Impairments of the primary afferent nerves in a rat model of diabetic visceral hyposensitivity

The most commonly used experimental model of diabetes in rodents is the elimination of pancreatic β islet cells by administration of STZ. STZ-induced diabetes produces early functional and biochemical abnormalities within peripheral nerves of rats and mice similar to those observed in human diabetic neuropathy [19]. Diabetes was induced in male Sprague–Dawley rats by an intraperitoneal (i.p.) injection of STZ (50 mg/kg, n = 40) dissolved in 0.01 M cold citrate buffer (pH 4.0). Control rats (n = 20) were injected with the same volume of vehicle (0.01 mol/L citrated buffer). From day 3 after injection, blood glucose level was checked weekly using the glucose-oxidase test strip and a reflectance meter (Roche Diagnostics GmbH, Ireland) on blood samples obtained from a tail prick. The animals showing a sustained hyperglycemia (27/40, 67.5 %) were used for comparison of the visceral sensitivity and the expression of neuronal markers 3–6 weeks after injection of STZ or vehicle.

CRD-induced VMR is a pseudoaffective response induced by noxious stimulation of the viscera and has been used extensively to evaluate visceral sensitivity in experimental animals [25]. To assess the changes in colorectal sensitivity in early diabetes, STZ-treated rats with sustained hyperglycemia (n = 9) were anesthetized by i.p. injection of pentobarbital sodium (60 mg/kg) and the rats’ body temperature were kept around 37.5 °C using a thermostatic blanket. A lubricated polyethylene balloon (3 cm long, 1.5 cm max diameter) was inserted into the colorectum (the distal end of the balloon was about 2 cm proximal to the anus) for gradual CRD. The balloon was connected to a pressure control device and a pressure sensor. Two silver wire electrodes were implanted into the musculus obliques externus abdominis and rectus abdominis and a reference electrode was inserted subcutaneously to record the eletromyogram (EMG) induced by CRD. The balloon pressure and EMG signals were fed via the Powerlab interface into a computer and analyzed by Chart software (Powerlab; AD Instrument, Bella Vista, Australia). Following the surgical preparation which is usually lasted about 1 h, the rat was kept under light anesthesia via inhalation of 1 % isoflurane and was allowed at least 30 min to acclimatize before the test began. CRD was performed via gradually inflating the balloon up to a pressure of 80 mmHg within 2 min and this was repeated at 10 min intervals. Generally, 6 trials were run to achieve a stable response (less than 20 % variability between the last two trials). The same experimental paradigm was completed with age-matched control rats (n = 7).

At 4–5 weeks after induction of diabetes mellitus, rats (n = 5) were deeply anaesthetized with pentobarbital (80 mg/kg, i.p.). The abdominal cavity was exposed by a midline incision on the abdominal wall. The colon 3 cm proximal to the anal sphincter with the attached pelvic ganglia and pelvic nerves was removed and was placed into a recording chamber (20 mL), where it was superfused continuously at a rate of 15 mL/min with oxygenated Krebs solution (composition in mmol/L: NaCl 113, KCl 5.9, CaCl₂ 1.25, MgSO₄ 1.2, NaH₂PO₄ 1.2, NaHCO₃ 25, Glucose 11.5). Chamber temperature was kept at 34 °C. A fine nerve branch emanating from the pelvic ganglion was dissected and recorded using a glass suction electrode filled with the buffer. The colon was catheterized at the both ends in order to apply repeated mechanical stimulation by ramp distension (0–60 mmHg) at an interval of 15 min. Neuronal signal was amplified (Neurolog NL102, Digitimer, UK), filtered (band pass 300–3000 Hz) and fed into a computer via Micro1401 interface (Cambridge Electronic Design, UK). Afferent discharge was saved and processed using spike 2 software (version 5.14, Cambridge Electronic Design, UK). The activity of single units in each recording was discriminated by performing the principal component analysis of the contour of individual spikes as described previously [26]. The same experimental paradigm was completed with vehicle-treated control rats (n = 4).

