OUP user menu

The regenerative deficit of peripheral nerves in experimental diabetes: its extent, timing and possible mechanisms

J. M. Kennedy , D. W. Zochodne
DOI: http://dx.doi.org/10.1093/brain/123.10.2118 2118-2129 First published online: 1 October 2000

Summary

Diabetes mellitus is reported to impair peripheral nerve regeneration, but the extent, timing and selectivity of the deficit is unclear. We studied regeneration of motor and sensory fibres in mice with experimental diabetes induced using streptozotocin (STZ). The mouse model featured several advantages over its counterpart in rats given STZ, while exhibiting the expected slowing of motor conduction velocity. Serial studies addressed fibre regrowth for up to 10 weeks after both sciatic nerve crush injury and complete sciatic nerve transection. Following nerve crush, there was a delay in motor fibre reinnervation of tibial innervated interosseous muscles of diabetics, manifest as a slow recovery of the M-wave recorded from these muscles. Despite an apparent recovery in M-waves by 6 weeks, this was not accounted for by restitution of tibial axon numbers in diabetic mice. Histological studies distal to crush or transection identified substantial delays in the regrowth of the numbers and calibre of regenerating myelinated fibres in diabetics for up to 8–10 weeks. Moreover, this delay was observed in both the tibial (largely motor) and sural (non-motor) distal sciatic branches. There was an associated delay in macrophage invasion and their later resorption in the diabetic nerves, indicating that a potential mechanism of impaired regeneration might be abnormal macrophage participation in nerve repair. Our findings indicate that during nerve regeneration, diabetic motor and sensory fibres have substantial and persistent deficits in regrowth associated with abnormalities in macrophage participation.

  • diabetes mellitus
  • neuropathy
  • regeneration
  • macrophage
  • peripheral nerve
  • ANOVA = analysis of variance
  • STZ = streptozotocin

Introduction

Regeneration of a peripheral nerve can be influenced by numerous factors. For instance, injuries that sever the nerve far from its target (Gordon and Fu, 1997) and an advanced age of an animal (Himes and Tessler, 1989) limit the regenerative potential. Alternatively, immediate surgical repair (Brown, 1972) and prior conditioning lesions (Oblinger and Lasek, 1984; Tetzlaff et al., 1996) enhance regrowth. Regeneration requires that extensive metabolic resources are available in order to bridge the gap between proximal and distal nerve stumps and to allow for maturation of the nerve fibres.

Diabetic patients are highly susceptible to acute nerve compression or entrapment injuries (Shahani and Spalding, 1969; Phalen, 1970; Fraser et al., 1979; Stevens et al., 1992; Chammas et al., 1995). Compression may result in demyelination and axonal degeneration (Dyck et al., 1990), and unfortunately there is a less favourable recovery in diabetics after decompression surgery (Haupt et al., 1993; Ward, 1997).

Regenerative success is noted to be inversely related to the duration of diabetes in rats (Ekstrom and Tomlinson, 1990). Bisby (Bisby, 1980) showed a delay in the onset of regeneration after crush injury that later also influences the elongation rate of axonal sprouts (Ekstrom et al., 1989). A delay in initial sprouting could lead to a lapse in nerve fibre maturation with a decrease in myelin thickness and axon diameter. Such a blunted maturation of fibres occurs as early as 14 days after injury in streptozotocin (STZ)-induced diabetic rats (Maxfield et al., 1995). Humans with diabetic sensory polyneuropathy also had decreased numbers of regenerating profiles a year following sural nerve biopsy, a deficit that was shown to be dependent on the severity of neuropathy (Sima et al., 1988; Bradley et al., 1995). An impaired neuronal regenerative programme could develop in diabetes as a result of a number of factors such as a failed upregulation of neurotrophins and their receptors (Ekstrom et al., 1989; Unger et al., 1998; Vo and Tomlinson, 1999) or neuropeptides (Calcutt et al., 1993) and impaired axonal transport or neuronal synthesis of critical cytoskeletal proteins such as neurofilaments and tubulin (Jakobsen et al., 1981; McQuarrie and Lasek, 1989; Mohiuddin and Tomlinson, 1997; McLean, 1997; Scott et al., 1999) .

STZ-rats typically have been examined in studies of diabetic nerve regeneration. Results obtained from such studies may be distorted by a prominent weight difference between diabetics and controls. In addition, it has been argued that STZ and subsequent hyperglycaemia may retard normal growth rates of the rats, resulting in deficient regeneration and a delayed conduction velocity unrelated to hyperglycaemia (Sharma et al., 1981; Dockery and Sharma, 1990). STZ-mice may serve to be a better model of diabetes since they have an earlier maturity with a negligible weight loss following diabetes initiation. Pancreatic lesions and progression of the disease in STZ-mice also mimic recent onset insulin-dependent diabetes mellitus (IDDM) in human patients (Gepts, 1965; Eisenbarth, 1986).

This study was undertaken to characterize the peripheral nerve regenerative response in an STZ-mouse model of diabetes. We examined diabetic axon regeneration following peripheral nerve transection or crush, two injuries with highly different regenerative success. Since diabetes preferentially may target sensory over motor nerves, both were examined in order to verify whether there is a difference in regeneration between them.

