• View in gallery
    Figure 1—

    Representative photomicrographs of pancreatic sections that were obtained from horses with clinically normal insulin sensitivity and were immunohistochemically stained for detection of insulin (A and B), glucagon (C and D), and somatostatin (E and F) within the pancreatic islets before (A, C, and E) and after (B, D, and F) an image analysis program was used to change the brown chromagen within hormone-expressing cells to red for quantification of immunoreactivity. The image analysis program was also used to outline the pancreatic islets (not shown) to determine the islet area and percentage of immunoreactivity for each hormone within the islet. Bar = 20 μm.

  • 1. Frank N, Geor RJ, Bailey SR, et al. Equine metabolic syndrome. J Vet Intern Med 2010;24:467475.

  • 2. Johnson PJ, Wiedmeyer CE, LaCarrubba A, et al. Diabetes, insulin resistance, and metabolic syndrome in horses. J Diabetes Sci Technol 2012;6:534540.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 3. Frank N, Tadros EM. Insulin dysregulation. Equine Vet J 2014;46:103112.

  • 4. Tóth F, Frank N, Martin-Jiménez T, et al. Measurement of C-peptide concentrations and responses to somatostatin, glucose infusion, and insulin resistance in horses. Equine Vet J 2010;42:149155.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5. Hoffman RM, Boston RC, Stefanovski D, et al. Obesity and diet affect glucose dynamics and insulin sensitivity in Thoroughbred geldings. J Anim Sci 2003;81:23332342.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6. Tóth F, Frank N, Elliot SB, et al. Optimisation of the frequently sampled intravenous glucose tolerance test to reduce urinary glucose spilling in horses. Equine Vet J 2009;41:844851.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 7. Freestone JF, Wolfsheimer KJ, Kamerling SG, et al. Exercise induced hormonal and metabolic changes in Thoroughbred horses: effects of conditioning and acepromazine. Equine Vet J 1991;23:219223.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8. Bergman RN, Phillips LS, Cobelli C. Physiologic evaluation of factors controlling glucose tolerance in man: measurement of insulin sensitivity and beta-cell glucose sensitivity from the response to intravenous glucose. J Clin Invest 1981;68:14561467.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9. Boston RC, Stefanovski D, Moate PJ, et al. MINMOD Millennium: a computer program to calculate glucose effectiveness and insulin sensitivity from the frequently sampled intravenous glucose tolerance test. Diabetes Technol Ther 2003;5:10031015.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10. Haschek WM, Rousseaux CG, Wallig MA. Pancreas. In: Fundamentals of toxicologic pathology. 2nd ed. London: Elsevier, 2010;237260.

  • 11. Steiner DJ, Kim A, Miller K, et al. Pancreatic islet plasticity: interspecies comparison of islet architecture and composition. Islets 2010;2:135145.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12. Hawkins KL, Summers BA, Kuhajda FP, et al. Immunocytochemistry of normal pancreatic islets and spontaneous islet cell tumors in dogs. Vet Pathol 1987;24:170179.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13. Bamford NJ, Baskerville CL, Harris PA, et al. Postprandial glucose, insulin, and glucagon-like peptide-1 responses of different equine breeds adapted to meals containing micronized maize. J Anim Sci 2015;93:33773383.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14. de Laat MA, McGree JM, Sillence MN. Equine hyperinsulinemia: investigation of the enteroinsular axis during insulin dysregulation. Am J Physiol Endocrinol Metab 2016;310:E61E72.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15. Chameroy KA, Frank N, Elliott SB, et al. Comparison of plasma active glucagon-like peptide 1 concentrations in normal horses and those with equine metabolic syndrome and in horses placed on a high-grain diet. J Equine Vet Sci 2016;40:1625.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16. Gosmanov NR, Szoke E, Israelian Z, et al. Role of the decrement in intraislet insulin for the glucagon response to hypoglycemia in humans. Diabetes Care 2005;28:11241131.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17. Ferrannini E, Muscelli E, Natali A, et al. Association of fasting glucagon and proinsulin concentrations with insulin resistance. Diabetologia 2007;50:23422347.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18. Chang-Chen KJ, Mullur R, Bernal-Mizrachi E. Beta-cell failure as a complication of diabetes. Rev Endocr Metab Disord 2008;9:329343.

  • 19. Yang K, Gotzmann J, Kuny S, et al. Five stages of progressive β-cell dysfunction in the laboratory Nile rat model of type 2 diabetes. J Endocrinol 2016;229:343356.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20. Dunbar LK, Mielnicki KA, Dembek KA, et al. Evaluation of four diagnostic tests for insulin dysregulation in adult light-breed horses. J Vet Intern Med 2016;30:885891.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21. Treiber KH, Kronfeld DS, Hess TM, et al. Use of proxies and reference quintiles obtained from minimal model analysis for determination of insulin sensitivity and pancreatic beta-cell responsiveness in horses. Am J Vet Res 2005;66:21142121.

    • Crossref
    • Search Google Scholar
    • Export Citation

Advertisement

Immunohistochemical expression of insulin, glucagon, and somatostatin in pancreatic islets of horses with and without insulin resistance

Kim M. NewkirkDepartment of Biomedical and Diagnostic Sciences, College of Veterinary Medicine, University of Tennessee, Knoxville, TN 37996.

