Reasons for performing study: Hyperinsulinaemia has been implicated in the pathogenesis of laminitis; however, laminar cell types responding to insulin remain poorly characterised.
Objectives: To identify laminar cell types expressing insulin receptor (IRc) and/or insulin-like growth factor-1 receptor (IGF-1R); and to evaluate the effect of dietary nonstructural carbohydrate (NSC) on their expression.
Methods: Mixed-breed ponies (n = 22) received a conditioning hay chop diet (NSC ∼6%); following acclimation, ponies were stratified into lean (n = 11, body condition score [BCS]≤4) or obese (n = 11, BCS ≥7) groups and each group further stratified to remain on the low NSC diet (n = 5 each for obese and lean) or receive a high NSC diet (total diet ∼42% NSC; n = 6 each for obese and lean) for 7 days. Laminar samples were collected at the end of the feeding protocol and stained immunohistochemically for IRc and IGF-1R. The number of IRc(+) cells was quantified; distribution of IGF-1R was qualitatively described. Laminar IRc content was assessed via immunoblotting.
Results: The number of IRc(+) cells was greater in the laminae of high NSC ponies than low NSC ponies (P = 0.001); there was a positive correlation between the change in serum insulin concentration and number of IRc(+) cells (r2= 0.74; P<0.0001). No epithelial IRc(+) cells were observed; IRc(+) cells were absent from the deep dermis. Analysis of serial sections identified IRc(+) cells as endothelial cells. The distribution of IGF-1R was more extensive than that of IRc, with signal in vascular elements, epithelial cells and fibroblasts.
Conclusions: Increased dietary NSC results in increased laminar endothelial IRc expression. Laminar keratinocytes do not express IRc, suggesting that insulin signalling in laminar epithelial cells must be mediated through other receptors (such as IGF-1R).
Potential relevance: Manipulation of signalling downstream of IRc and IGF-1R may aid in treatment and prevention of laminitis associated with hyperinsulinaemia.
Equine metabolic syndrome (EMS), a constellation of clinical findings currently accepted to include obesity, regional adiposity, systemic insulin resistance, and laminitis (historical or current) , is an increasingly common cause of laminitis in equids in developed nations. Pasture-associated laminitis (for which EMS-affected equids may be at increased risk ) is reported to be the most common cause of laminitis seen in general equine veterinary practice, accounting for approximately 46% of new clinical cases attended by equine veterinarians in the USA . This fact, along with the perennial identification of laminitis as a research priority for the equine veterinary community , has driven a vigorous research effort towards increasing understanding of laminitis associated with metabolic diseases.
Observational studies have associated abnormalities of insulin and glucose metabolism (particularly systemic insulin resistance and resultant hyperinsulinaemia) with increased risk of laminitis in ponies upon exposure to spring pasture , a common precipitating event in the development of the disease as reported by equine veterinary practitioners . Additional studies in light breed horses have also reported abnormalities of insulin and glucose metabolism in animals with obesity, regional adiposity and a predisposition to laminitis . While useful in many respects for characterising important risk factors for endocrinopathic laminitis and further galvanising the EMS phenotype, these studies are not mechanistic investigations and do not provide data supporting a pathophysiological link(s) between the reported abnormalities in endocrine/metabolic function and laminitis.
