Experimental Diabetes Attenuates Cerebral Cortical−Evoked Forelimb Motor Responses

This study was conducted using protocols approved by the Institutional Animal Care and Use Committee in accordance with the principles of laboratory animal care (National Institutes of Health Publ. 86-23, 1985). All animals were housed in pairs, allowed standard rat diet and water ad libitum, and maintained on a 10-/14-h light/dark cycle. Adult male SD rats (initial body weight 300 g) were divided into three experimental groups: nondiabetic vehicle-treated control animals (n = 14), untreated diabetic animals (n = 14), and insulin-treated diabetic animals (n = 6). Rats from each group were evaluated at 8 weeks after vehicle or streptozotocin (STZ) administration.

Experimental diabetes was induced in nonfasted rats (initial body weight 300 g) by a single intraperitoneal injection of freshly prepared STZ (60 mg/kg body wt; Sigma, St. Louis, MO) dissolved in citrate buffer (100 mmol/l; pH 4.5). For the insulin-treated group, two sustained-release insulin pellets (∼2 units · 24 h⁻¹ · implant⁻¹ for >40 days; Lin Shin Canada, Ontario, Canada) were implanted subcutaneously 4 weeks after STZ injection to normalize and maintain blood glucose levels near physiological concentrations of 100 mg/dl. Control rats received an equal volume of intraperitoneally injected, freshly prepared citrate buffer (100 mmol/l; pH 4.5). Nonfasting blood glucose levels were determined before vehicle or STZ was administered and at regular weekly intervals throughout the study. Blood samples were obtained by tail prick, and glucose content was quantitated using a commercial glucometer (Glucometer Dex; Bayer, Elkhart, IN) routinely standardized to a control test. STZ-administered rats exhibiting initial blood glucose levels <300 mg/dl were considered to be nondiabetic and were excluded from further study. Animals were weighed daily to monitor disease progression and severity.

Forelimb motor cortical representation was determined by intracortical microstimulation (ICMS), as previously described (23,24). At 8 weeks after vehicle or STZ administration, rats were anesthetized with ketamine (100 mg/kg) and xylazine (5 mg/kg); the plane of anesthesia was maintained with a constant intraperitoneal infusion of ketamine (80 mg · kg⁻¹ · h⁻¹). A heating pad was used to maintain body temperature at 37°C. Briefly, rats were placed in a stereotaxic frame, the cisterna magna was opened, and the cerebrospinal fluid was allowed to drain to minimize swelling of the cerebral cortex during the microstimulation procedure. A craniotomy was then performed over the region of the right sensorimotor cortex, and the dura was removed. The exposed hemisphere was kept moist throughout the procedure with mineral oil. The forelimb motor area was explored in 0.5-mm increments over a 30-mm² rostral-caudal and medial-lateral cortical area (23) with a custom-made stimulating electrode (a sharpened glass-insulated tungsten wire, noninsulated tip length of 100 μm) using the bregma as a cranial landmark. At each cortical coordinate, a stimulus current (2–50 μA) was applied as a 300-ms train of pulses (duration 0.2 ms) at 350 Hz at an electrode depth of 1.7 mm below the cortical pial surface (23). For each animal, ∼80 separate cortical coordinates were stimulated, and evoked contralateral forelimb movements were scored by two independent investigators. Central motor conduction times were not determined in this study. All cortical points were screened for movement using an initial current of 50 μA. The current threshold was determined as the lowest current necessary to reliably produce a visible movement. Current spreading and subsequent activation of subcortical motor pathways was minimized by limiting stimulating currents to ≤50 μA. Forelimb boundaries were defined as cortical coordinates that no longer elicited forelimb responses. When a forelimb movement was observed, the current threshold was recorded. The type of forelimb movement elicited (elbow/shoulder, digit/wrist) was not quantitated in this study. The net surface area of the forelimb motor cortex was calculated for each experimental group.

