Brain, Vol. 122, No. 4, 757-768,
April 1999
© 1999 Oxford University Press
Neurophysiological changes in the central and peripheral nervous system of streptozotocin-diabetic rats
Course of development and effects of insulin treatment
1 Department of Medical Pharmacology, Rudolf Magnus Institute for Neurosciences, Utrecht University and Departments of 2 Neurology and 3 Internal Medicine, University Hospital, Utrecht, The Netherlands
Correspondence to:
Dr G. Biessels, Department of Neurology, University Hospital Utrecht, PO Box 85500, 3508 GA, Utrecht, The Netherlands
 |
Abstract |
|---|
 Top Abstract IntroductionMethodResultsDiscussionReferences |
|---|
Diabetes mellitus can affect both the peripheral and the central nervous system. However, central deficits are documented less well than peripheral deficits. We therefore compared the course of development of neurophysiological changes in the central and peripheral nervous systems in streptozotocin-diabetic rats. Sciatic nerve conduction velocities and auditory and visual evoked potentials were measured prior to diabetes induction, and then monthly after diabetes induction for 6 months. In addition, the effect of insulin treatment was examined. Treatment was initiated after a diabetes duration of 6 months and was continued for 3 months. During treatment, evoked potentials and nerve conduction were measured monthly. In a third experiment, conduction velocities in ascending and descending pathways of the spinal cord were examined after 3 and 6 months of diabetes. Impairments of sciatic nerve conduction velocities developed fully during the first 2–3 months of diabetes. In contrast, increased latencies of auditory and visual evoked potentials developed only after 3–4 months of diabetes, and progressed gradually thereafter. Insulin treatment, initiated 6 months after induction of diabetes, improved both nerve conduction velocities and evoked potential latencies. Conduction velocities in the spinal cord tended to be reduced after 3 months of diabetes and were significantly reduced after 6 months of diabetes. The present study demonstrates that in streptozotocin-diabetic rats the course of development of peripheral and central neurophysiological changes differs. Peripheral impairments develop within weeks after diabetes induction, whereas central impairments take months to develop. Insulin can reverse both peripheral and central neurophysiological alterations.
diabetes mellitus; evoked potentials; nerve conduction; spinal cord; insulin
ANOVA = analysis of variance; ANOVAR = analysis of variance for repeated measurements; BAEP = brainstem auditory evoked potential; MNCV = motor nerve conduction velocity; SNCV = sensory nerve conduction velocity; STZ = streptozotocin; VEP = visual evoked potential
|
Introduction |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Peripheral nervous system disorders are one of the more frequent long-term complications of diabetes mellitus. The clinical features, epidemiology and pathophysiology of peripheral diabetic neuropathy have been studied extensively (for reviews, see Thomas and Tomlinson, 1993
; Yagihashi, 1995). Diabetes also affects the CNS (for review, see Biessels et al., 1994). However, compared with the disorders of diabetes in the PNS, those in the CNS are less well characterized and the pathophysiology is largely unknown.
In diabetic patients, deficits have been reported in neuropsychological, neuroradiological and neurophysiological studies. Neuropsychological studies report deficits in cognitive functions, in particular learning and memory and complex information processing (Tun et al., 1990). Neuroradiological studies report modest cerebral atrophy and an increased occurrence of subcortical and brainstem lesions (Dejgaard et al., 1991; Araki et al., 1994). Neurophysiological studies of the CNS in diabetic patients have mostly involved measurements of evoked potential latencies. Increases in the latencies of evoked potentials of different modalities, including visual evoked potentials (VEPs), brainstem auditory evoked potentials (BAEPs) and somatosensory evoked potentials, have often been reported (for review, see Di Mario et al., 1995).
Neurophysiological alterations have also been described in animal models of diabetes, in particular in rats. In the PNS of diabetic rats the time course of neurophysiological changes is well established. Deficits in both motor and sensory nerve conduction velocity (MNCV and SNCV, respectively) can be detected within weeks after the onset of diabetes and increase up to 2–3 months after diabetes onset, remaining relatively stable thereafter (e.g. Moore et al., 1980; Cameron et al., 1986; Brismar et al., 1987; Kappelle et al., 1993). Studies of MNCV and SNCV in diabetic rats have made important contributions to the elucidation of the pathogenesis of the effects of diabetes on the PNS, as well as in the development of putative pharmacotherapy (for review, see Cameron and Cotter, 1994). Neurophysiological alterations have also been reported in the CNS of diabetic rats (e.g. Carsten et al., 1989; Rubini et al., 1992; Sima et al., 1992; Terada et al., 1993; Morano et al., 1996), but the course of development is incompletely documented. Differences in test parameters, age at diabetes induction and diabetes duration complicate the comparison of different studies (see Discussion and Table 3). This is an important limitation for experimental studies into the pathophysiology of the CNS effects of diabetes. Therefore, the aim of the present study was to evaluate the course of development of neurophysiological alterations in the CNS in diabetic rats and to compare this with the course of development of neurophysiological alterations in the PNS. Moreover, since a previous study in diabetic rats indicated that neurophysiological alterations in the CNS, in contrast to the PNS, cannot be reversed with insulin treatment (Morano et al., 1996), a second aim of the study was to evaluate the effect of insulin treatment on established neurophysiological alterations in the CNS and PNS of diabetic rats.
