Gene expression analyses were performed using 8 arrays. Statistical analysis revealed that 11 of genes on the array were differentially expressed between the control and LPS-treated animals (p < 0.1) (Table 1). The relative expression of adrenocorticotropic hormone receptor (ACTHR, p < 0.07), interferon α receptor (IFNαR, p < 0.05), CD1 (p < 0.03), monocyte-chemoattractant protein 1 (MCP-1, p < 0.04) and growth hormone (GH, p < 0.04) genes was increased, while complement component-3 (C3, p < 0.04), myeloid membrane glycoprotein (CD14, p < 0.10), insulin-like growth factor binding protein-3 (IGFBP3, p < 0.01), interleukin 12 receptor (IL12R, p < 0.03), natural resistance-associated macrophage protein-1 (NRAMP1, p < 0.01) and superoxide dismutase (SOD, p < 0.08) gene expression was decreased in the LPS-treated animals. Overall, the fold change in gene expression for all of these genes was low (≤ 1.49), even though the signal intensity of MCP-1, SOD, ACTHR, IL12R and NRAMP1 was relatively high (>5000 pixels) from the microarray slides.

One of the principle complications in microarray analysis of gene expression is the relatively large amount of RNA required for each array. On average, 5–20 μg of total RNA are required per study. This is easily obtained from tissue samples collected from euthanized animals, but is more difficult to obtain from small volume biopsy samples collected from live animals. In this study, the SenseAmp kit (Genisphere Inc. Hatfield, PA) was chosen to amplify total RNA. Goff et al. 14 evaluated sense-strand mRNA amplification by quantitative real-time PCR analysis. Their results demonstrated that the SenseAmp kit yields the highest correlation between PCR products before and after amplification, and is also able to accurately amplify partially degraded samples.

Validation of the microarray results by qRT-PCR was performed on the CD14, IFNαR, C3, NRAMP1, SOD, MCP-1, and IGFBP3 genes (Table 2). Two housekeeping genes, GAPDH and RPLPO were also selected. Results from this analysis generally support the microarray data (Figure 1). Linear orthogonal polynomial contrasts (LOPCS) across dose were significant for CD14 (p = 0.06), NRAMP1 (p = 0.05), SOD (p=0.07), IGFBP3 (p = 0.03), and the GAPDH housekeeping gene (p = 0.05), indicating that the expression of these genes was reduced across LPS doses. GAPDH has also recently been reported to not be a reliable hepatic housekeeping gene for rats challenged with LPS 9. LOPCS across doses was also significant for MCP-1 (p=0.02), indicating that the expression of this gene was increased across LPS doses. A significant quadratic orthogonal polynomial contrast across dose was also noted for IFNαR (p = 0.02), indicating that the highest expression of this gene was observed at the 200 ng/kg LPS treatment level. A significant change in the expression of C3 and the RPLPO housekeeping gene across LPS doses was not observed.

The hepatic genes studied are either involved in LPS recognition, or in regulating the inflammatory response that occurs following LPS recognition. CD14 for example, plays a key role in LPS recognition during bacterial infection. LPS ligation to CD14-TLR4 complex subsequently activates numerous cell types to secrete pro-inflammatory cytokines including IL-6. Recent studies performed with bovine MAC-T cells 15, and rat lung 16 and liver tissues 9 have shown that CD14 expression levels were largely unaffected by LPS. An earlier rodent study however, reported up-regulation of CD14 3-hours, but not 6-hours post-challenge with 4 μg/kg of LPS administered i.v 6. Previously, our group demonstrated that ovine CD14 gene expression increased significantly 2 hours, but not 5 hours after systemic challenge with 200 and 400 ng/kg of LPS 17. In the present study, CD14 gene expression was reduced at the 5-hour sampling time. These results and others suggest that LPS induces tissue-specific and temporal differences in CD14 gene expression.

