Association of chronic diabetes and hypertension in sural nerve morphometry: an experimental study

For a long time, our laboratory has been dedicated to investigate abnormalities in the morphology and morphometry of peripheral nerves due to hypertension [2,3,7,24-27], pointing to the description of a so called “hypertensive neuropathy”. In fact, it was recently described that non-diabetic patients with chronic untreated hypertension may develop impairment of nerve function [28]. The positive correlation between hypertension and diabetes described in humans drove us to investigate this association under controlled experimental conditions using SHR and STZ-induced diabetes. Our study showed that the sural nerve myelinated fibers were highly affected by experimental diabetes, causing mainly the reduction of the fiber size. But the most striking result from this work is that the association between diabetes and hypertension caused an increase on the average size of the myelinated fibers, pointing to a small fiber loss, associated to axonal atrophy.

Abnormalities reported in human diabetic neuropathy include axonal degeneration, primary demyelination resulting from Schwann cell dysfunction, secondary segmental demyelination related to impairment of the axonal control of myelination, remyelination, proliferation of Schwann cells, atrophy of denervated bands of Schwann cells, onion-bulb formations and hypertrophy of the basal lamina [29,30]. Animal models of diabetes still do not reproduce all morphological alterations present in human nerves, but offer the possibility of investigating the disease under controlled risk factors environment, including the association with co-morbidities or risk factor for the disease.

De Visser et al. [31] studied the association of hypertension and type II diabetes in adult rats, comparing the Zucker Diabetic Fatty Rat (ZDF - a well stablished model of type II diabetes) and a genetic hybrid between the ZDF and the Spontaneously Hypertensive Heart Failure rat. The main finding for the diabetic rats was the presence of thinned myelin with a mild additive effect of hypertension that appeared only after 6 months of diabetes. Our results for the diabetic rats are somewhat similar to those from De Visser et al. [31] since we also demonstrated a small myelin area that was reflected on the total fiber size (smaller on WR + STZ compared to WR). On the other hand, on SHR animals, when compared to SHR + STZ, there was a tendency for increase in the average myelin area, associated with the increase in fiber size, suggestive of small myelinated fiber loss. In addition to these morphometric findings, SHR + STZ animals showed a great number of myelinated fibers with severe disruptions of the myelin sheath, particularly in large myelinated fibers (Figure 1).

Impairment of the motor conduction velocity in nerves from diabetic animals was described as related to large myelinated fibers dysfunction. Yamaguchi et al. [32] investigated the tail motor nerve conduction velocity in male Spontaneously Diabetic Torii-Lepr^fa rats (SDT Fatty). They demonstrated a reduction by 18% in the motor nerve conduction velocity in 24 weeks old SDT Fatty rats, compared to controls. In our study, the conduction velocity was not investigated because alterations in conduction velocity usually occur after a longer period then the experimental time we used (20 weeks of age, 12 weeks after diabetes induction). In our experimental model, because diabetes is severe (blood sugar levels sustained above 300 mg/dl without treatment), the animals do not survive under ethical conditions much longer than 12 weeks. Nevertheless, because larger fiber impairment is readily detectable in nerve conduction studies, and our study demonstrated that smaller fibers were more affected, we believe that the conduction studies in our experimental model would not add much.

It has been suggested that hyperglycemia is one factor for increasing Schwann cell metabolic stress. In fact, we recently have shown alterations of the Schwann cells mitochondria in nerves from acute STZ-diabetic rats [33]. The association between hypertension and hyperglycemia has also been suggested as a stress for the Schwann cells [9,31,34]. Our results showed that hypertension (without diabetes) significantly decreased the number and density of Schwann cells in the sural nerve compared to normotensive animals and, in the SHR + STZ, this number was reduced compared to SHR. The reduction in number of Schwann cells might be responsible, at least in part, for the severe myelin alterations we observed in SHR, with or without diabetes. Since the Schwann cells are the myelinating cells, an ideal relation or ratio between the number of the Schwann cells, the axon size and the myelin thickness is expected. The disruption of this ideal relation, like the reduction in the Schwann cells number can certainly cause myelin morphology alterations similar to the observed by us and others [9]. Metabolic stress in diabetic nerves was ascribed not only to hyperglycemia, but also to abnormal fatty acid metabolism, ischemic hypoxia and oxidative stress. The present study did not aim to investigate nerve metabolism but it has been shown that pharmacological interventions such as treatment with free radical scavenger can be effective in the prevention of the experimental diabetic neuropathy [35-37].