|
|
Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Animals
We chose to perform our experiments in young adult streptozotocin (STZ)-diabetic rats, the model that is generally used in neurophysiological studies of experimental diabetes. Although the occurrence of maturation effects is a potential disadvantage of the use of young adult rats (Kappelle et al., 1993
), in our opinion this disadvantage is outweighed by the importance of clear documentation of the neurophysiological characteristics of the model of choice of the vast majority of studies in this field. Male Wistar rats (starting weight ~350 g, aged 3 months; UWU-CPD, Harlan, Utrecht, The Netherlands) were housed on sawdust, maintained on a 12 h light : 12 h dark cycle and given food and water ad libitum. They were weighed weekly. Diabetes was induced by a single intravenous injection of STZ (Serva Feinbiochemica GMBH, Heidelberg, Germany) at a dose of 40 mg/kg body weight, dissolved in saline. Four days after the STZ injection, glucose was determined in blood samples obtained by pricking the tail, using a strip-operated blood glucose sensor (Companion2; Medisense, Birmingham, UK). Blood glucose levels were >15.0 mmol/l in all STZ-injected animals. All experiments were conducted according to the guidelines of the Utrecht University Committee for the welfare of experimental animals.
Experimental design
Experiment one: time-course study
Two groups of rats were used: a diabetic group (n = 11) and an age-matched, non-diabetic control group (n = 12). MNCV and SNCV in the sciatic nerve were used to monitor the effect of diabetes on the PNS. BAEPs and VEPs were used to monitor the effects on the CNS. MNCV, SNCV, BAEPs and VEPs were measured before induction of diabetes. After induction of diabetes these measurements were repeated every month for a period of 6 months.
Experiment two: insulin reversal study
To examine the effect of insulin on diabetes-induced changes in MNCV, SNCV, BAEPs and VEPs an additional experiment was performed in three groups of rats: an untreated diabetic group (n = 8), an insulin-treated diabetic group (n = 6) and a non-diabetic, age-matched control group (n = 8). Insulin treatment was initiated after 6 months of diabetes and continued for 3 months. Insulin was administered through subcutaneous implants at a dose of 2–4 U per day [Linplant; Møllegaard, Ejby, Denmark; release per implant, 2 i.u. insulin per day for >40 days (Kappelle et al., 1994; Stevens et al., 1994)]. MNCV, SNCV, BAEPs and VEPs were measured prior to the onset of the insulin treatment and repeated after 2 and 3 months.
Experiment three: conduction velocity in the spinal cord
The latencies of evoked potentials reflect the sum of the time for perception in sensory organs, axonal conduction in the peripheral and central components of the sensory pathways and the time for synaptic transmission. To examine specifically the effects of STZ-diabetes on axonal conduction in the CNS, conduction velocities in the ascending and descending pathways of the spinal cord were examined after 3 and 6 months of diabetes. After 3 months of diabetes, six diabetic rats and eight non-diabetic, age-matched controls were examined. After 6 months of diabetes, 11 diabetic rats and 12 non-diabetic, age-matched controls were examined (these same animals were used in Experiment 1).
Sciatic nerve electrophysiology
MNCV and SNCV were measured in the sciatic nerve according to the method described by De Koning and Gispen (1987). In short, the sciatic and tibial nerve were stimulated at the sciatic notch and ankle, respectively. The latencies of the responses of the musculature of the foot were measured. The MNCV and SNCV were calculated by dividing the distance between the two stimulation points by the differences in latencies of the M response and the H reflex after proximal and distal stimulation.
Evoked potentials
Placement of recording electrodes
One week prior to the baseline measurement, rats were anaesthetized with Hypnorm® (Janssen Pharmaceutica, Tilburg, Netherlands; containing fluanisone 10 mg/ml and fentanylcitrate 0.315; mg/ml; dose, 0.7 ml/kg intra- muscularly). Two stainless steel screws were implanted permanently into the skull, one over the left frontal region [coordinates, A 2.0, L 2.0 with bregma as reference point (Paxinos and Watson, 1986)] and one over the left occipital cortex (A –7.0, L 3.0) (adapted from H. J. Duckers, F. H. Lopes da Silva and W. H. Gispen, unpublished observations). Care was taken not to penetrate the dura. The animals were allowed to recover for 1 week.
Stimulation protocol BAEPs
For measurements of the evoked potentials, rats were slightly sedated with a low dose of Hypnorm (0.15 ml/kg subcutaneously) in order to prevent them from moving. The rat was placed in a soundproof, darkened room. For the recording of auditory evoked potentials a speaker was placed 30 cm above the head of the rat. Acoustic stimuli were presented as clicks (unfiltered square waves of 100 µs duration with constant polarity, applied at a frequency of 10 Hz). The threshold of the BAEP was determined, defined as the minimal sound pressure level to evoke a response of at least 0.5 µV at a latency of ~5 ms. BAEPs were then recorded at a sound pressure level of 60 dB above threshold.
Recording and analysis BAEPs
BAEPs were recorded from the posterior screw and referred to the anterior screw. An earth electrode was connected to the front paw. BAEPs were amplified, filtered (bandpass 216–3400 Hz) and stored in a computer. For analysis, 512 traces (sweep length 40 ms) were averaged. The latencies of waves I, III and V were determined (Fig. 3). Although some uncertainty remains as to the generators of these waves, wave I is generally assumed to be generated in the auditory nerve, wave III in the superior olivary complex and wave V in the lateral lemniscus or the inferior colliculus (Funai and Funasaka, 1983; Wada and Starr, 1983; Shaw, 1988). Hence, the latency of wave I and the interpeak latencies I–III and III–V were used as a measure of the function of the auditory nerve, the pontomedullary region and the rostral pontine and midbrain region, respectively.
|
Stimulation protocol VEPs
VEPs were evoked with flash stimuli (Mecablitz 40 MZ2 flashbulb; Metz Werke, Fürth, Germany; flash duration 70 µs, output per flash 3 J) delivered at an upward angle of 90°, 25 cm from the eyes at a frequency of 0.67 Hz. The ears of the animals were occluded.
