Laminitis is a painful and often life-threatening disease of the lamellar tissue that connects the axial skeleton of a horse with the inner surface of the hoof capsule. Depending on the underlying cause, the disease can be categorized as SIRS-induced laminitis, endocrinopathic laminitis, and supporting limb laminitis.1,2 A characteristic feature of SIRS-induced laminitis is damage of the lamellar dermoepidermal junction, which is believed to result from changes in the basal epidermal cells and degradation and dysadhesion of the adjacent basement membrane.3–5 This basement membrane is attached to the underlying dermal stroma via numerous anchoring fibrils mainly composed of type VII collagen.6,7 In experimentally induced SIRS-related laminitis, basement membrane damage is associated with loss of type VII collagen,3 and therefore, it appears plausible that degradation of this collagen type contributes to the lamellar failure characteristic of SIRS-induced laminitis in horses.
Like most constituents of the basement membrane, type VII collagen of the anchoring fibrils can be degraded by MMPs, which play a major role in the remodeling of the extracellular matrix in several physiologic and pathological processes.8,9 Of all known MMPs, only 2 collagenases (MMP-1 and MMP-8) and the 2 gelatinases (MMP-2 and MMP-9) are capable of cleaving type VII collagen9; therefore, these 4 MMPs may be relevant to the pathogenesis of SIRS-induced laminitis in horses.
Matrix metalloproteinase-1 (collagenase-1) is produced by several lamellar cell types, including fibroblasts, keratinocytes, and endothelial cells, with increased production on demand (eg, by inflammatory stimulation).8 This protein is capable of degrading type VII collagen of the anchoring fibrils and also other components of the basement membrane,8,9 and it is therefore of particular interest in laminitis research.5,10 The amount of gene expression and cellular production of MMP-1 was increased in horses with versus without CHO-induced laminitis (an often-used in vivo model of SIRS-related laminitis) in a previous study,5 but the contribution of MMP-1 to the pathological remodelling of the lamellar tissue remained doubtful because only the inactive form of MMP-1 was detected.
Matrix metalloproteinase-8 (collagenase-2) is predominantly associated with neutrophils, which store this protein in their specific granules and release it instantly after neutrophil activation,8 but MMP-8 can also be produced at lower amounts by other inflammatory cells as well as by keratinocytes, fibroblasts, smooth muscle cells, and endothelial cells.11,12 The MMP-8 can cleave type VII collagen, but it can also degrade other collagens of the adjacent dermal stroma.8,9 To the authors’ knowledge, the relevance of MMP-8 in the pathogenesis of laminitis has not yet been investigated, but owing to its association with activated neutrophils, one could reasonably anticipate that it may play a role in the pathogenesis of SIRS-induced laminitis.
Matrix metalloproteinase-2 (gelatinase A) is produced by a variety of lamellar cell types, including keratinocytes, fibroblasts, and endothelial cells.8 Matrix metalloproteinase-9 (gelatinase B) is constitutively produced in neutrophils, stored in their tertiary granules, and released from the granules on neutrophil activation, but its expression is also inducible in other cell types, including keratinocytes and endothelial cells with increased expression due to stimulation by inflammatory mediators or by hypoxia.8,13,14 Both MMPs can degrade type VII collagen and other components of the basement membrane.8,9,15 The role of MMP-2 and MMP-9 in the pathogenesis of SIRS-induced laminitis remains unclear; both enzymes have long been associated with degradation of the basement membrane in the course of laminitis, but their ability to initiate the typical lamellar basement membrane damage has been questioned given findings of in vivo laminitis experiments.4,16,17
An established risk factor for SIRS-induced laminitis is endotoxemia, which develops when LPS from gram-negative bacteria enter the circulation, which can happen as a result of various inflammatory diseases, such as gastrointestinal diseases, pleuropneumonia, retained placenta, or metritis.18–21 Direct exposure of the lamellar tissue to LPS through extracorporeal perfusion of isolated equine limbs has been shown to induce pathological changes consistent with those of CHO-induced and naturally acquired laminitis.22,23
The purpose of the study reported here was to use extracorporeal perfusion of isolated equine limbs with LPS-supplemented perfusate as an ex vivo model of SIRS-induced laminitis to investigate the effect of LPS on the expression of MMPs in the lamellar tissue that are capable of cleaving type VII collagen of the anchoring fibrils. Our hypothesis was that the perfusion of the isolated limbs with a clinically relevant concentration of LPS would result in a significant increase in lamellar presence of at least one of these MMPs.
Materials and Methods
Limb collection
Archived paraffin-embedded lamellar tissue blocks from previous studies22,24 were used in the present study. In those studies, right forelimbs (mean ± SD limb weight, 6.74 ± 1.69 kg) had been collected by disarticulation at the middle carpal joint from 10 horses of various breeds (mean ± SD age, 14.0 ± 6.1 years; mean body weight, 516 ± 70 kg) directly after death following routine slaughter processing at a local licensed abattoir. The left forelimbs of 3 of these horses had also been collected for use as control limbs. Prior to slaughter, investigators in the previous studies22,24 had examined the horses at walk and trot and assessed them for lameness. They had also performed a brief physical examination to detect obvious signs of a systemic disease. Horses were excluded from those studies if any signs of systemic or hoof disease were detected. Horses had been euthanized for nonresearch purposes by administration of a penetrating captive bolt and subsequent exsanguination in the routine slaughter process. Because only the limbs and blood of horses slaughtered at a licensed abattoir were used for the present study, no approval from an institutional care and use committee or research oversight entity was required.
