Laminar inflammatory gene expression in the carbohydrate overload model of equine laminitis




Reasons for performing study: There is a need to assess the laminar inflammatory response in a laminitis model that more closely resembles clinical cases of sepsis-related laminitis than the black walnut extract (BWE) model.

Objectives: To determine if a similar pattern of laminar inflammation, characterised by proinflammatory cytokine expression, occurs in the CHO model of laminitis as has been previously reported for the BWE model.

Methods: Sixteen horses administered 17.6 g of starch (85% corn starch/15% wood flour)/kg bwt via nasogastric (NG) tube were anaesthetised either after developing a temperature >38.9°C (DEV group, n = 8) or at onset of Obel grade 1 lameness (OG1 group, n = 8). Control horses (CON group, n = 8) were anaesthetised 24 h after NG administration of 6 l of deionised water. Laminar tissue was collected from horses while under anaesthesia, followed by humane euthanasia. Real time-quantitative PCR was used to assess laminar mRNA concentrations of genes involved in inflammatory signalling.

Results: Increased mRNA concentrations (P<0.05) for IL-1β, IL-6, IL-12p35, COX-2, E-selectin and ICAM-1 were present in laminae from horses with OG1 lameness but not at the DEV time, when compared to the CON horses. No differences between the groups were found for IL-2, IL-4, IL-10, TNF-α, IFN-γ or COX-1 at either the DEV or OG1 time points.

Conclusions: There was a notable difference in the temporal pattern of inflammatory events between the BWE and CHO models, with the majority of laminar inflammatory events appearing to occur at or near the onset of lameness in the CHO model, whereas many of these events peak earlier in the developmental stages in the BWE model. This suggests that, in addition to circulating inflammatory molecules, there may be a local phenomenon in the CHO model resulting in the simultaneous onset of multiple laminar events including endothelial activation, leucocyte emigration and proinflammatory cytokine expression.

Potential relevance: The similar (although somewhat delayed) inflammatory response in the CHO model of laminitis indicates that inflammatory signalling is a consistent entity in the pathophysiology of laminitis.


Extensive studies of equine laminitis using the black walnut extract (BWE) model indicate that equine laminitis, a common clinical sequela to many systemic diseases associated with Gram-negative sepsis, has similar pathophysiological events as organ injury/failure in human sepsis, including leucocyte extravasation and inflammatory mediator production. Increases in inflammatory gene expression reported with the BWE model include increases in proinflammatory cytokines, chemokines, cyclooxygenase-2 and endothelial adhesion molecules important in leucocyte adhesion and extravasation (Fontaine et al. 2001; Waguespack et al. 2004a; Blikslager et al. 2006; Belknap et al. 2007; Loftus et al. 2007). In these reports, the laminar inflammatory response was observed early in the course of development, occurring as early as 1.5 h post administration of BWE (Loftus et al. 2007). Whereas some of these laminar inflammatory events peak in the early developmental/prodromal period in the BWE model (Blikslager et al. 2006; Loftus et al. 2007), others continue to increase through the onset of lameness phase which occurs approximately 10–13 h post BWE administration (Belknap et al. 2007).

The carbohydrate (CHO) model of equine laminitis appears to more closely mimic clinical cases of laminitis that occur secondary to sepsis-related diseases (i.e. enterocolitis, large intestinal strangulation, acute metritis) than the BWE model due to: 1) the documented presence of endotoxaemia; 2) the extended time frame of systemic signs of sepsis prior to onset of lameness; and 3) a similar high incidence of structural failure of the laminae (which rarely occurs in the BWE model) (Carroll et al. 1987; Sprouse et al. 1987;van Eps and Pollitt 2006; Bailey et al. 2009; Belknap et al. 2009). Administration of an overload of CHOs results in a decrease in caecal pH, an increase in lactic acid producing bacteria, death of Gram-negative bacteria and substantial damage to the caecal mucosal barrier resulting in absorption of bacterial pathogen associated molecular pattern molecules (PAMPs) including endotoxin (Garner et al. 1978; Moore et al. 1979; Krueger et al. 1986; Sprouse et al. 1987; Bailey et al. 2009). In the oligofructose model, plasma endotoxin concentrations were found to peak at 8 h with increases in plasma TNF-α concentrations being present between 12 and 24 h (Bailey et al. 2009). The temporal pattern of these events presumably reflects the time necessary for the caecaldisruption to result in mucosal barrier dysfunction allowing bacterial products to enter the systemic circulation and cause a systemic inflammatory response that is similar to human sepsis.