Recording and analysis VEPs
VEPs were amplified, filtered (bandpass 1–586 Hz) and stored in a computer. For analysis, 128 traces (sweep length 450 ms) were averaged. Four waves could be identified, which were designated N1, P1, N2 and N3 (Fig. 4). The latencies of these waves were ~30, 38, 62 and 106 ms, respectively. These latencies and the general appearance of the VEP in the present study were similar to VEPs in the rat in previous studies (Sima et al., 1992; Schwarz and Block, 1994; Barth et al., 1995). Peaks N1 and P1 have been suggested to be generated in the primary and secondary visual cortex, whereas N2 and N3 may be generated in associative cortical areas beyond the classically defined visual cortex (Barth et al., 1995). In the present study, peaks P1 and N3 could be identified most reliably and therefore the latencies of these two peaks were used to monitor the effects of diabetes on the VEP.
|
In a pilot study, BAEPs and VEPs recorded in rats sedated with Hypnorm were compared with recordings of rats that were awake. Hypnorm injection did not affect the latency of peak I, nor did it affect the interpeak latencies I–III and III–V. The latencies of peak P1 and N3 in the VEP increased by 8 and 20%, respectively, in rats injected with Hypnorm.
Conduction velocity in the spinal cord
The procedure for the measurements of conduction velocity in the spinal cord was adapted from the method described by van de Meent et al. (1996). Rats were anaesthetized with Hypnorm (1.0 ml/kg intramuscularly), intubated and ventilated mechanically after neuromuscular blockade with suxamethonium chloride (1 mg/kg intravenously). Next, they were placed in a stereotact on a heating pad, and rectal temperature was maintained at 37.5 ± 0.5°C.
Placement of recording electrodes
Bipolar recording electrodes consisted of thin gold plates 1 mm in diameter and with an interpolar distance of 5 mm. A laminectomy was performed at the seventh cervical (C7) and the eleventh thoracic (T11) vertebra and the recording electrodes were placed in the epidural space beneath the vertebral arch of C6 and T10. The dura was left intact. An earth electrode was placed at the ear bars of the stereotact.
Stimulation protocol
For measurements of the conduction velocity in the descending spinal pathways, a stainless steel stimulation electrode, insulated except for the tip, was placed in the red nucleus on the left side, using the stereotact (Paxinos and Watson, 1986). A reference electrode was placed into the scalp next to the burr hole. Monophasic pulses of 100 µs duration were presented at a frequency of 5 Hz. First, the threshold stimulus intensity (the minimum intensity to evoke a response of at least 10 µV at T10) was determined. Next, recordings were made at a stimulus intensity of twice the threshold intensity. For measurements of conduction velocity in the ascending spinal pathways the left sciatic nerve was exposed at the thigh and a bipolar stainless steel stimulation electrode was placed around it. Monophasic pulses of 100 µs duration were presented at a frequency of 5 Hz. The threshold stimulus intensity (the minimum intensity to evoke a response of at least 2 µV at C6) was determined and further recordings were made at a stimulus intensity of twice the threshold intensity.
Recording and analysis
Evoked responses were recorded simultaneously at C6 and T10 and amplified, filtered (bandpass 1–3400 Hz) and stored in a computer. For analysis, 128 traces (sweep length 25 ms) were averaged. For measurements of the conduction velocity in the descending spinal pathways, the red nucleus was stimulated and the latency of the first negative peak at both C6 and T10 was measured (Fig. 1A); the first positive peak was discarded because in some rats it was not clearly separated from the stimulus artefact. Conduction velocity was calculated by dividing the distance between the recording sites by the difference between the peak latencies of C6 and T10. The general appearance of the evoked responses and the calculated conduction velocity (~75 m/s) was similar to previous studies in the rat (Fehlings et al., 1988; van de Meent et al., 1996). The early waves of the response are thought to be generated in the rubrospinal tracts (Zappulla et al., 1988; van de Meent et al., 1996), since these tracts present a major proportion of the extrapyramidal motor pathways of the rat and have a conduction velocity of at least 40–50 m/s (Kuypers, 1981). The response is not likely to be generated by the pyramidal tracts, since in rats these tracts consist of unmyelinated and small-diameter myelinated fibres, with an estimated conduction velocity of 8–18 m/s (Zappulla et al., 1988). For measurements of the conduction velocity in the ascending spinal pathways, the sciatic nerve was stimulated and the latencies of the first negative peak at both C6 and T10 were measured (Fig. 1B), and conduction velocity was calculated. The general appearance of the evoked responses and the calculated conduction velocity (~45 m/s) were similar to previous studies in the rat (Carsten et al., 1989). The response is thought to be generated in the dorsal column (Fehlings et al., 1988; Carsten et al., 1989).