NRAMP1 functions as a proton/divalent cation transporter in the membranes of the late endosomes/lysosomes, regulating cytoplasmic iron levels in macrophages, and plays a role in host innate immunity 18. NRAMP1 gene expression is dramatically increased in murine macrophages treated with LPS in vitro and in vivo 19, and its expression is both time- and dose-dependent 20. A study by Wyllie et al. demonstrated using NRAMP1 knockout mice that hepatic NRAMP1 expression is important for inducing early phase Kupffer cell activation and hepatic inflammation 21. The LPS dose-dependent down regulation of NRAMP1 gene expression observed in the current study may be part of a regulatory mechanism designed to control LPS-induced inflammation in the liver.

As a group of metal-containing enzymes, SODs have a vital anti-oxidant role conferred by their scavenging of the reactive oxygen species 22. A previous study using rats demonstrated that SOD activity decreased in the liver during the acute phase of an in vivo LPS challenge and then increased during the recovery phase. Similar findings were reported with hepatocytes exposed to LPS in vitro in the same study 23. A microarray study using liver tissue from rats challenged with LPS demonstrated induction of the SOD2 gene 24 hours post LPS challenge however, no assessment was made at earlier time points 9. The dose-dependent decrease in SOD gene expression that was observed in the present study 4–5 hours post LPS challenge, combined with these previously reported studies, suggest that SOD gene expression varies temporally in the liver following LPS challenge.

IGFBPs regulate the bioactivity of mitogenic IGF-I and may also inhibit the growth of certain cell lines by an IGF-I receptor-independent pathway 24. Priego et al. reported that LPS decreased the gene expression of IGFBP-3 in the rat liver following in vivo challenge 24. Our results confirm their results using an ovine LPS challenge model.

MCP-1 is an important leukocyte chemoattractant that is involved in recruiting neutrophils and monocytes/macrophages during inflammation. Several studies have shown that LPS induces MCP-1 gene expression in various tissues both in vivo and in vitro 252627. Two different LPS studies performed in vivo using the rat 9 and canine models 8 made no mention of hepatic MCP-1 induction using microarray analysis, although another chemokine, MIP-1, was reportedly induced 4 hours post LPS challenge in the canine model 8. A recent study however, reported hepatic MCP-1 protein expression in mice challenged with LPS 28. Our study demonstrates that LPS also induced hepatic MCP-1 gene expression in sheep.

All IFN subtypes are multifunctional cytokines that exhibit differential activities through a common receptor composed of the subunits IFNαR1 and IFNαR2 29. In this study we found that the expression of IFNαR1 gene was induced after LPS treatment, but the highest expression was observed at the 200 ng/kg LPS treatment level. A study by Severa et al., demonstrated that human mature dendritic cells modulate their sensitivity to IFN subtypes by differentially regulating the IFNαR subunits 30. Future studies on IFNαR may help us understand its role during LPS-induced hepatic inflammation.

C3 is a key molecule in both the classical and alternative pathways of the complement cascade. The expression of the C3 gene appears to be dependent on LPS dose, sampling time, and cell type. LPS has been reported to induce C3 gene expression for example, in a human hepatoma cell line in vitro 31 and in human mononuclear phagocytes and human polymorphonuclear leukocytes in vitro 3233. Others however, have reported that C3 gene expression is decreased in monocytes stimulated with LPS 34, and that C3 protein expression follows a bell shaped curve when monocytes are stimulated in vitro with LPS between 0.1–100 ng/ml 35. In the present study we report that C3 expression is suppressed in the ovine liver as determined by microarray analysis. Unfortunately, there was insufficient power to validate these results by qRT-PCR analysis.

Twelve yearling Riduea-Arcott ewe lambs were arbitrary assigned to three groups and challenged with LPS (0, 200 or 400 ng/kg) from E. coli serotype 0111:B4 (Sigma Chemical, St. Louis, MO) between 8 and 9 am. Liver biopsies (30–40 mg) were collected between 4 and 5 hours post-challenge, and tissues were immediately placed in RNAlater (Ambion, Austin, TX) and stored at -80°C until total RNA extraction was performed. The doses and biopsy sampling times were based on previous time trial experiment 17.