Search for other papers by Kim M. Newkirk in
Current site
Google Scholar
PubMed
Close
 DVM, PhD
,
Gordon EhrensingDepartment of Biomedical and Diagnostic Sciences, College of Veterinary Medicine, University of Tennessee, Knoxville, TN 37996.

Search for other papers by Gordon Ehrensing in
Current site
Google Scholar
PubMed
Close
 DVM
,
Agricola OdoiDepartment of Biomedical and Diagnostic Sciences, College of Veterinary Medicine, University of Tennessee, Knoxville, TN 37996.

Search for other papers by Agricola Odoi in
Current site
Google Scholar
PubMed
Close
 BVM, PhD
,
Raymond C. BostonDepartment of Clinical Studies–New Bolton Center, School of Veterinary Medicine, University of Pennsylvania, Kennett Square, PA 19348.

Search for other papers by Raymond C. Boston in
Current site
Google Scholar
PubMed
Close
 PhD
, and
Nicholas FrankDepartment of Clinical Sciences, Cummings School of Veterinary Medicine, Tufts University, North Grafton, MA 01536.

Search for other papers by Nicholas Frank in
Current site
Google Scholar
PubMed
Close
 DVM, PhD

Abstract

OBJECTIVE To assess insulin, glucagon, and somatostatin expression within pancreatic islets of horses with and without insulin resistance.

ANIMALS 10 insulin-resistant horses and 13 insulin-sensitive horses.

PROCEDURES For each horse, food was withheld for at least 10 hours before a blood sample was collected for determination of serum insulin concentration. Horses with a serum insulin concentration < 20 μU/mL were assigned to the insulin-sensitive group, whereas horses with a serum insulin concentration > 20 μU/mL underwent a frequently sampled IV glucose tolerance test to determine sensitivity to insulin by minimal model analysis. Horses with a sensitivity to insulin < 1.0 × 10−4 L•min−1•mU−1 were assigned to the insulin-resistant group. All horses were euthanized with a barbiturate overdose, and pancreatic specimens were harvested and immunohistochemically stained for determination of insulin, glucagon, and somatostatin expression in pancreatic islets. Islet hormone expression was compared between insulin-resistant and insulin-sensitive horses.

RESULTS Cells expressing insulin, glucagon, and somatostatin made up approximately 62%, 12%, and 7%, respectively, of pancreatic islet cells in insulin-resistant horses and 64%, 18%, and 9%, respectively, of pancreatic islet cells in insulin-sensitive horses. Expression of insulin and somatostatin did not differ between insulin-resistant and insulin-sensitive horses, but the median percentage of glucagon-expressing cells in the islets of insulin-resistant horses was significantly less than that in insulin-sensitive horses.

CONCLUSIONS AND CLINICAL RELEVANCE Results suggested that, in insulin-resistant horses, insulin secretion was not increased but glucagon production might be downregulated as a compensatory response to hyperinsulinemia.

Abstract

OBJECTIVE To assess insulin, glucagon, and somatostatin expression within pancreatic islets of horses with and without insulin resistance.

ANIMALS 10 insulin-resistant horses and 13 insulin-sensitive horses.

PROCEDURES For each horse, food was withheld for at least 10 hours before a blood sample was collected for determination of serum insulin concentration. Horses with a serum insulin concentration < 20 μU/mL were assigned to the insulin-sensitive group, whereas horses with a serum insulin concentration > 20 μU/mL underwent a frequently sampled IV glucose tolerance test to determine sensitivity to insulin by minimal model analysis. Horses with a sensitivity to insulin < 1.0 × 10−4 L•min−1•mU−1 were assigned to the insulin-resistant group. All horses were euthanized with a barbiturate overdose, and pancreatic specimens were harvested and immunohistochemically stained for determination of insulin, glucagon, and somatostatin expression in pancreatic islets. Islet hormone expression was compared between insulin-resistant and insulin-sensitive horses.

RESULTS Cells expressing insulin, glucagon, and somatostatin made up approximately 62%, 12%, and 7%, respectively, of pancreatic islet cells in insulin-resistant horses and 64%, 18%, and 9%, respectively, of pancreatic islet cells in insulin-sensitive horses. Expression of insulin and somatostatin did not differ between insulin-resistant and insulin-sensitive horses, but the median percentage of glucagon-expressing cells in the islets of insulin-resistant horses was significantly less than that in insulin-sensitive horses.

CONCLUSIONS AND CLINICAL RELEVANCE Results suggested that, in insulin-resistant horses, insulin secretion was not increased but glucagon production might be downregulated as a compensatory response to hyperinsulinemia.