Recently, a direct role for insulin in the pathogenesis of endocrinopathic laminitis has been suggested by the induction of laminitis in both normal ponies  and normal Standardbred horses  following 48–72 h of experimentally induced hyperinsulinaemia. While its effects on carbohydrate metabolism are among its best-characterised effects, insulin is a pleiotrophic hormone. Upon binding to its receptor (which is expressed on many cell types), insulin induces not only its well-characterised effects on transmembrane glucose uptake (mediated through activation of the PI3K signalling pathway), but also mitogenic and extracellular matrix metabolism effects (mediated through activation of the MAPK signalling pathways) [9,10]. Additionally, while the majority of insulin's effects on intermediary metabolism are attributed to activation of intracellular signalling downstream of insulin receptor (IRc), insulin is also known to bind and activate signal transduction through the insulin-like growth factor-1 receptor (IGF-1R), particularly at supraphysiological concentrations of insulin . Activation of the IGF-1R by insulin when insulin is present at supraphysiological concentrations may result in activation of the anti-apoptotic and mitogenic pathways commonly associated with IGF-1R signalling; abnormal laminar keratinocyte proliferation and maturation may result and contribute to endocrinopathic laminitis. While the ability of hyperinsulinaemia to induce equine laminitis has been increasingly well established, the pathway(s) through which insulin exacts its deleterious effects on the equine digital laminae are largely unknown. Therefore, the objectives of the study reported here were: 1) to identify and characterise the distribution of cells expressing IRc and IGF-1R within the digital laminae of ponies; and 2) to evaluate the effects of a dietary nonstructural carbohydrate challenge (one meant to mimic that which might be encountered during spring pasture exposure) on both the protein concentrations and cellular distribution of IRc and IGF-1R in the digital laminae.
Materials and methods
Twenty-two mixed-breed ponies (bodyweight 270.9 ± 74.4 kg; age [lean]= 9.2 ± 3.5 years, age [obese]= 11 ± 3.8 years) were used for this study. All animals received humane treatment in accordance with an animal care and use protocol approved by the Michigan State University Institutional Animal Care and Use Committee. Feedstuffs used in the protocol were analysed for nonstructural carbohydrate content (NSC, defined as the sum of measured starch and water-soluble carbohydrates) by a commercial laboratorya. All ponies obtained for the study were examined by a licensed veterinarian and deemed healthy based on the results of physical examination, complete blood count, and serum biochemical examination. Ponies were divided into 4 experimental groups based on body condition scoring results: lean, low NSC (n = 5); obese, low NSC (n = 5); lean, high NSC (n = 6); obese, high NSC (n = 6). All body condition scores (BCSs) were performed by 2 individuals (L.J.M. and R.J.G.); lean animals were those assigned a BCS of ≤4/9, and obese animals were those assigned a BCS of ≥7/9 .
All ponies were housed in dirt lots and conditioned to a diet of hay chop (7% starch and ethanol soluble carbohydrate on a dry matter basis) for 4 weeks prior to initiation of the experimental feeding protocol. Ponies were fed 2.5% of their bodyweight in hay chop per day, divided into 2 feedings (07.00 and 18.00 h Eastern Standard Time [EST]). Following the conditioning period, ponies either remained on the control diet (n = 10; lean and obese control groups) or received the same diet supplemented with sweet feed (1.5% bodyweight per day, fed 3 times daily at 07.00, 12.00 and 18.00 h EST) and oligofructoseb (2 g/kg bwt added to hay chop ration; lean and obese challenge groups, n = 12) for a period of 7 days. The mean NSC consumption of ponies in the control groups was approximately 1.8 g/kg bwt/day, while that of ponies in the challenged groups was approximately 8 g/kg bwt/day. All ponies were monitored 3 times daily during the experimental period.
All ponies underwent insulin-modified frequently-sampled intravenous glucose tolerance testing (FSIGTT) with minimal model analysis  during the first 2 weeks of the acclimation period. The FSIGTTs were performed between 07.00 h and 09.00 h EST following a 6–8 h period of withholding feed. Additionally, blood was collected into red top tubesc for measurement of basal serum insulin concentrations on Day 0 and Day 7 of the feeding protocol (between 07.00 h and 09.00 h prior to feeding the morning ration); serum insulin concentrations were measured with a radioimmunoassay validated for equine samples (Coat-A-Count)d).
Following the 7-day experimental period, ponies were subjected to euthanasia via intravenous overdose of pentobarbital sodium and phenytoin sodium (Fatal-Plus; 20 mg/kg bwt i.v.)e. The right front foot of each animal was removed by disarticulation of the metacarpophalangeal joint immediately following euthanasia, and 1.5 cm sagittal sections of the dorsal digit were cut with a band saw. After dissection of the digital laminae from the hoof wall and third phalanx, sections of laminae were snap-frozen in liquid nitrogen or fixed in 10% neutral buffered formalin; all laminar samples were processed within 15 min of euthanasia. Formalin-fixed samples were transferred to 70% ethanol after 48 h, where they were stored until embedding.