Methods

Animals

Procedures used in this study were approved by the Animal Care Committee of the University of Calgary and carried out in accordance with the `Guide to the Care and Use of Experimental Animals' issued by the Canadian Council on Animal Care. Adult male Swiss Wistar mice (20–30 g, n = 60) were used in this study (BioSciences, University of Calgary). Animals were housed two per plastic cage on a 12–12 h light–dark cycle with food and water available ad libitum. A number of animals were kept individually in cages to ensure that apparent autotomy injuries were self-inflicted and not due to conflicts with cage-mates. Mice were assigned randomly to either a diabetic or control group prior to the commencement of the experimental procedure. Diabetes was initiated by three consecutive injections of STZ [Zanosar (50 mg/ml); UpJohn, Don Mills, Ontario, Canada] in citrate buffer (pH 4.8) during the fasting state (one intraperitoneal injection/day; day 1, 85 mg/kg; day 2, 70 mg/kg; day 3, 55 mg/kg). Control animals received equivalent volume doses of the citrate buffer solution.

Hyperglycaemia was verified 1 week later by sampling from a tail vein. A fasting whole-blood glucose ≥ 16 mmol/l (normal 5–8 mmol/l) was our criterion for experimental diabetes (Zochodne et al., 1995). Whole-blood glucose tests were carried out using an Accuchek IIm (Boehringer Mannheim, Dorval, Quebec, Canada), while plasma glucose was checked with a glucose oxidase method (Ektachem DT-II Analyzer; Eastman Kodak Company, Rochester, NY, USA).

Surgery

Four-week diabetic and control mice were anaesthetized with sodium pentobarbital (65 mg/kg, i.p.). Left sciatic nerves were exposed by blunt dissection and either crushed or transected at mid-thigh level. Crush injury was delivered using a pair of watchmaker's forceps for 15 s while complete nerve transection was carried out using microscissors. A single suture (4–0 silk; Ethicon, Somerville, NJ, USA) in an adjacent muscle level served as a reference for the injury zone. Once injuries were created, muscle and skin were resutured in layers and the animals were allowed to recover.

Electrophysiology

Electrophysiological recordings were made under anaesthesia (Nicolet Viking I EMG machine; Nicolet, Madison, Wis., USA) as reported elsewhere in rats (Zochodne et al., 1992). Motor conduction in sciatic-tibial fibres was assessed by stimulating at the sciatic notch and knee while recording the M-wave (compound muscle action potential) from the tibial-innervated dorsal interossei foot muscles. Recordings were carried out at 0–10 weeks post-injury. All stimulating and recording used platinum subdermal needle electrodes (Grass Instruments; Astro-Med Inc., West Warwick, RI, USA) with near nerve temperature kept constant at 37°C using a subdermal thermistor connected to a temperature controller and heating lamp.

Morphometry

At 2, 4, 8 and 10 weeks post-injury, experimental animals were anaesthetized. Left and right sciatic and distal sural nerves were removed and processed for epon embedding (Zochodne et al., 1997). Animals subsequently were euthanized with an overdose of sodium pentobarbital. Briefly, samples were fixed in 2.5% glutaraldehyde in 0.025 M cacodylate buffer overnight, serially washed in 0.15 M cacodylate buffer, post-fixed in 2% osmium tetroxide in 0.12 M cacodylate, dehydrated using a series of graded alcohols and propylene oxide, and embedded in epon.

Transverse sections (1.0 μm thick) at fixed 10- and 15-mm distances distal to the site of injury were cut with an ultramicrotome (Reichert, Vienna, Austria) utilizing glass knives and were stained with toluidine blue. Morphological examination of specimens used a JAVA-based image analysis program (Jandel Scientific, San Rafael, Calif., USA) (Auer, 1994). Video images were obtained with a light microscope (Axioplan; Zeiss, North York, Ontario, Canada) and attached video camera (Sony AVC-D7; Toronto, Ontario, Canada) interfaced with a computer. The computer-assisted image analysis allowed for the determination of density, number and size frequency of intact myelinated fibres, axon diameter and area, and myelin thickness. All counting was performed with the microscopist blinded to the animal group. Degenerative axon profiles and macrophages were counted according to the criteria of Midroni and Bilbao (Midroni and Bilbao, 1995). Macrophages were identified as cells containing large amounts of amorphous debris (Midroni and Bilbao, 1995, pp. 61 and 62). Degradation of myelin into lipid and cholesterol esters often gave macrophages a foamy appearance under light microscopy (Midroni and Bilbao, 1995, p. 396).

Autotomy measurement

The autotomy measurement was based on previous scales developed to describe the self-inflicted loss of different parts of the foot or distal extremity following a sciatic nerve lesion. The scale used was a modification of those developed by Wall and colleagues (Wall et al., 1979) and Banõs and colleagues (Banõs et al., 1994). Mice with progressive autotomy were sacrificed by an overdose of pentobarbital.

Statistics

All data are presented as the mean ± standard error of the mean. Data were analysed by a one-way analysis of variance (ANOVA) with appropriate Bonferroni corrected post hoc Student's t-test comparisons. In all tests, statistical significance was set at α = 0.05.

