Materials and Methods

Horses—Six mature Standardbreds (2 mares and 4 geldings, 4 to 5 years old; mean ± SD weight, 443.2 ± 4.7 kg) that had been paddock rested for 1 month were used in the study. Horses were housed in box stalls with approximately 4 h/d of turnout into a small paddock. Horses were fed mixed grass hay at approximately 2.5% of body weight and had ad libitum access to fresh water and a salt block. Because an increase in dietary hydrolyzable carbohydrate is one factor that may contribute to alterations in insulin sensitivity¹⁷ and expression of GLUT-4 and insulin-signaling proteins, no grain concentrate was fed, and horses wore muzzles during turnout to prevent grazing. All procedures were approved by the Animal Care Committee of the University of Guelph.

Experimental design—A randomized, balanced crossover design was used. Horses received IV administered dexamethasone^a (dose, 0.08 mg/kg) or the equivalent volume of saline (0.9% NaCl) solution (control) every 48 hours for 3 weeks (11 treatments overall). After a 3-week washout period, the experimental protocol was repeated, with each group of horses receiving the other treatment. Treatments were administered at 8:00 AM. One day after the final treatment, a 2-hour EHC was administered, and muscle biopsy specimens were collected before and at the end of the EHC.

EHC—The EHC was performed according to described methodology.¹⁸ Feed was withheld overnight (approx 10 hours) prior to administration of each EHC. On the morning of the procedure (at 8:00 AM), horses were weighed to an accuracy of ± 0.5 kg on an electronic scale,^b and catheters^c were inserted into each jugular vein after aseptic preparation and desensitization of the overlying skin. One of the catheters was used for infusion of glucose and insulin, and all blood samples were obtained via the catheter in the opposite jugular vein. A muscle biopsy specimen was obtained, and blood was collected for determination of baseline whole blood glucose and serum insulin concentrations and to harvest serum (2 mL) to be used for preparation of the insulin infusate. Horses were returned to their stalls for a 1-hour recovery period, after which they were positioned in stocks for the EHC. A priming dose (18 mU/kg) of insulin^d was administered IV, and simultaneous infusions of 50% dextrose solution (variable rate) and insulin (3 mU/kg per minute) were started. Blood glucose concentration was measured every 5 minutes by use of an automated analyzer,^e and the rate of dextrose infusion was adjusted to maintain a glucose concentration of 5 mmol/L. Additional blood samples were collected every 15 minutes for subsequent measurement of serum insulin concentrations. Samples were placed in evacuated tubes containing no additive and centrifuged for 15 minutes at 1,600 × g, after which serum was harvested and stored at −20°C until analysis. A second biopsy sample was collected from the contralateral middle gluteal muscle at the end of the EHC, after which the glucose and insulin infusions were stopped.

Blood glucose and serum insulin data were used to calculate 2 measures of insulin sensitivity. Calculations were made from data collected during the last 60 minutes of the EHC, with the first hour used as an equilibration period to achieve steady-state euglycemia. The mean rate of glucose infusion (mg/kg per minute) was calculated for each 5-minute interval, and the mean infusion rate for the last 60 minutes of the EHC (M₆₀) was determined. The mean rate of dextrose infusion for the 60-to-90– and 90-to-120–minute periods was also calculated to evaluate the effect of time during the EHC on this measure of insulin sensitivity. A second measure of insulin sensitivity (M/I₆₀) was derived from the ratio of M to the prevailing insulin concentration during the last 60 minutes of the EHC (I₆₀). This ratio reflects the rate of glucose disposal per unit of insulin and is an index of tissue sensitivity to exogenous insulin.^19,20 Insulin clearance during the last 60 minutes of the EHC (MCR_I) was calculated on the basis of the insulin infusion rate and change in serum insulin concentration.¹⁹

Muscle biopsy—Muscle biopsy specimens were obtained at a standardized site and depth (6 cm) in the middle gluteal muscle by use of the percutaneous needle biopsy technique,²¹ with a second specimen collected from the opposite side. An approximately 100-cm square area over each middle gluteal muscle was shaved and surgically prepared, and the skin and underlying tissue were desensitized by infiltration of local anesthetic.^f The muscle specimen (wet weight, approx 500 mg) was quickly blotted to remove excessive blood and immediately immersed in liquid nitrogen. All specimens were stored at −80°C until analysis. Phenylbutazone^g (2 g, PO) was administered after the second biopsy to alleviate discomfort associated with the procedure.