Limb processing and perfusion
Isolated right forelimbs had been randomly assigned to 2 groups: perfusion under physiologic conditions (control-perfusion group; n = 5) or under the same conditions with addition of 80 ng of endotoxin (LPS from Escherichia coli type O55:B5)a/L of perfusate (LPS-perfused group; 5). This LPS concentration was chosen because it is within the range of the plasma endotoxin concentration reported to exist in horses with gastrointestinal or infectious (septicemic) diseases (5 to 825 ng/L)25–27 and is also at the upper end of the range detected in horses with CHO-induced laminitis (2.4 to 81 ng/L).28
Perfusion of each isolated right forelimb had been performed immediately after collection as previously reported.22,24 Briefly, the median artery was immediately flushed with ice-cold oxygenated and heparinized flushing solution, and the radial artery and the palmar branch of the median artery were ligated. For transportation to the laboratory (range in transportation times, 60 to 120 minutes), perfused right and nonperfused left forelimbs were placed on ice. Right forelimbs were then immediately connected to the perfusion system via medical-grade polyvinyl chloride tubing (3-mm internal diameter), which was directly inserted into the median artery by means of anatomic thumb forceps for stabilization. Venous return was collected via tubing inserted into the 2 main venous vessels (radial vein and palmar branch of the median vein). Perfusion was subsequently performed with heparinized autologous blood (diluted at a ratio of 3:2 with autologous blood plasma), which was oxygenated and warmed to 35°C. The perfusate was exchanged every 2 hours to prevent mechanically induced hemolysis. In 1-hour intervals, supplementation of the perfusate with glucose was performed to compensate for glucose use by cells in the perfused tissues. Total duration of perfusion was 10 hours, including an initial equilibration period of 30 minutes in which the blood flow was slowly increased to 12 mL/kg/min. For monitoring of the viability of the perfused tissues, glucose use as well as lactate and lactate dehydrogenase production were measured, and blood gas analysis was performed at 1-hour intervals.
Tissue preparation
Immediately after perfusion of right forelimbs or on receipt of control forelimbs at the laboratory, blocks of lamellar tissue (approx 10 × 10 × 10 mm) had been obtained from the proximal and distal portions of the dorsal aspect of the hoof. Specimens had been fixed in 4% formaldehyde, embedded in paraffin, and kept in a dark room-temperature (approx 20°C) storage unit until further processing.
Immunohistochemical analysis
Immunohistochemical staining of the lamellar tissue was performed in the present study by use of the following antibodies: polyclonal rabbit anti–human MMP-1 antibodyb (dilution, 1:600; this antibody is also recommended for detection of equine MMP-1), polyclonal rabbit anti–human MMP-2 antibodyc (dilution, 1:100; this antibody is also predicted to have cross-reactivity with equine tissue), polyclonal rabbit anti–human MMP-8 antibodyb (dilution, 1:100), and polyclonal rabbit anti– human MMP-9 antibodyd (dilution, 1:100). All 4 antibodies can be used to detect total MMP without differentiation between the inactive proenzyme and the active form as reported by the antibody manufacturers. Detection of active MMP-9 was performed with monoclonal mouse anti–human MMP-9 activated antibodye (clone 4A3; dilution, 1:50), which recognizes only the active form of MMP-9 and does not react with the MMP-9 proenzyme.29 To determine a suitable staining dilution with little background staining, dilution series were evaluated prior to the staining. For each antibody, all specimens were stained in a single batch.
For immunohistochemical staining, archived paraffin-embedded blocks of lamellar tissue were sectioned at 4-μm thickness. Afterwards, sections were mounted on silane-coated glass slides, deparaffinized, rehydrated, and incubated in 0.6% hydrogen peroxide for 15 minutes. For antigen retrieval, sections were heat treated (65°C) with a citrate buffer (pH, 6.0) for 2 hours (for MMP-2 and MMP-9) or with a Tris-EDTA buffer (pH, 9.0) for 30 minutes (for MMP-1 and activated MMP-9). According to information from the antibody manufacturer, no antigen retrieval was necessary for MMP-8. The sections were then incubated in 1.5% goat serum in PBS solution for 30 minutes. Afterward, slides were incubated overnight at 4°C with the primary antibody diluted in PBS solution. Additional sections (1 slide/antibody) were incubated with PBS solution alone to serve as negative control specimens. Slides were then incubated with a secondary antibody (horseradish peroxidase–conjugated anti-rabbitf or anti-mouseg antibody) for 30 minutes at room temperature. Finally, sections were incubated for 10 minutes in 3,3′-diaminobenzidine tetrahydrochloride to visually identify the antigen as a brown precipitate and counterstained with hemalum for 3 minutes. For each antibody, samples from a human placenta and samples from the forehooves of a horse that had naturally acquired acute clinical laminitis10,30 were treated with the same staining protocol and served as positive control specimens.