Experimental models of human sepsis have been used to help develop treatments to improve morbidity and mortality for clinical cases. The most common models used (most performed in laboratory rodents) are the i.v. infusion of endotoxin and the caecal ligation and puncture (CLP) model (Remick et al. 2000; Poli-de-Figueiredo et al. 2008). Like the experimental models used to induce laminitis, the different models can produce variable results making it difficult to determine effective treatments for sepsis. Although endotoxin infusion models have given valuable information regarding the response of animals/man to sepsis, the patterns of inflammatory responses do not usually accurately reflect the responses observed in human sepsis (Poli-de-Figueiredo et al. 2008). Similar to the BWE model of equine laminitis, LPS administration generally results in a more rapid but more transient cytokine response than observed in clinical sepsis cases, with cytokines peaking between 1.5–4 h and declining by 8 h post LPS administration (Remick et al. 2000).

The presence of laminar inflammation has been well documented in the early stages of BWE-induced laminitis and has been proposed to play a prominent role in laminar injury and failure (Fontaine et al. 2001; Waguespack et al. 2004a,b; Black et al. 2006; Blikslager et al. 2006; Belknap et al. 2007; Loftus et al. 2007; Riggs et al. 2007; Faleiros et al. 2009; Noschka et al. 2009). Although the pattern of the laminar inflammatory response has not been well described in the CHO model, the reported systemic response is more similar to the clinical case of laminitis in the septic horse, including a sustained response regarding immunological, cardiovascular and metabolic alterations, and the development of lameness (Sprouse et al. 1987; Allen et al. 1990; van Eps and Pollitt 2006; Eades et al. 2007; Bailey et al. 2009). Thus, it is critical to detail the laminar inflammatory response in the CHO model both to determine if these inflammatory events occur in multiple models of laminitis, and to establish the temporal pattern of inflammatory signalling in this clinically relevant model in order to design future anti-inflammatory treatment regimens.

Materials and methods


The experimental protocol was approved by the Institutional Animal Care and Use Committee. Twenty-four mature horses with a median bodyweight of 421 kg (range 341–524 kg) and median age 5 years (range 3–12 years) were used in this study. Each horse was determined to be healthy and free of digital pathology determined by physical and lameness examinations and radiographic evaluation of the distal phalanx. Horses were divided into 3 groups: a control group (CON, n = 8), a developmental period group (DEV, n = 8), and an Obel grade 1 lameness group (OG1, n = 8). Prior to the experiment, the horses were quarantined for 2 weeks at the University of Missouri, housed in stalls and fed a regular ration of free choice hay.

Carbohydrate overload model

A CHO gruel consisting of 85% cornstarch and 15% wood flour (17.6 g/kg bwt) was administered to each horse in the DEV and OG1 groups by nasogastric tube as previously described (Sprouse et al. 1987; Johnson et al. 1998). Six litres of deionised water were administered via nasogastric tube to each horse in the control group. All horses received complete physical examinations, consisting of rectal temperature, heart rate, respiratory rate, abdominal sounds, digital pulses, evaluation with hoof testers and gait evaluation, immediately prior to nasogastric tube passage, and at 2 h intervals following administration of either CHO or water. Blood was taken from an i.v. jugular catheter at 0, 2, 4, 8, 12, 16 h in all horses, and every 4 h after 20 h in the OG1 group until onset of Obel grade 1 lameness (horse lifts feet incessantly, short, stilted gait at trot). Anaesthesia was induced in the DEV group within 2 h of developing a temperature >38.9°C (occurring between 12 and 22 h) and in the OG1 group at the onset of lameness (occurring between 20 and 48 h). Although classically a 20% drop in central venous pressure is often used to determine the DEV time point in the CHO model (Allen et al. 1990), fever was used in this study as it occurs at approximately the same time as the drop in CVP and was a much more repeatable and accurate variable to measure. Due to the facts that: 1) it was anticipated from previous reports that up to 30% of the horses may not respond (exhibit any signs) to the CHO administration; and 2) the investigators were also interested in assessing laminar inflammatory signalling in these ‘nonresponders’, it was established prior to performing the animal protocols that animals in the DEV group which did not exhibit fever would be anaesthetised for sampling at 24 h after CHO administration, and horses in the OG1 group that did not exhibit lameness would be anaesthetised at 48 h after CHO administration. Two horses in each of DEV (no fever) and OG1 groups did not respond (no fever or lameness) to the CHO administration and were anaesthetised for sample harvesting at the appropriate times. Control horses were anaesthetised at 24 h after administration of water. Anaesthesia was induced with a combination of xylazine (1 mg/kg bwt i.v.) and ketamine (2.2 mg/kg bwt i.v.). Horses were placed in lateral recumbency and anaesthesia was maintained with isoflurane. While under general anaesthesia, a tourniquet was placed, followed by rapid removal of all 4 feet by disarticulation of the metacarpophalangeal joint. Laminar tissue was rapidly harvested after 1 cm thick sagittal sections of the digit were cut with a band saw. These tissues were immediately snap frozen in liquid nitrogen and later transferred to -80°C for storage until used for isolation of total RNA. Once the tissues were collected, each horse was subjected to euthanasia with pentobarbital sodium and phenytoin sodium (20 mg/kg bwt i.v.).

