| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Kolling Institute of Medical Research, University of Sydney, Royal North Shore Hospital (D.M., F.M., R.C.B., P.J.D.D.), St. Leonards, New South Wales 2065, Australia; Department of Intensive Care Medicine, University Hospital Gasthuisberg, Catholic University of Leuven (D.M., G.V.d.B.), B-3000 Leuven, Belgium; and Department of Clinical Pharmacology, Westmead Millennium Institute, University of Sydney, Westmead Hospital (C.L., S.C.), Westmead, New South Wales 2145, Australia
Address all correspondence and requests for reprints to: Dr. Dieter Mesotten, Department of Intensive Care Medicine, University Hospital Gasthuisberg, B-3000 Leuven, Belgium. E-mail: dieter.mesotten{at}med.kuleuven.ac.be.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Treatment of critically ill patients with human GH, generally regarded as an effective anabolic agent (5, 6, 7, 8, 9), has been shown to increase mortality due to septic shock and/or multiple organ failure (10). In the latter study GH administration was started 5–8 d from admission and for a maximum of 3 wk at a dosage of 0.1 mg/kg body weight·d (9). The explanation for the unexpected increased mortality associated with supraphysiological GH administration is still unclear. The lack of GH resistance, the modulation of the immune system, the prevention of glutamine mobilization, as well as the induction of insulin resistance and hyperglycemia by GH have all been suggested as potential mechanisms (11). Animal studies of endotoxinemia have demonstrated marked cholestatic and lethal effects of combined GH and LPS treatment (12). In view of a possible link between bile acid transport and endotoxin processing, we questioned whether GH may cause increased mortality in these previous studies by exacerbating LPS toxicity due to an impairment in endotoxin excretion.
The molecular determinants of cholestasis are a series of bile acid transporter molecules that are distributed among specific compartments within the hepatocyte (13, 14, 15, 16, 17). The bile acid transporters are generally grouped into four major categories based upon the driving force of the pathway and substrate specificity.
The Na+-dependent bile acid transport system encompasses the basolateral taurocholate transporter, Na+-taurocholate cotransporting polypeptide (Ntcp), which is responsible for the uptake of the majority of the conjugated bile acids (18). In addition, the sister gene of P-glycoprotein (Spgp), also known as the bile salt export pump, may be regarded as the apical counterpart of Ntcp and functions in an ATP-dependent manner (19).
The Na+-independent organic anion transporting polypeptides, Oatp1 (20) and Oatp2 (21), are polyspecific basolateral transporters. Whereas Oatp2 is confined to the midzonal to pericentral hepatocytes, Oatp1 is found in all lobular zones (22). These proteins have slightly different substrate preferences; for example, Oatp1 transports sulfobromophthalein and bilirubin monoglucuronide, whereas Oatp2 does not (22). Oatp4 (23) has been identified as the full-length isoform of the basolateral liver-specific transporter 1 (24). It has substrate specificity similar to that of Oatp1 and Oatp2 and is believed to make a major contribution to Na+-independent transport (25). The multispecificity of the basolateral organic anion transporters is also reflected in the transport characteristics of the apical ATP-dependent transporter, multidrug resistance-associated protein 2 (Mrp2). Also referred to as canalicular multispecific organic anion transporter (26, 27), its spectrum of transport includes glutathione conjugates, bilirubin diglucuronides, and divalent bile acids (28).
The multidrug resistance P-glycoproteins, Mdr1a and Mdr1b, initially characterized to confer resistance to chemotherapeutic drugs (29), have been grouped with multidrug resistance-associated protein, Mrp1, under the denomination escape transporters (16). The former are localized apically; the latter is confined to the basolateral membrane. Under normal conditions these transporters are expressed in hepatocytes at very low levels (30), whereas under conditions of stress, such as sepsis (31) and liver regeneration (32), they are markedly up-regulated, probably to protect the hepatocyte against cytotoxic compounds.
Mdr2 was categorized separately because of its high level of expression in normal liver (30), with minimal change of expression during situations of stress (31, 32), and because of its specific phosphatidylcholine flippase function (33, 34).