Formalin-fixed laminar samples were embedded in paraffin and sectioned at 5 µm for immunohistochemistry (IHC). Staining of laminar samples was performed as previously reported ; briefly, sections were deparaffinised and incubated in either a mouse monoclonal anti-IRc primary antibodyf (Abcam, ab54268, used undiluted, incubated overnight at 4°C; ab69508, diluted 1:1000, incubated overnight at 4°C) or a rabbit polyclonal anti-IGF-1R primary antibodyg (1:100, incubated overnight at 4°C) that has previously been validated for IHC on paraffin-embedded tissue . Two separate primary antibodies against IRc were tested for this IHC application to ensure specificity of the staining pattern, as no blocking peptide was commercially available to test fidelity; both antibodies yielded identical staining patterns when used for IHC on digital laminar tissue. Detection of immunoreactivity was performed using an immunoperoxidase systemh and DAB substrateh. After IHC staining, the number of IRc (+) cells in n = 10 (40x) light microscopy fields in both: 1) laminar tissue (restricted to dermal/epidermal laminae); and 2) deep dermis (dermal tissue situated between laminae and cortex of distal phalanx) was counted and recorded by a single investigator who was blinded to the source of the tissue section (T.A.B.). The distribution of IGF-1R (+) cells was assessed via light microscopy and qualitatively described. To determine the identity of the IRc (+) cells, serial laminar tissue sections (3 µm) were stained for IRcf, von Willebrand's factor (vWF)i (rabbit polyclonal, used at 1:1000; marker of endothelial cells) , CD163j (marker of macrophages) , and calprotectinf (marker of activated myeloid and epithelial cells) , again using an immunoperoxidase systemh and DAB substrateh.
The concentration of IRc in laminar tissue homogenates was assessed via western immunoblotting performed as described previously . Briefly, laminar protein samples (∼100 mg laminar tissue per sample) were prepared in 300 µl lysis bufferk and quantitated via the Bradford method . Protein samples (20 µg/sample) were denatured by boiling for 5 min in β-ME/SDS buffer, separated on an 8% polyacrylamide gel, and transferred to a PVDF membrane. The membrane was blocked for one hour with 5% milk in PBS-Tween 20 (0.1% v/v Tween-20 in PBS; PBST) at room temperature, rocking. The membrane was then incubated with primary antibodyf (ab69508, mouse monoclonal, 1:1000 in 5% milk) overnight at 4°C. The membrane was washed 5 times with 0.1% PBST as before. Goat anti-mouse IgG-HRP secondary antibodyk was diluted 1:5000 in 5% milk and incubated with the membrane for one hour at room temperature, rocking. The membrane was washed 5 times with 0.1% PBST and developed for 5 min using a chemiluminescent substrate (West Femto)k. The membrane was then stripped and reprobed for β-actin (primary, goat polyclonal, 1:1000; secondary, 1:20,000)g. Luminescence was measured using a computer software programl, and signal strength was determined using net IRc band intensity divided by β-actin band intensity.
Data analysis was performed using GraphPad Prismm. Data were assessed for normality by the Shapiro–Wilk and D'Agostino and Pearson omnibus normality tests and analysed with either a Student's t test (or nonparametric equivalent) or 2-way analysis of variance followed by a Bonferroni post test, as appropriate. Correlations were assessed using Pearson's statistic. Statistical significance was accepted at P≤0.05. Data are expressed as mean ± s.d. unless otherwise indicated.
Systemic insulin response to high-carbohydrate feeding
All ponies completed the feeding protocol; 4 ponies were observed to become clinically laminitic during the course of the feeding protocol (Obel Grade I–II/IV; one lean challenged pony, 3 obese challenged ponies). One of these animals was administered phenylbutazone (2.2 mg/kg bwt per os b.i.d.) for 3 days prior to euthanasia in accordance with the approved animal care and use protocol associated with the project; the 3 remaining animals were noted to be mildly laminitic (Obel Grade I/IV) on the day of euthanasia only and were not medicated.