RNA isolation and cDNA synthesis

Total RNA was extracted from 3 separate sections of dorsal lamina from each horse in the CON, DEV and OG1 groups using a kit (Absolutely RNA Miniprep)1, which includes a DNase treatment to remove genomic DNA contamination. PolyA mRNA was then isolated (mRNA extraction kit)2 and used to make complementary DNA (cDNA) for each sample via reverse transcription (Retroscript)3 using at total of 400 ng of mRNA. The cDNA was frozen at -20°C and stored until used for real-time quantitative PCR (RT-qPCR) analysis.

Real-time qPCR procedure

Real-time quantitative PCR was performed using a thermocycler (LightCyler)2 and quantified with external standards with the fluorescent format for SYBR Green I dye as previously described (Waguespack et al. 2004a,b). Primers were designed from equine-specific sequences for IL-1β, IL-2, IL-4, IL-6, TNF-α, IFN-γ, COX-1, COX-2, E-Selectin and the housekeeping genes (β-actin, β-2 microglobulin, glyceraldehyde-3 phosphate dehydrogenase and TATA-box binding protein) have been previously reported (Waguespack et al. 2004a,b). New primers for IL-12p35 (f: 5′-CGTTTTAGCCCGTCTCAA-3′ and r: 5′-AACTCCACCTGGTACATCTTCA-3′), ICAM-1 (f: 5′-GACGCCCCCAGAAGCCATCATA-3′ and r: 5′-GACCCCCTGCACTTCCTTCTTACTCA-3′), and IL-10 (f: 5′-GAGGCTGCGGCGCTGTCATC-3′ and r: 5′-TTTTTCATCTTCGTTGTCATATAGGCTTCT-3′) also were designed from equine-specific sequences. Concentrations of IL-4 mRNA were also determined using an additional set of previously reported primers (f: 5′-AAGGGCAAGAATTCGTGCAT-3′ and r: 5′-CGCTCAGGCATTCTTTGATCA-3′) designed from equine-specific sequences (Figueiredo et al. 2009). All primers were screened using gel electrophoresis and melt curve analysis (LightCyler)2 to confirm amplification of a single cDNA fragment of the correct melting temperature and size (Waguespack et al. 2004b). Amplified cDNA fragments of each gene were ligated into a vector (TOPO 010 E. Coli)4 and the vectors linearised with Hind III restriction enzyme4 for the use of templates to generate a standard curve for the RT-qPCR reaction (Waguespack et al. 2004a,b). Amplified cDNA fragments were sequenced after cloning to confirm correct DNA sequence for the products of each primer (Waguespack et al. 2004b).

Each cDNA sample was diluted 1:5 and 1:500 with 1x TE buffer to be used in the cytokine and housekeeping PCR reactions, respectively. PCR reactions were performed in glass capillaries containing 5 µl of diluted sample and 15 µl of PCR master mixture. Master mix included the following: 1 unit of Taq polymerase4, 0.2 units of uracil-N-glycosylase2, 1:10,000 dilution of SYBR Green stock solution, forward and reverse primers, PCR nucleotide plus2 and PCR buffer. The PCR buffer (20 mmol/l Tris-HCl) contained 0.05% each of Tween 20 and nonionic detergent. Primers for ICAM-1, IL-12p35 and IL-10 were used at a concentration of 2.5 µmol/l. All remaining primers were used at a concentration of 5 µmol/l.

