| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Section of Leukocyte Biology,2Critical Care Section,3.S. Department of Agriculture, Agricultural Research Service Children's Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, Houston, Texas 77030
Submitted 25 March 2003 ; accepted in final form 5 June 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
50% in portal blood samples of mice and rats but was not detectable in systemic blood. This low-grade portal endotoxemia was associated with hepatic injury in rats and C57BL/6 mice as indicated by inflammation and elevated serum transaminase activities. Blood levels of the cytokine TNF-
were increased by 50% in C57BL/6 mice; no significant elevation of this cytokine was detected in rats. Messenger RNA levels of the acute-phase proteins serum amyloid A, haptoglobin, and lipopolysaccharide binding protein were significantly enhanced after 3 wk of HU in endotoxin-sensitive rodents. In contrast, no histological changes or significant increases in serum enzyme activity were detected after HU in C3H/HEJ mice despite portal endotoxin levels of 222 ± 83.4 pg/ml. At the 3-wk time point, expression of acute-phase proteins was not elevated in C3H/HEJ mice; however, expression after 4 wk of HU was similar to endotoxin-sensitive rodents. In conclusion, these findings indicate that HU induced mild portal endotoxemia, which contributed to the observed hepatic injury in endotoxin-sensitive rodents. endotoxin; liver; hindlimb suspension
30-70% above ground-based controls (1, 2, 36); accumulation of fat droplets in hepatocytes was also noted (29). These findings were later supported by data collected from rats flown aboard the Cosmos 2044 and 1887 as well as the Spacelab 3 missions (29, 30). In addition to these biochemical alterations, increases in serum aspartate aminotransferase (AST) and alanine aminotransferase (ALT) activity were reported in rats after the COSMOS 2044 and Spacelab 3 missions. Complicating factors such as small sample sizes and operational constraints have made in-flight experiments difficult; therefore, animal models of simulated microgravity have been developed. Antiorthostatic bed rest in humans and hindlimb unloading (HU) in rodents are two of the most commonly used models. In the HU rat model, a castlike apparatus is fashioned on the tail and animals are suspended above the cage floor at heights sufficient to prevent weight bearing on the hindlimbs. This model mimics many of the physiological alterations caused by actual spaceflight including muscle atrophy, bone demineralization, fluid shifts, and depressed cellular immunity (15, 16, 28, 49, 50). In addition, HU reportedly contributes to the altered capacity of the liver and spleen to resist bacterial infection, a phenomenon most likely mediated by activated macrophages. HU of mice did not influence bacterial growth if they were infected at the start of suspension; however, HU of mice for 2, 4, or 7 days before infection resulted in an enhanced ability to eliminate the infection. Yet this enhanced resistance was accompanied by an inability to generate long-term protective immunological memory to the challenge organism (33). The mechanisms underlying macrophage activation and hepatic alterations have not yet been identified. Gut motility is believed to be slower during exposure to microgravity, which may result in the translocation of bacteria and bacterial products from the gut lumen into the systemic circulation. Moreover, previous findings in this laboratory suggest that prolonged HU elevated portal vein levels of endotoxin, a polymer in the outer membrane of gram-negative bacteria found predominantly in the ileum and colon.