Ponies were assigned to the experimental groups based solely on BCS with no stratification related to insulin sensitivity status. However, subsequent minimal model analysis indicated that insulin sensitivity was markedly lower in obese when compared with lean ponies (0.50 ± 0.26 × 10−4 l/min/miu in obese vs. 2.57 ± 2.18 × 10−4 l/min/miu in lean, P = 0.0005). Conversely, the acute insulin response to glucose was higher (P = 0.04) in obese (755.4 ± 340.9 miu/l/min) than in lean (478.4 ± 433.6 miu/l/min) ponies. There was no statistical difference in the age of the lean ponies compared with that of the obese ponies (P = 0.26).
No weight gain was observed in response to either the low NSC or high NSC feeding protocols. Basal serum insulin concentration was increased across the 7-day experimental feeding period in both lean (Day 0, 11.4 ± 7.5 miu/l; Day 7, 258.0 ± 306.2 miu/l) and obese (Day 0, 14.6 ± 6.9 miu/l; Day 7, 549.1 ± 277.3 miu/l) ponies fed the high-NSC diet when compared with low-NSC fed animals (P = 0.007). There was no difference in basal serum insulin concentration across the feeding protocol between the lean and obese control groups (P>0.05; Fig 1).
No difference in laminar IRc protein content was observed between the lean, low NSC ponies and the lean, high NSC ponies (P = 0.96). The laminar content of insulin receptor in obese, high NSC ponies, however, was lower than that of obese, low NSC ponies (P = 0.002; Fig 2).
Digital laminar IHC
Immunohistochemical staining for IRc with 2 separate antibodies (ab69508 and ab54268) resulted in an identical staining pattern in digital laminar tissue; one of these antibodies (ab69508) was also found to recognise the correct size band (∼95 kD, assessed with positive control 3T3 cell lysate) on western blot (see Immunoblotting above). The total number of IRc(+) cells was greater in the laminae of ponies fed a high NSC diet than in those fed the low-NSC diet (P = 0.01; Figs 3 and 4), and there was a positive correlation between the change in serum basal insulin concentration across the 7-day experimental period and number of laminar IRc(+) cells (r = 0.74; P < 0.0001). No IRc(+) epithelial cells were observed in any section, and IRc(+) cells were present in much lower numbers in the deep dermal tissue (including vasculature). The number of IRc(+) cells in the deep dermis did not change with diet in either lean or obese ponies. Analysis of serial section staining identified IRc(+) cells as endothelial cells, as all were also vWF(+) on IHC (Fig 5). The cellular distribution of IGF-1R expression within the digital laminae was more extensive than that of IRc, with signal appearing not only within vascular elements, but also on laminar epithelial cells and cells morphologically consistent with fibroblasts (Fig 6). Distinct staining of the cytoplasmic membrane and nucleus was observed in laminar epithelial cells (Fig 6). While not specifically quantitated (due to the questionable value of quantitation of immunohistochemical stain intensity), laminar IGF-1R expression did not appear influenced by diet, body condition or insulin sensitivity status.
Systemic insulin resistance and resultant hyperinsulinaemia have been implicated in the pathogenesis of endocrinopathic laminitis in equids [7,8]. Extrapolation from studies of metabolic syndrome in other species would suggest that an insulin resistant state with high circulating insulin concentrations would result in deleterious events downstream of the insulin receptor in the laminar cellular milieu, ranging from disruption of energy metabolism and possible cellular energy failure to aberrant regulation of the extracellular matrix (critical for adhesion of the laminar epidermis to the underlying dermis [9,22]). These events noted in the human literature have been of great interest in the study of laminitis at the level of the laminar basal epithelial cell (LBEC), as 2 events reported to possibly contribute to dysadherence of this cell layer from the underlying matrix (and laminar failure) are matrix degradation and energy failure [23–25]. An additional event documented in studies of human insulin resistance proposed to play a role in EMS-related laminitis is endothelial dysfunction leading to vasoconstriction of microvascular beds [26,27]. In this study, the only IRc(+) cell type in the digital laminae was the microvascular endothelial cell, whereas IGF-1R was present on multiple cell types (including the LBEC). These results raise the possibility that elevated insulin concentrations may play a pathophysiological role in the laminae by signalling through both IGF-1R on the LBEC and IRc on the laminar microvascular endothelial cell.