|
Statistical analysis
Data are presented as mean ± SEM (standard error of the mean). Between-group differences in body weight, blood glucose and spinal cord conduction velocities were analysed with two-tailed t tests for independent samples in Experiments 1 and 3 and by one-way analysis of variance (ANOVA) with post hoc Duncan's multiple range tests in Experiment 2. Analysis of variance for repeated measurements (ANOVAR) was used to study interactions between group and time on MNCV, SNCV and BAEP and VEP latencies in Experiment 1. The difference at individual time-points was analysed with two-tailed t tests for independent samples with Bonferroni correction (
= 0.01). The effect of insulin treatment on MNCV, SNCV and BAEP and VEP latencies in Experiment 2 was assessed by an ANOVA at the onset and the end of insulin treatment. If the ANOVA was significant (P < 0.05), post hoc Duncan's multiple range test was used to examine differences between the individual groups.
|
Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Experiment 1: time-course study
Animals, sciatic nerve conduction velocity
Diabetic animals failed to gain weight during the 6 months of the experiment and at the end of the experiment they had significantly reduced body weights and increased blood glucose levels compared with control animals (Table 1
).
|
MNCV and SNCV in non-diabetic rats increased gradually during the first 2–3 months of the experiment (Fig. 2
). In diabetic rats, MNCV and SNCV failed to increase. After 3 months, a deficit in MNCV of 8–10 m/s and in SNCV of 10–12 m/s had developed relative to control animals. These deficits remained stable during the final 3 months of the experiment. The interaction between group and time was significant for both MNCV and SNCV (ANOVAR: MNCV, P < 0.001; SNCV, P < 0.001).
|
Evoked potentials
BAEP (Fig. 3
)The minimal sound pressure level to evoke a detectable response was similar in diabetic and control rats and did not change during the course of the study (data not shown). Peaks I, III and V of the BAEP could be identified in all rats.
The latency of peak I was relatively stable in non-diabetic rats (Fig. 3). In contrast, in diabetic rats it appeared to decrease during the first month of the experiment. After 2 months of diabetes, the latency of peak I was significantly lower than in controls (P < 0.01). After 6 months of diabetes, the latency of peak I was similar in the two groups.
The interpeak latency I–III decreased gradually in control rats during the first 4 months of the study (Fig. 3). In diabetic rats this decrease was greater and after 1 month of diabetes the interpeak latency I–III was significant lower than in controls (P < 0. 005). After 2 months of diabetes the interpeak latency I–III started to increase gradually and at 5 and 6 months after diabetes induction it tended to be higher than in control rats (not significant).
The interpeak latency III–V was relatively stable in non-diabetic rats (Fig. 3). It was similar in control and diabetic rats during the first months of the experiment. After 2–3 months it started to increase in diabetic rats, and was statistically significant after 4 months of diabetes.
VEP (Fig. 4)
Peak P1 could be identified in all but one control rat and two diabetic rats. Peak N3 was identified in all but two control and two diabetic rats. In non-diabetic rats the latencies of peaks P1 and N3 were stable in time at a latency of ~37 and 105 ms, respectively. The latencies of peaks P1 and N3 were similar in control and diabetic rats during the first months of the experiment. After 3 months the latencies of both peaks started to increase in diabetic rats. For both peaks the increase in latency was statistically significant after 4 months of diabetes.
The eyes of all diabetic rats were carefully checked for the development of cataract. Cataracts were detectable after 4 months in two rats, after 5 months in four rats and after 6 months in five rats. The presence of cataracts in diabetic rats did not have an apparent effect on the latencies of peaks P1 and N3. Latencies were similar in diabetic rats with and without cataracts, and exclusion of all rats with cataracts did not affect the mean latencies of the diabetic group.
Experiment 2: effects of insulin reversal treatment
Animals, sciatic nerve conduction velocity
When insulin treatment was initiated, blood glucose levels dropped and body weight rose steadily. The blood glucose level in the insulin-treated animals averaged 8. 7 ± 0. 9 mmol/l in repeated measurements (Table 1). Prior to the onset of insulin treatment, SNCV and MNCV were markedly reduced in the diabetic rats compared with controls (Fig. 5). Both MNCV and SNCV in the non-treated diabetic rats were significantly impaired compared with non-diabetic controls at 6 and at 9 months. After 3 months of treatment both MNCV and SNCV had increased significantly in insulin-treated compared with untreated diabetic rats. The final values of MNCV and SNCV in the insulin-treated animals approached those of controls.
|
Evoked potentials (Fig. 6
)Peaks I, III and V of the BAEP could be identified in all rats. At the onset of Experiment 2, after a diabetes duration of 6 months, the latency of peak I and the interpeak latency I–III were 1. 84 ± 0.03 and 1. 51 ± 0.02 ms, respectively, in control rats and 1. 87 ± 0.01 and 1. 62 ± 0.02 ms in diabetic rats. Insulin treatment had no significant effects on these latencies (data not shown). The BAEP III–V interpeak latency in the diabetic rats was significantly increased compared with non-diabetic controls at the onset of treatment. After 3 months of treatment the untreated diabetic group was still significantly impaired compared with controls, whereas the BAEP III–V interpeak latency in the insulin-treated diabetic rats approached the non-diabetic control value (Fig. 6
).
|
Peak P1 of the VEP could be identified in all but two untreated diabetic rats. Peak N3 was identified in all but one control and one untreated diabetic rat. At the onset of treatment the VEP P1 and N3 latencies in both groups of diabetic rats were significantly increased compared with non-diabetic controls. After 3 months of treatment VEP P1 and N3 latencies were still impaired in the untreated diabetic group, whereas the latencies in the insulin-treated animals approached those of controls (Fig. 6
). The eyes of all diabetic rats were checked for the development of cataract. At the onset of treatment cataracts were present in one untreated rat and three insulin-treated diabetic rats. At the end of the experiment cataracts had developed in an additional two untreated diabetic rats. The presence of cataracts did not have an apparent effect on the latencies of peaks P1 and N3.