Evoked compound muscle action potential amplitudes and latencies of peripheral nerves were quantitated as previously described (25). Briefly, sciatic nerves were stimulated at the sciatic notch or ankle (tibial nerve) with unipolar pin electrodes using submaximal stimuli (7 mA, 0.05-ms duration) at low frequency (0.1 Hz) for the indirect measurement of sensory nerve (H-reflex) responses or supramaximal stimuli (25 mA, 0.05-ms duration) at high frequency (1 Hz) for the direct measurement of motor nerve (M-wave) responses. Surface ring and pin recording electrodes were used to record the evoked potentials (H-reflex or M-wave, respectively) from the plantar muscles of the foot. Individual responses were amplified and recorded using an Advantage electromyograph system. For each animal, 10 individual H-reflex responses were recorded and averaged. Evoked M-wave responses were recorded, in triplicate, as an average of 25 individual evoked responses. The sensory nerve conduction velocity was calculated as the distance from notch to ankle divided by the difference of the H-reflex latency. Motor nerve conduction velocity was calculated as the distance from notch to ankle divided by the difference of the M-wave latency. Body temperature was maintained at 37°C. At the conclusion of the nerve conduction studies, the animals were decapitated, and their brains were rapidly removed, snap frozen in liquid nitrogen, and stored at −70°C until used.

Changes in the content of insulin receptors or low-affinity p75 neurotrophic receptors expressed in the cerebral cortices of STZ-administered relative to vehicle-administered rats were quantitated by immunoblot. Brains were rapidly thawed and cerebral cortices were harvested, homogenized in ice-cold PBS supplemented with a commercial cocktail of protease inhibitors (Roche, Indianapolis, IN), and stored at −70°C until used. Thawed aliquots of cortical homogenates were diluted, and protein concentrations (∼10 mg/ml protein) were quantitated by the method of Lowry et al. (26) with BSA as the standard. Cortical proteins (30 μg/lane) were resolved by one-dimensional SDS-PAGE, transferred to a nitrocellulose membrane support, and immunostained overnight at 4°C, as previously described (27), using a 1:500 dilution of anti−p75 neurotrophic receptor polyclonal antibody (Promega, Madison, WI) or a 1:200 dilution of anti−insulin receptor (α-subunit specific) polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA). Immunostained blots were washed and incubated for 1 h at room temperature with a 1:10,000 dilution of horseradish peroxidase−conjugated anti-rabbit IgG secondary antibody (The Jackson Laboratories, West Grove, PA). Immunoblots were developed by enhanced chemiluminescence. Changes in arbitrary band density relative to vehicle-treated animals were quantitated from the same blot using Un-Scan-It software (Silk Scientific, Orem, UT).

Administration of STZ to SD rats (n = 14) produced a marked elevation of nonfasting blood glucose levels (>400 mg/dl) that was sustained throughout this 8-week study (Fig. 1). Subcutaneous implantation of insulin pellets at 4 weeks (n = 6) reversed the observed hyperglycemic effects of STZ and normalized blood glucose levels to near physiological concentrations, consistent with the experimental induction of insulin-dependent diabetes (28,29). By comparison, nondiabetic control rats (n = 14) receiving an equal volume of citrate buffer (100 mmol/l; pH 4.5) maintained nonfasting blood glucose levels near 100 mg/dl (Fig. 1A). The body weight of the nondiabetic rats steadily increased over the course of study. The STZ-induced diabetic rats, while consistently demonstrating sustained hyperglycemia, maintained their starting body weight (∼300 g) and failed to gain weight over the course of the study. In contrast, diabetic rats receiving insulin treatment experienced a rapid and steady weight gain that was nearly identical to that of the nondiabetic control rats at the time they were killed (Fig. 1B). The diminished capacity of the STZ-induced diabetic rat to gain body weight and the reversal of this with insulin treatment provided further evidence of the induction of clinically relevant diabetes in these animals. At no time throughout the course of this study did our STZ-induced diabetic rats lose body weight, suggesting that complications secondary to diabetes-induced nutritional deficiencies were negligible.

Microstimulation of the right sensorimotor cortex of ketamine-xylazine−anesthetized rats (Fig. 2A) elicited reproducible contralateral forelimb and hindlimb motor responses that were visually recorded by two independent observers. In a previous mapping study using nondiabetic Long-Evans hooded rats, Neafsey et al. (23) demonstrated the presence of at least two discontinuous cortical forelimb areas: a large caudal primary motor area beginning at the bregma and a small rostral supplementary area located near the frontal pole. In contrast, the topographical boundaries of the forelimb motor cortical areas in the nondiabetic SD rats were found to extend uniformly over an area 1.0–4.0 mm lateral from the midline (Fig. 2B) and from 4.0 mm rostral to 2.0 mm caudal of the bregma (Fig. 2C), suggesting possible strain-related differences in cortical forelimb topographical boundaries. The majority of cortical sites eliciting forelimb motor movements were located 3.0–0.0 mm rostral and 1.5–3.5 mm lateral to the bregma. Stimulation along the topographical areas bordering the forelimb motor cortex domain in these rats elicited various types of motor movements. Rostral borders produced jaw and tongue movements whereas medial limits caused vibrissae to move. The caudal borders showed tail, hindlimb, and toe representation.