Immunofluorescent staining of the general neuronal maker PGP 9.5 and the specific sensory marker CGRP for C- and Aδ-fibers [24] were performed in colon from diabetic rats (4–5 weeks after induction of diabetes, n = 7) and age-matched control rats (n = 5). Under anesthesia with pentobarbital, rats were transcardially perfused with saline followed by 4 % paraformaldehyde and 0.14 % picric acid in phosphate buffer (PB, 0.1 mol/L, pH 7.4). The colon were removed and post-fixed in the same fixative overnight at 4 °C and then cryoprotected with 30 % sucrose in 0.1 mol/L PB overnight at 4 °C. The samples were cut at 16 μm for staining as described previously [27]. The sections were first incubated with 0.05 mol/L phosphate-buffered saline (PBS) containing 10 % normal goat serum and 0.5 % TritonX-100 at room temperature for 1 h to block non-specific binding and this was followed by incubation with primary rabbit anti-PGP 9.5 antibody (1:1000, abcam, Cambridge, MA, USA) or rabbit anti-CGRP antibody (1:500, abcam, Cambridge, MA, USA) at 4 °C overnight. The sections were rinsed with PBS for four times and were then incubated with goat anti-rabbit Alexa fluor 488 secondary antibody (1:1000; Molecular Probes-Invitrogen, Eugene, OR, USA) at room temperature for 1 h. After washing with PBS the sections were mounted on glass slides and viewed under the fluorescent microscope (Leica DM2500, Leica Microsystems Limited, USA) and the digital images were analyzed using Leica application suite version 4.3 (Leica Microsystems Limited, USA). Nerve fiber density was evaluated by counting the number of neurites expressed PGP 9.5 or CGRP. The number of neurites/mm² was calculated and averaged from six randomized fields for each rat. Neurites appeared as thin, relatively straight green lines, distinct from the background, sometimes beaded with varying width, and sometimes branched. Each straight continuous or beaded line was counted as one neurite, while the tip of each branch was counted as another neurite, described as previous study [27]. Counts were done by two investigators who blinded to the group assignment and calculations were therefore done using averages of the two sets of counts.

Diabetic (n = 6) and control (n = 4) rats were deeply anesthetized and their colons were removed, frozen immediately on dry ice and stored at −80 °C. The samples were homogenized in lysis buffer containing 20 mmol/L Tris–HCl (pH 8.0), 150 mmol/L NaCl, 1 mmol/L EDTA, 1 % NP-40, 1 mmol/L PMSF, protease inhibitor cocktail (Sigma, St. Louis, MO, USA) and phosphatase inhibitor cocktail (Thermo scientific, Indianapolis, IN, USA) for 1 h at 4 °C. The lysates were centrifuged at 10,000g for 30 min at 4 °C and the concentration of protein in each supernatant was determined using a BCA assay (Pierce, Rackford, IL, USA). Thirty-microgram aliquots were separated on 4–20 % Tris–glycine ready gels (Bio-rad, Hercules, CA, USA), the separated proteins were transferred from the gel to the surface of Nitrocellulose membranes (Bio-rad). The membranes were blocked with 5 % fat-free dry milk in Tris-buffered saline (TBS) containing 0.1 % Tween-20 for 1 h, and were then incubated for 48 h at 4 °C with primary rabbit anti-PGP 9.5 antibody (1:200, abcam, Cambridge, MA, USA) and with mouse anti-β-actin antibody (1:2000, abcam) for 2 h at RT, respectively. Bound primary antibodies were detected with HRP-conjugated anti-rabbit or anti-mouse secondary antibody (1:5000, Bio-rad). Immunoreactive bands were visualized using enhanced chemiluminescence (Thermo scientific) and digital imaging was captured with Image Quant LAS 4000 mini (GE Healthcare, Life Science, USA). The density of specific bands was measured with NIH ImageJ (http://​rsb.​info.​nih.​gov/​ij/​) software and was normalized against a loading control (β-actin). We tried but failed to get the CGRP signal (band not at theoretic molecular weight, data not shown).

One day before VMR measurement, mechanical paw withdrawal thresholds (PWTs) were determined by probing the left hind paw with calibrated Von Frey hairs as described previously [28] in diabetic (n = 9) and control (n = 7) rats (same rats for VMR study). Von Frey hairs of increasing size (hence of increasing force) were applied to the plantar surface of the left hind paw until the animal withdrew the paw. The smallest size of the von Frey hairs required to produce a withdrawal reflex was recorded as the threshold.

Data were presented as mean ± SD. Statistical analysis of the data sets was carried out using Graphpad Prism version 3.0 (Graphpad software, La Jolla, CA, USA). Two-way ANOVA with Bonferroni’s posttest was used to assess the difference between experimental groups for blood glucose level and paw withdrawal threshold. To analyze VMR and pelvic nerve responses, an X–Y plot analysis was first conducted to measure the increases in afferent discharge rate from the baseline as the intraluminal pressure rose by every 1 mmHg. Such analysis gave rise to individual pressure-afferent response curves and then the group (average) pressure-afferent response curves. For area under the curve (AUC) analysis, the AUC of individual pressure-afferent response curve was determined and was then averaged. Unpaired t-test was used to assess the difference in averaged AUC between experimental groups. For PGP 9.5/CGRP expression analysis, unpaired t-test was used to assess difference between experimental groups. Significance was set at p < 0.05.