Experiment 3: conduction velocities in the spinal cord (Table 2)
Conduction velocity in the spinal cord was measured after 3 and 6 months of diabetes in separate groups of animals. After 3 months of diabetes, conduction velocities in both the descending and the ascending spinal pathways were reduced compared with controls, but not significantly. After 6 months of diabetes, there was a significant reduction in conduction velocities in both the descending (P < 0.01) and the ascending (P < 0.05) pathways in diabetic rats compared with controls (Table 2).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
In this longitudinal study, we show that the course of development of neurophysiological alterations differs between the PNS and CNS in diabetic rats. In the PNS, impairments of MNCV and SNCV relative to control values developed during the first 2–3 months of diabetes and remained relatively stable thereafter. In contrast, neurophysiological impairments in the CNS were not observed during the first months of diabetes. Impairments of the BAEP III–V interpeak latencies and VEP P1 and N3 latencies became detectable after 3–4 months of diabetes, and showed gradual progression thereafter. Likewise, conduction velocities in descending and ascending spinal pathways tended to be decreased after 3 months of diabetes, and were significantly decreased after 6 months of diabetes. Insulin treatment, initiated 6 months after induction of diabetes, significantly improved MNCV, SNCV and the VEP N3 latency and tended to improve the BAEP III–V interpeak and VEP P1 latency.
Course of development of peripheral and central neurophysiological impairments
The time-course of impairments in sciatic nerve conduction velocities corresponds with previous studies in STZ-diabetic rats (e.g. Moore et al., 1980; Brismar et al., 1987; Kappelle et al., 1993). The increase in MNCV and SNCV in the non-diabetic control rats (Fig. 2) is commonly observed in longitudinal studies in young adult rats and is related to maturation of the nerve (Moore et al., 1980; Kappelle et al., 1993). In diabetic rats MNCV and SNCV remained stable throughout the experiment, leading to a conduction deficit relative to controls of ~10 m/s after 3 months of diabetes. It should be noted that the deficits in MNCV and SNCV in the diabetic rats do not reflect just an arrest of maturation: if diabetes is induced in fully grown rats, deficits in MNCV and SNCV develop in a manner similar to those in young adult rats, although the rate of development may be slower (Cameron et al., 1986; Kappelle et al., 1993; Wright and Nukada, 1994). These latter findings emphasize that the conduction deficits that are observed in young adult diabetic rats are at least partially due to diabetes-induced pathology, the possible mechanisms of which will be discussed below.
The interpeak latency III–V was stable in time in non-diabetic rats. In diabetic rats, during the first 2 months after diabetes induction no changes were observed in the interpeak latency III–V. Thereafter, the latency increased gradually compared with controls. A similar time-course was observed for changes in the VEP P1 and N3 latencies. These findings correspond with previous studies of VEPs in diabetic rats, which report no increases in latencies after 2 months of diabetes (Apaydin et al., 1993), but a significant increase in latencies from 4 months to 8 months after diabetes onset (Sima et al., 1992; Morano et al., 1996) (Table 3). In diabetic rats, the changes in VEP latencies appeared to be unrelated to the development of cataract, since increases in the VEP latencies had usually developed after 4 months of diabetes, whereas cataracts were detected in a significant proportion of the diabetic rats only after 5 months of diabetes. Moreover, in individual rats the development of cataracts was not associated with a shift in peak latencies. These findings are in line with the notion that the flash VEP is relatively insensitive to the effect of even severe opacification of the lens (Halliday and Kriss, 1993). Although rats were not routinely checked for retinopathy, retinopathy is unlikely to explain the latency shift in the diabetic rat, since in this animal model the development of significant retinopathy may take more than 1 year (Engerman and Kern, 1995) and since the effect of even severe retinal dysfunction on flash VEP latencies appears to be limited (Halliday and Kriss, 1993).
The study of the time-course of the early components of the BAEP provided unexpected results. In diabetic rats, the latency of wave I and the interpeak latency I–III tended to decrease rather than increase compared with controls. Since the latencies of the BAEP are determined by the condition of the components of the auditory pathway, as well as the length of the pathway, the differences between control and diabetic rats may reflect changes at various levels. It is possible that maturation and growth effects are involved in the difference between the two groups. The distance between the left and right external acoustic meatus of the control rats increased by ~15% during the first 3 months of the experiment (data not shown), whereas it remained unchanged in diabetic rats. Although this distance provides an indirect measure, this may indicate that the length of the auditory pathways increased in control animals relative to diabetics, possibly explaining the difference in latencies between the two groups. This hypothesis has been tested in a separate experiment, using fully grown rats aged 8 months at the time of induction of diabetes, thus excluding growth effects as a possible confounder. During the first 2 months of this experiment no differences between diabetic and control rats were observed in the latency of wave I and the interpeak latency I–III (unpublished observations). These growth-related observations may explain part of the heterogeneity of the findings in previous studies on the effects of experimental diabetes on the early components of the BAEP, which generally used rats that were younger than the rats used in the present experiment (Table 3).