Insulin has been shown to bind and activate IGF-1R at supraphysiological concentrations in cell cultures ; based on the results of the study reported here, IGF-1Rs cellular distribution within the laminae is much more extensive than that of the IRc. While IRc appears to be limited to the laminar microvasculature, IGF-1R is present on multiple cell types, including laminar epithelial cells, endothelial cells and dermal constituents (probably fibroblasts and tissue macrophages). Whereas IGF-1 is reported to be important in the normal maturation of epithelial cells in human skin , insulin signalling through the IGF-1R in the setting of hyperinsulinaemia is thought to underlie the keratinocyte dysregulation of acanthosis nigricans and skin tag, 2 cutaneous manifestations of human metabolic syndrome characterised by epithelial hyperplasia and hyperpigmentation . The nuclear localisation of IGF-1R in laminar epithelial cells reported here is consistent with this receptor's documented role in mitogenesis, as nuclear translocation of surface-activated receptor has been reported in both normal and neoplastic cells [30,31]. We also observed a similar surface localisation of IGF-1R in the laminar keratinocyte in vivo as has been reported in cultured equine laminar keratinocytes . Insulin-like growth factor-1 receptor signalling events in the laminar epidermis, particularly its reported effects on regulation of extracellular matrix [33,34], cytoskeletal dynamics [35–37], and cellular phenotype , may play a role in the dysregulation/dysadhesion of the LBEC from the underlying dermis; further investigation of these signalling events in the LBEC are warranted.
Insulin signalling through laminar IRc in hyperinsulinaemic equids is likely to exact its most profound direct effects on laminar perfusion, as IRc signal appears restricted primarily to the laminar dermal microvascular endothelium. Endothelial insulin resistance has been associated with increased vasomotor tone in response to repeated insulin exposure, classically thought to result from decreased elaboration of nitric oxide and increased elaboration of endothelin-1. These endothelial events have been suggested to be involved in the hypertension reported in laminitis-prone ponies exposed to pasture . In horses with insulin resistance, changes in vasomotor tone to a tissue with poor collateral circulation (the foot) may promote hypoxia/ischaemia, both of which have been suggested to play roles in the pathogenesis of laminitis [24,38–40]. Recent work utilising in vitro modelling of induced vascular insulin resistance in equine digital vessel explants suggests that insulin, normally vasodilatory, may have vasoconstrictor effects in the digital vasculature of the insulin-resistant equid . However, the concentrations of insulin used to induce insulin resistance in the vascular explants in this study were several orders of magnitude higher than that likely to be encountered in the insulin-resistant equid (∼1.4 × 106 miu/l) [6,42]; consequently, the results may not accurately reflect events occurring in vivo.
Hyperinsulinaemia itself reportedly contributes to endothelial insulin resistance, as prolonged exposure of endothelial cells to increased concentrations of insulin in vitro has been demonstrated to alter their insulin responsiveness, with induction of cellular insulin resistance over a relatively short period of time (24–48 h). Rapid downregulation of surface expression of IRc has been documented on endothelial cells in response to insulin treatment, which may partially explain the endothelial insulin resistance observed in response to hyperinsulinaemia in vivo[43,44]. Interestingly, it is shortly after the reported time required for downregulation of IRc that horses and ponies are observed to become laminitic during experimental hyperinsulinaemia [7,8]. A similar general decrease of tissue IRc protein concentrations as reported in these studies may be responsible for the decrease in IRc noted on western hybridisation of protein extracted from the digital tissue (laminar tissue with underlying deep dermis) in obese ponies placed on a high NSC diet. The immunohistochemical results of our study, however, do not support IRc downregulation as a general effect in the digital vasculature, as the number of endothelial cells expressing IRc in laminar microvasculature in response to hyperinsulinaemia following high-NSC feeding was observed to increase. These seemingly discordant findings between immunoblotting and IHC results are most likely to reflect a general decrease in cellular expression of IRc in the digital tissue, but an increase in the number of laminar microvascular endothelial cells expressing detectable IRc. Alternatively, as different antibodies generated against different epitopes of the IRc were used for IHC and WB (and they were used to detect native vs. denatured IRc protein, respectively), it is possible that the discrepancy represents these methodological differences. Nonetheless, IRc expression in the digital laminar microvasculature appears to be uniquely regulated compared with both: 1) studies of vascular IRc regulation in the face of hyperinsulinaemia in other species; and 2) the vasculature in the deep dermis adjacent to the laminae in the current study.