Insulin resistance is defined as a reduction in the action of insulin on target tissues, and it is an important component of insulin dysregulation in horses. Insulin dysregulation, abnormally increased adiposity, altered adipokine concentrations, and dyslipidemia are associated with the development of laminitis in horses and can be collectively referred to as equine metabolic syndrome.1,2 Horses with insulin resistance have hyper insulinemia, which is attributed to both abnormally increased insulin secretion and abnormally decreased insulin clearance by the liver and peripheral tissues.3,4 Although the increase in insulin secretion has been clinically evaluated in insulin-resistant horses, changes in the functional mass of β cells (cells that secrete insulin) within the pancreatic islets of horses with and without insulin resistance have not been investigated. Glucagon can also increase insulin secretion3; therefore, the functional mass of a cells (cells that produce glucagon) in pancreatic islets also warrants investigation. Somatostatin is secreted from δ cells within pancreatic islets and exerts paracrine effects on adjacent α and β cells to regulate glucagon and insulin secretion, respectively. Thus, a decrease in the expression of somatostatin might allow for an increase in insulin secretion, and the functional mass of somatostatin-expressing cells in pancreatic islets of horses also needs to be assessed.

Immunohistochemistry is the preferred method for evaluation of glucagon and somatostatin expression because referral laboratories do not measure those hormones in the plasma or serum of horses, and paracrine signaling is important within pancreatic islets. Understanding changes in the production of insulin, glucagon, and somatostatin in insulin-resistant horses is vital to elucidation of the complex pathogenesis of insulin dysregulation in horses and may lead to the development of more effective treatment strategies for affected horses.

The purpose of the study reported here was to evaluate insulin, glucagon, and somatostatin expression within pancreatic islets of horses with and without insulin resistance. Our hypothesis was that insulin expression by β cells of insulin-resistant horses would be greater than that in the insulin-sensitive horses.

Materials and Methods

Animals

All study procedures were approved by the University of Tennessee Institutional Animal Care and Use Committee. Horses assigned to the insulin-resistant group were part of a research herd maintained at the University of Tennessee. The insulin-resistant horses were humanely euthanized over a period of 12 months for various reasons including laminitis, other medical problems, and financial constraints associated with maintaining the herd. All horses with insulin resistance were fed grass hay with a nonstructural carbohydrate content of < 12% on a dry-matter basis in an amount calculated to meet maintenance requirements, in addition to a balanced supplemental vitamin and mineral mix and housed in dirt enclosures without access to pasture for a minimum of 6 months prior to euthanasia.

Horses assigned to the insulin-sensitive group were donated to the University of Tennessee Veterinary Teaching Hospital after their owners opted to euthanize them. Reasons for donation included behavioral problems, lameness, blindness, and other chronic medical conditions. Dietary histories were not obtained for those horses.

Serum insulin concentration

For the horses in the insulin-sensitive group, a blood sample (10 mL) was collected by jugular venipuncture for measurement of serum insulin concentration, the results of which were used to determine study eligibility. Initial blood samples were also collected from horses in the insulin-resistant group, and 4 horses had an additional blood sample (10 mL) collected at a later time when insulin concentrations were measured again for reasons unrelated to this study. The most recent blood sample for insulin measurement was obtained at a mean ± SD of 48 ± 37 days prior to euthanasia for horses in the insulin-resistant group and 1 to 7 days prior to euthanasia for horses in the insulin-sensitive group. For each horse, access to food but not water was restricted beginning at 10 pmthe night before the blood sample was collected (analogous to fasting in human medicine), and all blood samples were collected between 8 am and 9 am. Horses had to have an unfed (ie, fasting) serum insulin concentration > 20 μU/mL to be considered for the insulin-resistant group and < 20 μU/mL to be assigned to the insulin-sensitive group. Owner-donated horses with fasting serum insulin concentrations ≥ 20 μU/mL were not eligible for study enrollment.

FSIGTT

Results of the FSIGTT in conjunction with a fasting serum insulin concentration were used to assign horses to the insulin-resistant group; the FSIGTT was not performed for horses assigned to the insulin-sensitive group. To be assigned to the insulin-resistant group, horses had to have a fasting serum insulin concentration > 20 μU/mL and an SI < 1.0 × 10−4 L•min−1•mU−1 when minimal model analysis of FSIGTT results was performed as described.5,6

To perform the FSIGTT, access to food but not water was restricted beginning at 10 pm the night before and until the test was completed. A catheter was aseptically placed in a jugular vein for blood sample collection and dextrose and insulin administration. Blood samples were collected at −10, −5, and 0 minutes (immediately before) before infusion of an IV bolus of 50% dextrose solution (50 mg/kg) and at 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 14, 16, and 19 minutes after dextrose infusion for determination of glucose and insulin concentrations. An IV bolus of regular insulina (30 mU/kg) was administered 20 minutes after dextrose infusion, and additional blood samples were collected at 22, 23, 24, 25, 27, 30, 35, 40, 50, 60, 70, 80, 90, 100, 120, 150, and 180 minutes after dextrose infusion. All blood samples were obtained via the jugular catheter, and during each sample acquisition, blood (10 mL) was collected into a blood collection tube containing sodium fluoride-potassium oxalate for determination of plasma glucose concentration and a serum separator tube for determination of serum insulin concentration.