It is concluded that evoked potentials can be used to examine the effects of experimental diabetes on the CNS. There are, however, certain limitations. Since experimental diabetes invariably leads to growth arrest, the latencies of evoked potentials may provide unreliable results in pathways that still increase in length in young adult rats. Therefore, studies of, for example, the peripheral components of the BAEP should be performed in fully grown rats. Pathways that are located more centrally, like the central components of the BAEP and the structures that generate the VEP, are affected much less by growth effects in young adult rats, since their growth is completed at a relative young age.
To assess directly the effects of diabetes on axonal conduction in the CNS, conduction velocities were measured in the ascending and descending pathways of the spinal cord. Conduction velocities tended to be reduced after 3 months of diabetes and were significantly reduced after 6 months of diabetes. These findings are largely in line with a previous study on dorsal column function in diabetic rats, which showed no conduction abnormalities 1 month after diabetes induction but a modest reduction 3 months after diabetes induction [Terada et al. (1993); but see Carsten et al. (1989) in Table 3].
Effects of insulin treatment
In Experiment 2, it was shown that, even after 6 months of untreated hyperglycaemia, insulin treatment that led to near-normalization of blood glucose levels restored sciatic nerve conduction velocities towards control levels. MNCV and SNCV improved gradually during the 3-month treatment period. This gradual improvement corresponds with previous observations in diabetic rats (Brismar et al., 1987): after 6 months of uncontrolled hyperglycaemia, 3 weeks of vigorous insulin therapy restored MNCV in spontaneously diabetic BB rats by 40%.
Insulin treatment restored the interpeak latency III–V of the BAEP, as well as the latencies of peaks P1 and N3 of the VEP towards control values. However, although the evoked potential latencies of insulin-treated rats were not significantly different from those of controls after 3 months of treatment, the improvement compared with untreated diabetic rats was statistically significant only for the VEP N3 latency (Fig. 6). To our knowledge only one other study on the effects of restoration of normoglycaemia on evoked potentials in chronically hyperglycaemic rats has been published (Morano et al., 1996). In that study, islet transplantation was performed 4 months after onset of diabetes, in order to achieve near-normal glycaemia. Transplanted diabetic rats showed a further increase, rather than a decrease, in the VEP P1 latency compared with untreated diabetic rats (Morano et al., 1996). Whether this different outcome is related to the different method of achieving normoglycaemia remains to be determined. Since both the study by Morano et al. (1996) and our study used a limited number of rats in each experimental group, studies using larger numbers of animals are needed to examine further the potential reversibility of evoked potential abnormalities in diabetic rats in more detail.
Underlying mechanisms
The pathogenetic mechanisms underlying the neurophysiological alterations in the PNS in diabetic rats have been studied extensively (for reviews, see Cameron and Cotter, 1994; Arezzo, 1997). Metabolic and vascular processes have been proposed to change the microenvironment of the nerve, leading to early reductions in nerve conduction velocities (Greene and Lattimer, 1983; Tuck et al., 1984). Further slowing of nerve conduction velocity may be due to structural changes, which can be detected as early as 1 month after diabetes onset and are slowly progressive (Yagihashi, 1995).
Less is known about the underlying mechanisms of alterations in the CNS in diabetic rats (for review, see Biessels et al., 1994). Cerebral metabolic (e.g. Knudsen et al., 1989; Kumar and Menon, 1993) and vascular (e.g. Duckrow et al., 1987; Jakobsen et al., 1990) disturbances have been demonstrated within weeks after diabetes induction. However, the severity of these disturbances appears to be limited compared with the PNS (Biessels et al., 1994), possibly leading to a less hostile neuronal microenvironment. This could explain an important part of the difference in the course of development of functional changes in the CNS and PNS in diabetic rats.
Conclusion
The course of development of neurophysiological changes in diabetic rats differs between the PNS and CNS. In the PNS, neurophysiological changes can be detected within weeks after diabetes induction. In the CNS, changes in the central auditory, visual and spinal pathways develop concomitantly from 3 months after diabetes induction onwards. The present observations have important implications for experimental studies into the pathogenesis of the effects of diabetes on the CNS. Studies that aim to link putative pathogenetic factors to neurophysiological deficits and studies that assess the functional benefits of pharmacological interventions aimed at correcting possible pathogenetic factors should use rats with a diabetes duration of at least several months.
|
Acknowledgments |
|---|
We wish to thank Dr Henk van de Meent for his assistance in establishing the method for the measurement of conduction velocity in the spinal cord, Dr Paul Westers of the Utrecht University Centre of Biostatistics for advice concerning analysis of the data and Dr Hessel Franssen of the Department of Clinical Neurophysiology for advice on the evoked potential measurements
|
References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Apaydin C, Oguz Y, Agar A, Yargicoglu P, Demir N, Aksu G. Visual evoked potentials and optic nerve histopathology in normal and diabetic rats and effect of ginkgo biloba extract. Acta Ophthalmol (Copenh) 1993; 71: 623–8.[Medline]
Araki Y, Nomura M, Tanaka H, Yamamoto H, Yamamoto T, Tsukaguchi I, et al. MRI of the brain in diabetes mellitus. Neuroradiology 1994; 36: 101–3.[Web of Science][Medline]
Arezzo JC. The use of electrophysiology for the assessment of diabetic neuropathy. Neurosci Res Commun 1997; 21: 13–23.