Additionally as a defensive mechanism in response to cholestasis, 6- and 7-hydroxylation of bile acids to more polar and less toxic metabolites occurs. Increased excretion of 6
-hydroxylated bile acids has been reported in patients with chronic cholestasis (35, 36). In rats, the 6ß-hydroxylated bile acid, ß-muricholic acid (ß-MCA) is increased after bile duct ligation (37, 38). Rat CYP3As are the most likely enzymes to catalyze this 6ß-hydroxylation to form ß-MCA as well as -MCA (39, 40). Recent work has pointed to the role of the nuclear pregnane X receptor, which functions as a bile acid sensor, in the induction of CYP3A and several bile acid transporters (41, 42).
Our hypothesis for this study was that GH enhances the cholestatic potential of LPS, which, in turn, might also impair the processing of bile acids and endotoxin through modulation of CYP3A and bile acid transporter expression. To study this, high dose GH was continuously infused into rats over a period of 4 d, with administration of a sublethal, cholestatic dose of LPS on the fourth day. Gene and protein expression levels of the major bile acid transporters and bile acids were measured and related to markers of serum cholestasis. Our results indicate a mechanism by which GH exacerbates cholestasis and suggest molecular mechanisms by which the hepatocyte might decrease intracellular toxicity from bile products and endotoxin.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
90 µg/100 g body weight·d). This mode and total amount of GH administration have been shown to result in greater metabolic effects during endotoxinemia (43). The remaining 40 rats were sham-operated by making an incision in the skin without implanting a pump. On the morning of d 4, food was withdrawn, and the rats were administered a single ip injection of either a nonlethal, but cholestatic, dose of LPS (5 mg/kg Escherichia coli 055B5; L-2880, Sigma-Aldrich Corp., St Louis, MO) or nonpyrogenic saline. The rats were then killed under halothane anesthesia 6 or 12 h after the LPS or saline injection. This resulted in 8 experimental groups of 10 rats each: sham-operated/saline-injected 6 h, sham-operated/LPS-injected 6 h, GH/saline-injected 6 h, GH/LPS-injected 6 h, sham-operated/saline-injected 12 h, sham-operated/LPS-injected 12 h, GH/saline-injected 12 h, GH/LPS-injected 12 h. One rat was omitted from the GH/LPS 12 h group due to faulty ip LPS injection. One animal in the sham/LPS 12 h group was left out of the serum analyses because of hemolysis caused by sample collection, but was used for analysis of gene and protein expression. Blood was collected by cardiac puncture, clotted at room temperature, and spun at 2600 x g for 20 min at 4 C to collect serum.
Serum samples were stored, protected from light, at –20 C until analysis. The liver was quickly removed and snap-frozen in liquid nitrogen for the extraction of total RNA. Tissues were stored at –80 C until analysis. All animal procedures were conducted with the approval of the Royal North Shore Hospital animal care and ethics committee.
RNA isolation and Northern blot analyses
Total RNA was isolated from freeze-clamped, whole liver tissue using the guanidine thiocyanate/cesium chloride method (44). Total RNA (20 µg) was denatured, electrophoresed on a 1% agarose/formaldehyde gel, transferred to a nylon membrane (Zetaprobe GT, Bio-Rad, Hercules, CA) by overnight capillary blotting, and baked at 80 C for 2 h in a gel dryer. Membranes were prehybridized at 42 C for 1 h in 50% formamide, 1% sodium dodecyl sulfate (SDS), 5x Denhardt’s, and 5x sodium chloride/sodium phosphate/EDTA (SSPE) with the addition of 100 µg/ml salmon sperm DNA (45). Specific cDNAs for Ntcp (GenBank accession no. M77479), Spgp (NM012690), Oatp1 (L19031), Oatp2 (U95011), Mrp2 (L49379), Mdr2 (NM012690), Mdr1a (S66618), Mdr1b (M81855), and Mrp1 (X96394) were generated by RT-PCR from liver total RNA using primers with published sequences (31, 32). A cDNA probe for Oatp4 (AJ271682) was generated by RT-PCR from total liver RNA using the following primers: forward, 5'-TGGCCTAACCTTGACCTACG-3'; and reverse, 5'-CCACAGCTGGTGACAGACC-3'. The cDNAs were cloned into pGEM-T Easy (Promega, Madison, WI) and sequenced to confirm their identity. The rat cDNA probes were labeled with [32P]deoxy-CTP by random priming (Ready-To-Go, Amersham Pharmacia Biotech, Piscataway, NJ). Hybridization was performed at 42 C for 16 h. Blots were washed in 1x standard saline citrate/0.1% SDS for 30 min at 42 C and subsequently in 0.1x standard saline citrate/0.1% SDS for 30 min at 50 C.
