Materials and Methods

Animals—Poland China–Landrace crossbred barrows were obtained from 2 commercial vendors.^a,b Animal husbandry conditions used for the study have been described elsewhere.^8,9 The swine in experiment 1 were part of a large group purchased at a body weight of 20 kg and raised to market weight (approx 90 kg); swine were individually housed in pens in our facility, which was accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International. During the finishing stage (60 to 90 kg), swine were fed a corn-soy diet ad libitum at 90% of the calculated energy requirement as determined in accordance with the Agriculture Research Council equation, and dietary amounts were adjusted weekly as previously reported.⁸ Feed restriction results in a uniform rate of weight gain and minimizes variation in basal metabolic variables, with each pig serving as its own control animal⁹; water was available ad libitum. The swine in experiment 2 were purchased^b at 35 and 85 kg, which were designated as growing and finishing swine, respectively. They were initially housed in groups (3 swine/pen). Swine were fed a corn-soy diet and had access to water ad libitum. After a 7-day acclimation period, swine were individually housed in pens. Animal study protocols were approved by our institutional animal care and use committee.

Experiments—Two experiments were conducted. In experiment 1, 5 market-weight swine were assigned to each of the 2 treatment groups by use of a randomization procedure^8,9 (injected IV with LPS^c at 20 μg/kg or sham-injected with sterile saline [0.9% NaCl] solution). Swine were slaughtered 24 hours after treatment. After slaughter, approximately 500 g of liver tissue was collected from various lobes. Tissue specimens were divided into 50-g aliquots and quickly frozen (within 20 minutes after organ removal) at −80°C. The liver weights were recorded at time of slaughter; body weights were recorded prior to LPS treatment.

In experiment 2, 3 growing swine and 3 finishing swine were assigned to each of 2 treatments (treatments were 2 different doses of LPS established on the basis of results from another study⁹). Each pig received a single IV injection of LPS at 2 or 20 μg/kg and was slaughtered 24 hours after treatment. The liver weights were recorded at time of slaughter; body weights were recorded prior to LPS treatment. Liver samples were collected and processed as described for experiment 1. Livers from 30 untreated, healthy finishing (> 85 kg) swine from several unrelated, unpublished experiments conducted by our laboratory group were used as the control liver samples.

Enzyme assays—Preparation of homogenates and subcellular fractions for determination of hepatic DME activities has been described elsewhere.^10–12 Measurements of microsomal protein concentration and the classical DME activities associated with specific forms of the CYP were performed as described elsewhere.^10–12 Assays based on spectrophotometric rate determinations (NADPH- and NADH-dependent reductases and GSH-S-transferase) were performed with a UV-visible diode-array spectrophotometer^d and the enzyme kinetic program supplied with the instrument. The difference spectra used in the determination of CYP and cytochrome b₅ were also obtained by use of this instrument. Protein concentrations of each tissue subfraction (S10, microsomes, and cytosol) were determined with the bicinchoninic acid^e protein procedure in microtiter plates with BSA^f as the standard.¹³ All CYP-dependent enzymatic assays were conducted with the S10 fractions. Absorbance measurements were performed by use of a microtiter plate reader.^g Assays requiring the use of fluorescence measurements were performed via a fluorescence microtiter plate reader.^h These included the O-dealkylations of ERF and PRF^e that were assayed without modifications from the procedures described elsewhere,^10–12 except for the instrument used to measure fluorescence. Other coumarinⁱ-based fluorescence assays were performed in accordance with the procedure described in another study,¹⁴ but the excitation (360 nm) and emission (485 nm) filters were changed to allow use of the fluorescent microtiter plate reader with the stock filters supplied by the manufacturer.

Western blot analysis—Western blot analysis was performed on swine liver microsomes (10 μg of protein/lane) as described elsewhere.¹⁰ Sources of antibody preparations used for western blot analysis of CYP isoforms were also described in another study.¹⁰ The primary antibodies were used at final titers of 1:500 or 1:1,000, diluted in Dulbecco PBS solution containing 5% BSA (calcium and magnesium free), in accordance with the manufacturers' recommendations. The secondary alkaline phosphatase–conjugated antibodies^j were used at a concentration of 1:1,000 (diluted in Dulbecco PBS solution supplemented with 5% BSA). Developed western blots were examined by use of densitometric analysis, and relative staining intensities of the western blots were determined with a gel scanner.^k Resulting images were quantitated by use of vendor-supplied software.^l The raw densitometry analysis results were reported in RAUs. Results of duplicate assays were normalized against results for 1 pig used as an internal control sample, and mean values for the duplicates were calculated. To facilitate comparison of the western blot analyses with the enzyme results, RAUs were normalized to RAUs per milligram of S10 protein.