All slides were labeled with a randomly assigned unique alphanumeric code and then evaluated via light microscopy at 25X magnification by 1 investigator (BP-Z), who was blinded to the identity of specimens. The intensity of staining was subjectively scored by use of a semiquantitative scoring system that ranged from 0 to 3 in increments of 0.5.31–33 Score 0 was assigned when no stained cells were visible, and scores 1, 2, and 3 were assigned when the degree of staining was assessed as mild, moderate, or intense, respectively. To examine the distribution of the MMPs in the lamellar tissue, 9 locations were evaluated in each section (Figure 1): basal and suprabasal epidermal cells (stratum basale and spinosum) at the tip, in the middle, and at the base of the primary epidermal lamellae; dermal tissue at the tip, in the middle, and at the base of the primary dermal lamellae as well as in the supralamellar region (above the tip of the primary epidermal lamellae); and endothelial cells and vascular smooth muscle cells in the supralamellar region. For this evaluation, the entire lamellar tissue section was considered with evaluation of 1 slide/archived tissue block and antibody. Because 2 lamellar tissue specimens (proximal and distal portion of the dorsal aspect of the hoof) were obtained per limb, a total of 2 slides were evaluated per limb. The total score per slide (sum of all scores for the 9 locations; minimum, 0; maximum, 27) was used for statistical evaluation.
Photomicrograph of an equine lamellar tissue section showing the locations of 9 regions evaluated for the presence of 4 MMPs capable of cleaving type VII collagen of the anchoring fibrils. Images in panel A are enlargements of the boxed regions in panel B. Numbered areas represent basal and suprabasal epidermal cells (stratum basale and spinosum) at the tip (1), in the middle (2), and at the base (3) of the primary epidermal lamellae and dermal tissue at the tip (4), in the middle (5), and at the base (6) of the primary dermal lamellae as well as in the supralamellar region (above the tip of the primary epidermal lamellae; 7). Numbered areas also represent endothelial cells (8) and vascular smooth muscle cells (9) in the supralamellar region. Immunohistochemical stain for MMP-1; bars = 100 μm in panel A and 200 μm in panel B.
Citation: American Journal of Veterinary Research 79, 9; 10.2460/ajvr.79.9.986
Statistical analysis
Data analyses were performed by use of a commercial software program.h Because of the small sample size, all data analyses were performed with nonparametric tests. Data were considered to differ significantly at values of P < 0.05. For comparison of the immunohistochemical scores for lamellar tissue sections (n = 2/limb) from the proximal versus distal portion of the dorsal aspect of the hoof, the Wilcoxon test was used. For all used antibodies, scores for the proximal and distal specimens did not differ significantly; therefore, the mean of the scores for the 2 sections was calculated for each limb. These mean scores were used for further statistical analyses.
To control for differences caused by the perfusion, comparisons between perfused (right) and nonperfused (left) forelimbs were performed to include data only for the 3 horses that provided both forelimbs. This comparison was performed with the Wilcoxon test. Comparisons between the LPS-perfused and control-perfused right forelimbs were performed with the Mann-Whitney U test. Descriptive statistics such as median and IQR were used to summarize the results.
Results
Immunohistochemical analyses
All of the antibodies used in the immunohistochemical analyses had a distinct reaction in the positive control samples, whereas no reaction was visible in the negative control samples. No significant difference was identified in immunohistochemical scores between nonperfused control limbs and control-perfused equine limbs for any antibody (Figure 2).
Immunohistochemical scores for detection of total (active and latent) MMP-1, MMP-2, MMP-8, and MMP-9 as well as of active MMP-9 in equine lamellar tissue sections. A—Comparison of mean scores for 2 sections (1 each from the proximal and distal portions of the dorsal aspect of the hoof) between isolated equine forelimbs that underwent perfusion with autologous blood for 10 hours (control-perfused group; gray circles) and nonperfused contralateral forelimbs (white circles; n = 3/group). B—Comparison of the mean scores for the same 2 sections between the control-perfused group (circles) and for 2 similar sections from forelimbs that underwent perfusion as described for panel A with addition of 80 ng of LPS/L of perfusate (diamonds; n = 5/group). In both situations, the amount of visible staining was scored on a scale from 0 to 3 for each of the 9 locations in Figure 1 (maximum score/stain, 27). *,†Median values for the bracketed data points differ significantly (*P < 0.05; †P < 0.01) between LPS-perfused and control-perfused groups.
Citation: American Journal of Veterinary Research 79, 9; 10.2460/ajvr.79.9.986
Comparison of the scores of lamellar tissue sections from LPS-perfused limbs versus control-perfused limbs revealed no significant differences in results for MMP-1, MMP-2, and MMP-8. However, scores for total and active MMP-9 were significantly different between these 2 groups (Figure 2). For total MMP-9, the median score for LPS-perfused limbs was significantly (P = 0.01) higher (14.5; IQR, 13.5 to 16) than that for control-perfused limbs (11; IQR, 10 to 11.5). A similar result was obtained for the active form of MMP-9, for which the median score for LPS-perfused limbs was also significantly (P = 0.009) higher (10.5; IQR, 10 to 11.5) than that for control-perfused limbs (7.5; IQR, 4 to 9).
Distribution of MMPs
In all groups, mild to moderate staining for MMP-1 was detected in endothelial cells, vascular smooth muscle cells, keratinocytes (particularly basal epidermal cells), and fibroblasts in the entire lamellar region (Figure 3). The staining pattern for MMP-2 was also similar in all groups, with mild to moderate staining of the basal and suprabasal epidermal cells as well as of the dermal tissue, including fibroblasts, and dermal extracellular matrix, endothelial cells, and vascular smooth muscle cells. For MMP-8, all groups had mild to moderate staining of intravasal neutrophils, keratinocytes, endothelial cells, fibroblasts, and dermal extracellular matrix. The dermal tissue in the secondary dermal lamellae and basal epidermal cells singularly or in groups had a moderate to intense reaction.