Results

Animals and diabetes

Diabetic animals demonstrated features of hyperglycaemia including polydipsia, polyuria and a decrease in activity. Whole-blood glucose measurements verified significant hyperglycaemia beginning 1 week following the final STZ injection. Weights of diabetic and control mice were similar throughout the experimental period (Table 1).

View this table:
Table 1

Physical characteristics of mice

WeekNon-diabeticDiabetic
Weight (g)Glucose (mmol/l)Weight (g)Glucose (mmol/l)
Data are presented as mean ± SEM. Diabetics and non-diabetics were compared using a one-way ANOVA with Bonferroni post hoc tests. Week 1 glucose measurements are of whole blood while other time-points are comparing plasma glucose levels. All measurements were taken following an overnight fast. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001. (n = 60 total; 6/group.)
129.7 ± 0.48.8 ± 0.229.6 ± 0.318.8 ± 0.3***
635.2 ± 0.88.2 ± 1.333.0 ± 0.639.8 ± 4.2***
835.3 ± 1.18.9 ± 1.232.0 ± 0.6*27.4 ± 1.6***
1237.6 ± 1.99.4 ± 0.633.3 ± 1.3**35.1 ± 5.4***
1435.5 ± 0.67.2 ± 0.836.2 ± 1.341.1 ± 5.0***

Intact nerve

Electrophysiology

Baseline motor conduction velocity and M-wave amplitudes in the sciatic–tibial nerve territory at the beginning of the experiment were similar in diabetics and controls (Table 2). A significant slowing of conduction velocity in the uninjured sciatic-tibial nerve was evident in diabetic mice at 6 weeks of diabetes duration and persisted throughout the experimental period. M-wave amplitudes were not altered by diabetes.

View this table:
Table 2

Electrophysiological properties of the sciatic-tibial nerve following injury

Conduction velocityM-wave amplitude
Non-diabeticDiabeticNon-diabeticDiabetic
Data are presented as mean ± SEM. Non-diabetic controls and diabetics were compared at each time point using a one-way ANOVA with Bonferroni post hoc Student's t-tests. *P ≤ 0.05; **P ≤ 0.01. (n = 60 total; 6/group.)
Weeks post-crush
043.19 ± 0.5742.17 ± 0.813.41 ± 0.383.32 ± 0.30
2 0 000
312.57 ± 3.85 7.25 ± 4.190.65 ± 0.290.40 ± 0.23
423.67 ± 1.51 13.57 ± 1.76**1.02 ± 0.130.49 ± 0.11*
625.48 ± 1.6919.51 ± 0.62*1.66 ± 0.20 1.57 ± 0.31
827.75 ± 0.9021.41 ± 2.10*1.92 ± 0.291.93 ± 0.23
Weeks post-transection
040.78 ± 1.0241.41 ± 0.613.45 ± 0.413.53 ± 0.31
4 0 000
6 8.96 ± 3.06 3.17 ± 1.760.12 ± 0.050.08 ± 0.04
1025.38 ± 1.6620.12 ± 1.62*0.76 ± 0.190.73 ± 0.12

Morphometry

Myelinated fibre density and number in uninjured tibial fascicles were not influenced by diabetes at any time point. Fibre and axonal diameter and area, along with myelin thickness were also unchanged (Table 3). Elevated numbers and densities of endoneurial macrophages were, however, seen in diabetic tibial fascicles at later stages of diabetes (12 and 14 weeks; P < 0.05). This rise in macrophage number coincided with an increased number and density of degenerating axon profiles at the same time points (12 weeks P < 0.05; 14 weeks P < 0.01). However, since the actual number of degenerating profiles was quite small, the findings were not associated with a detected decline in numbers or density of intact myelinated fibres.

View this table:
Table 3

Morphological properties of uninjured tibial and sural fascicles at 12 weeks of diabetes duration

PropertyTibialSural 1Sural 2
Non-diabeticDiabeticNon-diabeticDiabeticNon-diabeticDiabetic
Sural 1 denotes a sampling site proximal to the sciatic trifurcation (same level as tibial) while sural 2 is 5 mm distal to the trifurcation. At least six animals per group were sampled. Diabetic and control animals were compared on each parameter using a one-way ANOVA with Bonferroni corrected post hoc Student's t-test comparisons (α = 0.05). Macro = macrophage; Deg. profile = degenerating profile. *P ≤ 0.05; **P ≤ 0.01.
Endo area (mm2) 0.093 ± 0.003 0.095 ± 0.007 0.012 ± 0.001 0.013 ± 0.002 0.017 ± 0.003 0.013 ± 0.003
Fibre no. 1626 ± 34 1659 ± 89 286 ± 13 290 ± 55 312 ± 10 220 ± 27*
Fibre density (no./mm2)17 790 ± 46017 680 ± 106723 750 ± 150923 130 ± 99418 322 ± 67016 860 ± 712*
Fibre diameter (μm) 9.2 ± 0.2 9.1 ± 0.2 7.8 ± 0.1 7.1 ± 0.2 8.0 ± 0.4 7.9 ± 0.3
Axon diameter (μm) 6.9 ± 0.2 6.8 ± 0.2 5.9 ± 0.2 5.6 ± 0.1 6.3 ± 0.6 6.0 ± 0.3
Fibre area (μm2) 71.1 ± 3.7 69.3 ± 2.9 51.1 ± 1.9 43.1 ± 0.9 54.5 ± 8.4 49.9 ± 3.9
Axon area (μm2) 42.5 ± 2.5 41.6 ± 1.9 29.1 ± 0.9 26.8 ± 4.5 31.2 ± 1.5 28.4 ± 0.8
Myelin (μm2) 1.10 ± 0.03 1.10 ± 0.04 1.00 ± 0.04 0.78 ± 0.03** 1.03 ± 0.03 0.80 ± 0.02**
Macro no. 0.3 ± 0.3 4.0 ± 1.4** 0 0 0 1.2 ± 0.2**
Macro density 3 ± 3 40 ± 11* 0 0 0 86 ± 27*
Deg. profile no. 3 ± 1 8 ± 2** 0 2 ± 1* 1 ± 0 3 ± 1**
Deg. profile density (no./mm2) 29 ± 8 87 ± 14* 0 103 ± 36* 34 ± 13* 252 ± 60**