Northern blots were quantified using a PhosphorImager screen and ImageQuant software (Molecular Dynamics, Sunnyvale, CA). Bile acid transporter gene expression was corrected for lane to lane loading variation by expressing data as a ratio with 18S rRNA.
Preparation of liver membranes
Liver plasma membrane fractions were isolated by homogenizing 0.5 g frozen rat liver with 20 strokes of a loose Dounce in 2 ml homogenization buffer (250 mmol/liter sucrose, 0.1 mmol/liter phenylmethanesulfonylfluoride, and 10 mmol/liter Tris-HCl, pH 7.6). The homogenate was centrifuged for 5 min at 5,000 rpm. Subsequently, the supernatant was centrifuged for 1 h at 100,000 x g at 4 C. The pellet was resuspended in 500 µl buffer (300 mmol/liter sucrose and 10 mmol/liter HEPES, pH 7.5) by passage through a 22-gauge needle. Protein concentrations were determined using a bicinchoninic acid colorimetric protein assay kit (Pierce Chemical Co., Rockford, IL).
Analysis of protein levels by Western immunoblotting
Protein analysis was performed on five randomly chosen samples from each experimental group. Liver plasma membrane homogenates were suspended in reducing SDS sample buffer [62.5 mM Tris (pH 6.8), 2% SDS, 8.7% glycerol, and 100 mM dithiothreitol]. Samples (100 µg, without boiling) were separated on a 1-mM 7% Tris-Acetate NuPAGE gel (Ntcp, Spgp, Mrp1, Mrp2, and Mdr1) or a 10% bis-Tris NuPAGE gel (Oatp1, Oatp2, and CYP3A2; Invitrogen Life Technologies, Carlsbad, CA) and transferred to a Hybond enhanced chemiluminescence membrane (Amersham International, Little Chafont, UK) for Ntcp, Spgp, and CYP3A2 or a polyvinylidene difluoride membrane (Roche, Mannheim, Germany) for Oatp1, Oatp2, Mdr1, Mrp1, and Mrp2.
Equal protein loading was confirmed by Coomassie staining of gels and Ponceau S staining of the membranes after transfer. Membranes were blocked in a Tris-buffered saline solution with 5% nonfat dry milk and 0.1% Tween (Ntcp, Spgp, Mrp1, Mdr1, and CYP3A2) or with 1% casein (Oatp1, Oatp2, and Mrp2). The membranes were incubated with the primary antibodies for either 1 h at room temperature (Ntcp, 1:5,000; Spgp, 1:2,000; Oatp1, 1:500; Oatp2, 1:10,000; Mrp2, 1:3,000) or overnight at 4 C (Mdr, 1:2,500; Mrp1, 1:1,000; CYP3A2, 1:2,000). Primary antibodies against Ntcp, Spgp, Oatp1, Oatp2, and Mrp2 were provided by Drs. Bruno Stieger and Bruno Hagenbuch (University Hospital, Zurich, Switzerland), and those against Mrp1 and Mdr1 (P-glycoprotein) were obtained from Alexis Biochemicals (Lausen, Switzerland). However, the latter antibodies were not specific. Although the Mrp1 antibody cross-reacted with Mrp2, the P-glycoprotein antibody detects not only Mdr1b, but also Mdr1a and Mdr3. Antirat CYP3A2 was purchased from Daiichi Pure Chemicals (Tokyo, Japan). Immunoreactive signals were detected by a 1-h incubation with horseradish peroxidase-conjugated rabbit antimouse (1:3,000) for Mdr1 or goat antirabbit (1:5,000) for all other proteins (both from DAKO, Glostrup, Denmark), followed by chemiluminescent detection (Western Lightning, PerkinElmer Life Sciences, Boston, MA). After exposing the blots to Hyperfilm MP (Amersham International), band intensities were determined with the Ultrascan XL enhanced laser densitometer (LKB, Bromma, Sweden) and Gelscan XL 3.0 software (Amersham Pharmacia Biotech, Bromma, Sweden).