Representative photomicrographs of results of immunohistochemical staining for MMP-1, MMP-2, and MMP-8 in equine lamellar tissue sections from the nonperfused control limbs and control- and LPS-perfused limbs described in Figure 2, with the brown color indicating the presence of the tested MMP. The MMP-1 staining in a nonperfused control section (A) and control-perfused (B) and LPS-perfused (C) sections has a similar pattern, with a mild (score = 1) reaction of the dermal extracellular matrix and moderate (score = 2) reaction of keratinocytes, fibroblasts, and endothelial and vascular smooth muscle cells. The MMP-2 staining of similar sections (D, E, and F, respectively) reveals a mild to moderate (score = 1.5) reaction for dermal tissue and moderate reaction for keratinocytes and endothelial cells. The MMP-8 staining of similar sections (G, H, and I, respectively) reveals a mild reaction of keratinocytes with intense (score = 3) staining of single basal epidermal cells (arrowheads) and a moderate reaction of the dermal tissue with distinct (score = 3) staining in some secondary dermal lamellae (arrows). Bar = 50 μm in all panels. See Figure 2 for remainder of key.
Citation: American Journal of Veterinary Research 79, 9; 10.2460/ajvr.79.9.986
For total MMP-9 (Figure 4), moderate staining of the epidermal basal cells and mild staining of the suprabasal epidermal cells were detected in the entire lamellar region in tissue sections from nonperfused control and control-perfused limbs. In both groups, the dermal tissue (fibroblasts and dermal extracellular matrix) had a mild staining pattern, with a decrease from the base to the tip of the primary dermal lamellae. The endothelial and vascular smooth muscle cells in the supralamellar region also had mild staining. In the sections from LPS-perfused limbs, mild MMP-9 staining of the dermal extracellular matrix, moderate staining of fibroblasts and suprabasal epidermal cells, and intense staining of the basal epidermal cells were visible in the entire lamellar region. In the supralamellar region, endothelial and vascular smooth muscle cells as well as intravasal neutrophils had a moderate MMP-9 staining reaction.
Representative photomicrographs of results of immunostaining for total (active and latent) MMP-9 (A to D) and active MMP-9 (E to I) in the types of sections described in Figure 3. The brown color indicates the presence of MMP-9 in the lamellar tissue. The total MMP-9 staining reveals a mild (score = 1) reaction of the dermal tissue and suprabasal epidermal cells and a moderate (score = 2) reaction of the basal epidermal cells in a nonperfused control (A) and control-perfused (B) section. Also revealed is a moderate reaction in fibroblasts, endothelial, and suprabasal epidermal cells and an intense (score = 3) reaction in basal epidermal cells (C) in an LPS-perfused section as well as a moderate reaction in fibroblasts and endothelial and vascular smooth muscle cells in the supralamellar region (D) of a similar section. The active MMP-9 staining reveals a mild reaction of keratinocytes and endothelial cells in a nonperfused control (E) and control-perfused (F) section; mild to moderate (score = 1.5) staining of keratinocytes, dermal tissue, and intravasal neutrophils; and moderate reaction in endothelial cells in an LPS-perfused section (G). Also revealed is a moderate reaction in fibroblasts and endothelial cells in the supralamellar region of an LPS-perfused section (H) as well as an intense reaction in the suprabasal epidermal cells at the tip of the primary epidermal lamellae in a similar section (I). Bars = 100 μm in panels A and B, 50 μm in panels C and E to I, and 25 μm in panel D. See Figures 2 and 3 for remainder of key.
Citation: American Journal of Veterinary Research 79, 9; 10.2460/ajvr.79.9.986
Active MMP-9 had a similar distribution in tissue sections, but the staining intensity was much lower than that for total MMP-9 (Figure 4). In tissue sections from nonperfused control and control-perfused limbs, staining for active MMP-9 was absent or only mild in the basal and suprabasal epidermal cells in the entire lamellar region. Fibroblasts had mild staining at the base, but in most sections, no staining was visible in the middle and at the tip of the primary dermal lamellae. In the supralamellar region, the extravasal dermal cells as well as the endothelial and vascular smooth muscle cells had mild staining. In tissue sections from LPS-perfused limbs, the dermal tissue as well as the basal and suprabasal epidermal cells in the entire lamellar region had mild to moderate active MMP-9 staining, whereas the fibroblasts and endothelial cells in the supralamellar region had a moderate reaction. In sections from 2 of 5 LPS-perfused limbs, the suprabasal epidermal cells had moderate active MMP-9 staining at the base of the primary epidermal lamellae and an intense reaction at the tip of the primary epidermal lamellae.
Discussion
Endotoxemia is regarded as a risk factor for equine laminitis,18–20 but its role as a direct cause of laminitis is controversial, given that short-term infusion of LPS is not successful in inducing laminitis in otherwise healthy horses.34–37 However, that finding can be explained by the rapid clearance and the resulting brief half-life of circulating endotoxin (2 to 10 minutes in horses after experimental induction of endotoxemia).25,35 After jugular LPS infusion with a bolus of 30 ng/kg, the peak plasma endotoxin concentration is only 13.2 ng/L and decreases within 20 minutes to a concentration < 4 ng/L,35 which is a concentration regarded as the upper limit at which a horse can remain clinically normal.27 Therefore, the duration of effective endotoxemia after bolus jugular LPS infusion at 20 to 30 ng/kg (a dose often used to induce short-term endotoxemia in vivo)34–37 may be too brief to induce pathological changes in the lamellar tissue.