It has long been hypothesized that endotoxemia plays a role in liver pathogenesis (38). In support of this hypothesis, several studies in human alcoholics and rodent alcohol models have demonstrated a direct correlation between blood endotoxin levels and the severity of liver injury (21, 35). Treatment of rats with nonabsorbable antibiotics that selectively destroyed gram-negative bacteria and endotoxin blunted the hypermetabolic state due to acute ethanol administration as well as early alcoholic hepatitis (3, 42). Although these findings provide evidence that endotoxin contributes to liver injury, there is still no direct evidence that chronic release of endotoxin into the portal vein, in the absence of ethanol intoxication, is sufficient to cause injury. Accordingly, the present study was designed to document the occurrence of portal endotoxemia in rodent models of HU and to determine whether there is associated liver inflammation and injury. Rats were exposed to HU for up to 4 wk. Initial studies indicated that portal, but not systemic, endotoxemia was present and correlated with increased markers of liver injury. Because rats on the HU protocol displayed enhanced toxicity to antibiotics commonly used to destroy gram-negative organisms and endotoxin, subsequent experiments were performed using mice to determine whether the observed injury was endotoxin dependent. The mouse strains used in these experiments included endotoxin-sensitive C57BL/6 and the endotoxin-resistant C3H/HEJ mouse strains.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
Endotoxin measurement. After 4 wk, heparinized blood samples were drawn from the portal vein or vena cava. An improved method of detection developed previously was used to measure endotoxin (42). Briefly, samples were centrifuged at 50 g for 10 min, and platelet-rich plasma (PRP) was stored at -80°C. Plasma samples were diluted 1:10 and heated to 75°C for 10 min to denature proteins that interfere with the assay. Subsequently, endotoxin was measured by use of a kinetic test with a chromogenic substrate based on the Limulus amebocyte lysate assay (BioWhittaker). The concentration of endotoxin was calculated from standards prepared for each assay in plasma from untreated animals. To prevent contamination by exogenous endotoxin, strict nonpyrogenic technique was used for sample collection and the assay procedure.
Histopathology. At the end of the experiment, a small section of liver was dissected and preserved in zinc-buffered formalin. Subsequently, tissues were processed for routine hematoxylin and eosin staining to assess general architecture and injury. Hepatic neutrophil content was determined by immunohistochemical staining with an anti-neutrophil antibody clone 7/4 (serotec). The number of positively stained neutrophils in a 70-mm2 area was counted.
Enzyme activity. Immediately before death, blood samples were collected from the inferior vena cava and serum stored at -80°C. ALT and AST activities were measured as markers of injury according to standard enzymatic assay procedures (8).
Tumor necrosis factor-. Measurement of tumor necrosis factor- (TNF-) was performed in blood samples of control and HU rodents using the cytoscreen immunoassay kit (Bio-source International, Camarillo, CA).
Cytochrome P-450 ELISA. Microsomal fractions were isolated from liver homogenates by the calcium chloride precipitation method (12). Total protein content of the microsomal fraction was determined by using the Bradford method (9). Subsequently, protein levels of cytochrome P-450 2B and 3A were measured at 37°C on microplates according to a standard ELISA kit procedure (Amersham Biosciences, Buckinghamshire, UK).
Molecular probes. A specific cDNA probe for the acute-phase reactant serum amyloid A (SAA) was prepared by RT-PCR using the following primers: sense primer, 5'-TATGATGCTGCYMAAAGGGG-3'; antisense primer, 5'-CTCAGACAAATACTTCCATG-3'. Reverse-transcription protocols were performed with 5 µg of total RNA. Aliquots of the reverse-transcription reaction were amplified by using 5 U Taq DNA polymerase (Promega, Madison, WI) for 30 cycles at 93°C for 1 min, 55°C for 2 min, and 72°C for 3 min. The resulting 222-base pair fragment was purified and cloned into the pGEM-T easy vector (Promega) before sequencing. The cDNA clones for haptoglobin and lipopolysaccharide binding protein (LBP) were obtained commercially (ATCC, Manassas, VA).
RNA isolation. Livers obtained from control and HU animals were immediately frozen in liquid nitrogen. Total RNA was extracted from frozen liver via the acid guanidium thiocyanate-phenol-chloroform method (11). Subsequently, 10 µg of total RNA containing ethidium bromide was electrophoresed on 1% agarose gels containing formaldehyde and then transferred to a nylon membrane (Gene Screen Plus; New England Nuclear, Boston, MA) by standard procedures.
Northern hybridization. Membranes were hybridized in QuikHyb (Stratagene, La Jolla, CA) at 68°C for 2 h with 1 x 106 dpm random hexamer 32P-labeled cDNA probes, and filters were processed as described previously (44). Quantitation of the Northern hybridization results was performed by densitometry. Density was normalized to the intensity of the 18S ribosomal RNA subunit as described previously (17).