Blood sample processing and analysis

Blood samples collected in the sodium fluoride-potassium oxalate tubes were immediately cooled on ice, and blood samples collected in serum separator tubes were allowed to clot at room temperature (approx 22°C) for 1 hour. All blood samples were centrifuged at 1,000 × g for 10 minutes, and the plasma or serum was harvested from each sample and stored frozen at −20°C until analyzed. The plasma glucose concentration in each sample was measured in duplicate by use of a colorimetric assayb on an automated discrete analyzer.c Serum insulin concentration was measured in duplicate with a radioimmunoassayd previously validated for use with equine serum and plasma in our laboratory and by others.7 An intra-assay coefficient of variation < 5% was required for acceptance of glucose assay results, and < 10% was required for acceptance of insulin assay results. Minimal model8 parameters for SI, glucose effectiveness, acute insulin response to glucose, and disposition index were calculated as described9 with computer software.e,f

Pancreatic specimens

All horses were euthanized with an IV infusion of a barbiturate overdose and necropsied as soon as possible after death was confirmed. Pancreatic tissue specimens collected during necropsy were fixed in neutral-buffered 10% formalin. The specimens were then processed in a routine manner and prepared with H&E stain for histologic evaluation or by immunohistochemical methods for determination of insulin, glucagon, and somatostatin expression in pancreatic islets. All pancreatic specimens were evaluated by the same board-certified veterinary pathologist (KMN).

Immunohistochemical analysis

Pancreatic tissue from each horse was sent to the Histology Laboratory at the University of California-Davis Veterinary Medical Teaching Hospital for immunohistochemical staining. For each horse, each of 3 pancreatic sections was stained with antibodies against insuling (dilution, 1:600), glucagonh (dilution, 1:100), or somatostatini (dilution, 1:1,000). Each of the stains used was validated for use in canine and human pancreatic tissue. For all 3 antibodies, the chromogen was present only in the islets; no chromogen was present in the negative control specimens.

Following immunohistochemical staining, digital images of 10 islets/slide were obtained at 600X magnification for each hormone (insulin, glucagon, and somatostatin) by use of a microscope-mounted digital cameraj and microimaging software.k The images were digitally analyzed with image processing softwarel to determine the area of each islet and extent of immunoreactivity for each hormone in the islets. Briefly, the images were first calibrated to the size of the 600X field of view by use of a micrometer measurement bar. Each islet was outlined to identify and measure the area of interest. Then, with the software set to a sensitivity of 4 and pixel expansion of 1, the area of immunoreactivity (brown chromogen) was digitally selected, and the percentage of immunoreactivity for the hormone of interest within each islet was calculated.

Statistical analysis

The data distributions for continuous variables were assessed for normality by means of the Shapiro-Wilk test. The mean ± SD was reported for normally distributed variables, whereas the median (IQR) was reported for variables that were not normally distributed. The Mann-Whitney U test was used to compare age, BCS, and fasting serum insulin concentration between the insulin-resistant and insulin-sensitive groups. Expression of each hormone (insulin, glucagon, and somatostatin) was determined by the islet area and percentage of immunohistochemical staining (immunoreactivity) for that hormone within the islets. The islet area and percentage of immunoreactivity for each hormone within the islets were compared between the insulin-resistant and insulin-sensitive groups by use of 2-tailed t tests when data were normally distributed and Mann-Whitney U tests when data were not normally distributed. Then, logistic regression models were used to evaluate whether the expression of each hormone varied significantly between the insulin-resistant and insulin-sensitive groups when adjusted for age. All analyses were performed with statistical software,m and values of P < 0.05 were considered significant.

Results

Horses

A total of 23 horses were enrolled in the study; 13 horses were assigned to the insulin-sensitive group, and 10 were assigned to the insulin-resistant group. The insulin-sensitive group consisted of 6 mares and 7 geldings with a median age of 12 years (IQR, 3 to 25 years) and BCS (scored on a scale of 1 to 9) of 3.5 (IQR, 2.5 to 8) and included 5 Tennessee Walking Horses, 3 Quarter Horses, 2 American Paint Horses, and 1 each of Arabian, National Show Horse, and Saddlebred. The insulin-resistant group consisted of 4 mares and 6 geldings with a median age of 15.5 years (IQR, 7 to 22 years) and BCS of 8 (IQR, 7 to 9) and included 3 Arabians, 2 Paso Finos, and 1 each of Kentucky Mountain Saddle Horse, Missouri Fox Trotter, Morgan, Tennessee Walking Horse, and Tennessee Walking Horse–Quarter Horse crossbred. History and physical examination findings for horses in the insulin-resistant group included regional adiposity, obesity, lameness, and hoof changes consistent with laminitis.

The median age did not differ significantly (P = 0.304) between the insulin-sensitive and insulin-resistant groups. Median BCS for the insulin-resistant group was significantly (P < 0.001) greater than that for the insulin-sensitive group.

Serum insulin concentration and FSIGTT

A fasting serum insulin concentration was determined a second time for 4 of the 10 horses in the insulin-resistant group for reasons unrelated to the present study. When the initial fasting serum insulin concentrations for those 4 horses were used in the analysis, the median fasting serum insulin concentration for the insulin-resistant group (77 μU/mL; IQR, 11 to 622 μU/mL) was significantly (P < 0.001) greater than that for the insulin-sensitive group (7 μU/mL; IQR, 2 to 17 μU/mL). When the second fasting serum insulin concentrations for those 4 horses were used in the analysis, the median fasting serum insulin concentration for the insulin-resistant horses (104 μU/mL; IQR, 10 to 622 μU/mL) was still significantly (P < 0.001) greater than that for the insulin-sensitive horses. The most recent fasting serum insulin concentration (concentration determined closest to euthanasia) was > 20 μU/mL for 9 of the 10 horses in the insulin-resistant group and 10 μU/mL for the remaining horse.