The effectiveness of endotoxemia for laminitis induction can be deduced from another study,38 in which hepatic portal LPS infusion with a concentration of 1 μg of LPS/kg/h over a 24-hour period was successful in inducing hoof discomfort and weight shifting,38 which are clinical signs that suggest that laminitis was induced. Additionally, direct exposure of the lamellar tissue to LPS through extracorporeal perfusion of isolated equine limbs was shown to induce an inflammatory reaction in the lamellar tissue with intra- and extravascular accumulation of leukocytes, migration of activated neutrophils to the dermoepidermal basement membrane, damage of the basement membrane, and invasion of neutrophils into the epidermal cell layer.22,23 These pathological changes are similar to those described for horses with CHO-induced and naturally acquired laminitis.10,39,40
The technique of isolated limb perfusion used in the present study has the advantage that the effect of continuous 10-hour LPS exposure on the lamellar tissue can be investigated under controlled conditions and without systemic influences. Extracorporeal perfusion of isolated equine limbs offers the possibility of examining complex relationships and interactions in the lamellar tissue that are neglected in simplified in vitro studies, without the ethical dilemma of animal experiments of in vivo studies. However, an important limitation of the setup used in the present study was the time required to transport collected limbs to the laboratory, providing for a period of up to 2 hours of cold ischemia of the limb. The effects of ischemia and reperfusion as well as the effects of extracorporeal perfusion were assessed by comparison of nonperfused control samples and samples obtained after control perfusion, which revealed no significant differences in immunohistochemical scores for any antibody used.
For LPS-perfused limbs, only immunohistochemical scores for total and active MMP-9 were significantly increased, compared with scores for control-perfused limbs, whereas scores for MMP-1, MMP-2, and MMP-8 were not significantly different between these 2 groups. We suggest that these results may indicate that the latter MMPs are of less relevance for degradation of type VII collagen and other components of the lamellar basement membrane in the developmental phase of endotoxin-induced laminitis in horses. However, a possible LPS-induced activation of latent MMPs preexisting in the lamellar tissue would not be detectable given that the antibodies used in our study react with both the latent and active forms of the MMP. The probability of such an activation is fairly high because LPS exposure results in an inflammatory reaction in the lamellar tissue,23 and it is known that the investigated MMPs can be activated directly by endotoxin or indirectly by inflammatory factors such as reactive oxygen species.8,15,41
Findings in the present study for MMP-2 were not surprising because this MMP is produced constitutively in most cell types and its expression is not induced by inflammation.8 However, MMP-1 production increases on demand after stimulation (eg, by inflammatory mediators) by initiating gene transcription.5,8 Because LPS perfusion is associated with lamellar inflammation,23 we expected an increase in staining for MMP-1. But activation of the transcriptional apparatus needs time, and therefore, accumulation of new MMPs in the tissue is delayed by 6 to 12 hours.8 In the present study, the duration of LPS exposure may have been too brief to cause a significant increase in gene expression and new synthesis of MMP-1, but such an increase may be expected with a longer duration, similar to that in the early stage of CHO-induced laminitis in vivo.5 However, MMP-9 is also released from activated neutrophils,8 and these inflammatory cells accumulate in the lamellar tissue after LPS exposure.23 The combination of new synthesis by lamellar cells with the release of MMP-9 from migrated neutrophils may be the cause for the significant increase in MMP-9 staining that we observed.
To our knowledge, the present study represented the first involving evaluation of MMP-8 in an equine laminitis model, and the results were quite surprising. Similar to MMP-9, this MMP is predominately released from activated neutrophils,8 and we presumed that its presence would be greater in the inflamed lamellar tissue from LPS-perfused limbs than in the noninflamed tissue from control-perfused limbs. Nevertheless, the amount of MMP-8 associated with LPS perfusion was not significantly different than after control perfusion. A possible explanation is that neutrophils contain fairly low quantities of MMP-8 in their granules and that, when activated, neutrophils release only 15% to 20% of the stored MMP-8.11 Because MMP-8 is believed to be expressed only in small quantities in keratinocytes and fibroblasts,11,12 it was also interesting to find that the lamellar dermal tissue (particularly in the secondary dermal lamellae) and also some keratinocytes had such a distinct MMP-8 staining pattern even in perfused and nonperfused control specimens. The observed staining patterns suggested that in equine lamellar tissue, MMP-8 may play a physiologic role in migration of keratinocytes along the basement membrane in hoof wall growth.