Data evaluation and statistical analysis—Contents for CYP and cytochrome b₅ were reported as nanomoles per milligram of microsomal protein; subcellular fraction protein concentrations were reported as milligrams of protein per milliliter. The GSH-S-transferase and reductase activities were reported as nanomoles of product formed per minute per milligram of protein. Appropriate external standard curves were used to convert absorbance and fluorescence units, which were the units of measure for all assays, to nanomoles of product formed. The nanomoles of product formed per unit of time were normalized by dividing the value by the amount of S10 protein included in the assay mixtures to determine the specific activity. Mean values were calculated for data from appropriate animal replicates. Results are reported as mean ± SD for swine within groups. All data transformations were performed by use of a spreadsheet.^m Significance was determined by use of a 1-way ANOVA and the Student-Newman-Keuls multiple range test or Kruskal-Wallis ANOVA on ranks and the Dunn multiple comparison test.ⁿ Values were considered significant at P ≤ 0.05.

Results

Experiments 1 and 2—Results of experiment 1 were summarized (Tables 1–3). Results of experiment 2 also were summarized (Tables 4 and 5).

Liver weights and protein concentrations of hepatic subcellular fractions—Swine in experiment 1 treated with LPS (20 μg/kg) had significantly higher liver weights (1.46 ± 0.05 g [n = 5] vs 1.85 ± 0.14 g [4], respectively) and significantly lower protein concentrations in subcellular fractions than did control swine (Table 1). The administered dose induced such a strong inflammatory response in this treatment group in experiment 1 that 1 pig in the LPS-treated group died. Swine in experiment 2 receiving LPS (2 or 20 μg/kg) had a significant decrease in protein concentrations in the various subcellular fractions (Table 4).

Microsomal DME components and cytosolic GSH-S-transferase activity—In experiment 1, LPS treatment reduced the CYP and cytochrome b₅ contents and enzymatic activities of hepatic microsomal DME components in treated swine to less than those of the corresponding control swine. The LPS-treated swine also had lower GSH-S-transferase activities (Table 2). In experiment 2, CYP contents were unchanged in the LPS-treated swine, although cytochrome b₅ content typically was reduced after LPS treatment, but this may have been a reflection of the low number of treated swine because these values did not differ significantly (Tables 4 and 5). Administration of LPS in experiment 2 caused a consistent increase (> 2-fold) in cytochrome c reductase activity, whereas the cytochrome b₅ reductase activity was unaffected.

CYP-dependent DME activities—After LPS administration in experiment 1, enzymatic activities usually associated with CYP1A (ie, ERF) and CYP2A (ie, CoH and EtOC) were significantly reduced. However, CYP2B-mediated reactions had a range of responses to LPS treatment, as indicated by the fact that EtOMC O-deethylase activity was significantly inhibited, PRF O-dealkylase activity was totally eliminated, MOMC O-demethylase activity was unaffected, and aminopyrine N-demethylase activity was increased almost 5-fold. For the most part, results obtained in experiment 2 were similar to the effects seen in experiment 1. In experiment 2, we included reactions associated with CYP2E (ie, ANH and PNP-OH). As mentioned previously, CYP2B-mediated PRF O-dealkylase activity was eliminated in all LPS-treated swine (Tables 4 and 5). Although CYP1A-associated ERF O-deethylase and CYP2E-dependent ANH activities were significantly reduced when evaluated on the basis of LPS dose (2 or 20 μg/kg) or both dose and age, reactions catalyzed by CYP2A (ie, CoH and EtOC O-deethylase) and CYP2B (MOMC O-demethylase) were not significantly affected. On the other hand, LPS-treated finishing swine had higher amounts of CYP2B-associated aminopyrine N-demethylase activity than did control swine. This result was similar to the effects of LPS seen in experiment 1 (Table 3).

CYP isoforms—The LPS treatment completely eliminated all detectable CYP2A and CYP2B isoforms in liver samples obtained from the growing and finishing swine (Tables 4 and 5); this effect was irrespective of the dose of LPS administered to the swine. In addition, LPS appeared to induce a dose-dependent decrease in CYP2D content when test groups were combined; the results were observed mostly in growing swine. Lipopolysaccharide treatment also completely eliminated detectable CYP1A content in growing swine but had no effect on this isoform in finishing swine. Concentrations of hepatic CYP2E were elevated in growing swine treated with 20 μg of LPS/kg. In contrast to the aforementioned observations, LPS treatment did not affect the relative concentrations of CYP2C11, CYP2C13, and CYP3A in either of the LPS-treated test groups, irrespective of the age of swine.

Discussion

High doses of LPS caused hepatic edema, which resulted in increased liver weight with concomitant depression of protein concentrations in hepatic subcellular fractions. The 20 μg/kg dose of LPS also inhibited hepatic DME activities, which reduced the activity of several CYP-mediated reactions. In fact, CYP2B-mediated PRF O-dealkylase activity was essentially eliminated in these swine. In experiment 1, both reductase activities were inhibited by approximately 70%, whereas in experiment 2, there was a > 2-fold enhancement of cytochrome c reductase activities but cytochrome b₅ reductase activities were unaffected. Whether these conflicting results reflected differences in breeding stock between the 2 suppliers or were attributable to physiologic variations in this measure in swine cannot be determined. However, because this is the only measure in which we detected differences in the direction of the response between these 2 experiments, these results support the interpretation that we were measuring physiologic variations.