Sural fascicles were studied at both the same level as the tibial fascicle (proximal to trifurcation) and also 5 mm distal to the trifurcation. Axonal calibre (diameter and area) in diabetes was similar to control values (Table 3). Myelin was thinner at both sample sites in diabetic mice at later stages of diabetes (12 and 14 weeks; P < 0.01) but overall fibre area and diameter had only a non-significant trend toward lower values in diabetics. There was a distal decrease in overall fibre density and number in diabetics (12 and 14 weeks; P < 0.05). Numbers and densities of degenerating axon profiles were also increased (12 and 14 weeks; P < 0.05) at both sural levels, with the most distal sampling point also having an elevated number and density of macrophages (Fig. 1).

Fig. 1

Uninjured sural nerves (1.0 μm thick) 5 mm distal to the sciatic trifurcation in non-diabetic and diabetic mice. (A) Non-diabetic sural nerve with no evidence of pathology. (A′ and B) Diabetic sural nerves demonstrating the presence of degenerating profile units (arrows) and occasional endoneurial macrophages (m). Magnification bar = 20 μm for A and A′, 16 μm for B.

Injured nerve

Electrophysiology

A mid-thigh crush of the sciatic nerve resulted in delays in the recovery of motor conduction velocity (Fig. 2) and of the sciatic-tibial territory M-wave in diabetic mice. These findings were prominent at 3–6 weeks following crush injury (P < 0.05), then recovered. Transection injury produced a slowed recovery of conduction velocity in the reinnervating fibres of diabetics, but M-wave recovery in both groups was significantly delayed (Table 2).

Fig. 2

Recovery of M-potentials following crush injury to the left sciatic nerve in diabetic and control mice. Note that diabetes was initiated 4 weeks prior to injury. Data are presented as a percentage of the baseline value ± standard error of the mean. Comparisons between diabetics (filled squares) and controls (open squares) were made using a one-way ANOVA with Bonferroni post hoc tests. *P ≤ 0.05.

Morphometry

Diabetic tibial fascicles did not show any apparent regenerative delay at 2 weeks following nerve injury (transection or crush) but did have significantly thicker myelin sheaths (Fig. 3). Fibre number and density were decreased in diabetics at 8 weeks (final time point) following crush and at 10 weeks (final time point) following transection (Fig. 4). Diabetic fibre and axon diameters were smaller at 4 and 8 weeks (P < 0.05) after crush and at 10 weeks following transection (P < 0.05). Fibre and axon areas were also reduced in diabetics (P < 0.05) following crush (4 and 8 weeks) and transection (10 weeks). Diabetic tibial fascicles had significantly thinner myelin sheaths at 4 and 8 weeks after crush (P < 0.05). Fewer macrophages were recruited into the diabetic tibial fascicle 2 weeks post-crush, but macrophage numbers and densities increased later (8 weeks post-crush and 10 weeks post-transection), as control values declined, and persisted during the attempted regenerative process (Fig. 5). The number and density of degenerating axon profiles were also increased in diabetic tibial fascicles at 4 and 8 weeks after crush (Table 4).