Given the controversy surrounding the importance of MMP-9 in the pathogenesis of laminitis, it was also interesting to note that of all examined collagen type VII–degrading MMPs, only amounts of total and also active MMP-9 were significantly greater in LPS-perfused versus control-perfused tissues. This result suggested that MMP-9 may play a role in the degradation of collagen type VII in the initial phase of endotoxin-induced laminitis. Our findings were supported by those of other studies,30,42,43 in which a relationship was identified between high amounts of MMP-9 and laminitis. On the other hand, some research4,10,16 regarding equine laminitis has shown that the amount of latent MMP-9 proenzyme is predominantly increased, whereas the amount of active MMP-9 is unaffected, in horses with experimentally SIRS-induced laminitis and those with naturally acquired acute and aggravated chronic laminitis. This discrepancy may be due to methodological differences because the latter studies involved gelatin zymography for detection of MMP-9, and although gelatin zymography is considered the technique of choice for detection of active MMPs in fluids, it can be problematic for the determination of the actual activity of MMPs in tissues.44,45 With regard to active MMP-9, gelatin zymography may yield unreliable results owing to possible inactivation of active MMP-9 by the protein extraction procedure and a possible lack of sensitivity to identify focal areas of MMP-9 activity in the lamellar tissue.10
In this context, it is noteworthy that MMP-9 gene expression also increases in lamellar tissue of horses with experimentally CHO-induced and naturally acquired acute laminitis.30 However, no increased expression of this gene was detectable in lamellar basal epidermal cells isolated by laser capture microdissection and analyzed by broad transcriptome analysis technique in horses with CHO-induced laminitis,17 whereas in the present study, immunohistochemical scores for total and active MMP-9 were higher in basal epidermal cells of the lamellar tissues from LPS-perfused limbs than in those from control-perfused limbs. An explanation may be that examination of MMP-9 gene expression allows assessment of the potential for new synthesis of MMP-9 but not activation of preexisting zymogen nor an increase in the amount of MMP-9 release by activated neutrophils (which migrate toward and even into the basal epidermal layer after LPS exposure23).
Another explanation for the discrepancy between findings in previous studies and those of the present study may be that tissue was directly exposed to LPS in the present study, whereas other studies involved evaluation of laminitis in vivo, in which other factors may be more important in triggering the disease. Endotoxin is known to provoke the release of neutrophil elastase from primary granules of stimulated neutrophils,46 and neutrophil elastase is an activator of MMP-9.16 Therefore, the ex vivo model of laminitis used in the present study may have led to a higher amount of active MMP-9 than in other settings in which endotoxin may be of less relevance. A lack of specific MMP-9 activators in some laminitis research was also discussed in a report43 of experimental intestinal obstruction associated with an increase in the amount of active MMP-9 in lamellar tissue at a point corresponding to the developmental stage of colic-induced laminitis.43 To clarify the conditions underlying the discrepancy between our results and those of other studies, it would be interesting to use the technique of the extracorporeal perfusion of isolated equine limbs with different detection methods such as gelatin zymography and MMP-9 gene expression analysis with immunohistochemistry.
Another consideration regarding the results of the present study was the use of antibodies tested for specificity only in human tissues. We could not exclude the existence of cross-reactivity between the used antibodies and other proteins or MMPs in the evaluated equine lamellar tissue. In human tissues, the antibody against active MMP-9 reacts effectively with the active form and does not appear to react with the proenzyme of MMP-9.29 Furthermore, equine and human MMP-9 are 100% homologous for the sequence of the immunogen used for the activated MMP-9 antibody (comparison of sequences by use of a bioinformatics search tooli; protein sequence of the immunogen, FQTFEGDLK). Therefore, it is likely, but not established, that this antibody also reacts in the equine tissue exclusively with the active form.
An additional limitation of the present study was that it involved an analysis at a single point in time, and no evidence exists that the increase in active MMP-9 causes the damage at the dermoepidermal interface that occurs in lamellar tissue with LPS perfusion in the same model of laminitis as used in the presentstudy.23 A possible next step could be a longitudinal study to confirm the correlation between both events in extracorporeally LPS-perfused isolated equine limbs.
Matrix metalloproteinase-9 is capable not only of degrading type VII collagen but also of cleaving other components of the basement membrane, such as type IV and XVIII collagen, nidogen (entactin), laminin, fibrillin, and fibronectin as well as type I, III, V, and XIV collagen of the underlying dermal stroma.9,15,47,48 Therefore, the distribution of MMP-9 in the LPS-perfused lamellar tissue may be relevant to the damage incurred by both types of basement membrane in endotoxin-induced laminitis. The increase in MMP-9 that was identified after LPS perfusion in the lamellar basal epidermal cells and adjacent fibroblasts in the present study may contribute to the damage of the dermoepidermal basement membrane and to dermoepidermal separation that occurs in horses with SIRS-induced laminitis. Damage of the vascular basement membrane of small veins and capillaries observed in horses with experimentally CHO-induced laminitis3 could be induced by an increase in the release of MMP-9 from the adjacent endothelial and vascular smooth muscles cells, given that these cell types had an increased amount of MMP-9 production after LPS exposure in the present study. Both types of basement membrane may also be damaged by neutrophil-derived MMP-9 during migration of neutrophils into the extravasal tissue, given that neutrophils were also observed as a source of MMP-9 after LPS perfusion.
Of the 4 MMPs that are capable of degrading type VII collagen, only amounts of total and active MMP-9 were significantly increased in the lamellar tissue after extracorporeal perfusion of the isolated equine limbs with a clinically relevant concentration of LPS, whereas amounts of total MMP-1, MMP-2, and MMP-8 were not significantly different from those of control-perfused limbs. These results suggested that of the examined MMPs, only MMP-9 may be involved in initial pathological processes in the lamellar tissue during the development of endotoxin-induced laminitis. Because the present study involved only subjective, semiquantitative evaluation of immunohistochemical staining patterns, additional research will be needed to confirm the relevance of MMP-9 in the pathogenesis of endotoxin-induced laminitis.
Acknowledgments
Supported by a research grant from the Vienna University of Veterinary Medicine and an award for the development of alternatives to animal experiments from the International Federation Against Animal Experiments (Internationaler Bund der Tierversuchsgegner).
The authors declare that there were no conflicts of interest.