Uracil-N-glycosylase activation, to prevent PCR product carryover, was conducted at 50°C for 2 min and was followed by denaturation at 95°C for 2 min. Amplification was for 40–45 cycles, with the annealing temperature set at 1–5°C below the melting temperature for each specific set of primers, extension was set at 72°C for 5 s, and fluorescence acquisition for 10 seconds in the SYBR Green format. Single fluorescence acquisition in each cycle was set at either 80 or 82°C, depending on the melting temperature of the cDNA product of interest as previously described (Waguespack et al. 2004a,b). After amplification cycling, melting curves of the PCR product were acquired through a stepwise increase in temperature from 65–95°C.

Standard and target samples were prepared in separate capillaries. Standard curves and water for negative control were performed for each gene of interest and amplified with each series of reactions. Standards were made up of 10-fold serial dilutions of linearised plasmids containing the different gene-specific cDNA inserts. All samples were run in duplicate.

Data analysis

Average copy number from each sample was determined for each gene (housekeeping and inflammatory marker). As previously reported, RT-qPCR data from the 4 housekeeping genes for each sample was evaluated by the computer software program geNorm5 to determine which 2 genes received the best acceptable score to be used for normalisation (Blikslager et al. 2006). β-actin and glyceraldehyde-3 phosphate dehydrogenase were selected and then used by the geNorm software to create a normalisation factor for each sample. To determine the corrected copy number value for each sample the amplification data obtained by RT-qPCR for each gene was divided by the normalisation factor of the selected housekeeping gene for the same sample. After normalisation, the fold change from the average control value was calculated for each sample. Data were analysed nonparametrically using Kruskal-Wallis and Dunn's multiple comparisons test to compare groups. Although RT-qPCR data were obtained for the 2 nonresponders (horses not developing fever i.e. >38.6°C) in the DEV group and 2 nonresponders (horses not displaying signs of fever or Obel grade 1 lameness) in the OG1 group, the data from these horses were not included in statistical analyses. However, the data from nonresponders are included on the graphs to demonstrate the pattern of inflammatory gene expression in these animals. Only 6 control horses were used in the analysis (2 randomly removed) to allow for equal groups. Statistical significance was set at P<0.05 to maintain type I error for all tests.


Six of 8 horses in the DEV group developed fevers with an average high temperature of 39.8°C (range 39.6–40.6°C), which occurred between 12 and 22 h post CHO administration (median 17 h). Six of 8 horses in the OG1 group developed signs of Obel grade 1 lameness (lifting feet incessantly and demonstrating a short, stilted gait at trot), which occurred between 20 and 48 h post CHO administration (median 25 h). Two horses in the DEV group and 2 in the OG1 group did not develop a fever (≥38.6°C) or lameness, respectively (nonresponders in OG1 group also did not develop a fever at any point). Laminar mRNA concentrations of IL-1β, IL-6, IL-12p35, COX-2, E-Selectin and ICAM-1 were significantly increased in the OG1 horses when compared to control horses; however, no significant changes were present in the DEV horses for these inflammatory factors (Fig 1; Tables 1–4). No change in laminar mRNA concentrations of TNF-α, IFN-γ, COX-1, IL-2 and IL-10 were present in either the DEV or OG1 groups (Tables 1, 2, 4). Measureable concentrations of IL-4 mRNA were not present in any of the laminar tissue (CON, DEV and OG1) as the levels were below detection for our PCR. The PCR was validated with equine tissue from spleen, liver and kidney as well as sequenced plasmid standards that amplified with the use of the equine specific primers designed in our laboratory and from Figueiredo et al. (2009). Although laminar mRNA concentrations from the nonresponders were not included in the statistical analysis, they are displayed on the graphs as open circles to demonstrate the absence of a laminar inflammatory response in these horses (Fig 1).

Figure 1.

Median fold changes in laminar mRNA concentrations after CHO administration in horses (DEV and OG1). Significantly increased inflammatory mediator when compared to control (CON) horses (*P<0.05;**P<0.01;***P<0.001).

Table 1. Laminar cytokines of innate immunity mRNA concentrations in the CHO model of laminitis
Cytokines of innate immunityCONDEVOG1
  1. *mRNA values expressed as median cDNA copies per normalisation factor. ∧Fold increase expressed in medians from control. NS = nonsignificant fold change from control.