Data analysis. Statistical analysis of the data presented here was performed by using Student's t-test or two-way ANOVA with Tukey's multiple comparisons test where appropriate; P < 0.05 was selected as the level of significance. For each parameter tested, data are expressed as means ± SE.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
Hepatic injury due to HU. Hematoxylin and eosin-stained liver sections collected from control animals at the time of death appeared normal (Fig. 2, A-C). After 4 wk of HU in rats and C57BL/6 mice, fat accumulation was found consistently in the cytoplasm of hepatocytes. There were minimal neutrophils present in livers of control animals (0.5 ± 0.3 neutrophils/70 mm2 area). Neutrophil content increased to 77.3 ± 15.6 cells/70 mm2 area in C57BL/6 mice exposed to HU; there was a similar response in the rat. The observed steatosis and inflammation were primarily confined to areas around the central vein (Fig. 2, D and E). In contrast, these histological changes were not detectable in liver sections from C3H/HEJ mice (Fig. 2F). Consistent with histological findings, activity of the injury marker AST increased significantly after HU in endotoxin-sensitive rodents but not in C3H/HEJ mice (Fig. 3). The liver-specific marker ALT was increased after HU in C57BL/6 mice but was not statistically different from controls in rats or C3H/HEJ mice. Because endotoxin is known to stimulate the release of cytokines from macrophages and inflammatory cells, TNF- was measured in serum samples collected at the time of death. In rats, serum TNF- levels were not increased after HU. Serum TNF- was increased significantly by approximately threefold in C57BL/6 mice but not in C3H/HEJ mice (Fig. 4).
|
|
|
Evidence of an acute-phase response. It is well known that endotoxemia can stimulate the expression of acute-phase reactants (7, 22); therefore, mRNA expression of the representative reactants SAA, haptoglobin, and LBP were analyzed. Haptoglobin and LBP were 1.06 ± 0.4 and 1.60 ± 0.05, respectively, in control rats (n = 4). Expression of these proteins was elevated significantly by 50 and 60%, respectively, after 3 wk of HU (n = 5; P < 0.05); rats do not express SAA. Figure 5 summarizes the mRNA expression of these acute-phase reactants in mice. Minimal SAA expression was detected in C57BL/6 control mice, and similar results were obtained in C3H/HEJ control mice. Clearly, mRNA levels of this acute-phase reactant were elevated severalfold over control levels in C57BL/6 mice after 3 or 4 wk of HU with mean relative density values of 99.7 ± 43.1 and 75.0 ± 18.3, respectively (Fig. 5A). In contrast, expression of SAA was only enhanced after 4 wk of HU in C3H/HEJ mice (84.3 ± 45.8); results obtained at the 3-wk time point (1.0 ± 0.1) were similar to control levels but were significantly lower than expression levels of this acute-phase protein in C57BL/6 mice (99.7 ± 43.1; P < 0.05 using two-way ANOVA and Tukey's multiple-comparison test) (Fig. 5B). In C57BL/6 mice, mRNA levels of haptoglobin were significantly increased after 3 wk (4.1 ± 2.1) or 4 wk (2.3 ± 0.4) of HU; LBP was also increased to 1.7 ± 0.1 and 2.0 ± 0.1 after 3 or 4 wk, respectively (P < 0.05 using two-way ANOVA and Tukey's multiple-comparison test). Comparable to C57BL/6 mice, haptoglobin and LBP were increased by approximately twofold in C3H/HEJ mice after HU at each of the time points measured.