After 8 weeks of untreated diabetes, affected rats showed a significant attenuation in the number of individual motor cortical sites eliciting forelimb motor movements (Figs. 2B and C). In addition, the net area of the motor cortex that represented the forelimb was also significantly reduced compared with nondiabetic control rats (Fig. 3). The overall topographical organization of the forelimb cortical region was reduced in diabetic rats. These findings demonstrated that an 8-week course of STZ-induced uncontrolled diabetes reduces but does not shift the motor cortical topographical map in the SD rat. Cortical neurons in STZ-induced diabetic rats that were capable of responding to stimuli exhibited near normal activation thresholds (Table 1). The mean activation thresholds for vehicle-treated (20.8 ± 1.5 μA; n = 13), STZ-induced diabetic (24.8 ± 2.0 μA; n = 14), and STZ-induced diabetic and insulin-treated (23.5 ± 1.4 μA; n = 6) rats were not significantly different. These findings suggest that uncontrolled hyperglycemia can significantly attenuate the activation of some but not all forelimb motor cortical neurons. Cerebral cortical forelimb motor neurons that were responsive in STZ-induced diabetic rats exhibited normal activation thresholds, suggesting that these neurons are relatively resistant to the effects of uncontrolled hyperglycemia.

Diabetic rats treated at 4 weeks with insulin implants exhibited cortical-evoked forelimb motor responses that were nearly identical to those of nondiabetic control rats (Figs. 2B and C). The net cortical area representing the rat forelimb was also spared or restored by insulin therapy (Fig. 3), suggesting that central insulin/euglycemia is important for the maintenance of motor cortical responsiveness.

Peripheral nerve disease is a chronic complication of human and experimental diabetes (30) that may confound observational or quantitative measurements of evoked motor responses. To determine the degree to which peripheral nerve disease influenced our measurements of evoked motor responses, we quantitated peripheral nerve function in vehicle- and STZ-administered animals. The conduction velocity of the sciatic nerves in STZ-induced diabetic rats in this study was found not to differ significantly from those of nondiabetic vehicle-treated control rats (43 ± 6 [n = 8] vs. 44 ± 6 m/s [n = 10]; P > 0.05). In addition, evoked compound muscle action potential amplitudes, quantitated as the ratio of proximal (sciatic notch) to distal (ankle) responses (27), were similarly unaffected by the relatively short-term (8-week) duration of diabetes (0.80 ± 0.04 [n = 8] vs. 0.89 ± 0.07 [n = 10]; P > 0.05) used in this study. From these data, we concluded that 8 weeks of uncontrolled diabetes was insufficient to produce statistically significant changes in peripheral nerve function. These findings do not, however, address the possibility of the presence of subclinical neuropathology. We suggest that the functional changes we observed in cerebral cortical motor responses of STZ-induced diabetic rats most likely occurred via a central mechanism that is independent of overt peripheral nerve disease.

This study demonstrated for the first time that experimentally induced insulin-dependent diabetes produces a marked attenuation of cerebral cortical function in the SD rat as measured by ICMS-evoked forelimb motor responses. Cortical-evoked forelimb motor deficits observed in diabetic rats were prevented by insulin therapy. Secondary complications of peripheral nerve disease were not evident during the course of this study, suggesting a direct effect of hyperglycemia, insulin deficiency, or both on corticospinal motor dysfunction. We propose that acute or subacute periods of hyperglycemia, deficiencies of central insulin, or a combination of both may contribute to measurable deficits of corticospinal motor function in the insulin-dependent diabetic patient.

The comorbidity and mortality attributed to diabetes and its associated complications are well established. Impairment of the CNS in diabetic patients, referred to in some studies as diabetic encephalopathy, is relatively unknown and as a result its pathogenesis is understudied. Subtle neuropsychological deficits observed in some children with type 1 diabetes have raised reasonable concerns about the effect this disease may have on cognitive development in school-age children (31,32). Mental and motor slowing have also been reported among both type 1 and type 2 adult diabetic subjects (33,34,35), and impaired central motor conduction, assessed by magnetic stimulation, in diabetic patients has been described (6,7). CNS motor systems appear uniquely sensitive to the effects of poorly controlled diabetes. Nonketotic hyperglycemia in primary diabetes has even been associated in rare cases with hemichorea-hemiballismus (8).