 (25–75% percentile)(49,850–127,500)(45,650–126,000)(68,800–162,000)
 Fold increase 0.081.5
 P value NSNS
 (25–75% percentile)(925–2735)(1490–6745)(5820–21,750)
 Fold increase 2.411.2
 P value NSP<0.01
 (25–75% percentile)(40–235)(2121–52,600)(58,750–570,500)
 Fold increase 43.32089
 P value NSP<0.001
Table 2. Laminar cytokines of adaptive immunity mRNA concentrations in the CHO model of laminitis
Cytokines of adaptive immunityCONDEVOG1
  1. *mRNA values expressed as median cDNA copies per normalisation factor. ∧Fold increase expressed in medians from control. NS = nonsignificant fold change from control. IL-4-values below the level of detection and not included in this chart.

 (25–75% percentile)(7.1–55.9)(7.4–30.4)(10.4–122.2)
 Fold increase 0.061.74
 P value NSNS
 (25–75% percentile)(3.4–167)(11.8–191)(25.2–241)
 Fold increase 0.420.66
 P value NSNS
 (25–75% percentile)(41–236)(126–495)(190–678)
 Fold increase 3.85.3
 P value NSP<0.05
 (25–75% percentile)(1530–5185)(1635–6105)(2335–9575)
 Fold increase 1.021.35
 P value NSNS
Table 3. Laminar adhesion molecule mRNA concentrations in the CHO model of laminitis
Adhesion moleculeCONDEVOG1
  1. *mRNA values expressed as median cDNA copies per normalisation factor. ∧Fold increase expressed in medians from control. NS = nonsignificant fold change from control.

 (25–75% percentile)(32–309)(54–574)(767–3280)
 Fold increase 0.910.4
 P value NSP<0.05
 (25–75% percentile)(3725–11,350)(9200–65,100)(19,180–54,250)
 Fold increase 1.75.8
 P value NSP<0.05
Table 4. Laminar COX mRNA concentrations in the CHO model of laminitis
  1. *mRNA values expressed as median cDNA copies per normalisation factor. ∧Fold increase expressed in medians from control. NS = nonsignificant fold change from control.

(25–75% percentile)(4230–14,150)(1330–5555)(223–6195)
Fold increase 0.460.29
P value NSNS
(25–75% percentile)(1530–5660)(2050–99,600)(56,500–120,000)
Fold increase 1.231.2
P value NSP<0.05


The studies of laminar inflammation in the BWE model of laminitis have supplied us with many inflammatory events to assess in a different laminitis model that more closely replicates clinical laminitis, the CHO model. To our knowledge, the current study is the first to document that a robust proinflammatory cytokine response also occurs in the CHO model. However, the results also suggest a very different temporal pattern of cytokine expression to that reported in the BWE model.

Early increases in laminar mRNA concentrations of both IL-1β and IL-6 have been reported in the BWE model with a sustained increased in IL-6 mRNA concentration through the onset of lameness (Fontaine et al. 2001; Belknap et al. 2007; Loftus et al. 2007). In the present study, laminar mRNA concentrations of IL-1β and IL-6 were only increased in the OG1 horses. Although the response was delayed in comparison to the BWE model, there was still a robust response with laminar IL-6 mRNA concentrations increasing >2000-fold from control values at the onset of lameness. These proinflammatory cytokines, which are an important part of the innate immune response, are also characteristic of the cytokine response reported with human sepsis/SIRS (Cohen 2002). Increased plasma concentrations of IL-1β and IL-6 have been negatively correlated with outcome in human patients with sepsis (Song and Kellum 2005).