Barth DS, Goldberg N, Brett B, Di S. The spatiotemporal organization of auditory, visual, and auditory-visual evoked potentials in rat cortex. Brain Res 1995; 678: 177–90.[Web of Science][Medline]
Biessels GJ, Kappelle AC, Bravenboer B, Erkelens DW, Gispen WH. Cerebral function in diabetes mellitus [see comments]. [Review]. Diabetologia 1994; 37: 643–50. Comment in: Diabetologia 1995; 38: 251–2, Comment in: Diabetologia 1995; 38: 1004–5.[Web of Science][Medline]
Brismar T, Sima AA, Greene DA. Reversible and irreversible nodal dysfunction in diabetic neuropathy. Ann Neurol 1987; 21: 504–7.[Web of Science][Medline]
Buller N, Laurian N, Shvili I, Laurian L. Delayed brainstem auditory evoked responses in experimental diabetes mellitus. J Laryngol Otol 1986; 100: 883–91.[Web of Science][Medline]
Cameron NE, Cotter MA. The relationship of vascular changes to metabolic factors in diabetes mellitus and their role in the development of peripheral nerve complications. [Review]. Diabetes Metab Rev 1994; 10: 189–224.[Web of Science][Medline]
Cameron NE, Cotter MA, Harrison J. Effect of diabetes on motor conduction velocity in different branches of the rat sciatic nerve. Exp Neurol 1986; 92: 757–61.[Web of Science][Medline]
Carsten RE, Whalen LR, Ishii DN. Impairment of spinal cord conduction velocity in diabetic rats. Diabetes 1989; 38: 730–6.[Abstract]
De Koning P, Gispen WH. Org.2766 improves functional and electrophysiological aspects of regenerating sciatic nerve in the rat. Peptides 1987; 8: 415–22.[Web of Science][Medline]
Dejgaard A, Gade A, Larsson H, Balle V, Parving A, Parving HH. Evidence for diabetic encephalopathy. Diabet Med 1991; 8: 162–7.[Web of Science][Medline]
Di Mario U, Morano S, Valle E, Pozzessere G. Electrophysiological alterations of the central nervous system in diabetes mellitus. [Review]. Diabetes Metab Rev 1995; 11: 259–77.[Web of Science][Medline]
Duckrow RB, Beard DC, Brennan RW. Regional cerebral blood flow decreases during chronic and acute hyperglycemia. Stroke 1987; 18: 52–8.
Engerman RL, Kern TS. Retinopathy in animal models of diabetes. [Review]. Diabetes Metab Rev 1995; 11: 109–20.[Web of Science][Medline]
Fehlings MG, Tator CH, Linden RD, Piper IR. Motor and somatosensory evoked potentials recorded from the rat. Electroencephalogr Clin Neurophysiol 1988; 69: 65–78.[Web of Science][Medline]
Funai H, Funasaka S. Experimental study on the effect of inferior colliculus lesions upon auditory brain stem response. Audiology 1983; 22: 9–19.[Web of Science][Medline]
Greene DA, Lattimer SA. Impaired rat sciatic nerve sodium-potassium adenosine triphosphatase in acute streptozocin diabetes and its correction by dietary myo-inositol supplementation. J Clin Invest 1983; 72: 1058–63.
Halliday AM, Kriss A. The visual evoked potential and electroretinogram in the investigation of diseases of the eye. In: Halliday AM, editor. Evoked potentials in clinical testing. 2nd ed. Edinburgh: Churchill Livingstone; 1993. p. 141–93.
Jakobsen J, Nedergaard M, Aarslew-Jensen M, Diemer NH. Regional brain glucose metabolism and blood flow in streptozotocin-induced diabetic rats. Diabetes 1990; 39: 437–40.[Abstract]
Kappelle AC, Bravenboer B, van Buren T, Traber J, Erkelens DW, Gispen WH. Amelioration by the Ca2+ antagonist, nimodipine, of an existing neuropathy in the streptozotocin-induced, diabetic rat. Br J Pharmacol 1993; 108: 780–5.[Web of Science][Medline]
Kappelle AC, Biessels G, Bravenboer B, van Buren T, Traber J, de Wildt DJ, et al. Beneficial effect of the Ca2+ antagonist, nimodipine, on existing diabetic neuropathy in the BB/Wor rat. Br J Pharmacol 1994; 111: 887–93.[Web of Science][Medline]
Knudsen GM, Jakobsen J, Barry DI, Compton AM, Tomlinson DR. Myo-inositol normalizes decreased sodium permeability of the blood–brain barrier in streptozotocin diabetes. Neuroscience 1989; 29: 773–7.[Web of Science][Medline]
Kumar JS, Menon VP. Effect of diabetes on levels of lipid peroxides and glycolipids in rat brain. Metabolism 1993; 42: 1435–39.[Web of Science][Medline]
Kuypers HGJM. Anatomy of the descending pathways. In: Brookhart TM, Mountcastle VB, editors. Handbook of physiology, Sect. I; Vol. 2, Pt 1. Bethesda (MD): American Physiological Society; 1981. p. 597–666.
