Peripheral Neuropathy in Mouse Models of Diabetes

Abstract

Peripheral neuropathy is a frequent complication of chronic diabetes that most commonly presents as a distal degenerative polyneuropathy with sensory loss. Around 20% to 30% of such patients may also experience neuropathic pain. The underlying pathogenic mechanisms are uncertain, and therapeutic options are limited. Rodent models of diabetes have been used for more than 40 years to study neuropathy and evaluate potential therapies. For much of this period, streptozotocin-diabetic rats were the model of choice. The emergence of new technologies that allow relatively cheap and routine manipulations of the mouse genome has prompted increased use of mouse models of diabetes to study neuropathy. In this article, we describe the commonly used mouse models of type 1 and type 2 diabetes, and provide protocols to phenotype the structural, functional, and behavioral indices of peripheral neuropathy, with a particular emphasis on assays pertinent to the human condition. © 2016 by John Wiley & Sons, Inc.

Keywords: allodynia; corneal confocal microscopy; hyperalgesia; hypoalgesia; nerve conduction velocity; nerve morphometry; peripheral neuropathy; skin biopsy; type 1 diabetes; type 2 diabetes.

Figure 1

Typical traces recorded after stimulation at the ankle (solid line) and at the hip (dashed line), depicting latencies between stimulations (Δ latencies).

Figure 2

A: Cross section, cut at 6 μm thickness, of a resin-fixed mouse sciatic nerve. Note the multiple fascicles embedded in fat, suggesting a distal portion of the nerve trunk and that one fascicle (red box) shows evidence of mechanical artifact. B: Mechanical damage causes distortion of axons and myelin that precludes morphometry. C: Well preserved region of nerve showing fibers viable for morphometry (examples within the white circle are indicated by a red *) and also fibers that are not quantified due to being paranodal sections (blue box and blue X), containing Schmidt-Lanterman incisures (green diamond and red X) or having a Schwann cell nucleus present (orange circle). The yellow arrow points to an endoneurial blood vessel.

Figure 3

Hargreaves apparatus calibration curve. Mean±sem of N=6 individual curves combined to give a daily average used to convert response time to response temperature.

Figure 4

A: Image of mouse skin processed in paraffin, cut at 6μm, and stained with PGP9.5. The white boxes enclose IENF that would be counted as crossing the dermal:epidermal junction (although the entire axon is only visible upon adjusting the fine focus of the microscope). Note the punctate PGP9.5 staining. The red arrows point to additional profiles included using the fragment counting method. The black circle encloses a presumptive Langerhans cell and its processes, which exhibit more intense and uniform staining that IENF. B: Image of skin processed as a frozen section, cut at 50μm, stained with PGP9.5 and imaged using a program (Image-Pro Plus, Media Cybernetics Inc.) that stacks serial images to produce a composite that facilitates tracking of individual SNP and IENF across planes of section. Within each white box is a progression of SNP to IENF. Bar = 10μm.

Figure 5

Placement of a mouse on the custom designed platform for corneal imaging. The anesthesia delivery system is visible to the left of the image and the objective of the corneal confocal microscope with tomocap installed is at the center of the image.

Figure 6

Confocal images of mouse cornea. (A) Sub-basal nerves (arrowheads), (B) stromal nerve (arrow) and (C) stromal nerve image with a counting grid overlay that gives an occupancy score of 10/64 grids.

Figure 7

Indices of large (A) and small (B) fiber neuropathy in adult female C57Bl/6J mice after 8 weeks of STZ-induced diabetes (black bars) compared to values in age-matched adult female C57Bl/6J mice (control: white bars). Data are group (N=9–10) mean ± SEM with statistical comparison by unpaired 1 tailed t test.