Increases in mRNA concentrations of TNF-α, a prominent inflammatory cytokine present in human sepsis/SIRS, have not been found in the laminar tissue of horses after BWE administration (Belknap et al. 2007). Similar findings were also present in this study where no increase in TNF-α mRNA concentrations were present after CHO administration at either the DEV or OG1 time point. Previously, the absence of a laminar TNF-α response was purported to be due to the lack of mononuclear cells present in the laminar tissue; however, recent findings from our laboratory provide evidence of a population of macrophages within the primary and secondary dermis in both normal horses and those with laminitis (Faleiros et al. 2010). Although TNF-α is not present in laminar tissue during the development of laminitis, increases in serum concentrations of TNF-α have been reported in the oligofructose model of laminitis (Bailey et al. 2009). The source of the serum TNFα may be due to expression in visceral organs as increases in mRNA concentrations of TNF-α have been reported in the lung and liver from horses administered BWE (Stewart et al. 2009); however, in pigs administered LPS no correlation between plasma cytokine concentrations and organ or peripheral blood cells mRNA concentrations were found (Ebdrup et al. 2008). No increases in mRNA concentration of TNFα from peripheral blood mononuclear cells have also been reported after administration of oligofructose to horses (Tadros et al. 2009). Therefore it is possible to have a systemic increase in plasma cytokine response during the DEV and OG1 time without demonstrating a significant increase in other tissues such as the lamina. The lack of TNF-α response in the laminar tissue may be similar to reports in rats after endotoxin administration where systemic increases in TNF-α were present but local production of TNF-α in the interstitial fluid of the rat paw was not present (Nedrebøet al. 2004). Compartmentalisation of the inflammatory response in regards to expression of specific cytokines is another possible explanation for increases in TNF-α in some tissues and not in others. This is believed to occur in many models of sepsis/SIRS including a baboon model of lung sepsis where TNF-α was also not up-regulated whereas increases in IL-1β and IL-6 were found in the identical tissue samples (Cavaillon and Annane 2006; Zhu et al. 2007).

Increases in mRNA concentration of IL-12p35 were present in laminar tissue at the onset of lameness and is similar to reports in the BWE model of laminitis where increases in IL-12 were only reported for the late (12 h) time point (Belknap et al. 2007). IL-12 is known to be important for stimulation of natural killer cells and T cells to produce IFN-γ, which is part of the Th1/cell mediated response of adaptive immunity (Watford et al. 2003). However, similar to previous reports in the BWE model (Belknap et al. 2007), no change in IL-2 or IFN-γ mRNA concentrations was present at either time point in the current study. A possible explanation for these findings may be that a prominent Th1 response does not occur in the laminar tissue of horses as seen with human sepsis/SIRS (Cohen 2002; Miller et al. 2007). Another possibility may be that the Th1 response is slower in onset in this model, with only the up-regulation of IL-12p35 occurring in the times evaluated. Perhaps later samples after OG1 may have an increase in the production of IFN-γ or IL-2 mRNA. A delayed Th1 response has been reported where the stimulation of monocytes, natural killer cells and T cells did not peak until after the innate inflammatory response occurred in a baboon model of pulmonary sepsis (Zhu et al. 2007).

The results from this study also suggest that the lamina does not mount an anti-inflammatory response during the DEV or OG1 phases of laminitis, as no changes in IL-10 or IL-4 mRNA concentrations were found. These 2 cytokines are well known in the adaptive Th2 response and IL-10 in particular is not only known to be up-regulated in human sepsis, but also believed to be prognostic for outcome in human patients with sepsis/SIRS (Scumpia and Moldawer 2005; Miller et al. 2007). Laminar IL-10 mRNA concentrations were found to be decreased early in the BWE model at the 3 h DEV time, but were back to control values by the time lameness occurred (Belknap et al. 2007). The fact that IL-10 is not up-regulated in the CHO or BWE model of laminitis, when it is known to be a prominent component of human sepsis/SIRS, could be explained by the absence of a TNF-α response in these same horses. TNF-α is known to be a potent inducer of systemic IL-10 release and immunisation of chimpanzees with an anti-monoclonal TNF-α antibody resulted in significant attenuation (by >60%) of the plasma IL-10 response to endotoxin (van der Poll et al. 1994; Scumpia and Moldawer 2005). No change in laminar IL-10 mRNA concentrations may also be explained by compartmentalisation of the inflammatory response similar to what occurs with sepsis/SIRS, where the pathophysiolgical events differ from organ to organ or from organ to peripheral blood (Cavaillion and Annene 2006). Therefore, although IL-10 does not change in the lamina in response to CHO overload, it is possible that other organs may have increased or decreased IL-10 expression in response to the CHO challenge. The lack of IL-4 response should not be due to a lack of lymphocytes, as both T- and B-lymphocytes are present in the laminar tissue (Faleiros et al. 2010). Finally, it is possible that, as expression of anti-inflammatory cytokines is a late event in very sick individuals in human sepsis studies (Scumpia and Moldawer 2005), laminar IL-4 or IL-10 response may not occur until much later in the disease process or only occur in a more severe disease process than that assessed in this model.

