Repeated monitoring of corneal nerves by confocal microscopy as an index of peripheral neuropathy in type-1 diabetic rodents and the effects of topical insulin.

Animals

All animal protocols were approved by the Institutional Animal Care and Use Committee (IACUC) of University of California, San Diego. In order to assess assay reproducibility and establish any between-strain differences, 9 female albino Sprague-Dawley rats (245-295 g) and 8 age-matched female, pigmented Long-Evans rats (260-289 g) were monitored every 4 weeks for 3 months. Female Sprague-Dawley rats (221-253 g) were then used for a longitudinal study of the impact of type-1 diabetes on corneal nerves while female Swiss Webster mice (25-30 g) were used to assess impact of topical insulin to the eye. Rats were made diabetic with a single dose of streptozotocin (STZ: 55 mg/kg i.p.) ( Calcutt, 2004 ) . Mice were made diabetic by injection of STZ (90 mg/kg, i.p.) on two consecutive days ( Davidson et al., 2009 ) . Blood glucose levels were measured 4 days after injection of STZ and animals with blood glucose levels of >15 mmol/l were considered diabetic. Age-matched animals served as controls. For insulin treatment, 0.1 IU of regular U-100 Humulin (Lilly, Indianapolis, IN) in 10 μl saline ( Guo et al., 2011 ) was applied directly on the eye of 8 control and 8 diabetic mice daily for 4 weeks and corneal nerve occupancy compared to control and diabetic mice receiving saline treatments. Plasma insulin, blood glucose and HbA1c Blood was collected from the tail by venipuncture and glucose concentration measured using the OneTouch ultra mini system (LifeScan, Inc., Milipitas, CA, USA). Blood was centrifuged and 50 μl of plasma used to measure insulin concentration by ELISA (Ultrasensitive Rat Insulin kit, Mercodia, Uppsala, Sweden). HbA1c was measured using the A1CNow system (Bayer, Sunnyvale, CA, USA).

Animal imaging platform

A small animal platform ( Fig. 1 ) was developed for use with the Heidelberg Retina Tomograph 3 with Rostock Cornea Module (Heidelberg Engineering, Heidelberg, Germany). The chin rest was removed and the animal platform placed on the chin rest attachment of the microscope. The imaging platform consists of a base platform supporting two independent moving parts: a) nose cone for the continuous application of isoflurane anesthesia and b) body platform with Velcro straps for body and head restraint. The relative distance between nose cone and body platform can be adjusted for different sized animals. A small attachment to the body platform raises the head for additional adjustment for optimal imaging of the cornea. The base platform can swivel to orientate the animal such that the laser light enters perpendicular to the corneal surface at the apex. The entire platform can be removed, turned 180° and placed back on the chin rest attachment to image both eyes.

Imaging procedures

Under isoflurane (2% in oxygen) anesthesia, rats or mice were placed on the imaging platform and secured with body and head straps such that the eyes were open and one eye was facing the objective of the CCM. GenTeal gel (Novartis Pharmaceuticals Corp., East Hanover, NJ, USA) was placed on both eyes for laser light coupling and to prevent the eye from drying. The microscope objective was positioned to the center apex of the cornea using the laser reflection on the eye and real-time images. By using the objective focus, depth was set to zero at the internal reflection of the tomocap, being careful not to position the objective such that the pressure from the tomocap begins to wrinkle the corneal surface ( Kobayashi et al., 2006 ) . The depth was adjusted to 15 μm and one volume stack of 40 images (384× 384 pixels, 1 μm lateral resolution) was collected. Under the volume scan option, the microscope automatically refocuses 10 μm superficial to the set depth and collects volumes up to 80 μm in depth (2 μm depth resolution), resulting in a volume stack from approximately 5-80 μm. Images were collected every 4 weeks and nerve occupancy in the sub-basal plexus and stromal layers of the cornea quantified using a custom designed graphical user interface developed from MATLAB (Natick, MA, USA).