The interval between FSIGTT and euthanasia varied among the horses in the insulin-resistant group, and the mean ± SD interval between FSIGTT and euthanasia was 7.5 ± 9.1 months. Minimal model analysis of FSIGTT results indicated that the median SI for horses in the insulin-resistant group was 0.11 × 10−4 L•min−1•mU−1 (IQR, 0.02 × 10−4 L•min1•mU−1 to 0.94 × 10−4 L•min−1•mU−1).

Insulin, glucagon, and somatostatin expression in pancreatic islets

The mean total area of pancreatic islets did not differ significantly between the insulin-resistant and insulin-sensitive groups. Subjectively, insulin-expressing cells (β cells) tended to be located peripherally whereas glucagon-expressing cells (α cells) were located centrally within the pancreatic islets (Figure 1). Somatostatin-expressing cells (δ cells) were randomly scattered throughout the islets.

Figure 1—
Figure 1—

Representative photomicrographs of pancreatic sections that were obtained from horses with clinically normal insulin sensitivity and were immunohistochemically stained for detection of insulin (A and B), glucagon (C and D), and somatostatin (E and F) within the pancreatic islets before (A, C, and E) and after (B, D, and F) an image analysis program was used to change the brown chromagen within hormone-expressing cells to red for quantification of immunoreactivity. The image analysis program was also used to outline the pancreatic islets (not shown) to determine the islet area and percentage of immunoreactivity for each hormone within the islet. Bar = 20 μm.

Citation: American Journal of Veterinary Research 79, 2; 10.2460/ajvr.79.2.191

Image analysis data for pancreatic specimens immunohistochemically stained for detection of insulin, glucagon, and somatostatin were summarized (Table 1). The mean percentage of immunoreactivity for insulin, glucagon, and somatostatin was approximately 64%, 18%, and 9%, respectively, of the islet area for the insulin-sensitive group and 62%, 12%, and 7%, respectively, of the islet area for the insulin-resistant group. When data for the insulin-sensitive and insulin-resistant groups were combined, the mean percentage of immunoreactivity for insulin, glucagon, and somatostatin was approximately 63%, 15%, and 8%, respectively, of the islet area. Neither insulin nor somatostatin expression differed significantly between the insulin-resistant and insulin-sensitive groups, even when the analysis was adjusted for age. However, the median percentage of the islet area that stained positive for glucagon for the insulin-resistant group (12%; IQR, 5% to 13%) was significantly (P = 0.001) less than that for the insulin-sensitive group (18%; IQR, 16% to 22%).

Table 1—

Summary of image analysis data for pancreatic specimens that were obtained from 10 insulin-resistant horses and 13 insulin-sensitive horses and were immunohistochemically stained to detect insulin, glucagon, and somatostatin within the islets.

HormoneVariableInsulin-resistant horsesInsulin-sensitive horsesP value
InsulinIslet area stained (μm2)1,261 ± 5981,368 ± 3220.618
Total islet area (μm2)2,231 ± 1,0352,406 ± 4760.594
Percentage of islet area stained62 ± 564 ± 60.449
GlucagonIslet area stained (μm2)414 ± 255612 ± 2150.063
Total islet area (μm2)3,551 ± 1,1433,112 ± 9120.317
Percentage of islet area stained12 (5–13)18 (16–22)0.001
SomatostatinIslet area stained (μm2)223 ± 154253 ± 2110.712
Total islet area (μm2)3,236 ± 9252,629 ± 5480.063
Percentage of islet area stained7 ± 49 ± 80.388

Values represent mean ± SD or median (IQR) unless otherwise specified. Values of P < 0.05 were considered significant. Horses were euthanized with a barbiturate overdose and necropsied as soon as possible after death was confirmed to collect pancreatic specimens for immunohistochemical evaluation for insulin, glucagon, and somatostatin expression. For each immunohistochemically stained specimen, digital images of 10 islets/slide were obtained at 600X magnification by use of a microscope-mounted camera and analyzed with image-processing software to determine the mean islet area and mean percentage of immunoreactivity for the hormone of interest within those islets.

Discussion

Results of the present study indicated that immunohistochemical staining could be used to identify and distinguish cells expressing insulin (β cells), glucagon (α cells), and somatostatin (δ cells) in the pancreatic islets of horses. Insulin-expressing cells tended to be located at the periphery of islets, whereas glucagon-expressing cells were centrally located. Results also indicated that cells expressing insulin, glucagon, and somatostatin make up approximately 63%, 15%, and 8%, respectively, of the pancreatic islet area of horses. To our knowledge, this study was the first to evaluate all 3 hormones (insulin, glucagon, and somatostatin) in the pancreatic islets of horses. Neither insulin nor somatostatin expression within the pancreatic islets differed significantly between the insulin-resistant and insulin-sensitive horses; however, glucagon expression was significantly less for insulin-resistant horses (12%) than for insulin-sensitive horses (18%).