Upregulation of IRc expression on the digital microvascular endothelium in response to high-carbohydrate feeding was a somewhat unexpected result of this study. It is commonly held that exposure to insulin results in IRc downregulation on the cell surface of many cell types; however, this effect has been reported to be somewhat cell type- and tissue-specific [45,46]. Insulin-mediated upregulation of IRc expression has been reported in human lymphocyte cell cultures , among others. Further, increased cell surface expression of IRc has been reported in response to several other common physiological stimuli, such as exposure to glucocorticoids [48–50] and growth arrest ; these factors may be involved in the regulation of IRc expression within the laminar microvasculature of the equine digit. Overall, the unique upregulation of the IRc in the laminar microvasculature may possibly play a role in the greater sensitivity of laminar tissue (compared with other tissues/organs) to injury in EMS.
The total amount of NSC consumed by the ponies in the high-NSC group during the 7-day experimental feeding period was approximately 8 g/kg bwt/day, an amount that approximates the amount administered to horses to induce laminitis via the oligofructose model (10 g/kg bwt) [52,53] but far less than that used to induce laminitis via the enteral carbohydrate model (17.6 g/kg bwt corn and wood starch) . The daily NSC dose administered to the ponies in this study was divided into 3 daily feedings, not administered as a single bolus, and each pony received only ∼2 g/kg bwt/day of NSC as oligofructose. The experimental diet was meant to mimic the NSC consumption that might occur in a pony following acute pasture exposure  and was not composed to create a laminitis model, per se; however, it was anticipated that the diet might induce laminitis in a certain number of the animals.
In conclusion, the results of the work reported here support a role for insulin in the pathophysiology of endocrinopathic equine laminitis; however, direct effects on LBECs mediated through IRc appear unlikely due to its limited expression on this cell type. Further work is needed to more discretely characterise the mechanism(s) by which insulin exacts its effects in the equine digit; delineating the pathways mediating its effects are likely to lead to logical points of intervention for the prevention and treatment of laminitis associated with EMS.
Authors' declaration of interests
No competing interests have been declared.
Source of funding
Funding for this study was provided by the Morris Animal Foundation (grant # D12EQ-027) and Michigan State University College of Veterinary Medicine internal funds.
The authors would like to acknowledge the staff of the MSU Diagnostic Center for Population and Animal Health for their skilled assistance in tissue processing, and the staff and students of the MSU Department of Large Animal Clinical Sciences for caring for the ponies used in this study.
a Equi-Analytical Laboratories, Ithaca, New York, USA.
b Beneo-ORAFTI, Tienen, Belgium
c Becton, Dickinson and Company, Franklin Lakes, New Jersey, USA.
d Siemens Medical Solutions Diagnostics, Los Angeles, California, USA.
e Vortech Pharmaceuticals Ltd, Dearborn, Michigan, USA.
f Abcam PLC, Cambridge, Massachusetts, USA.
g Santa Cruz Biotechnology Inc, Santa Cruz, California, USA.
h Vector Laboratories Inc, Burlingame, California, USA.
i DAKO North America Inc, Carpinteria, California, USA
j Cosmo Bio. Co. Ltd, Tokyo, Japan
k Pierce, Rockford, Illinois, USA.
l Carestream Health Inc, Rochester, New York, USA.