The authors thank Magdalena Helmreich and Claudia Höchsmann for technical assistance and Dr. Alexander Tichy for help with the statistical analyses.
ABBREVIATIONS
| CHO | Carbohydrate overload |
| IQR | Interquartile (25th to 75th percentile) range |
| LPS | Lipopolysaccharide |
| MMP | Matrix metalloproteinase |
| SIRS | Systemic inflammatory response syndrome |
Footnotes
Sigma-Aldrich Handels GmbH, Vienna, Austria.
Santa Cruz Biotechnology Inc, Santa Cruz, Calif.
Abcam, Cambridge, England.
Abnova/EMBLEM, Heidelberg, Germany.
AbDSerotec, Oxford, England.
BrightVision poly-HRP-anti-rabbit antibody, ImmunoLogic, Duiven, Netherlands.
BrightVision poly-HRP-anti-mouse antibody, ImmunoLogic, Duiven, Netherlands.
IBM SPSS Statistics, version 24, IBM Corp, Armonk, NY.
BLAST, National Center for Biotechnology Information, National Institutes of Health, Bethesda, Md. Available at: blast.ncbi.nlm.nih.gov. Accessed Mar 8, 2016.
References
1. Orsini JA. Science-in-brief: Equine laminitis research: milestones and goals. Equine Vet J 2014;46:529–533.
2. Wells-Smith L. Laminitis in the 21st century. Vet Rec 2015;176:70–71.
3. Pollitt CC, Daradka M. Equine laminitis basement membrane pathology: loss of type IV collagen, type VII collagen and laminin immunostaining. Equine Vet J Suppl 1998;26:139–144.
4. Visser MB, Pollitt CC. The timeline of metalloprotease events during oligofructose induced equine laminitis development. Equine Vet J 2012;44:88–93.
5. Wang L, Pawlak EA, Johnson PJ, et al. Expression and activity of collagenases in the digital laminae of horses with carbohydrate overload-induced acute laminitis. J Vet Intern Med 2014;28:215–222.
6. Keene DR, Sakai LY, Lunstrum GP, et al. Type VII collagen forms an extended network of anchoring fibrils. J Cell Biol 1987;104:611–621.
7. Villone D, Fritsch A, Koch M, et al. Supramolecular interactions in the dermoepidermal junction zone–anchoring fibril-collagen VII tightly binds to banded collagen fibrils. J Biol Chem 2008;283:24506–24513.
8. Birkedal-Hansen H, Moore WGI, Bodden MK, et al. Matrix metalloproteinases: a review. Crit Rev Oral Biol Med 1993;4:197–250.
9. Chakraborti S, Mandal M, Das S, et al. Regulation of matrix metalloproteinases: an overview. Mol Cell Biochem 2003;253:269–285.
10. Loftus JP, Johnson PJ, Belknap JK, et al. Leukocyte-derived and endogenous matrix metalloproteinases in the lamellae of horses with naturally acquired and experimentally induced laminitis. Vet Immunol Immunopathol 2009;129:221–230.
11. Owen CA, Hu Z, Lopez-Otin C, et al. Membrane-bound matrix metalloproteinases-8 on activated polymorphonuclear cells is a potent, tissue inhibitor of metalloproteinase-resistant collagenase and serpinase. J Immunol 2004;172:7791–7803.
12. Pirilä E, Korpi JT, Korkiamäki T, et al. Collagenase-2 (MMP-8) and matrilysin-2 (MMP-26) expression in human wounds of different etiologies. Wound Rep Regen 2007;15:47–57.
13. Mungall BA, Pollitt CC, Collins R. Localisation of gelatinase activity in epidermal hoof lamellae by in situ zymography. Histochem Cell Biol 1998;110:535–540.
14. Medina-Torres CE, Mason SL, Floyd RV, et al. Hypoxia and a hypoxia mimetic up-regulate matrix metalloproteinase 2 and 9 in equine laminar keratinocytes. Vet J 2011;190:e54–e59.
15. Visse R, Nagase H. Matrix metalloproteinases and tissue inhibitors of metalloproteinases: structure, function and biochemistry. Circ Res 2003;92:827–839.
16. Loftus JP, Belknap JK, Black SJ. Matrix metalloproteinase-9 in laminae of black walnut extract treated horses correlates with neutrophil abundance. Vet Immunol Immunopathol 2006;113:267–276.
17. Leise BS, Watts MR, Roy S, et al. Use of laser capture microdissection for the assessment of equine lamellar basal epithelial cell signaling in the early stages of laminitis. Equine Vet J 2015;47:478–488.
18. Belknap JK, Moore JN, Crouser EC. Sepsis—from human organ failure to laminar failure. Vet Immunol Immunopathol 2009;129:155–157.
19. Neuder LE, Keener JM, Eckert RE, et al. Role of p38 MAPK in LPS induced proinflammatory cytokine and chemokine gene expression in equine leukocytes. Vet Immunol Immunopathol 2009;129:192–199.
20. Parsons CS, Orsini JA, Krafty R, et al. Risk factors for development of acute laminitis in horses during hospitalization: 73 cases (1997–2004). J Am Vet Med Assoc 2007;230:885–889.
21. Taylor S. A review of equine sepsis. Equine Vet Educ 2015;27:99–109.
22. Patan-Zugaj B, Gauff FC, Licka TF. Effects of the addition of endotoxin during perfusion of isolated forelimbs of equine cadavers. Am J Vet Res 2012;73:1462–1468.