Image analysis and quantification

We defined the progression from sub-basal plexus to stromal layers by the disappearance of the fine, linear corneal nerves of the sub-basal plexus and the highly reflective background along with the appearance of bright keratocytes and large nerve fibers on a dark background ( Labbe et al., 2006 ) . Once the sub-basal:stromal junction was determined, inter-animal concordance was achieved using the last image of the sub-basal plexus and the first image of the stroma as fixed anatomical points. Each image was viewed by eye and nerves traced using a WACOM Bamboo tablet (Saitama, Japan) in Image J (NIH) software. Composite images were made by layering all images from sub-basal plexus using Adobe PhotoShop (San Jose, CA), with color-coding of the traces from each image ( Fig. 2 ). In the absence of continuous images of nerves in the sub-basal plexus, even after stacking in a volume scan, we developed an alternative quantification system based on occupancy. Each image of a volume scan was loaded into MATLAB and a 5×5 grid overlaid for a total of 25 possible occupancies per image. In preliminary modeling studies, we compared 3×3, 5×5, 8×8 and 10×10 grids (data not shown). If a nerve was observed anywhere inside a box, that box was counted as occupied ( Fig. 2 ). For each image, the number of boxes in the grid containing one or more nerves was counted. Nerve occupancy was calculated for both the sub-basal plexus and stromal layers of the cornea. Equation 1 shows the formula for calculating % nerve occupancy per layer, per animal: (Eq.1). % n e r v e o c c u p a n c y = ( t o t a l # b o x e s o c c u p i e d b y n e r v e s i n l a y e r # o f i m a g e s i n l a y e ∗ t o t a l # o f b o x e s i n g r i d ) ∗ 100 Average nerve occupancy was measured at each image depth and then organized by distance from the sub-basal plexus:stromal junction. Volume average analysis was conducted by calculating the average nerve occupancy of the entire volume of sub-basal plexus or stromal layers. Epidermal nerves Hind paw skin (plantar surface) was fixed overnight at 4°C in 4% paraformaldehyde in 0.1 M sodium phosphate buffer. The skin was processed, hemisectioned and embedded in paraffin. Sections were cut at a thickness of 6 μm and collected onto glass slides then treated with 3% hydrogen peroxide for 15 minutes followed by normal goat serum (Vectastain Rabbit IgG ABC Kit, Vector Laboratories #PK4001, Burlingame, CA) for 30 minutes. Sections were incubated in a primary antibody against rabbit anti-human Protein Gene Product 9.5 (PGP9.5) (1:1000; AbD Serotec #7863-0504, Raleigh, NC) overnight at 4°C. They were then washed and incubated with biotinylated goat anti-rabbit secondary antibody (Vectastain ABC Kit) for 1 hour, followed by a wash and incubation with an avidin-biotin complex solution (Vectastain ABC Kit) for an additional 1 hour. The reaction product was demonstrated by NovaRed staining (NovaRed Peroxidase Substrate Kit, Vector Laboratories #SK4800). Gill’s hematoxylin was used as a counterstain. Skin was viewed using a light microscope and the number of nerve profiles of intra-epidermal nerve fibers (IENFs) and sub-epidermal nerve plexi (SNP) counted and quantified per unit length of the dermal:epidermal junction in the section ( Beiswenger et al., 2008b ) .

Imaging procedures

Under isoflurane (2% in oxygen) anesthesia, rats or mice were placed on the imaging platform and secured with body and head straps such that the eyes were open and one eye was facing the objective of the CCM. GenTeal gel (Novartis Pharmaceuticals Corp., East Hanover, NJ, USA) was placed on both eyes for laser light coupling and to prevent the eye from drying. The microscope objective was positioned to the center apex of the cornea using the laser reflection on the eye and real-time images. By using the objective focus, depth was set to zero at the internal reflection of the tomocap, being careful not to position the objective such that the pressure from the tomocap begins to wrinkle the corneal surface ( Kobayashi et al., 2006 ) . The depth was adjusted to 15 μm and one volume stack of 40 images (384× 384 pixels, 1 μm lateral resolution) was collected. Under the volume scan option, the microscope automatically refocuses 10 μm superficial to the set depth and collects volumes up to 80 μm in depth (2 μm depth resolution), resulting in a volume stack from approximately 5-80 μm. Images were collected every 4 weeks and nerve occupancy in the sub-basal plexus and stromal layers of the cornea quantified using a custom designed graphical user interface developed from MATLAB (Natick, MA, USA).