Serum analyses
Serum total bilirubin, conjugated bilirubin, alanine aminotransferase (ALT), aspartate aminotransferase (AST), and 
-glutamyltransferase (-GT) were determined by routine clinical assays using commercial kits on an automated analyzer (Roche Hitachi pModular).
All samples in this study met the criteria of the Roche serum indexes for absence of hemolysis and lipemia. Total bile acids were determined by measuring total 3-hydroxy bile acids in an enzymatic colorimetric assay (46). Additionally, individual bile acids were quantified by HPLC/mass spectrometry. Authentic bile acid standards were purchased from Sigma-Aldrich Corp., except for hyodeoxycholic acid (HDCA), -MCA, and ß-MCA, which were purchased from Steraloids (Newport, RI); taurocholic acid, which was from Calbiochem (San Diego, CA); and the deuterated assay internal standard (2,2,4,4-d4 cholic acid), which was from C/D/N Isotopes (Québec, Canada). One hundred microliters of serum were diluted with 100 µl PBS to which 41 ng deuterated assay internal standard were added. Solid phase extraction was carried out using Oasis HLB cartridges (Waters Corp., Milford, MA) according to the manufacturer’s directions. Typical recoveries of extracted bile acids exceed 85%. Chromatographic separations were carried out with a Waters 2695 pump equipped with an autoinjector. The analytes were separated on a Waters X-Terra MS C18 column (3.5 µm, 2.1 x 150 mm). The mobile phase consisted of solvent A (water), solvent B (methanol), and solvent C (100 mM ammonium acetate, pH 4.5) delivered as a gradient: 0–15 min solvent B, 67%; 15–25 min solvent B, 67–85%; and 25–35 min solvent B, 85–67%, with solvent C constant at 10% at a flow rate of 0.2 ml/min. The HPLC was coupled with a Waters ZQ quadrapole mass detector via an electrospray ionization interface operating in the negative ion mode.
Quantitative determination of bile acids was performed by time-scheduled single ion recordings using [M-H]- ions. We determined that the following tune parameters are optimal for bile acid detection: capillary voltage, 3 kV; cone voltage, 40 V; extractor voltage, 5 V; RF lens, 0.3 V. Source temperature was 100 C, and desolvation temperature was 300 C. Desolvation gas flow was set at 350 liters/h, and cone gas flow rate was 60 liters/h.
Serum IL-6 levels were measured using a rat specific ELISA kit (Cytoscreen, BioSource International, Camarillo, CA). Serum endotoxin activity was analyzed by AMS Laboratories Pty. Ltd. (Rockdale, Australia) using a quantitative kinetic assay (Kinetic-QCL, BioWhittaker, Walkersville, MD). In brief, endotoxin activates the Limulus amebocyte lysate coagulation cascade and consequently releases the added chromogenic substrate (para-nitroaniline), imparting a yellow color to the solution, which is quantitated by spectrophotometer (47).
Statistics
Statistical analyses were performed using StatView version 5.0 (SAS Institute, Inc., Cary, NC). All data are presented as the mean ± SEM. Serum IL-6 data were normalized by log transformation before analysis. Differences among treatment groups were analyzed by one-way ANOVA with Fisher’s protected least significant difference test. Due to the heterogeneity of the endotoxin serum activity and the opposing effects of GH over time, the results were square root-transformed and analyzed by t test. P < 0.05 was considered statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Without LPS treatment, serum levels of IL-6 were undetectable. At 6 and 12 h after LPS injection, IL-6 levels were, respectively, 7- and 23-fold higher in the GH-treated rats than in the sham group. To assess whether the increased inflammatory status of the GH-primed, LPS-treated animals was also reflected by increased parenchymal liver damage, AST and ALT serum levels were measured (Fig. 1, B and C). LPS treatment increased levels of AST in the sham-operated group at 6 h and in GH-treated animals at 6 and 12 h. Twelve hours after LPS treatment, significantly increased serum AST levels were seen in the GH-treated animals compared with the sham-operated group. ALT was only significantly increased in the GH/LPS groups, which was evident at both 6 and 12 h.