Appropriate glycemic control plays a significant role in minimizing diabetes-associated complications (36,37). As evidenced in this study, glycemic control is similarly effective at minimizing diabetes-related CNS deficits. The mechanism by which glycemic control protects against peripheral nervous system or CNS injury is unclear. Recurrent episodes of hypoglycemia due to insulin therapy can result in severe brain dysfunction (38), but a clear correlation has not been demonstrated among adult diabetic patients (39,40). Conversely, increasing experimental evidence (19,41) suggests that CNS deficits are more likely to occur after extended periods of hyperglycemia in affected patients (i.e., those with poorly controlled diabetes).

Chronic hyperglycemia is strongly implicated in the development of vascular complications of diabetes, including gradual damage to the CNS (42). A vascular cause for diabetes-impaired, cortical-evoked responses observed in this subacute (8-week) study, however, was considered remote, and thus diabetes-dependent effects on peripheral and central vascular morphology were not quantitated.

An alternative explanation for the attenuated cortical-evoked responses in our STZ-induced diabetic rats could have been the indirect effects of altered electrolyte balance. Although we did not measure osmolarity or electrolyte levels, sustained plasma glucose levels >500 mg/dl, as observed in this study, would be expected to produce notable osmotic diuresis, with possible changes in electrolyte homeostasis. Egleton et al. (43), however, reported that electrolyte levels within the plasma and CSF of hyperglycemic rats with STZ-induced diabetes of 4 weeks’ duration were similar to those of nondiabetic euglycemic controls. Although ion transporter dysfunction was seen in their study, Egleton et al. (43) reported that the net concentration of electrolytes in the rat CSF was clearly not affected during acute STZ-induced diabetes. Extended periods of uncontrolled hyperglycemia, however, may adversely affect CSF electrolyte levels to a degree that ultimately influences CNS function.

A direct effect of elevated blood glucose (possibly ketogenesis)/hypoinsulinemia on CNS function must also be considered. In this regard, changes in local glucose concentrations can alter the firing rate of some types of CNS neurons (i.e., hypothalamic). However, in one study, elevating local glucose concentrations from 5 to 30 mmol/l had no effect on long-term potentiation in healthy CA1 hippocampal neurons, suggesting that hyperglycemia alone may be insufficient to initiate cortical decline in the diabetic patient (44). Decreases in central insulin content or compensatory changes of cortical insulin (or other) neurotrophic receptor expression may contribute to CNS decline in the insulin-dependent diabetic patient. Differential regulation of insulin receptor expression in retina from diabetic patients (45) and STZ-induced diabetic rats (46) has been reported. Earlier studies by Pacold and Blackard (47), however, found no change in insulin receptor expression in brain homogenates from STZ-induced diabetic rats when compared with vehicle-treated controls. In our hands, insulin therapy was effective at preventing cerebral cortical−evoked forelimb motor response deficits, possibly because of changes in central insulin receptor expression or signaling. We found that the relative content of the insulin receptor protein expressed in cerebral cortical homogenates of STZ-induced diabetic rats was not, however, altered when compared with vehicle-treated animals and quantitated by Western immunoblot (data not shown). The cerebral cortical expression of low-affinity p75 neurotrophic receptors, recognized as a 75-kDa doublet on immunoblot, was similarly unaffected by STZ-induced diabetes (data not shown). These findings contrast with the well-established changes of neurotrophic factors implicated in the pathogenesis of diabetic neuropathy (48), a common peripheral complication of chronic diabetes. We suggest that cerebral cortical–evoked forelimb motor response deficits in diabetes occur through a mechanism(s) that does not involve changes in the expression of cerebral cortical insulin or p75 neurotrophic receptors. Altered expression of other growth factor receptors, including brain-derived neurotrophic factor, remains an intriguing possibility. Diabetes-induced deficits in cerebral corticospinal motor function involving altered postsynaptic receptor signaling (49) or neurotransmitter metabolism/receptor dysfunction (50) remain to be established.

In summary, we have reported that uncontrolled hyperglycemia/hypoinsulinemia secondary to STZ-induced diabetes produces a marked attenuation of cerebral cortical−evoked forelimb motor responses in SD rats. Deficits in evoked motor responses were preventable with insulin therapy. We propose that CNS deficits related to insulin-dependent diabetes, including impairment of motor responses, may be preventable with appropriate glycemic control.

This study was supported in part by grants from the Department of Veteran Affairs (Veterans Health Administration, Medical Research Service and Rehabilitation Research and Development Service; B3413R) and the Potts Foundation (Loyola University Chicago).