View this table:
Table 4

Morphological properties of injured tibial fascicles

PropertyWeeks post-crush injuryWeeks post-transection
24810
NDNDNDND
Six animals per group at each time point were used in the experiment. Data are presented as mean ± SEM. Diabetics (D) and non-diabetic controls (N) were compared on each parameter using a one-way ANOVA with Bonferroni post hoc Student's t-tests. Macro = macrophage; Deg. profile = degenerating profile. *P ≤ 0.05.
Endo area (mm2) 0.11 ± 0.01 0.10 ± 0.01 0.09 ± 0.01 0.07 ± 0.01 0.11 ± 0.01 0.11 ± 0.02 0.08 ± 0.01 0.10 ± 0.01
Fibre no. 987 ± 118 915 ± 156 1662 ± 87 1264 ± 230* 2021 ± 64 1535 ± 87* 1224 ± 96 1057 ± 67*
Fibre density (no./mm2) 8802 ± 567 9050 ± 108118 760 ± 119118 830 ± 240918 620 ± 91113 570 ± 1325*15 710 ± 124510 260 ± 970
Fibre diameter (μm) 4.50 ± 0.10 4.76 ± 0.18 5.86 ± 0.15 4.98 ± 0.10 6.78 ± 0.22 5.90 ± 0.08* 5.62 ± 0.33 4.89 ± 0.16*
Axon diameter (μm) 3.52 ± 0.09 3.58 ± 0.13 4.58 ± 0.11 3.90 ± 0.10* 5.40 ± 0.13 4.58 ± 0.13* 4.11 ± 0.21 3.56 ± 0.14*
Fibre area (μm2)14.92 ± 0.6917.54 ± 1.22 26.74 ± 1.58 18.96 ± 0.76* 36.90 ± 1.97 28.72 ± 0.69* 29.13 ± 7.64 21.33 ± 1.66*
Axon area (μm2) 9.48 ± 0.40 10.50 ± 0.6 17.42 ± 1.05 12.08 ± 0.57* 24.92 ± 1.67 18.54 ± 0.68* 19.18 ± 5.33 13.44 ± 0.97*
Myelin (μm) 0.48 ± 0.02 0.60 ± 0.03* 0.66 ± 0.02 0.53 ± 0.03* 0.72 ± 0.02 0.60 ± 0.03* 0.65 ± 0.05 0.59 ± 0.03
Macro no. 159 ± 22 99 ± 13* 32 ± 4 34 ± 3 19 ± 4 76 ± 18* 9 ± 3 51 ± 14*
Macro density (no./mm2) 1382 ± 152 962 ± 133* 376 ± 57 463 ± 99 180 ± 34 624 ± 164* 101 ± 35 465 ± 106*
Deg. profile no. 348 ± 28 318 ± 23 89 ± 14 150 ± 12* 49 ± 5 91 ± 11* 55 ± 21 87 ± 22
Deg. profile density (no./mm2) 3059 ± 245 3167 ± 394 996 ± 84 2142 ± 149* 501 ± 81 813 ± 145* 648 ± 264 812 ± 131
Fig. 3

Tibial fascicles (1.0 μm thick) 10 mm distal to nerve injury in non-diabetic (A, B, C and D) and diabetic mice (A′, B′, C′ and D′) at various time points after injury (A and A′, 2 weeks; B and B′, 4 weeks; C and C′, 8 weeks; D and D′, 10 weeks). The 2-, 4- and 8-week time points are following sciatic nerve crush while 10 weeks is post-transection. Note the reduced numbers of regenerating myelinated fibres and increased numbers of persistent degenerative axon profiles following both types of injury in diabetics. (Bar = 20 μm for all panels.)

Fig. 4

Fibre density of tibial fascicles following nerve injury and in relation to the contralateral tibial fascicle. Crush injury is represented at 2, 4 and 8 weeks, with transection injury demonstrated at the 10-week time point. Data are presented as mean ± standard error of the mean. Diabetics (black bars) and non-diabetics (cross-hatched bars) were compared using a one-way ANOVA with Bonferroni post hoc Student t-tests. Contra; contralateral tibial fascicle. *P ≤ 0.05.

Fig. 5

The number of persistent endoneurial macrophages in the tibial fascicle following ipsilateral sciatic nerve injury. Crush injury is demonstrated at 2, 4 and 8 weeks. Transection injury is specified at the 10-week time point. Data are presented as mean ± standard error of the mean. Diabetics (black bars) and non-diabetics (cross-hatched bars) were compared using a one-way ANOVA with Bonferroni post hoc Student's t-tests. Contra; contralateral tibial fascicle. *P ≤ 0.005, **P ≤ 0.01.

Diabetic sural nerves at both 10 and 15 mm below the lesion site demonstrated decreased fibre and axon diameters (4 and 8 weeks post-crush, P < 0.05; and 10 weeks post-transection, P < 0.05). Fibre and axon areas were similarly undersized in diabetic mice at both sample sites (8 weeks after crush and 10 weeks after transection; P < 0.05 and P < 0.01, respectively). Fibre number and density were compromised in diabetics following crush at 4 and 8 weeks, and after transection at 10 weeks (P < 0.05). Myelin was thicker in diabetic sural nerves immediately after crush (2 weeks), but regenerating fibres later fell below control levels (4 and 8 weeks; P < 0.05). Transection injury also yielded thinner myelin in diabetics after 10 weeks (P < 0.01). Macrophages were once again elevated in control animals at 2 weeks post-injury (transection and crush) but to a lesser extent in diabetic mice. The increase in the number of macrophages at 2 weeks coincided with an increased number and density of distal (15 mm) degenerating units in control animals compared with diabetics (P < 0.01). Macrophage number and density rose later in diabetics and tended to remain above control levels following crush (4 and 8 weeks; P < 0.05) and transection (10 weeks; P < 0.05). Degenerating unit density also increased at 4 and 8 weeks after crush (Tables 5 and 6).