|
, A–D). Continuous high dose GH did not change serum total bilirubin, serum total bile acids, or serum -GT compared with levels in the sham-operated animals. In all non-LPS-treated animals, -GT levels remained below the limit of detection, whereas among the LPS-treated animals, -GT levels were only significantly elevated at 12 h, with 4.5-fold higher -GT levels in the GH-treated animals. Total bilirubin levels were elevated by LPS only in the GH treatment groups (2.5- and 3-fold the levels in saline-injected animals at 6 and 12 h, respectively).
|
The analysis of the individual bile acids reflected the changes in total serum bile acids, with LPS or GH treatment on their own not affecting any of the individual bile acids measured (Table 1). Serum levels of cholic acid (CA), the predominant bile acid, were 3-fold increased in GH-primed LPS-treated animals. This was accompanied by an even greater increase in its conjugated metabolite taurocholic acid in GH/LPS-treated animals at 6 h. Serum concentrations of deoxycholic acid, originating from CA 7-dehydroxylation by intestinal bacteria, and its taurine conjugate (taurodeoxycholic acid) showed a similar increase in the GH/LPS treatment group. The level of chenodeoxycholic acid (CDCA), the other primary bile acid, was about 10-fold lower under all treatment conditions than that of CA. Neither CDCA nor its 7-dehydroxylated metabolite, lithocholic acid, was affected by treatment schedule. However, tauro-CDCA levels were increased in GH/LPS-treated animals compared with control rats. Glycine conjugates of bile acids did not differ between the treatment groups (data not shown). Serum concentrations of -MCA and particularly ß-MCA, the 6ß-hydroxylated bile acids, were markedly increased in the GH-primed, LPS-treated animals at 6 h.
|
).
|
). Serum endotoxin activity remained relatively stable from 6–12 h post LPS injection. Endotoxin activity in the sera of non-LPS-injected animals was below the level of detection.
Gene and protein expression levels of the major transporters
As serum markers of cholestasis were significantly altered in GH-primed animals relative to saline controls, and altered gene and protein expression of the bile acid transporters has been linked to impaired bile flow (31), we examined steady state hepatic mRNA and protein levels of the four groups of bile acid transporters.
Na+-dependent bile acid transporter system.
The basolateral transporter Ntcp mRNA was expressed as a single 1.7-kb band on Northern blot analysis (Fig. 4A). When averaged for all LPS-treated animals, Ntcp mRNA levels were markedly suppressed by LPS to approximately 25% of those in the sham/saline 6 h group, which was used as the reference group in all of the gene and protein expression studies (Fig. 4B). There was no statistically significant incremental effect of GH with LPS beyond that of LPS alone despite an apparently 60% suppressive effect at 12 h. In its protein expression, Ntcp (51 kDa) was significantly suppressed by LPS only at 12 h (Fig. 4, C and D). Ntcp gene and protein expression levels were strongly correlated (r = 0.57; P = 0.0008). By comparison, mRNA levels of the apically located Spgp (5.2 kb) showed relatively small treatment effects and were only suppressed upon LPS injection in the sham-operated animals to 75% of the level in the reference group (Fig. 4, E and F). Spgp was expressed as a 160-kDa protein (Fig. 4G). The relatively small treatment effects on Spgp gene expression were reflected by the lack of changes in protein levels (Fig. 4H).