Pancreatic islets are predominately composed of 3 cell types (α, β, and δ cells), which are indistinguishable on tissue specimens stained with H&E stain. Alpha cells secrete glucagon, β cells secrete insulin, and δ cells secrete somatostatin. The abundance and location of each of those immunoreactive hormones can be determined with immunohistochemical staining methods. The primary role of insulin is to stimulate glucose transport into cells, whereas glucagon facilitates glucose availability during times of need.10 Glucagon can also stimulate insulin secretion.3 Somatostatin inhibits secretion of both insulin and glucagon through paracrine signaling.10 In rodents and dogs, β cells are located centrally, whereas α and δ cells tend to be located along the periphery of the pancreatic islets.11,12 In humans, α and β cells are randomly distributed throughout the pancreatic islets.11,12 Interestingly, for the horses of the present study, β cells were located along the periphery and a cells were located centrally in the pancreatic islets; that distribution was the same for horses in both the insulin-resistant and insulin-sensitive groups. The clinical importance of the fact that the distribution of a and P cells within the pancreatic islets of horses is opposite that in dogs and rodents is unclear. Similar to other mammals, δ cells were randomly distributed throughout the pancreatic islets of the horses of the present study.

In rats and mice, β cells comprise 60% to 80%, whereas α cells comprise 15% to 20% and δ cells comprise < 10% of pancreatic islet cells.11 In humans, pancreatic islets are comprised of approximately 50% P cells, 40% α cells, and 10% δ cells.11 In dogs, pancreatic islets are comprised of 60% to 80% β cells, 5% to 30% α cells, and < 5% δ cells, which are randomly distributed throughout the islets.12 For the horses of the present study, the mean percentage of β cells in the pancreatic islets (63%) was slightly lower than that for other nonhuman mammals, whereas the mean percentages of α cells (15%) and δ cells (8%) were consistent with those for other species.10–12

The immunohistochemical staining method used must be considered when percentages of insulin-, glucagon-, and somatostatin-expressing cells are assessed. In the present study, the immunoreactivity of each hormone was quantitatively assessed on digital images, and the immunohistochemically stained area was compared with the total islet area. Pancreatic islets are histologically well-defined in horses and were manually outlined before the total islet area was calculated. In the present study, each pancreatic section was stained to detect the immunoreactivity of only 1 hormone. Thus, for each horse, 3 different sections of the pancreas were evaluated to determine the immunoreactivity of the 3 hormones. Consequently, the total islet area differed among the 3 hormones, and the percentages of islet area stained for insulin, glucagon, and somatostatin did not always sum to 100%. Although there were small areas of the islet that were not immunoreactive for any of the 3 hormones evaluated, this was an expected phenomenon associated with immunohistochemical staining and does not detract from the overall findings of this study. An immunohistochemical staining procedure in which all 3 hormones could be distinguished on the same section or slide would be ideal to assess relative hormone expression and may be developed for future studies.

Insulin expression did not differ significantly between the insulin-resistant and insulin-sensitive groups, but glucagon expression for the insulin-resistant group was significantly lower than that for the insulin-sensitive group, regardless of subject age. Thus, our hypothesis that insulin expression of P cells in insulin-resistant horses would be greater than that in the insulin-sensitive horses was not supported. Also, there was no evidence that β-cell hyperplasia or hypertrophy was associated with hyperinsulinemia in insulin-resistant horses. Abnormally increased insulin secretion and decreased insulin clearance can contribute to hyperinsulinemia, and hepatic insulin clearance is abnormally decreased in horses with insulin resistance.4 Incretin hormones, such as glucagon-like peptide-1 and glucose-dependent insulinotropic peptide (formerly known as gastric inhibitory peptide), also increase the release of insulin from the pancreas.13–15 Enteroendocrine cells release those 2 incretin hormones following exposure to glucose. Horses and ponies with insulin resistance have abnormally increased concentrations of glucagon-like peptide-1, which is expected to increase insulin secretion and lead to β-cell hypertrophy or hyperplasia over time.13–15 The immunohistochemical methods used in the present study were used to assess the abundance of insulin within the pancreatic islets, but the insulin secretion rate was not measured. The methods used in the present study were insufficient to detect β-cell hypertrophy or hyperplasia, but horses and ponies with insulin resistance may secrete insulin at a high rate without a concurrent increase in the amount of insulin stored within the pancreatic islets.

The apparent decreased glucagon expression in insulin-resistant horses relative to that in insulin-sensitive horses warrants further study. Glucagon is an insulin secretagogue,3 and a decrease in glucagon expression in the pancreatic islets of insulin-resistant horses may be a response to hyperinsulinemia. In humans, it has been hypothesized that insulin inhibits glucagon-secreting α cells.16 Thus, hyperinsulinemia would be expected to decrease glucagon secretion, and blood glucagon concentration does in fact decrease in healthy human subjects who receive IV infusions of insulin.16 However, abnormally increased glucagon secretion and hyperglucagonemia during periods of fasting have been associated with insulin resistance in humans and contribute to the development of hyperglycemia in human patients with type 2 diabetes mellitus.17 Human patients with type 2 diabetes mellitus develop β-cell hypertrophy and hyperplasia early in the disease process, followed by β-cell loss as the disease progresses.18 In Nile rats (Arvicanthis niloticus) in which diabetes mellitus was experimentally induced with a high-calorie, low-fiber diet, α-cell area and fasting blood glucose concentrations increased during the first 12 months on the experimental diet.19 Therefore, the apparent decrease in glucagon expression in the insulin-resistant horses of the present study suggested that α-cell area was likewise decreased and contrary to findings in rats and humans.17–19 That conflicting finding may be a result of interspecies differences. Chronic hyperinsulinemia is the most frequently detected glucose-insulin disorder of horses, whereas blood insulin concentrations tend to decrease with disease progression in human patients with type 2 diabetes mellitus.3 Further research is necessary to determine whether a decrease in α-cell mass represents a cause or effect of insulin resistance in horses.