|
Cytochrome P-450 2B and 3A expression. Protein expression of CYP 2B and CYP 3A, two representative isoforms of cytochrome P-450, was measured by ELISA to determine whether HU-induced liver injury altered the hepatic metabolic capacity. Protein expression of CYP 2B and CYP 3A was 231.0 ± 85 and 29.6 ± 1.7 µg/mg microsomal protein, respectively, in control rats. After 4 wk of HU, protein expression of these isoforms was diminished by 30-70% (Fig. 6). In mice, on the other hand, protein expression of these enzymes was not affected by HU (data not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
Data presented here indicated that endotoxin levels in portal blood samples were increased by >50% after HU in rats and mice (Fig. 1). Endotoxin was not detected in systemic blood samples, suggesting that endotoxin is cleared effectively by the liver. Because blood leaving the gut empties directly into the portal vein, the liver is exposed to gut-derived endotoxin. Kupffer cells, the resident hepatic macrophages, are primarily responsible for endotoxin clearance (51). Kupffer cell activation by endotoxin requires CD14 and Toll-like receptors on the cell surface, and the presence of LBP (52), a class I acute-phase protein upregulated in response to systemic endotoxemia. It has been shown that LBP binds to the lipid A portion of endotoxin and facilitates binding to CD14 (45, 46). Intracellular signaling events in response to CD14/Toll-4 interaction with endotoxin result in the release of a variety of proinflammatory mediators including free radicals and cytokines such as TNF- that induce hepatic injury and dysfunction (13). The consequence of HU-induced portal endotoxemia on the liver was addressed in endotoxin-responsive C57BL/6 mice, and results were compared with the response of C3H/HEJ mice that do not respond to endotoxin owing to a mutation in the Toll-like receptor 4. Consistent with portal endotoxemia, evidence of hepatic injury was observed in endotoxin-sensitive rodents (Figs. 2 and 3). Although endotoxin values were elevated significantly in endotoxin-resistant mice, no histological changes or increases in enzyme markers of injury were detected after 4 wk of HU. As expected, the elevations in serum levels of TNF- were blunted in C3H/HEJ mice. These findings support the hypothesis that endotoxin plays a causal role in HU-induced liver injury.
The mechanism underlying portal endotoxemia has not been identified. Vascular changes including increased blood flow to the gastrointestinal tract have been reported after HU (14, 25). Gut motility is believed to be slower during exposure to microgravity (5). These alterations may result in the translocation of bacteria and bacterial products such as endotoxin from the gut lumen into the circulation.
As a mechanism of reconstituting the homeostatic state after trauma and infection, glucocorticoids and macrophage-derived cytokines initiate the synthesis and secretion of acute-phase reactants; a major portion of these reactants is produced by the liver. Endotoxemia is a potent inducer of the acute-phase response (7), a phenomenon most likely mediated by Kupffer cell-derived cytokines. In support of HU as a model of endotoxemia, expression of the class I acute-phase proteins haptoglobin, SAA, and LBP was enhanced in C57BL/6 mice after just 3 wk of HU (Fig. 5). In C3H/HEJ mice, SAA expression was only elevated at the 4-wk time point. This observed delay in elevated expression of SAA in endotoxin-resistant mice demonstrates that endotoxin plays a role in the early acute-phase response after HU. However, additional factors most likely contribute to the resulting acute-phase response at later time points.
HU-induced injury results in altered hepatic function. Proinflammatory cytokines such as TNF- are known to induce hepatic injury and dysfunction (13). In fact, TNF- has been shown to play an important role in the mechanisms underlying liver injury and can directly reduce protein levels of cytochrome P-450 (34, 39). As reported here, TNF- was increased in C57BL/6 mice after HU, and there was a trend toward increased serum levels of this cytokine in HU rats compared with nonsuspended controls (Fig. 4). It is likely that increased circulating TNF- contributed to the development of injury and hepatic malfunction due to HU.
In rats, chronic HU decreased hepatic protein expression of CYP 2B and there was also a trend toward decreased CYP 3A expression (Fig. 6). HU did not affect these enzymes in mice (data not shown). Previous findings on the hepatic metabolic capacity after spaceflight or HU are equivocal. For example, protein levels of the enzymes responsible for lipid and glycogen metabolism were lower in flight rats compared with ground controls, suggesting that the accumulation of hepatic lipid and glycogen occurred because of decreased mobilization and not increased synthesis (29). In addition, protein levels of cytochrome P-450 were also decreased during exposure to microgravity (29, 30). On the other hand, a ground-based study by Brunner et al. (10) reported an increase in total body clearance of antipyrine, a compound metabolized by several P450 isoforms. Rats in the Brunner study lost weight, suggesting that dietary intake was less than adequate. It is well known that food restriction stimulates cytochrome P-450 activity (40), which may explain the stimulation in antipyrine metabolism observed in the previous study. In the present study, HU decreased hepatic protein expression of CYP 2B and 3A in rats (Fig. 6) but had no effect on these enzymes in mice (data not shown). Food consumption during the 4-wk study was 23.6 ± 1.8 g/day in control rats and 20.5 ± 2.6 g/day in tail-suspended rats. Although food intake was similar between the two groups, body weight was significantly lower in suspended rats (208.2 ± 3.0 g) compared with controls (268.7 ± 1.9) at the end of the study; however, no decrements in body weight were observed. Thus alterations in P450 expression cannot be explained by dietary intake.
