participate in lipid peroxidation through interaction with ROS or lipid peroxyl groups.

An alternative explanation is that eNOS offers protection from acute CCl4-induced toxicity by a mechanism unrelated to ROS production, such as changes in hepatic blood flow. Previous data have shown that acute CCl4 toxicity results in increased hepatic blood flow, both in the portal vein and hepatic artery (5,6). Increase in hepatic artery blood flow is mediated by increased NO production via up-regulation of NOS, resulting in reduced hepatic damage. Inhibition of NOS activity dampens hepatic artery flow exacerbating hepatic damage (6). NO has also been shown to improve hepatic microvascular perfusion through reducing neutrophil accumulation and platelet aggregation (27). Beneficial effects of increased hepatic blood flow in CCl4-induced injury may be related to improved oxygen delivery to the damaged liver (6). Inflammation results in increased arterial blood flow due to increased oxygen demand for synthesis of DNA and proteins needed for tissue repair in hepatic regeneration (28). In CCl4 toxicity, increased hepatic arterial blood flow provides greater oxygen supply to promote liver regeneration following acute hepatic injury (6). Although the exact protective mechanism afforded against CCl4 toxicity by increased hepatic artery blood flow remains to be determined, further study on hepatic hemodynamic changes in CCl4 toxicity will strengthen the current findings.

The principle control of hepatic blood flow under normal conditions is at the level of the sinusoid and is mediated by eNOS (8). In newborn rats, eNOS mRNA, protein, and activity increase in parallel reaching adult levels by 20 d of life (9). Immunohistochemical staining reveals that eNOS is present in the sinusoids at birth, albeit in low amounts and scattered across the lobule. Sinusoidal eNOS increases in intensity and distribution with age, reaching an adult pattern by 20 d of life (9). There is also data showing that the hepatic artery contributes more to sinusoidal blood flow maintenance in the immature liver than in the adult liver (7). At baseline, the hepatic artery is under the control of NOS. Supportive evidence includes vasoconstriction, and resulting increased hepatic artery resistance, following inhibition of NOS by L-NAME and partial reversal of vasoconstriction following administration of L-arginine (29). These findings may help explain why, despite paucity of eNOS in the sinusoids of the newborn rat, eNOS offers protection in CCl4 toxicity. Additionally, the protective effect of NO may be related to its anti-apoptosis activity and protection against cytokine mediated cell death (30).

In this model of hepatotoxicity, eNOS activity was increased after exposure to CCl4, without changes in mRNA and protein levels. This increase in eNOS activity might be associated with the posttranslational modification of eNOS due to phosphorylation mediated by Akt, a down stream effector in the PI3 K/Akt signaling pathway. Previous studies have shown that Akt directly phosphorylates eNOS, activating it to produce NO both in transfected COS-7 cells and cultured endothelial cells (17). Akt has no effect on iNOS or nNOS (17), making any contribution of these forms of NOS unlikely in our model. Activation of Akt is dependent on the degree of shear force stress, with direct correlation between increased stress and increased Akt phosphorylation (18). In the current study, treatment of newborn rats with CCl4 resulted in increased phosphorylation of Akt and eNOS associated with increased eNOS activity. Administration of the PI3 K inhibitor wortmannin resulted in reduction of Akt and eNOS phosphorylation both in control and CCl4-treated animals. Co-administration of CCl4 and wortmannin, led to a significant decrease in eNOS activity and increase in hepatic damage. Thus, posttranslational eNOS activation is responsible for increased eNOS activity following CCl4 exposure.

In conclusion, we have confirmed that newborn animals exposed to CCl4 have hepatic injury, which is associated with increased levels of eNOS activity. Co-administration of CCl4 and L-NAME results in worsening of hepatic damage, suggesting a protective role for eNOS in this model. We propose that the protective effect of eNOS is accomplished by its activation through the posttranslational modification via PI3 K/Akt signaling pathway.