Studies^8,9 conducted by our laboratory group have revealed that LPS treatment of swine generates different cytokines, depending on the age (metabolic state) of the treated swine. In addition to our observations discussed previously, the differential effects of LPS on DME activities in the 20 μg/kg treatment group might suggest that LPS-generated cytokines were targeting expression of specific CYP isoforms and were not universally downregulating all CYP isoforms. The diverse effects on CYP2B-mediated reactions also suggest that cytokine regulation of CYP expression in swine differs from that observed in humans and rodents^15–20 because almost all CYPs in these species (except for CYP2B in rodents) are downregulated during an inflammatory response. This is not a universal observation because it appears that even in rats, the impact of infection or inflammation on DME activity is dependent on the route and type of infecting entity.²¹ Although CYP isoforms are recognized as a type of acute-phase protein whose values decrease during an inflammatory response, this conclusion might not be applicable to swine because of the selective decrease of CYP isoforms following LPS exposure.^6,22 Whether this is true for other aggravating causes of cytokine production (eg, vaccinations or infections) remains to be determined.

Analysis of results of the present study revealed that the pattern of swine CYP isoforms affected by LPS treatment differed from that described in studies in rodents. The reduction in swine liver CYP content was isoform specific and dependent on the metabolic status of the swine and not necessarily the dose of LPS. Growing swine have a daily rate of protein deposition that is greater than the daily rate of fat deposition, whereas finishing swine have a higher rate of fat deposition relative to that of protein deposition.²³ Comparing these 2 metabolically different groups, only CYP2A and CYP2B values were decreased in both growing and finishing swine whereas the amount of CYP1A was decreased only in growing swine. The underlying reason for the differences in CYP1A values in LPS-treated growing versus finishing swine cannot currently be determined.

In contrast to the results observed in swine, different CYP isoforms are downregulated in rats during an inflammatory process. Endotoxin challenge decreases CYP1A2, CYP2B1, CYP2C11, CYP2C13, and CYP2E1 expression in rats.^15–19 Although expression of rat CYP3A1 is not affected by the IFN inducer, polyinosinicpolycytidylic acid, alone, CYP3A1 is decreased in polyinosinic-polycytidylic acid–treated rats when CYP3A production is subsequently induced with pregnenolone-16 α-carbonitrile or troleandomycin.²⁴ This suggests that CYP3A concentrations can be affected by inflammation only under conditions in which there is extensive mRNA production. It was subsequently determined that mice had downregulation of CYP3A expression at all levels of induction, whereas CYP3A inhibition in rats was evident only at low levels of induction.²⁵ In another study,²⁶ investigators found that LPS works through the TLR4 to decrease CYP3A11 concentrations because TLR4-mutant mice do not have loss of CYP3A11 expression. The types of metabolic reactions affected in swine following LPS treatment provide further support for the contention that swine CYPs have a unique spectrum of metabolic activities. Analysis of these results suggests that some changes in CYP expression are attributable to a more direct effect of LPS, rather than being indirectly mediated by inflammatory cytokines.

The observed changes in swine CYP proteins and metabolic reactions following LPS treatment are consistent with our observations that swine CYPs do not catalyze the same reactions as do human and rodent CYPs.¹⁰ The LPS treatment eliminated detectable CYP2B1 proteins but did not affect MOMC O-dealkylase or EtOMC O-dealkylase activity and actually increased the amount of aminopyrine N-demethylase activity in finishing swine. Conversely, LPS treatment eliminated detectable PRF O-dealkylase activity usually associated with CYP2B1 isoforms and decreased the metabolism of ANH, although CYP2E1 values were unchanged or increased. Metabolism of another CYP2E substrate, PNP, was increased by LPS exposure.

Minimal information exists on the in vivo half-lives of specific rat or human CYP isoforms. Assessment of the sparse information available indicates that different CYP isoforms have vastly different half-lives.^27–30 Because of practical considerations, those studies were conducted in vitro. Differences in the half-lives of swine CYPs could help explain differences in their regulation following endotoxin challenge. If a particular CYP isoform has a relatively long half-life, the relatively short period of inflammation in the study reported here might not have been sufficient to cause an effect on rates of synthesis (or degradation).

Analysis of the results of the present study indicates that the concentrations and metabolic activities of CYP isoforms in LPS-treated swine differ from those reported for LPS-treated rats. These results also provide further evidence for our earlier observations that swine CYPs catalyze reactions other than those that would have been predicted on the basis of results in other species, including rats, mice, and humans.¹⁰ A direct consequence of these observations is that swine infected by LPS-generating bacteria would most likely have a markedly reduced ability to metabolize drugs that are in the pathways of the susceptible CYP isoform categories identified here. Drugs metabolized by non-CYP pathways, by CYP3A, or by CYP2C might not be affected. If metabolism of drugs is affected, there would be prolonged clearance of drugs or xenobiotics to which the sick animals are exposed. This could result in unexpected toxic effects, extended drug withdrawal times, or both.