Findings reported here may provide an explanation for the observed enhancement in the hepatic immune function during HU. For example, Miller and Sonnenfeld (31, 33) reported an increased resistance to infection with Listeria monocytogenes in HU mice. Listeria monocytogenes is a gram-positive intracellular pathogen that is cleared primarily by the liver and spleen (27). Subsequently, there is a rapid influx of neutrophils into these organs, and previous studies indicate that neutrophils and macrophages play important roles in minimizing the hepatic bacterial burden during the first 2-4 h postinfection (19, 20, 26, 41, 43). Elimination of bacteria by neutrophils at infective foci is believed to occur via phagocytosis and production of toxic metabolites such as cytokines and free radicals (47). In a study by Galleli et al. (18), intravenous administration of endotoxin before Listeria infection enhanced the resistance of mice to this organism. Concurrent with portal endotoxemia, the present study demonstrated that HU significantly enhanced the accumulation of neutrophils in the liver as well as TNF- protein levels in serum of endotoxin-sensitive mice. Taken together, these data suggest that endotoxin may play a role in HU-induced alterations in immune function by priming defense mechanisms.
In summary, data presented herein support the hypothesis that HU-induced portal endotoxemia results in hepatic injury and malfunction, most likely owing to the release of toxic mediators from activated Kupffer cells. Although the absence of gravity cannot accurately be simulated on Earth, HU mimics many of the physiological alterations caused by actual spaceflight. On the basis of the findings reported here, HU may also be used to model the release of endogenous endotoxin in the portal vein. In previous attempts to model chronic endotoxemia, miniosmotic pumps have been surgically implanted in the peritoneal cavity or subscapular area. However, various problems associated with this technique detract from its usefulness as an endotoxemia model. For example, the physical presence of the pump in the peritoneal cavity may stimulate an inflammatory response (6). Chronic infusion of exogenous endotoxin has been shown to induce tolerance to the development of pathophysiological conditions (23), which most likely results from the artificially high concentrations of endotoxin achieved by this method. The degree of portal venous endotoxemia observed after 4 wk of HU was comparable to that described after ethanol exposure for similar times (21). Moreover, results presented here provide evidence of histological changes as well as elevated serum markers of injury. Thus HU may serve as a more appropriate model of portal endotoxemia because it avoids the complications associated with artificially high endotoxin concentrations and surgical manipulation.
|
DISCLOSURES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. D. Cani, R. Bibiloni, C. Knauf, A. Waget, A. M. Neyrinck, N. M. Delzenne, and R. Burcelin Changes in Gut Microbiota Control Metabolic Endotoxemia-Induced Inflammation in High-Fat Diet-Induced Obesity and Diabetes in Mice Diabetes, June 1, 2008; 57(6): 1470 - 1481. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. A. Rivera, S. H. Abrams, M. H. Tcharmtchi, M. Allman, T. T. Ziba, M. J. Finegold, and C. W. Smith Feeding a corn oil/sucrose-enriched diet enhances steatohepatitis in sedentary rats Am J Physiol Gastrointest Liver Physiol, February 1, 2006; 290(2): G386 - G393. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Li, C. A. Rivera, A. R. Burns, and C. W. Smith Hindlimb unloading depresses corneal epithelial wound healing in mice J Appl Physiol, August 1, 2004; 97(2): 641 - 647. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