Thickening of the Schwann cell basement membrane is one of the alterations in diabetes, shown by ultrastrucutral investigations of peripheral nerves [14,33,38]. In our material it is evident that unmyelinated fibers are easily visible in diabetic animals (normotensive and hypertensive) compared to controls and also in SHR (Figure 1). An ultrastrucutral study of this material will help us to clarify this issue.

A reduction of myelinated nerve fiber size is the most characteristic and reproducible abnormality of peripheral nerve in experimentally induced diabetic rats [19,39-47]. This reduction is claimed to be due to a reduction in axon size caused by metabolic alterations in the axons [39,41,42,45,46], retarded growth rate [40,44], or increased hyperosmolarity [43]. More recently, the alterations of the endoneural microvasculature is also pointed out as a factor for the axon size reduction [12,13].

Our results of myelinated fiber alteration in hypertensive rats are very likely to be explained by the morphological changes in endoneural microvessels associated with hypertension, as described before [2,4]. Besides the larger number of collapsed blood vessels (with thickening of the walls and absence of lumen) and/or vessels with thick walls in SHR and SHR + STZ, a smaller fascicular area occupied by the capillary was statistically significant in these animals. Gregory et al. [9] observed reduction of endoneural blood vessels in SHR, associated to a decrease in sciatic nerve blood flow. Associated to these findings, they also showed reduced axonal diameter in SHR and diabetic SHR in the absence of changes to other structural parameters [9]. In line with our results, they affirmed that large fiber neuropathy in SHR resembled that of STZ-diabetic rats, but includes axons with thin myelin, suggestive of segmental demyelination/remyelination. The large fiber neuropathy present in our results for SHR can be identified by the reduced percentage of the large fibers peak on the distribution histograms (Figure 2), which is accentuated by the association between hypertension and diabetes. Adding to their results, we also show the small fiber neuropathy in SHR, through the myelinated fiber distribution shift to the right, also blunted by the association between hypertension and diabetes (Figure 2). The small fiber neuropathy in hypertension, similar to the one described in this study, was previously characterized in SHR though results obtained in our laboratory [2], by the decrease in number of myelinated fibers, the significant increase in myelinated fiber average size and the shift in fiber distribution histograms.

Experimental models of arterial infarct suggested that small nerve fibers are more vulnerable to ischemia than the large nerve fibers [48,49]. Thus, small fiber neuropathy in SHR can be associated to the damage of endoneural vessels [2,4]. In favor of the small fibers loss, an altered g ratio was also observed in the current study (Table 3 and Figure 2). The shift to the right observed in the g ratio distributions in SHR (diabetic and non-diabetic) could be suggestive of demyelination, as pointed out by Gregory et al. [9]. On the other hand, it is also described that larger fibers present larger g ratios compared to smaller fibers [50,51]. If the smaller fibers were lost, larger fibers will shift the g ratio distribution to the larger values.

The data presented here resulted from a comparison between SHR and Wistar rats, with or without experimentally induced diabetes. Spontaneously hypertensive rats (SHR) were first inbred from their normotensive counterparts, the Wistar-Kyoto rats (WKY). Thus, it would be expected for WKY to be used as the normotensive controls of SHR in diverse experimental protocols [7]. Nevertheless, Wistar rats (WR) have been recently used as the SHR controls as often as the WKY [52,53]. Peripheral nerve function is significantly affected by maturation and aging but rats described as adults have weights that vary in a range of about 180 g to 1000 g or more [17,54]. In the present study, animals with absolutely the same age but significantly different body weight presented similar fascicular area and diameter of the sural nerves. This indicates that the nerves developed in a similar way and the stress due to hypertension did not impact the overall SHR development. Actually, SHR are described as presenting smaller body weight compared to age matched WKY [9] as well as to Wistar rats [7].