View this table:
Table 5

Morphological properties of injured sural fascicles 10 mm distal to sciatic injury

PropertyWeeks post-crush injuryWeeks post-transfection
24810
NDNDNDND
Six animals per group were used at each time point. Data are presented as mean ± SEM. Diabetics (D) and non-diabetic controls (N) were compared on each parameter using a one-way ANOVA with Bonferroni post hoc Student's t tests. Macro = macrophage; Deg. profile = degenerating profile. *P ≤ 0.05; **P ≤ 0.01.
Endo area (mm2)0.018 ± 0.003 0.018 ± 0.001 0.019 ± 0.004 0.014 ± 0.001 0.020 ± 0.004 0.023 ± 0.004 0.015 ± 0.000 0.017 ± 0.003
Fibre no. 180 ± 30 207 ± 31 320 ± 27 175 ± 27* 366 ± 48 258 ± 43 231 ± 9 210 ± 11*
Fibre density (no./mm2) 9890 ± 183111 190 ± 115816 890 ± 146113 250 ± 1006*18 250 ± 116811 270 ± 689*15 180 ± 797 12 540 ± 1316
Fibre diameter (μm) 4.32 ± 0.22 4.52 ± 0.16 5.40 ± 0.21 4.73 ± 0.08* 5.84 ± 0.18 5.14 ± 0.27* 5.78 ± 0.10 4.64 ± 0.25*
Axon diameter (μm) 3.42 ± 0.21 3.40 ± 0.13 4.36 ± 0.14 3.75 ± 0.07* 4.86 ± 0.11 4.14 ± 0.25* 4.56 ± 0.15 3.53 ± 0.18*
Fibre area (μm2) 14.22 ± 1.60 15.18 ± 1.02 22.28 ± 2.11 24.98 ± 4.14 29.12 ± 1.35 21.40 ± 1.86* 27.38 ± 1.70 17.83 ± 1.95**
Axon area (μm2) 9.22 ± 1.23 9.02 ± 0.61 14.44 ± 1.45 15.90 ± 3.01 19.79 ± 1.45 13.96 ± 1.55* 18.06 ± 1.01 11.46 ± 1.30**
Myelin (μm) 0.46 ± 0.02 0.56 ± 0.02* 0.60 ± 0.03 0.50 ± 0.00* 0.58 ± 0.02 0.50 ± 0.00* 0.64 ± 0.04 0.51 ± 0.03*
Macro no. 21 ± 2 14 ± 3* 4 ± 1 10 ± 1* 3 ± 1 10 ± 2* 1 ± 0 8 ± 2*
Macro density (no./mm2) 1078 ± 134 764 ± 147* 247 ± 64 714 ± 96 159 ± 47 441 ± 102* 71 ± 20 471 ± 55*
Deg. profile no. 48 ± 4 45 ± 10 27 ± 6 25 ± 3 7 ± 2 12 ± 4 5 ± 3 7 ± 2
Deg. profile density (no./mm2) 2694 ± 251 2514 ± 550 1443 ± 263 1828 ± 364* 354 ± 81 532 ± 68* 307 ± 170 405 ± 99
View this table:
Table 6

Morphological properties of injured sural fascicles 15 mm distal to sciatic injury

PropertyWeeks post-crush injuryWeeks post-transection
24810
NDNDNDND
Six animals per group were used at each time point in the experiment. Data are presented as mean ± SEM. Diabetics (D) and non-diabetic controls (ND) were compared on each parameter using a one-way ANOVA with Bonferroni post hoc Student's t-tests. Macro, macrophage; Deg. profile, degenerating profile. P ≤ 0.05; **P ≤ 0.01.
Endo area (mm2)0.023 ± 0.0030.027 ± 0.004 0.022 ± 0.005 0.016 ± 0.006 0.023 ± 0.005 0.022 ± 0.001 0.020 ± 0.003 0.022 ± 0.003
Fibre no. 88 ± 21 139 ± 9* 280 ± 70 154 ± 43 353 ± 31 275 ± 39* 269 ± 14 230 ± 10*
Fibre density (no./mm) 3920 ± 461 5095 ± 74914 560 ± 46911 220 ± 537*15 400 ± 159812 410 ± 966*13 540 ± 44210 248 ± 996*
Fibre diameter (μm) 4.62 ± 0.23 4.78 ± 0.11 5.78 ± 0.48 4.83 ± 0.39* 6.05 ± 0.39 4.93 ± 0.24* 5.43 ± 0.09 4.59 ± 0.18*
Axon diameter (μm) 3.66 ± 0.21 3.58 ± 0.12 4.60 ± 0.44 3.76 ± 0.39* 4.90 ± 0.33 3.73 ± 0.14* 4.30 ± 0.04 3.48 ± 0.13*
Fibre area (μm2)15.38 ± 2.7517.58 ± 0.74 25.83 ± 4.95 25.90 ± 3.32 29.27 ± 3.33 20.20 ± 1.50* 23.33 ± 0.72 16.31 ± 1.39**
Axon area (μm2) 9.94 ± 2.0810.62 ± 0.67 17.68 ± 3.42 18.10 ± 2.75 19.62 ± 2.42 12.47 ± 0.75* 15.98 ± 0.50 10.11 ± 0.93**
Myelin (μm) 0.50 ± 0.0 0.60 ± 0.0* 0.63 ± 0.03 0.58 ± 0.03* 0.64 ± 0.04 0.56 ± 0.04* 0.60 ± 0.0 0.52 ± 0.02*
Macro no. 43 ± 6 27 ± 5* 6 ± 1 9 ± 2* 3 ± 1 6 ± 1* 3 ± 1 14 ± 3**
Macro density (no./mm2) 1881 ± 430 1096 ± 240* 262 ± 16 565 ± 155* 123 ± 23 261 ± 40* 158 ± 51 669 ± 102**
Deg. profile no. 76 ± 8 49 ± 11** 21 ± 2 8 ± 4 3 ± 1 7 ± 2* 8 ± 3 11 ± 2
Deg. profile density (no./mm2) 3580 ± 601 1922 ± 400** 1146 ± 310 1593 ± 580 122 ± 60 281 ± 88** 396 ± 262 521 ± 121