|
, the basolateral transporter Oatp1 was expressed as two transcripts of 4.3 and 1.4 kb. Oatp2 and Oatp4 were expressed as single transcripts of 4 and 2.8 kb, respectively (Fig. 5, C and E), whereas the apically located transporter Mrp2 was expressed as several transcripts of 8.5, 6.5, and 5.5 kb (Fig. 5G). GH alone significantly decreased gene expression levels of Oatp2 at 6 and 12 h, but also increased Oatp4 mRNA levels to 160% of that in saline control animals at 12 h. All four transporters showed down-regulation of mRNA levels after LPS treatment (Fig. 5, B, D, F, and H). However, an additional 20% suppressive effect of GH on LPS administration was observed in Oatp1 mRNA levels at 12 h (Fig. 5B).
|
, A and C), whereas the apically located transporter, Mrp2, had a molecular weight of 180 kDa (Fig. 6E). Oatp1 protein levels changed similarly with treatment compared with mRNA levels (r = 0.49; P = 0.004; Fig. 6B). However, the statistically significant incremental effect of GH with LPS beyond that of LPS alone on Oatp1 gene expression at 12 h was absent in protein expression. At 6 h, GH treatment increased Oatp2 protein levels, contrary to its effects on Oatp2 gene expression (Fig. 6D). As Oatp2 protein levels increased over time in the sham/saline group, a separate analysis at 12 h was performed. At the latter time point Oatp2 protein levels were significantly decreased in LPS-treated animals and correlated significantly with Oatp2 mRNA levels (r = 0.67; P = 0.005). Compared with the sham/saline 6 h reference group, LPS did not suppress Mrp2 (Fig. 6F). Mrp2 protein levels were significantly increased in GH-treated animals at 12 h. Unexpectedly, Mrp2 levels were 4-fold higher in the GH/LPS-treated animals at 6 h. There was no correlation between Mrp2 gene and protein expression levels.
|
). Mdr1b mRNA (single transcript of 4 kb) appeared increased by LPS alone at 6 and 12 h, but this was not significant by ANOVA. However, there was a very strong effect of combined GH and LPS treatment, with 7- and 22-fold inductions in the GH/LPS-treated animals at 6 and 12 h, respectively (Fig. 7, A and B). However, using an antibody that detects Mdr1a, Mdr1b, and Mdr3, no changes in Mdr1 protein (150 kDa) could be detected, possibly due to the lack of specificity of the antibody (Fig. 7, C and D). Mrp1 mRNA levels (single transcript of 6 kb) were at the limit of detection by Northern blot analysis at the 6 h point in all groups (Fig. 7E). At 12 h after LPS injection, there was a significant induction of Mrp1 mRNA in the sham-operated animals, whereas the GH-treated rats revealed a 2-fold higher induction (Fig. 7, E and F). Mrp1 protein levels (185 kDa) were 4- and 2.5 fold induced upon LPS injection in the GH-treated animals only at 6 and 12 h, respectively (Fig. 7, G and H). GH treatment on its own increased Mrp1 at 12 h. However, there was no significant correlation between Mrp1 gene and protein expression due to the different temporal effect of the LPS treatment. Mdr1a mRNA could not be detected in any of the treatment groups (data not shown).
|
Protein expression of CYP3A2
GH treatment suppressed hepatic CYP3A2, expressed as a 50-kDa protein, to levels 50% of those in the sham/saline treatment group (Fig. 8). Although there was no decrease in CYP3A2 protein upon LPS treatment, the combination of GH and LPS resulted in an incremental effect of LPS with GH beyond that of GH alone, resulting in a 80% fall in protein expression.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our model of endotoxinemia, GH treatment led to a more severe inflammatory response after LPS injection, as demonstrated by markedly higher serum IL-6 levels in the GH-primed rats. Among the cytokines, IL-6 has been described as having the best correlation with mortality from septic shock (50). Importantly, GH has been shown not to increase IL-6 release from isolated human monocytes after endotoxin stimulation (51), suggesting that the primary mechanism for GH toxicity is not through immunomodulation.
It has been reported that cholestasis could be the trigger to an exacerbated inflammatory response and endotoxinemia (52, 53). In our study GH priming before ip LPS administration clearly induced changes consistent with impaired bile acid transport. The augmented cholestatic response to endotoxin after GH priming was clearly demonstrated by increased serum total and conjugated bilirubin, total bile acid, and -GT levels. To identify a mechanism by which GH might augment the cholestatic effects of endotoxin, we studied changes in the hepatic bile acid transporters and related these to alterations in serum markers of cholestasis.