Moore SA, Peterson RG, Felten DL, O'Connor BL. A quantitative comparison of motor and sensory conduction velocities in short- and long-term streptozotocin- and alloxan-diabetic rats. J Neurol Sci 1980; 48: 133–52.[Web of Science][Medline]
Morano S, Sensi M, Di Gregorio S, Pozzessere G, Petrucci AF, Valle E, et al. Peripheral, but not central, nervous system abnormalities are reversed by pancreatic islet transplantation in diabetic Lewis rats. Eur J Neurosci 1996; 8: 1117–23.[Web of Science][Medline]
Notvest RR, Inserra JJ. Tolrestat, an aldose reductase inhibitor, prevents nerve dysfunction in conscious diabetic rats. Diabetes 1987; 36: 500–4.[Abstract]
Paxinos G, Watson C. The rat brain in stereotaxic coordinates. San Diego (CA): Academic Press; 1986.
Rubini R, Biasiolo F, Fogarolo F, Magnavita V, Martini A, Fiori MG. Brainstem auditory evoked potentials in rats with streptozotocin-induced diabetes. Diabetes Res Clin Pract 1992; 16: 19–25.[Web of Science][Medline]
Schwarz M, Block F. Visual and somatosensory evoked potentials are mediated by excitatory amino acid receptors in the thalamus. Electroencephalogr Clin Neurophysiol 1994; 91: 392–8.[Web of Science][Medline]
Shaw NA. The auditory evoked potential in the rat—a review. [Review]. Prog Neurobiol 1988; 31: 19–45.[Web of Science][Medline]
Sima AA, Zhang WX, Cherian PV, Chakrabarti S. Impaired visual evoked potential and primary axonopathy of the optic nerve in the diabetic BB/W-rat. Diabetologia 1992; 35: 602–7.[Web of Science][Medline]
Stevens EJ, Carrington AL, Tomlinson DR. Nerve ischaemia in diabetic rats: time-course of development, effect of insulin treatment plus comparison of streptozotocin and BB models. Diabetologia 1994; 37: 43–8.[Web of Science][Medline]
Terada M, Yasuda H, Kikkawa R, Koyama N, Yokota T, Shigeta Y. Electrophysiological study of dorsal column function in streptozotocin-induced diabetic rats: comparison with 2,5-hexanedione intoxication. J Neurol Sci 1993; 115: 58–66.[Web of Science][Medline]
Thomas PK, Tomlinson DR. Diabetic and hypoglycemic neuropathy. In: Dyck PJ, Thomas PK, Griffin JW, Low PA, Poduslo JF, editors. Peripheral neuropathy. 3rd ed. Philadelphia: W.B. Saunders; 1993. p. 1219–50.
Tuck RR, Schmelzer JD, Low PA. Endoneurial blood flow and oxygen tension in the sciatic nerves of rats with experimental diabetic neuropathy. Brain 1984; 107: 935–50.
Tun PA, Nathan DM, Perlmuter LC. Cognitive and affective disorders in elderly diabetics. [Review]. Clin Geriatr Med 1990; 6: 731–46.[Medline]
van de Meent H, Hamers FP, Lankhorst AJ, Buise MP, Joosten EA, Gispen WH. New assessment techniques for evaluation of post-traumatic spinal cord function in the rat. J Neurotrauma 1996; 13: 741–54.[Web of Science][Medline]
Wada SI, Starr A. Generation of auditory brain stem responses (ABRs). III. Effects of lesions of the superior olive, lateral lemniscus and inferior colliculus on the ABR in guinea pig. Electroencephalogr Clin Neurophysiol 1983; 56: 352–66.[Web of Science][Medline]
Wright RA, Nukada H. Vascular and metabolic factors in the pathogenesis of experimental diabetic neuropathy in mature rats. Brain 1994; 117: 1395–1407.
Yagihashi S. Pathology and pathogenetic mechanisms of diabetic neuropathy. [Review]. Diabetes Metab Rev 1995; 11: 193–225.[Web of Science][Medline]
Yildiz O, Özata M, Özkardes A, Ciorakcii A, Gündogan N, Gündogan MA. L-carnitine treatment improves brain stem auditory evoked potentials in diabetic rats. Neuroreport 1996; 7: 2957–9.[Web of Science][Medline]
Zappulla RA, Hollis P, Ryder J, Moore FM, Adamson J, Moustakis W, et al. Noncortical origins of the spinal motor evoked potential in rats. Neurosurgery 1988; 22: 846–52.[Web of Science][Medline]
Received August, 1998. Revised November 8, 1998. Accepted November 11, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. M. Lechuga-Sancho, A. I. Arroba, L. M. Frago, C. Garcia-Caceres, A. D.-R. de Celix, J. Argente, and J. A. Chowen Reduction in the Number of Astrocytes and Their Projections Is Associated with Increased Synaptic Protein Density in the Hypothalamus of Poorly Controlled Diabetic Rats Endocrinology, November 1, 2006; 147(11): 5314 - 5324. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Price, C. Dent, B. Duran-Jimenez, Y. Liang, L. Zhang, E. J. Rebar, C. C. Case, P. D. Gregory, T. J. Martin, S. K. Spratt, et al. Gene Transfer of an Engineered Transcription Factor Promoting Expression of VEGF-A Protects Against Experimental Diabetic Neuropathy Diabetes, June 1, 2006; 55(6): 1847 - 1854. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. A.F. Sima and Z.-g. Li The Effect of C-Peptide on Cognitive Dysfunction and Hippocampal Apoptosis in Type 1 Diabetic Rats Diabetes, May 1, 2005; 54(5): 1497 - 1505. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Verrotti, T. Giuva, G. Morgese, and F. Chiarelli Topical Review : New Trends in the Etiopathogenesis of Diabetic Peripheral Neuropathy J Child Neurol, June 1, 2001; 16(6): 389 - 394. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||