The selection criteria used to define the 2 study groups and insulin status of horses at the time of euthanasia were limitations of the present study. To be included in the insulin-resistant group, horses had to have an SI < 1.0 × 10−4 L•min−1•mU−1. Although an SI cutoff for the diagnosis of insulin resistance in horses has not been established, the SI cutoff selected as a criterion for the insulin-resistant group in the present study has been used by other investigators20 and is similar to the lowest quintile range for the SI (0.14 × 10−4 L•min1•mU−-1 to 0.78 × 10−4 L•min1•mU−1) of 46 healthy horses.21 The use of the selected SI cutoff was supported by the fact that the fasting serum insulin concentration for the insulin-resistant group was significantly greater than that for the insulin-sensitive group.

The interval between determination of insulin status by means of the FSIGTT and euthanasia for the horses in the insulin-resistant group was also a concern, which we attempted to address by evaluating the most recent fasting serum insulin concentration (concentration determined closest to the time of euthanasia) for those horses. The mean ± SD interval between determination of the most recent fasting serum insulin concentration and euthanasia (1.6 ± 1.2 months) was much shorter than the interval between the FSIGTT and euthanasia (7.5 ± 91 months), and the most recent fasting serum insulin concentration was > 20 μU/mL for 9 of the 10 horses in the insulin-resistant group. The most recent fasting serum insulin concentration was 10 μU/mL for the remaining horse in the insulin-resistant group, and it is possible that the insulin sensitivity increased over time for that horse. The interval between testing to determine insulin status and euthanasia should be standardized for horses of future studies.

Another limitation of the present study was the fact that horses assigned to the insulin-sensitive group did not undergo an FSIGTT and minimal model analysis. To be included in the insulin-sensitive group, horses had to have a fasting insulin concentration < 20 μU/mL, which was assumed to be indicative of clinically normal insulin sensitivity. However, it is possible that some horses in the insulin-sensitive group were insulin resistant and would have been classified as such had the FSIGTT been performed.

The insulin status of horses may also be affected by season and diet; however, further investigation is required to confirm or refute that. Although evidence of β-cell hypertrophy or hyperplasia was not observed within the pancreatic islets of the insulin-resistant horses of the present study, alterations might develop across different seasons or in response to different diets. The insulin-resistant horses of this study were fed hay with a low nonstructural carbohydrate content (< 12%), instead of pasture grass or grain.

Finally, > 10 breeds of horses from multiple sources were represented in the present study, which caused substantial variation among horses. It is possible that insulin metabolism and hormone production within pancreatic islets may indeed differ between insulin-resistant and insulin-sensitive horses of the same breed, but the large variation among the horses of the present study limited our ability to detect those differences.

The pancreatic mass should be measured, the number of total islets should be estimated, and the pancreatic site selected for sampling should be standardized in future studies of pancreatic hormone expression in horses with and without insulin dysregulation. Although the expressions of insulin, glucagon, and somatostatin relative to the islet area were calculated for the horses of the present study, changes in the total number of pancreatic islets were not determined. It is possible that insulin-resistant horses have greater pancreatic mass or more islets than healthy horses, and this should be investigated further.

In the present study, insulin expression within the pancreatic islets did not differ between insulin-resistant and insulin-sensitive horses, which suggested that insulin resistance did not increase insulin secretion in horses. However, glucagon expression in the islets of insulin-resistant horses was significantly less than that of insulin-sensitive horses, which might indicate that glucagon production is downregulated as a compensatory response to hyperinsulinemia in insulin-resistant horses.

Acknowledgments

Supported by a grant from the University of Tennessee Faculty Education Advancement and Research Fund.

The authors declare that there were no conflicts of interest.

ABBREVIATIONS

BCS

Body condition score

FSIGTT

Frequently sampled IV glucose tolerance test

IQR

Interquartile (25th to 75th percentiles) range

SI

Sensitivity to insulin

Footnotes

a.

Humulin R, Eli Lilly and Co, Indianapolis, Ind.

b.

Glucose, Roche Diagnostic Systems Inc, Somerville, NJ.

c.

Cobas Mira, Roche Diagnostic Systems Inc, Somerville, NJ.

d.

Coat-A-Count Insulin, Diagnostic Products Corp, Los Angeles, Calif.

e.

MinMod Millennium, version 6.10, Raymond Boston, University of Pennsylvania, Kennett Square, Pa.

f.

Stata/MP 9.2, Parallel Edition, version 13.1, Stata Corp, College Station. Tex.

g.