Endotoxin-induced cholestasis has been attributed in part to down-regulation of canalicular membrane proteins Mrp2 (54, 55) and Spgp (31) and probably a secondary effect on the basolateral transporters Ntcp (56, 57, 58) and the Oatps (31).
Although we confirmed a marked LPS effect on Ntcp and the Oatps, with only a limited effect on Spgp, changes in these proteins did not point to the mechanism of GH-enhanced cholestasis, as GH did not significantly exacerbate these LPS-induced changes. However, we found that GH priming strongly enhanced the previously described (31) up-regulation of the escape transporter Mrp1 by endotoxin. A marked up-regulation of Mdr1b transcription was also noted. Although the exact transport properties of Mdr1b for endogenous substrates remains to be elucidated, Mdr1b is of prime importance in the protection of the hepatocytes against toxic stress (59, 60). In contrast to the study of Cao et al. (61), who described GH induction of Ntcp gene expression in isolated rat hepatocytes, we observed no inductive effect of GH treatment on this transporter in vivo. Although the explanation for this disparity is unclear, the difference in response may be related to the rapid dedifferentiation that occurs in hepatocytes in primary culture (62, 63).
Under conditions of elevated serum conjugated bilirubin, as in the GH-primed, endotoxin-treated animals, the basolateral membrane of the hepatocyte appears to be of prime importance. The basolateral transporters, Oatp1 and Oatp4, may be involved in the uptake of bilirubin glucuronides from the blood into the hepatocyte (17, 64). The uptake of unconjugated bilirubin into the hepatocyte may remain unaffected, because this process probably works through diffusion (65). We found that the ratio of conjugated to unconjugated bilirubin remained unaltered in our animals, suggesting that the mechanism of conjugation was unaffected, as shown in previous studies (66). Thus, unconjugated bilirubin may still be processed in the hepatocyte, and subsequently these mono- and diglucuronide bilirubin conjugates could be exported back to the liver sinus, and thus into the blood, by Mrp1, which is markedly overexpressed in GH-primed, endotoxin-treated rats.
The significance of the basolateral transporters during cholestasis has only recently been appreciated with the recognition of Mrp1 and notably its homologs, Mrp3 (67) and Mrp4 (68), as significant backflow transporters for conjugated bilirubin and bile acids (69, 70, 71). The major changes at the basolateral site are reflected in the relative preservation of the apical transporters (72) during cholestasis, as was shown in our study. In our data the assessment of canalicular transporter expression was not always clear-cut. The poor correlation between Mrp2 mRNA and membrane-bound protein levels is probably posttranslational in origin and may be attributed to the rapid redistribution of Mrp2 to an intracellular subapical compartment (55, 73), which can only be appreciated by microscopic immunolocalization. The dramatic increase in Mdr1b gene expression was also not reflected in elevated Mdr1 protein levels. This may be explained by the cross-reactivity of the polyclonal P-glycoprotein antibody for Mdr1a, Mdr1b, and Mdr3.
The temporal effect of GH priming was further reflected by changes in serum endotoxin levels in LPS-treated animals. Although at 6 h there was the trend toward worsening endotoxinemia (200% induction), at 12 h this was reversed to an 80% fall in serum endotoxin activity. A potential mechanism for GH toxicity may be by aggravating cholestasis-induced intrahepatic accumulation of bile acids through impaired clearance in the bile (74, 75). Concomitantly, because of the altered bile flow, Kupffer cells might be exposed to higher levels of endotoxin, thereby contributing to the elevated IL-6 production. An alternative explanation may be that endotoxin activity is increasingly suppressed by high levels of serum bile acids and bilirubin (47). However, whether the suppressive activity of serum bile acids on endotoxin activity is protective (76) cannot be determined from our study.