Autotomy

Autotomy scores in diabetic mice were higher and occurred earlier than in non-diabetics following injury. The scores, however, did vary among individual animals, and autotomy did not occur in all. Scores were zero for the contralateral paw.

Discussion

The major findings of the present work are: (i) STZ-mice have a slowed motor conduction velocity in uninjured sciatic-tibial nerves, resembling human diabetic neuropathy, but with relatively preserved nerve morphometry in tibial fascicles; (ii) uninjured sural nerves of diabetics exhibit decreased myelin thickness and a distal loss of fibres; (iii) there is a delay in the recovery of the sciatic-tibial M-wave, indicating slower target muscle reinnervation, and in motor conduction velocity in diabetic mice following nerve injury; (iv) nerve injury is associated with slowed regrowth and maturation of regenerating fibres in diabetic mice, observed as decreased fibre density and number as well as decreased fibre and axon calibre distal to the injury zone; (v) diabetes influences the infiltration and resorption of macrophages following nerve injury; and (vi) autotomy behaviour was elevated in diabetic mice.

It is important to note that STZ-mice developed severe hyperglycaemia without the weight loss characteristic of STZ-rats. Control and diabetic mice were of the same age, and diabetes was initiated after animals had reached maturity. The STZ-mouse model eliminates sources of variability that may have arisen in previous studies of STZ-rat models.

As in human diabetes, our STZ-mice had reduced motor conduction velocity in the sciatic-tibial nerve despite normal fibre calibre. STZ-rat models are also observed to develop significant slowing of motor and sensory conduction velocity early after the development of hyperglycaemia (Kalichman et al., 1998; Cameron et al., 1999) but with no evidence of fibre loss or degeneration (Sharma and Thomas, 1987). Conduction slowing without changes in fibre calibre may be accounted for by abnormalities in nerve polyol flux (Gabbay et al., 1966; Greene et al., 1987, 1992). Polyol accumulation produces decreased Na+,K+-ATPase activity, with intra-axonal sodium accumulation leading to slowed conduction of action potentials (Gould, 1976; Mandersloot et al., 1978). Paranodal and segmental demyelination that is seen in STZ and Bio Breeding/Wistar rat models theoretically may also contribute to conduction slowing in the mouse model (Sima and Sugimoto, 1999).

In our model, intact diabetic sural axons had a reduced myelin thickness and there was a distal loss of fibres at the end point of the study. Axonal atrophy and degeneration of both myelinated and unmyelinated fibres are often noted in nerve biopsies from diabetic patients, with the most severe fibre loss occurring in distal sensory nerves such as the sural (Dyck et al., 1986). Distal loss of fibres in sural nerve may support the concept that defects in growth factor synthesis, uptake or signalling contribute to diabetic neuropathy. The synthesis of growth factors such as nerve growth factor (NGF) (Faradji and Sotelo, 1990; Hellweg and Hartung, 1990) and insulin-like growth factor (IGF-1) (Ishii, 1995; Zhuang et al., 1997) may be decreased in target cells of diabetic sympathetic, sensory and motor (IGF-1 only) nerves. This growth factor deficiency may be compounded by slowed retrograde axonal transport limiting delivery of growth factors to proximal neurones (Jakobsen et al., 1981; Mayer and Tomlinson, 1983; Larsen and Sidenius, 1989). Diabetic dorsal root ganglion neurones may also be prone to oxidative stress-related damage, leading to loss through apoptosis (Greene et al., 1999).

As expected, regeneration was more robust, both electrophysiologically and histologically, following crush injury than after transection. This is due to the preservation of endoneurial tubes after crush injury, thereby providing a guide for regenerating axons. The apparent recovery of the M-wave by 6 weeks post-crush injury in diabetics despite incomplete restitution of tibial fibre numbers may indicate that robust collateral sprouting occurs at the interosseous motor endplates. Following denervation, nerve sprouts previously have been noted to appear at nerve terminals and nodes of Ranvier of remaining intramuscular nerves, forming accessory or double endings on muscle fibres (for a review, see Wernig and Herrera, 1986) in an attempt to compensate for loss of adjacent nerve fibres (Rafuse et al., 1992).