Serum immunoreactive insulin concentration—Serum insulin concentrations were measured in duplicate with a commercially available radioimmunoassay kit^h validated for use in the horse.²² Intra- and inter-assay CVs for the insulin assay were 2.9% and 4.8%, respectively.

Muscle enzyme activities—Glycogen synthase activity was determined according to a previously described method.^18,23 Duplicate homogenates of frozen muscle were incubated at 37°C with media containing 0 or 10 mmol of glucose-6-phosphate/L for determination of active and total GS activity, respectively. Uridine diphosphate-glucose was added to the samples to catalyze the incorporation of glucose from UDP-glucose to glycogen, in turn releasing UDP. Fluorescence was measured in the presence of UDP reagent, prior to and after addition of pyruvate kinase. Active and total GS activities were determined by the difference in NAD⁺ fluorescence measured in samples before and after addition of pyruvate kinase. Fractional velocity of GS was calculated as active GS activity divided by total GS activity. Intra- and interassay CVs were 5.9% and 6.4%, respectively.

Hexokinase activity was assayed according to described techniques¹⁸ by use of procedures outlined by earlier investigators.²⁴ In brief, muscle homogenates were incubated for 1 hour at room temperature (20°C) after addition of glucose-6-phosphate dehydrogenase. The reaction was stopped, phosphogluconate dehydrogenase was added to the homogenates, and fluorescence readings were taken after 15 minutes of incubation at room temperature. Intra- and interassay CVs were 1.7% and 2.1%, respectively.

Muscle glycogen determination—Frozen muscle specimens weighing 50 to 60 mg were lyophilized, pulverized, and dissected free of visible blood, connective tissue, and fat. Glycogen concentration (as glucosyl units) was determined in duplicate after acid hydrolysis following procedures described by Passonneau and Lauderdale.²⁵

Western immunoblot analysis of GLUT-4, PKB, and GSK-3—Western immunoblot was used to assess GLUT-4, PKB, and GSK-3α/β protein abundance and the phosphorylation state of PKB Ser⁴⁷³ and GSK-3α/β Ser^21/9 in muscle homogenates.^26,27 In brief, the total protein concentration of muscle lysates was determined by use of a commercially available kit,ⁱ and equal quantities of protein (40 μg) were separated by SDS-PAGE and transferred to nitrocellulose membranes.^j After blocking, membranes were incubated with a polyclonal, anti-GLUT-4,^k anti-PKB, anti-phospho PKB-Ser⁴⁷³ and anti-phospho GSK-3α/β Ser^21/9,^g or with monoclonal anti-GSK-3α/β^l followed by incubation with an antirabbit horseradish peroxidase–linked immunoglobulin-G antibody.^m Protein bands were viewed after application of a chemiluminescence reagent,ⁿ and signals were quantified by use of densitometry.° Intensity of GLUT-4, PKB, and GSK-3α/β protein bands was corrected to tubulin^p protein bands on the same membrane.

Statistical analysis—Descriptive statistics of continuous variables are expressed as mean ± SEM. Normality was tested with the Kolmorgorov-Smirnov statistic. Non-normally distributed variables were log₁₀ transformed to achieve normality (baseline insulin concentration and western immunoblot densitometry data). Paired variables (eg, basal insulin concentration, M₆₀, M/I₆₀, and MCR_I) were analyzed by paired t test. Muscle variables (eg, glycogen concentration, GLUT-4 protein abundance, and GS and hexokinase activities) were analyzed by 2-way ANOVA with repeated measures to evaluate the effects of treatment (dexamethasone-treated vs control horses), time (before vs after EHC), and their interaction. Pairwise comparisons were performed with the Bonferroni t test. Values of P < 0.05 were considered significant. A software program^q was used for statistical computations.

Results

EHC and insulin sensitivity—Resting (baseline) serum insulin concentration before the EHC was significantly (P < 0.001) higher in the dexamethasone treated horses than in control horses. Mean resting blood glucose concentration did not differ between treatments (P = 0.10; Table 1). Blood glucose concentrations during the 60-to-120–minute period of the EHC did not differ between treatments, whereas serum insulin concentration was significantly higher in treated horses than in control horses at most time points. Mean I₆₀ was higher (P < 0.05) in dexamethasone-treated horses than in control horses. Because the increment from basal insulin concentration was similar during the EHC, insulin clearance (MCR_I) did not differ between treatments. For each treatment, mean glucose infusion rate in the 60-to 90–minute period did not differ from that measured in the 90-to-120–minute period (Figure 1). Mean values for glucose infusion rate, M₆₀, and M/I₆₀ were approximately 70% lower (P < 0.001) in the dexamethasone group, compared with control horses.