Another important contributor to the enhanced cholestatic potential of LPS in GH-primed rats could have been the suppression of CYP3A-mediated detoxification of bile acids. In rats, the male-predominant CYP3A subfamily members, CYP3A18 (77) and CYP3A2 (78), are suppressed by continuous GH treatment, as was confirmed for CYP3A2 in our study. LPS treatment alone did not decrease CYP3A2 protein expression, in contrast to other studies in which higher doses of LPS were administered, and animals were studied over an extended time course (79, 80).
Nevertheless, LPS treatment significantly enhanced the suppression of CYP3A2 by GH. To unravel the physiological impact of GH/LPS treatment on the conjugating and hydroxylating capacity of the liver for bile acids, individual bile acids were quantitated in serum by mass spectrometry. Similar to bilirubin glucuronidation, taurine conjugation of the bile acids remained intact. Moreover, the 10-fold elevation of tauro-CA compared with unconjugated CA could point to impaired basolateral uptake of conjugated bile acids, consistent with diminished expression of Ntcp and Oatp1. The second pathway for the detoxification of hydrophobic bile acids is hydroxylation. A highly significant increase in the trihydroxy bile acid metabolites -MCA and ß-MCA was observed. Interestingly, the increase in bile acid 6ß-hydroxylation, a protective detoxification mechanism (81), remained intact despite the suppression of CYP3A2. As the male-predominant CYP3A subfamily members, CYP3A2 and CYP3A18, are all likely to be down-regulated in this model, it is possible that the female-predominant CYP3A9 (77, 78) or an as yet undescribed enzyme could catalyze this reaction. Continuous GH infusion is known to induce sex-independent hepatic metabolic feminization (82, 83). However, feminization of the CYP3A expression pattern did not seem to affect the bile acid-detoxifying capacity of the liver. However, it is important to note that the impact of hepatic feminization in the enhanced endotoxin-induced cholestasis by GH treatment through other metabolic pathways could not be excluded.
In conclusion, in this model of endotoxinemia (Fig. 9) we have shown that GH greatly exacerbates the cholestatic potential of LPS in conjunction with preserved bile acid detoxification and sustained elevation of serum bilirubin, bile acid, and IL-6 levels. The aggravated liver damage, increased cholestasis, and inflammation contrasted with a drop in serum endotoxin activity, which may suggest toxicity of the bile acids or local endotoxin-stimulated activity of hepatic monocytes. The incremental effect of prolonged GH exposure on the elevated serum bilirubin and bile acid levels involves modulation of basolateral transporters, which suggests impaired hepatic uptake and, more importantly, increased Mrp1-mediated basolateral efflux of their substrates into the circulation.
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: ALT, Alanine aminotransferase; AST, aspartate aminotransferase; CA, cholic acid; CDCA, chenodeoxycholic acid; -GT, -glutamyltransferase; HDCA, hyodeoxycholic acid; LPS, lipopolysaccharide; MCA, muricholic acid; Mdr, multidrug resistance P-glycoprotein; Mrp, multidrug resistance-associated protein; Ntcp, Na+-taurocholate cotransporting polypeptide; Oatp, organic anion transporting polypeptide; SDS, sodium dodecyl sulfate; Spgp, sister gene of P-glycoprotein; UDCA, ursodeoxycholic acid.
Received January 28, 2003.
Accepted for publication May 29, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-hydroxylation of bile acids in man. Biochem J 154:507–516[Medline]
-Hydroxylation of taurochenodeoxycholic acid and lithocholic acid by CYP3A4 in human liver microsomes. Biochim Biophys Acta 1438:47–54[Medline]
, interleukin-6, and interferon-) release from activated peripheral blood mononuclear cells or after minimal to moderate surgical stress. J Clin Endocrinol Metab 85:3383–3390 and IL-6 and mortality after Escherichia coli endotoxemia. Am J Respir Crit Care Med 157:1550–1558
, P < 0.01 vs. sham/LPS at same time point; #, P < 0.02 vs. 6 h point in each treatment group. In non-LPS-treated animals IL-6 was not detectable, i.e. below the lowest standard of the ELISA. B and C, Serum levels of AST (B) and ALT (C) are shown. Data are the mean ± SEM. *, P < 0.05 vs. sham/saline 6 h; 