Increases in mRNA concentrations of the leucocyte adhesion molecules E-selectin and ICAM-1 were present in this study at the OG1 time point when compared to controls. In the BWE model, these adhesion molecules were markedly up-regulated very early, peaking at 1.5 h after administration of BWE and decreasing precipitously in expression by the onset of lameness (Belknap et al. 2007; Loftus et al. 2007). The endothelium is highly active and quite dynamic, responding rapidly to its local environment as indicated by the induction of ICAM-1 and E-Selectin by increased concentrations of proinflammatory cytokines IL-1β and IL-6 (Stratowa and Audette 1995; Aird 2003). Endothelial activation in the lamina, demonstrated by increased mRNA concentrations of adhesion molecules, supports additional preliminary findings from our laboratory, in which an influx of leucocytes is seen throughout the laminar dermis at the developmental and OG1 time points after BWE administration (Faleiros et al. 2009). The delayed increases in mRNA concentrations of adhesion molecules observed in the present study when compared to previous reports in the BWE model may be explained by the delayed production of local cytokines, such as IL-1β and IL-6, which have a positive effect on the transcriptional regulation of these adhesion molecules (Stratowa and Audette 1995; Loftus et al. 2007). Furthermore, presence of up-regulated adhesion molecules during the onset of lameness may be indicative of the severity of laminar injury in this model, as a similar study comparing nonsurvivors of sepsis to survivors found that continued elevation in ICAM-1 was found more commonly in nonsurvivors (Hein et al. 2005).

Similar to the difference in cytokine and adhesion molecule gene expression between the BWE and CHO models, laminar COX-2 mRNA concentration was only increased at the onset of lameness (vs. increases at both developmental and lameness time points in the BWE model; Waguespack et al. 2004a; Blikslager et al. 2006). No change in COX-1 mRNA concentrations were present at any time point in this study, which is consistent with other reports and is not unexpected considering COX-1 is constitutively expressed unlike COX-2 which is inducible in most tissues by inflammation (Waguespack et al. 2004a; Tsatsanis et al. 2006). Although increases in mRNA expression of the COX isozymes may not directly correlate to protein concentrations and enzyme activity (Wilson et al. 2004), increases in COX-2 protein concentrations have been documented in addition to an increase in mRNA concentrations in the BWE model of laminitis (Waguespack et al. 2004a; Blikslager et al. 2006). Additionally, the presence of cytokines (such as IL-1β), TLR ligands and hypoxia (all of which are known or suspected to occur during the development of laminitis) can affect several post transcriptional modulations in the stabilisation of COX-2 mRNA (Tsatsanis et al. 2006; El-Achkar et al. 2007). The importance of COX-2 in models of sepsis is demonstrated by the use of COX-2 null mice exposed to endotoxin in which increased survival, decreased leucocyte influx into the kidneys and lungs, reduced NF-κB activation, and greatly increased IL-10 production occurred compared to the wild type mice (Ejima et al. 2003). Further research is warranted in evaluating the use of the use of the new COX-2 equine specific inhibitor firocoxib to determine if it may be more beneficial in these cases. It will, however, be important to determine if the negative vascular effects observed in some human patients on coxibs (such as increase in thrombus formation or vasoconstriction) will adversely affect patients with laminitis (Mitchell and Warner 2006).

It has been documented that not all horses are ‘responders’ to the CHO model, with up to 25% of horses not exhibiting clinical signs of laminitis while the majority of animals in the same studies developed Obel grades 1–2 laminitis (Weiss et al. 1997). The reason that some horses are nonresponsive to CHO administration resulting in their protection from laminar injury is unknown. Although these nonresponsive horses have usually been removed from previous studies, we were interested in comparing laminar inflammatory events in the nonresponsive animals to those animals that responded to the clinical model. Similar to previous studies, approximately 25% (2 in each of the DEV and OG1 groups) of the horses did not respond to the administration of CHO. This lack of response was not only present in the laminar tissue, but was present at the systemic level with none of the 4 ‘nonresponders’ exhibiting a febrile response. Interestingly, none of the laminar inflammatory events present in the responsive horses in each group were present in these ‘nonresponders’ (see Fig 1). Although there were not enough ‘nonresponders’ for statistical analyses, the consistent lack of inflammatory signalling in the 4 nonresponsive horses suggests that the protection from CHO-induced laminar injury is upstream of a central event inducing the expression of diverse inflammatory molecules including a wide array of proinflammatory cytokines and chemokines, endothelial adhesion molecules, and COX-2. It is possible that this lack of response may simply be explained by a different bacterial flora in the intestinal tract that does not result in the decrease in caecal pH, endotoxin production or the subsequent damage to the caecal mucosa observed in CHO overload (Garner et al. 1978); differences in bacterial flora have recently been demonstrated to affect the response to sepsis in models of human sepsis (Opal and Cohen 1999; Sriskandan and Cohen 1999). Another explanation may be that the nonresponders have genetic polymorphism(s), as has been reported in human sepsis for such genes as TNF-α and TLR4, that result(s) in their protection and lack of laminar inflammation (Stuber et al. 1996; Cohen 2002; Dahmer et al. 2005). Further research involving the protective mechanisms in these horses that keep them from developing clinical signs associated with CHO administration will be valuable in discovery of effective treatments for laminitis.

