Neurotrophic modulation of myelinated cutaneous innervation and mechanical sensory loss in diabetic mice

Abstract

Human diabetic patients often lose touch and vibratory sensations, but to date, most studies on diabetes-induced sensory nerve degeneration have focused on epidermal C-fibers. Here, we explored the effects of diabetes on cutaneous myelinated fibers in relation to the behavioral responses to tactile stimuli from diabetic mice. Weekly behavioral testing began prior to streptozotocin (STZ) administration and continued until 8 weeks, at which time myelinated fiber innervation was examined in the footpad by immunohistochemistry using antiserum to neurofilament heavy chain (NF-H) and myelin basic protein (MBP). Diabetic mice developed reduced behavioral responses to non-noxious (monofilaments) and noxious (pinprick) stimuli. In addition, diabetic mice displayed a 50% reduction in NF-H-positive myelinated innervation of the dermal footpad compared with non-diabetic mice. To test whether two neurotrophins nerve growth factor (NGF) and/or neurotrophin-3 (NT-3) known to support myelinated cutaneous fibers could influence myelinated innervation, diabetic mice were treated intrathecally for 2 weeks with NGF, NT-3, NGF and NT-3. Neurotrophin-treated mice were then compared with diabetic mice treated with insulin for 2 weeks. NGF and insulin treatment both increased paw withdrawal to mechanical stimulation in diabetic mice, whereas NT-3 or a combination of NGF and NT-3 failed to alter paw withdrawal responses. Surprisingly, all treatments significantly increased myelinated innervation compared with control-treated diabetic mice, demonstrating that myelinated cutaneous fibers damaged by hyperglycemia respond to intrathecal administration of neurotrophins. Moreover, NT-3 treatment increased epidermal Merkel cell numbers associated with nerve fibers, consistent with increased numbers of NT-3-responsive slowly adapting A-fibers. These studies suggest that myelinated fiber loss may contribute as significantly as unmyelinated epidermal loss in diabetic neuropathy, and the contradiction between neurotrophin-induced increases in dermal innervation and behavior emphasizes the need for multiple approaches to accurately assess sensory improvements in diabetic neuropathy.

Figure 1. Diabetic mice develop reduced behavioral responses to non-noxious mechanical stimulation

Behavioral responses to mechanical stimuli were evaluated weekly in non-diabetic and diabetic mice using von Frey monofilaments of different strengths. A) Using a 0.16 g monofilament, significant reductions in the number of withdrawals following von Frey application were evident in diabetic mice 5 weeks post-STZ injection. Similar reductions were evident with 0.6 g (B), 1.4 g (C) and 2.0 g (D) monofilaments, although the differences between non-diabetic and diabetic mice were visible at 4 weeks using the larger monofilaments. Data plotted as means ± SEM. * denotes P < 0.05 from non-diabetic values.

Figure 2. NGF and Insulin treatments increase mechanical sensitivity

Behavioral responses of non-diabetic and treated diabetic mice to non-noxious mechanical stimuli were evaluated using a 1.4 g von Frey monofilament. Treatments began after testing on the 6^th week after diabetes induction (arrows). NGF and insulin treatments both significantly increased behavioral responses of diabetic mice to non-noxious mechanical stimuli. In contrast, treating mice with CSF, NT-3, or combined NGF and NT-3 did not improve the behavioral responses of diabetic mice to von Frey stimulation of the footpad. Data plotted as means ± SEM. # denotes significant differences between neurotrophin- or insulin-treated diabetic mice and CSF-treated diabetic mice (P < 0.05). Non-diabetic mice were significantly different from CSF-treated diabetic mice at all time points.

Figure 3. Reduced MBP-positive fibers in the dermis of diabetic mice

Six weeks after STZ injection, footpad skin from non-diabetic (A) and diabetic mice (B) was processed for visualization of myelin basic protein (MBP)-positive axons. A, B) Images of MBP-positive axons in non-diabetic (A) and diabetic (B) mouse footpad. Arrows denote fascicles of MBP-positive axons, some of which extend up into the dermal papillae. Compared to healthy non-diabetic mice, diabetic mice displayed a significant reduction in MBP-positive fascicles within the dermis. e, epidermis; d, dermis. Scale bar equals 50 μms for each image.

Figure 4. Neurotrophin and insulin restore diabetes-induced reduction of NF-H-positive axons in the footpad

Footpad skin from non-diabetic and treated diabetic mice was processed for visualization of neurofilament heavy chain (NF-H)-positive axons. Diabetic mice treated with CSF exhibited a severe reduction in the innervation pattern of NF-H-positive axons. Following treatment with NGF, NT-3, NGF and NT-3 combined, or insulin for two weeks, the innervation of the footpad skin by NF-H-positive axons appeared markedly improved in diabetic mice. Scale bar = 100μm.

Figure 5. Neurotrophin and insulin treatments restore myelinated innervation in the footpad skin of diabetic mice

Area of footpad skin occupied by NF-H-positive axons was measured in non-diabetic and treated diabetic mice. CSF-treated diabetic mice had significantly less myelinated innervation in the footpad than non-diabetic footpad skin. Following treatment with NGF, NT-3, NGF and NT-3, or insulin, myelinated innervation was significantly improved in diabetic footpad skin. Data plotted as means ± SEM. * denotes significant differences between CSF-treated diabetic mice and all other groups (P < 0.05).

Figure 6. NT-3 treatment increases Merkel cell numbers associated with axons in the footpad of diabetic mice

A, B) Photomicrographs of immunocytochemical staining of sections from the mouse footpad for Merkel cells (arrows). Included are sections from a non-diabetic and a diabetic mouse. Diabetic mice had only slightly fewer Merkel cells (P > 0.05, see quantification below). C) Photomicrograph of immunocytochemical staining of footpad skin from a C57Bl/6 mouse visualizing both myelinated axons (NF-H-positive, green) and Merkel cells (cytokeratin 18, red). Arrow indicate Merkel cells innervated by myelinated axons. D) Merkel cells were quantified using conventional fluorescence microscopy and were counted if they were adjacent to the epidermis and innervated by NF-H-positive axons. Merkel cell numbers in CSF-treated diabetic mice were not significantly different than non-diabetic mice. NT-3 treatment of diabetic mice significantly increased the number of Merkel cells in the footpad. Data plotted as means ± SEM. * denotes P < 0.05; vs. non-diabetic, CSF-treated diabetic, and insulin-treated diabetic mice. Scale bar = 100 μms for A–B, and 50 μm for C.