Total RNA was isolated with Trizol reagent (Invitrogen, Burlington, ON) 17, and amplified using the Genisphere's SenseAmp kit (Genisphere Inc. Hatfield, PA) according to the manufacturer's instructions. Briefly, 0.25 μg of total RNA was used to synthesize first strand cDNA using an oligo (dT) primer and random primer. First strand cDNA was purified then tailed with dTTP using Terminal Deoxynucleotidyl Transferase. The T4 template Oligo was annealed to the 3' end of the cDNA. Klenow enzyme fills in the 3'end of first strand cDNA to produce a double-stranded T7 promoter. Sense-strand RNA (sense RNA) copies of the original starting material were generated during in vitro transcription. Amplified sense RNA was quantified using Agilent 2000 Bioanalyzer.

A set of 109 immune, endocrine and inflammation-associated genes was selected for triplicate spotting onto Corning GAPS II slides using a VIRTEK Chip Maker Pro spotter (BioRad, Mississauga, Canada). Positive controls included 5 housekeeping genes (β-actin, GAPDH, HPRT, PRLPO and β2-microgobulin), and a serial dilution of pooled bovine genomic DNA. Negative controls included a bacterial gene (VapA) and a partial plasmid sequence of pACYC177. All gene products were PCR amplified from either bacterial clones, or liver total RNA. The original clone sets and gene-specific primers were donated by Tao et al. 12.

For each sample, Alexa Fluor 555 or Alexa Fluor 647-labeled cDNAs were generated from 2~2.5 μg of SenseRNA using the SuperScript Indirect cDNA Labeling system (Invitrogen, Burlington, ON). Labelled control animal cDNA was then mixed with labeled cDNA of an animal from either the 200, or 400 ng/kg LPS dose groups, and then hybridized to the array for 18 h in a GeneTAC HybStation (Genomic Solutions, Ann Arbor, MI) using step-down temperatures from 65°C to 50°C in sealed chambers. Following hybridization, the station applies three washes, one with medium stringency buffer, one with high stringency buffer and one with post wash buffer (Genomic Solutions). Slides were finally rinsed briefly at room temperature in 2 × SSC and once in ddH₂O. Washed microarrays were dried by centrifugation at 1700 rpm for 2 min in a cushioned 50 ml tube. Dye swapping was performed on half of the samples to prevent dye bias. Dried Slides were scanned using GenePix™ 4000 (Axon Instruments Inc. Union City, USA). The images were analyzed and tabulated using GenePix Pro 3.0.

Microarray data were normalized using LOWESS (Locally Regression and Smoothing Scatterplots) procedure of SAS. The program was obtained from Dr. P. Coussens (Department of Animal Science, Michigan State University). Normalized data were imported into Microsoft Excel, log transformed, and the median blank intensity on a microarray for each dye was subtracted from the respective normalized spot intensity values. These blank corrected values were then used to calculate a mean log expression difference between LPS-treated and control samples. The significance of the values was determined using a two-tailed Student's t-test. The antilog of the mean log expression difference for an individual gene on the array yielded the approximate fold change in expression between cDNA from the LPS-dose groups and control group.

To confirm gene expression differences observed from microarray results, qRT-PCR was performed on 9 genes. The primers for housekeeping gene, GAPDH and RPLPO were developed by Tellam 36. Other primers were developed using Primer 3 software by our group (Table 2).

qRT-PCR was performed in triplicate for each sample on ABI 7000 Sequence detection system (Applied Biosystems, Streetsville, ON) using default two-step amplification procedures and 2 × SYBR Green Master Mix (Invitrogen, Burlington, ON) in a 25 μl reaction volume according to manufacture instructions. The conditions for the PCR reaction were: 50°C for 2 min, 95°C for 2 min followed by a maximum of 50 cycles of 95°C for 15 sec, annealing temperature for 30 sec and 72°C for 30 sec. The annealing temperatures for genes of interest are included in Table 2. The standard curve method was used to determine relative quantitation of mRNA abundance 37. Statistical analysis was carried out on the qRT-PCR data using GLM of SAS (SAS 2002, SAS Institute Inc., Gary, NC). Orthogonal polynomial contrasts were performed on the least squares means to identify both linear and quadratic responses across dose. Residual plots were examined to assess homogeneity of variance.