Table 1—

Values for measures of glucose metabolism and insulin sensitivity during an EHC in 6 horses that received dexamethasone or an equivalent volume of saline (0.9%) solution (control).

View Table

View Full Size

Plots depicting mean ± SEM serum insulin concentrations (panel A) and glucose infusion rates (panel B) during an EHC after control (0.9% NaCl solution) and dexamethasone treatment in 6 Standardbreds. The priming dose of insulin and infusions of dextrose and insulin solutions commenced at time point zero. One blood sample was collected 60 minutes before time point zero. *Serum insulin concentrations were significantly (P < 0.05) higher in dexamethasone-treated horses than in control horses at baseline and at most time points during the EHC. #In the dexamethasone treatment group, mean glucose infusion rate was significantly (P < 0.001) lower than in control horses during the 60-to-90–minute (black bars) and 90-to-120–minute (white bars) periods. CON = Control horses. DEX = Dexamethasone treatment group.

Citation: American Journal of Veterinary Research 69, 1; 10.2460/ajvr.69.1.51

• Download Figure
• Download figure as PowerPoint slide

Muscle enzyme activities, immunoblot analyses, and muscle glycogen determination—No difference was detected in total GS activity between treatments or sampling times (Table 2). Mean basal GS activity and GS fractional velocity in post-EHC specimens were decreased in dexamethasone-treated horses, compared with control horses. In control horses, mean basal GS activity and fractional velocity were significantly (P < 0.01) higher in post-EHC specimens, compared with pre-EHC specimens, but were unchanged in dexamethasone-treated horses. Mean hexokinase activity was higher (P < 0.05) in treated horses than in control horses in both the pre- and post-EHC specimens, with no difference seen after insulin stimulation in either treatment. No difference was detected in total GLUT-4 protein abundance between treatments or sample times (ie, pre- vs post-EHC time points; Figure 2). Muscle glycogen concentration also did not differ between treatments or sample times.

View Full Size

Plot depicting mean ± SEM GLUT-4 protein abundance in lysates of middle gluteal muscle from the same 6 horses as in Figure 1 measured before (pre; black bars) and after (post; white bars) an EHC. Notice that there was no significant difference between treatments or sample times. For each treatment, representative immunoblots for GLUT-4 and tubulin (loading control) are indicated in the lower panel. See Figure 1 for remainder of key.

Citation: American Journal of Veterinary Research 69, 1; 10.2460/ajvr.69.1.51

• Download Figure
• Download figure as PowerPoint slide

Table 2—

Glycogen concentration and the activities of hexokinase and GS in middle gluteal muscle before (pre) and after (post) an EHC in the same 6 horses as in Table 1.

View Table

Expression of total PKB, GSK-3A, and GSK-3B, as determined by densitometric scanning of immunoblots, was similar in treated and control horses (Figures 3 and 4). Phosphorylation of PKB Ser⁴⁷³ was more than 5-fold higher (P < 0.05) in postversus pre-EHC specimens in both treated horses and control horses. However, no effect of treatment on PKB Ser⁴⁷³ phosphorylation was detected. In both control and treatment horses, phosphorylation of GSK-3A Ser²¹ or GSK-3B Ser⁹ did not change after the 2-hour period of hyperinsulinemia (ie, in post- vs pre-EHC samples), but phosphorylation of GSK-3A Ser²¹ and GSK-3B Ser⁹ was significantly (P < 0.05) lower in dexamethasone-treated horses, compared with control horses (Figure 5).

View Full Size

Plots depicting muscle tissue protein lysates resolved by SDS-PAGE and immunoblotted with anti-PKB total protein antibodies (A) and anti-phospho PKB Ser⁴⁷³ antibodies (B) and expressed as the ratio of PKB Ser⁴⁷³ to tubulin area density, before (pre; black bars) and after (post; white bars) an EHC in control horses and in dexamethasone-treated horses. Representative immunoblots (including tubulin as loading control) are indicated. *Within a treatment, values for PKB Ser⁴⁷³ were significantly (P < 0.05) different from pre-EHC values. Notice that values for total PKB and PKB Ser⁴⁷³ did not differ between treatments. See Figure 1 for remainder of key.

Citation: American Journal of Veterinary Research 69, 1; 10.2460/ajvr.69.1.51

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Plots depicting muscle tissue protein lysates resolved by SDS-PAGE and immunoblotted with monoclonal anti-GSK-3A (A) and anti-GSK-3B (B) antibodies, expressed as the ratio of GSK-3α/β to tubulin area density, before (pre; black bars) and after (post; white bars) an EHC in control and dexamethasone-treated horses. Representative immunoblots are indicated below the graphs. Notice that there was no difference in total GSK3-α/β protein abundance between treatments or sample times. See Figure 1 for remainder of key.