Diabetics had delayed fibre regrowth and maturation, characterized by decreased fibre numbers, density, diameter, area and myelin thickness distal to the injury site. Sensory (sural) and mixed (tibial) nerves were similarly involved. An impairment in regeneration could be due to several factors at the time points we studied, that involve components of the regeneration process. Diabetic nerves may simply lag behind in nerve fibre outgrowth from the outset, leading to dramatic differences at later sampling points. Nerve injury typically upregulates various growth-associated genes and neuropeptides through the retrograde cell body reaction. Enhanced expression of growth-associated protein-43 (GAP-43/B-50) is one critical component for neuronal outgrowth (Meiri et al., 1998) and has been reported to be depleted in diabetics (Maeda et al., 1996; Scott et al., 1999). Impaired regeneration cannot be explained completely by a deficient expression of GAP-43/B-50 since even high levels of this protein do not guarantee prolonged axonal outgrowth (Chong et al., 1994; Buffo et al., 1997). GAP-43/B-50 may be more important in later axonal guidance (Strittmatter et al., 1995). Diabetic regenerative clusters may also have difficulty elongating because their original basal laminae persist even at advanced stages of regeneration (Bradley et al., 1995).

Deficits in the nerve trunk content or action of vasoactive neuropeptides such as calcitonin gene-related peptide (CGRP) and substance P in diabetics may limit the hyperaemic flare in nerve microvessels following nerve injury, perhaps rendering a relative ischaemic state. Zochodne and Ho found a blunting in the hyperaemic response to capsaicin in diabetic rats (Zochodne and Ho, 1993). A diminished flare reaction was also evident in human diabetes in response to intra-epidermal injection of substance P, histamine or capsaicin into the medial forearm (Aronin et al., 1987). Excessive accumulation of advanced glycation end-products (AGEs) and low density lipoproteins (LDLs) in the vascular matrix combined with endothelial vascular wall thickening in diabetic rats will produce a narrowing of vessel luminal area (for a review, see Brownlee, 1997) that may also contribute to relative ischaemia (Maeda et al., 1999).

Macrophages assist Schwann cells in phagocytosing axon and myelin debris following nerve injury (Beuche and Friede, 1984Beuche and Friede, 1990; for a review, see Perry et al., 1993). Wallerian degeneration is delayed if there is an impairment in the invasion of macrophages (Perry and Brown, 1992) while enhancement of macrophage influx results in a greater regenerative capacity (Miyauchi et al., 1997; Prewitt et al., 1997). He and colleagues (He et al., 1996) noted that in mice, macrophages typically are enlisted to the endoneurium 3–5 days post-transection or post-crush until 15–20 days. Conti and colleagues (Conti et al., 1997) demonstrated immunocytochemically that macrophage activation is abnormal in diabetic rats and that they inappropriately infiltrate the endoneurium 2 weeks after diabetes induction. Acute hyperglycaemia has been noted to impair the respiratory burst of alveolar macrophages (Kwoun et al., 1997) and to inhibit their release of interleukin-1 (IL-1) (Hill et al., 1998), important in the upregulation of NGF (Lindholm et al., 1987; Horie et al., 1997). A comparatively reduced number of macrophages observed in diabetic mice early after injury may indicate that there is a deficit in Wallerian degeneration. For example, enhanced phosphorylation of neurofilaments may produce a lag in Wallerian degeneration by promoting resistance to calpain-mediated proteolysis (Terada et al., 1998).

Macrophages must be inactivated once their role in the regenerative process is complete. This inactivation may be mediated by high levels of cAMP secondary to the increased expression of CGRP and vasoactive intestinal polypeptide (VIP) (Cheng et al., 1995) or through the expression of interleukin-10 (IL-10) and tumour necrosis factor-β (TGF-β). Even if macrophage invasion is somehow impaired, the diabetic nerve would have the opportunity eventually to catch up to the control regenerative end point. However, the persistence of macrophages in our results might impair Schwann cell viability (Skoff et al., 1998) or function (e.g. myelin splitting) (Tamura and Parry, 1994).

Autotomy behaviour is sometimes seen in animals following peripheral nerve injury (Wall and Gutnick, 1974; Banõs et al., 1994). These behaviours are directed to the denervated areas after a delay of several days to weeks (Wall et al., 1979; Asada et al., 1996). Diabetic mice in our study exhibited earlier and higher autotomy levels than controls. It is possible that they experience enhanced neuropathic pain since autotomy scores may be an appropriate measure of the degree of pain or dysaesthaesia (for a review, see Coderre et al., 1986). In fact, noxious stimuli after nerve injury will enhance autotomy presumably by enhancing pain perception (Coderre and Melzack, 1986). These painful perceptions could arise from the activation of protein kinase C (PKC) following nerve injury (Mao et al., 1992; Malmberg et al., 1997). Diabetes is already known to alter the expression of PKC isozymes (Roberts and McLean, 1997) which may ultimately also lead to impaired Na+,K+-ATPase activity (Hermenegildo et al., 1992) and conduction deficits.

Our results provide evidence that peripheral nerve motor and sensory regeneration is slowed substantially in experimental diabetes mellitus. The regenerative defect was not complicated by a major difference in body weight or maturational level between diabetics and controls which have complicated diabetic rat animal models in the past. It is possible that the regenerative deficit is related to the observed slow infiltration and later relative persistence of endoneurial macrophages.

Acknowledgments

We wish to thank Brenda Boake who provided expert secretarial assistance. D.W.Z. is a Medical Scholar of the Alberta Heritage Foundation for Medical Research. J.M.K. is supported by an Alberta Heritage Foundation for Medical Research studentship. The work was supported by an operating grant from the Medical Research Council of Canada.

References

View Abstract