Dako A0564 (guinea pig origin), Carpinteria, Calif.

h.

ThermoFisher Scientific, Waltham, Mass.

i.

ImmunoStar (rabbit origin), Hudson, Wis.

j.

DP25, Olympus Corp, Center Valley, Pa.

k.

Cell Sens Entry, Olympus Corp, Center Valley, Pa.

l.

Media Cybernetics Inc, Bethesda, Md.

m.

Stata, version 13, Stata Corp, College Station, Tex.

References

  • 1. Frank N, Geor RJ, Bailey SR, et al. Equine metabolic syndrome. J Vet Intern Med 2010;24:467475.

  • 2. Johnson PJ, Wiedmeyer CE, LaCarrubba A, et al. Diabetes, insulin resistance, and metabolic syndrome in horses. J Diabetes Sci Technol 2012;6:534540.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 3. Frank N, Tadros EM. Insulin dysregulation. Equine Vet J 2014;46:103112.

  • 4. Tóth F, Frank N, Martin-Jiménez T, et al. Measurement of C-peptide concentrations and responses to somatostatin, glucose infusion, and insulin resistance in horses. Equine Vet J 2010;42:149155.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5. Hoffman RM, Boston RC, Stefanovski D, et al. Obesity and diet affect glucose dynamics and insulin sensitivity in Thoroughbred geldings. J Anim Sci 2003;81:23332342.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6. Tóth F, Frank N, Elliot SB, et al. Optimisation of the frequently sampled intravenous glucose tolerance test to reduce urinary glucose spilling in horses. Equine Vet J 2009;41:844851.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 7. Freestone JF, Wolfsheimer KJ, Kamerling SG, et al. Exercise induced hormonal and metabolic changes in Thoroughbred horses: effects of conditioning and acepromazine. Equine Vet J 1991;23:219223.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8. Bergman RN, Phillips LS, Cobelli C. Physiologic evaluation of factors controlling glucose tolerance in man: measurement of insulin sensitivity and beta-cell glucose sensitivity from the response to intravenous glucose. J Clin Invest 1981;68:14561467.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9. Boston RC, Stefanovski D, Moate PJ, et al. MINMOD Millennium: a computer program to calculate glucose effectiveness and insulin sensitivity from the frequently sampled intravenous glucose tolerance test. Diabetes Technol Ther 2003;5:10031015.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10. Haschek WM, Rousseaux CG, Wallig MA. Pancreas. In: Fundamentals of toxicologic pathology. 2nd ed. London: Elsevier, 2010;237260.

  • 11. Steiner DJ, Kim A, Miller K, et al. Pancreatic islet plasticity: interspecies comparison of islet architecture and composition. Islets 2010;2:135145.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12. Hawkins KL, Summers BA, Kuhajda FP, et al. Immunocytochemistry of normal pancreatic islets and spontaneous islet cell tumors in dogs. Vet Pathol 1987;24:170179.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13. Bamford NJ, Baskerville CL, Harris PA, et al. Postprandial glucose, insulin, and glucagon-like peptide-1 responses of different equine breeds adapted to meals containing micronized maize. J Anim Sci 2015;93:33773383.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14. de Laat MA, McGree JM, Sillence MN. Equine hyperinsulinemia: investigation of the enteroinsular axis during insulin dysregulation. Am J Physiol Endocrinol Metab 2016;310:E61E72.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15. Chameroy KA, Frank N, Elliott SB, et al. Comparison of plasma active glucagon-like peptide 1 concentrations in normal horses and those with equine metabolic syndrome and in horses placed on a high-grain diet. J Equine Vet Sci 2016;40:1625.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16. Gosmanov NR, Szoke E, Israelian Z, et al. Role of the decrement in intraislet insulin for the glucagon response to hypoglycemia in humans. Diabetes Care 2005;28:11241131.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17. Ferrannini E, Muscelli E, Natali A, et al. Association of fasting glucagon and proinsulin concentrations with insulin resistance. Diabetologia 2007;50:23422347.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18. Chang-Chen KJ, Mullur R, Bernal-Mizrachi E. Beta-cell failure as a complication of diabetes. Rev Endocr Metab Disord 2008;9:329343.

  • 19. Yang K, Gotzmann J, Kuny S, et al. Five stages of progressive β-cell dysfunction in the laboratory Nile rat model of type 2 diabetes. J Endocrinol 2016;229:343356.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20. Dunbar LK, Mielnicki KA, Dembek KA, et al. Evaluation of four diagnostic tests for insulin dysregulation in adult light-breed horses. J Vet Intern Med 2016;30:885891.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21. Treiber KH, Kronfeld DS, Hess TM, et al. Use of proxies and reference quintiles obtained from minimal model analysis for determination of insulin sensitivity and pancreatic beta-cell responsiveness in horses. Am J Vet Res 2005;66:21142121.

    • Crossref
    • Search Google Scholar
    • Export Citation

Contributor Notes

Dr. Ehrensing's present address is Department of Pathobiology and Diagnostic Investigation, College of Veterinary Medicine, Michigan State University, East Lansing, MI 48824.

Address correspondence to Dr. Newkirk (knewkirk@utk.edu).