Citation: American Journal of Veterinary Research 69, 1; 10.2460/ajvr.69.1.51

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Plots depicting muscle tissue protein lysates resolved by SDS-PAGE and immunoblotted with anti-phospho GSK-3A Ser²¹ antibodies (A) and anti-phospho GSK-3B Ser⁹ antibodies (B) and expressed as the ratio of GSK-3α/β^21/9 to tubulin area density in control and dexamethasone-treated horses. Representative immunoblots are shown in the lower panels. *Values for dexamethasone-treated horses are significantly (P < 0.05) different from those in control horses. See Figure 1 for remainder of key.

Citation: American Journal of Veterinary Research 69, 1; 10.2460/ajvr.69.1.51

• Download Figure
• Download figure as PowerPoint slide

Discussion

In this study, the effects of 3 weeks of dexamethasone treatment on insulin-mediated glucose disposal and selected components of insulin signaling and glucose metabolism in skeletal muscle of horses were evaluated. The main effects of prolonged dexamethasone administration were induction of insulin resistance, manifested by an increase in resting serum insulin concentration and a decrease in the glucose infusion rate required to maintain euglycemia during the EHC; abrogation of the insulin-stimulated increase in muscle GS activity; a decrease in phosphorylation of GSK-3A Ser²¹ and GSK-3B Ser⁹; and a 30% to 40% increase in HK activity after dexamethasone treatment. Dexamethasone treatment did not affect muscle glycogen concentration; expression of total GLUT-4, PKB, and GSK-3α/β; or insulin-stimulated phosphorylation of PKB Ser⁴⁷³.

The dexamethasone-induced insulin resistance observed in the present study is consistent with findings from previous studies^5,6 in horses in which the same dexamethasone treatment regimen was used. When normalized for insulin concentrations, this insulin resistance was confirmed as a 70% decrease in whole-body glucose disposal. In addition, after dexamethasone treatment, we observed a 10-fold increase in resting serum insulin concentrations but unchanged blood glucose concentrations in the horses in the present study. Maintenance of glucose homeostasis requires that the product of insulin secretion and insulin sensitivity remain constant.²⁸ In the dexamethasone-treated horses in the study reported here, the hyperinsulinemia evident in the dexamethasone-treated group compensated for decreased tissue sensitivity to insulin such that glucose homeostasis was maintained.

The means by which glucocorticoids induce insulin resistance in horses are not known. In humans, treatment with dexamethasone results in increased hepatic gluconeogenesis and glucose output and decreased glucose utilization in insulin-sensitive tissues.²⁹ We investigated the effects of dexamethasone treatment on aspects of insulin signaling and glucose metabolism in skeletal muscle because, at least in humans, this tissue accounts for 70% to 90% of insulin-stimulated glucose uptake.³⁰ In skeletal muscle, the interaction of insulin with its receptor initiates a cascade of protein-protein interactions that mediates glucose uptake and storage (ie, glycogen synthesis). Activation of the kinase PKB is required for translocation of the insulin-sensitive glucose transporter GLUT-4 to the plasma membrane. Activated PKB also phosphorylates and deactivates GSK-3 (GSK-3A at Ser²¹ and GSK-3B at Ser⁹), promoting activation of GS and therefore glycogen synthesis.

In the study reported here, GS activity was not augmented in response to hyperinsulinemia after dexamethasone treatment. This finding is in agreement with observations in rats, in which dexamethasone treatment is associated with nearly complete inhibition of insulin's action in increasing GS fractional activity and a 70% to 80% reduction in insulin-stimulated glycogen synthesis.^31,32 Furthermore, reduced GSK-3 phosphorylation (ie, increased GSK-3 activity) appears to contribute to the decrease in insulin-stimulated GS activation in skeletal muscle from dexamethasone-treated rats, as indicated by partial restoration of insulin-stimulated GS fractional activity with pharmacologic inhibition of GSK-3.³² Consistent with these findings, there was a reduction in GSK-3 phosphorylation (GSK-3A Ser²¹ and GSK-3B Ser⁹) that was not altered by insulin stimulation in the dexamethasone-treated horses in the present study. Thus, the absence of an insulin-stimulated increase in GS fractional activity in the dexamethasonetreated horses may reflect the reduction in GSK-3α/β phosphorylation in the middle gluteal muscle. Failure to detect an increase in GSK-3A phosphorylation after hyperinsulinemia in the control group or the dexamethasone treatment group may reflect the training state of the horses in this study (untrained and sedentary). In previous work in our laboratory, an increase in GSK-3A, but not GSK-3B, phosphorylation in response to hyperinsulinemia was observed in equine skeletal muscle.²⁷ In that study, horses completed 2 months of physical conditioning prior to the onset of the study, and control samples were obtained from trained but nonexercised horses.