23. Patan-Zugaj B, Gauff FC, Plendl J, et al. Effect of endotoxin on leukocyte activation and migration into laminar tissue of isolated perfused equine limbs. Am J Vet Res 2014;75:842–850.
24. Patan B, Budras KD, Licka TF. Effects of long-term extracorporeal blood perfusion of the distal portion of isolated equine forelimbs on metabolic variables and morphology of laminar tissue. Am J Vet Res 2009;70:669–677.
25. Fessler JF, Bottoms GD, Coppoc GL, et al. Plasma endotoxin concentrations in experimental and clinical equine subjects. Equine Vet J Suppl 1989;7:24–28.
26. Senior JM, Proudman CJ, Leuwer M, et al. Plasma endotoxin in horses presented to an equine referral hospital: correlation to selected clinical parameters and outcomes. Equine Vet J 2011;43:585–591.
27. Steverink PJGM, Salden HJM, Sturk A, et al. Laboratory and clinical evaluation of a chromogenic endotoxin assay for horses with acute intestinal disorders. Vet Q 1994;16(suppl 2):117–121.
28. Sprouse RF, Garner HE, Green EM. Plasma endotoxin levels in horses subjected to carbohydrate induced laminitis. Equine Vet J 1987;19:25–28.
29. Duncan ME, Richardson JP, Murray GI, et al. Human matrix metalloproteinase-9: activation by limited trypsin treatment and generation of monoclonal antibodies specific for the activated form. Eur J Biochem 1998;258:37–43.
30. Coyne MJ, Cousin H, Loftus JP, et al. Cloning and expression of ADAM-related metalloproteinases in equine laminitis. Vet Immunol Immunopathol 2009;129:231–241.
31. de Laat MA, van Eps AW, McGowan CM, et al. Equine laminitis: comparative histopathology 48 hours after experimental induction with insulin or alimentary oligofructose in Standardbred horses. J Comp Pathol 2011;145:399–409.
32. Fedchenko N, Reifenrath J. Different approaches for interpretation and reporting of immunohistochemistry analysis results in the bone tissue—a review. Diagn Pathol 2014;9:221.
33. Grosche A, Morton AJ, Graham AS, et al. Mucosal injury and inflammatory cells in response to brief ischaemia and reperfusion in the equine large colon. Equine Vet J Suppl 2011;43:16–25.
34. Kwon S, Moore JN, Robertson TP, et al. Disparate effects of LPS infusion and carbohydrate overload on inflammatory gene expression in equine laminae. Vet Immunol Immunopathol 2013;155:1–8.
35. Menzies-Gow NJ, Bailey SR, Katz LM, et al. Endotoxin-induced digital vasoconstriction in horses: associated changes in plasma concentrations of vasoconstrictor mediators. Equine Vet J 2004;36:273–278.
36. Tadros EM, Frank N. Effects of continuous or intermittent lipopolysaccharide administration for 48 hours on the systemic inflammatory response in horses. Am J Vet Res 2012;73:1394–1402.
37. Tadros EM, Frank N, Donnell RL. Effects of equine metabolic syndrome on inflammatory responses of horses to intravenous lipopolysaccharide infusion. Am J Vet Res 2013;74:1010–1019.
38. Duncan SG, Meyers KM, Reed SM, et al. Alterations in coagulation and hemograms of horses given endotoxins for 24 hours via hepatic portal infusions. Am J Vet Res 1985;46:1287–1293.
39. French KR, Pollitt CC. Equine laminitis: loss of hemidesmosomes in hoof secondary epidermal lamellae correlates to dose in an oligofructose induction model: an ultrastructural study. Equine Vet J 2004;36:230–235.
40. Nourian AR, Baldwin GI, van Eps AW, et al. Equine laminitis: ultrastructural lesions detected 24–30 hours after induction with oligofructose. Equine Vet J 2007;39:360–364.
41. Gaffney J, Solomonov I, Zehorai E, et al. Multilevel regulation of matrix metalloproteinases in tissue homeostasis indicates their molecular specificity in vivo. Matrix Biol 2015;44–46:191–199.
42. Johnson PJ, Tyagi SC, Katwa LC, et al. Activation of extracellular matrix metalloproteinases in equine laminitis. Vet Rec 1998;142:392–396.
43. Laskoski LM, Valadao CAA, de Oliveira Vasconcelos R, et al. Expression of matrix metalloproteases-2 and -9 in horse hoof laminae after intestinal obstruction, with or without hydrocortisone treatment. Cienc Rural 2013;43:66–72.
44. Prescimone T, Tognotti D, Caselli C, et al. Reappraisal of gel zymography for matrix metalloproteinases. J Clin Lab Anal 2014;28:374–380.
45. Snoek-van Beurden PAM, Von den Hoff JW. Zymographic techniques for the analysis of matrix metalloproteinases and their inhibitors. Biotechniques 2005;38:73–83.
46. Dagleish MP, Brazil TJ, Scudamore CL. Potentiation of the extracellular release of equine neutrophil elastase and alpha-a-proteinase inhibitor by a combination of two bacterial cell wall components: fMLP and LPS. Equine Vet J 2003;35:35–39.
47. Martins VL, Caley M, O'Toole EA. Matrix metalloproteinases and epidermal wound repair. Cell Tissue Res 2013;351:255–268.
48. Ortega N, Werb Z. New functional roles for non-collagenous domains of basement membrane collagens. J Cell Sci 2002;115:4201–4214.