Overall, the CHO administration results in a laminar inflammatory response delayed in onset when compared to the early fulminant inflammatory response in the BWE model (Fontaine et al. 2001; Waguespack et al. 2004a,b; Belknap et al. 2007; Loftus et al. 2007). Whereas animals in the BWE model undergo a rapid and aggressive inflammatory response by 1.5 h following BWE administration, the inflammatory gene expression in the majority of animals in the current study were no higher than control values at the DEV time point for most genes assessed. Interestingly, the same 2 animals in the DEV group were responsible for the 2 highest values for ICAM-1, COX-2, IL-1β and IL-6, indicating that the majority of variability in laminar inflammatory gene results observed between horses is due to these 2 undergoing inflammatory signalling changes more rapidly than the other animals exposed to the same stimulus. As a more consistent inflammatory response is observed in the OG1 group, it is likely that a laminar inflammatory response is initiated in many of the animals in the time period between the onset of clinical signs of inflammation (fever in this case) and onset of lameness. It is likely that the CHO model simply differs from the BWE model as demonstrated by the similarities and differences between the 2 models used to induce experimental sepsis in rodents, the CLP and endotoxin models (Villa et al. 1995; Remick et al. 2000; Poli-de-Figueiredo et al. 2008). In the BWE model, the rapid absorption of a substance which induces a profound systemic inflammation is the likely cause of laminar inflammation in this model, similar to the rapid and marked systemic inflammatory response following the administration of LPS in experimental animals. In the CHO model, the absorption of an overwhelming amount of inflammatory trigger molecules most likely takes several hours, due to the time it takes for the intraluminal changes in flora and injury to the intestinal epithelium with resulting loss of the barrier function of the hindgut to occur. This is similar to the CLP rodent model of sepsis where there is a more gradual build-up but more sustained presence of bacterial products in the circulation than endotoxin models, leading to a similar organ injury as seen in clinical sepsis. It is unknown if the laminar inflammatory response is sustained for a longer period of time after CHO administration as no samples were collected after the OG1 time point. However, if sustained, the inflammatory response is likely to play an important role in the protracted and increased severity of lameness and laminar injury that often occurs after grain overload.

In conclusion, results from this study suggest that, although delayed when compared to the BWE model, a marked laminar inflammatory response occurs in the CHO model of laminitis. Furthermore, with the numerous redundant inflammatory mediators produced in the lamina after CHO administration, targeting multiple inflammatory signalling pathways may be necessary to stop the development of laminitis. However, the fact that all laminar inflammatory events documented to occur in affected laminae did not occur in the horses which did not respond clinically to CHO overload leaves open the possibility that there still may be a central inflammatory signalling mechanism, which, if blocked, may inhibit all deleterious inflammatory events likely to lead to severe laminar injury. Research evaluating new anti-inflammatory therapies, particularly those aimed at signal transduction and transcriptional control of inflammation will be essential to the search of finding successful treatments/preventions for this devastating disease (Ivanenkov et al. 2008).


Supported by USDA NRI CSREES 2007-35204-18563. Presented in part at the American College of Veterinary Internal Medicine Conference, Montreal, Canada, June 2009 and at the Fifth International Equine Conference on Laminitis, West Palm Beach, Florida, November 2009.

Manufacturers' addresses

1 Stratagene, LaJolla, California, USA.

2 Roche, Indianapolis, Indiana, USA.

3 Ambion Inc, Austin, Texas, USA.

4 Invitrogen, Carlsbad, California, USA.

5 Ghent University, Ghent, Belgium.