Reduction in skeletal muscle glycogen content can be a strong regulator of GS activity, given the inverse relationship between muscle glycogen concentration and GS fractional velocity.³³ Increased GS activity has been observed in horses after 3 days of glycogen-depleting exercise when glycogen was reduced by approximately 50%.³⁴ However, this relationship is not evident when there is no depletion or even supercompensation of muscle glycogen content.⁷ Horses used in the present study were sedentary; muscle glycogen concentration was not depleted by exercise and did not differ between the dexamethasone treatment group and control horses. It is therefore unlikely that glycogen content can account for the absence of increase in insulin-stimulated GS activity measured in the horses treated with dexamethasone. The lack of effect of dexamethasone on muscle glycogen concentration was also reported in a study of Quarter Horses with polysaccharide storage myopathy.⁵

Studies³² in rats have revealed a 40% to 60% decrease in insulin-stimulated glucose uptake by soleus muscle after dexamethasone treatment. This decrease in glucose uptake may reflect impairment of insulinstimulated GLUT-4 translocation to the sarcolemma,¹⁵ because dexamethasone treatment does not decrease GLUT-4 protein abundance or gene expression in skeletal muscle.^32,35–37 In the present study, there also was no change in GLUT-4 protein abundance after dexamethasone treatment. However, quantification of the membrane-associated fraction of total GLUT-4 to determine whether changes in GLUT-4 translocation occurred after dexamethasone treatment was not undertaken. Nonetheless, it is possible that decreased GLUT-4 translocation to the plasma membrane during the EHC contributed to the lower glucose disposal rate in the dexamethasone-treated horses.

Decreased insulin-stimulated PKB Ser⁴⁷³ phosphorylation in skeletal muscle without alteration in total PKB protein abundance has been a consistent finding after dexamethasone treatment in rats.^32,37 Because activation of PKB is important for insulin-stimulated glucose uptake, it has been proposed that this decrease in PKB Ser⁴⁷³ phosphorylation contributes to the dexamethasone-induced insulin resistance in skeletal muscle. However, in the study presented here, dexamethasone treatment did not alter phosphorylation of PKB Ser⁴⁷³ before or after insulin stimulation. Similar to findings from previous studies^31,32 in rats, data from the present study suggest that administration of dexamethasone induces skeletal muscle insulin resistance by impairing insulin-signaling proteins rather than by reducing expression of GLUT4, PKB, or GSK-3. In that context, it is more likely that insulin resistance in the dexamethasone-treated horses can be attributed to alterations or impairment in the signaling pathway distal to PKB.

The phosphorylation of glucose by hexokinase, along with glucose transport (mediated by GLUT-4) and storage (mediated by GS), is a potential rate-controlling step regulating insulin-stimulated muscle glucose metabolism, and defects in hexokinase function have been implicated in the development of insulin resistance.^38,39 Therefore, the increase in hexokinase seen after dexamethasone treatment was surprising in light of other observations consistent with development of insulin resistance. Nonetheless, 1 study¹⁵ in rats also revealed increased capacity for glucose phosphorylation in muscle after dexamethasone treatment. In human skeletal muscle, 2 isoforms of hexokinase activity (HKI and HKII) have been identified, with the expression and activity of HKII regulated by insulin.³⁹ As such, it is possible that the chronic state of hyperinsulinemia induced by administration of dexamethasone could account for the higher activity of this enzyme observed at the end of the treatment period.

The present study revealed that 21 days of dexamethasone treatment induced insulin resistance in Standardbreds as indicated by the substantial decrease in glucose infusion rate during the EHC. Although several alterations in proteins and enzymes involved in skeletal muscle glucose metabolism were observed after the period of induced insulin resistance, the potential effects of dexamethasone on other physiologic functions, including regulation of vascular function, should also be considered. In humans, 48 hours of dexamethasone treatment abrogated insulin-mediated vasodilation of the soleus and gastrocnemius muscle vascular bed during an EHC.^7,40 It is therefore possible that alterations in glucose and insulin delivery to skeletal muscle as a result of reduced blood flow represent another mechanism in dexamethasone-induced insulin resistance.