Diabetic peripheral neuropathy affects up to half of diabetic patients. This neuronal damage leads to sensory disturbances, including allodynia and hyperalgesia. Many growth factors have been suggested as useful treatments for prevention of neurodegeneration, including the vascular endothelial growth factor (VEGF) family. VEGF-A is generated as two alternative splice variant families. The most widely studied isoform, VEGF-A165a is both pro-angiogenic and neuroprotective, but pro-nociceptive and increases vascular permeability in animal models. Streptozotocin (STZ)-induced diabetic rats develop both hyperglycaemia and many of the resulting diabetic complications seen in patients, including peripheral neuropathy. In the present study, we show that the anti-angiogenic VEGF-A splice variant, VEGF-A165b, is also a potential therapeutic for diabetic neuropathy. Seven weeks of VEGF-A165b treatment in diabetic rats reversed enhanced pain behaviour in multiple behavioural paradigms and was neuroprotective, reducing hyperglycaemia-induced activated caspase 3 (AC3) levels in sensory neuronal subsets, epidermal sensory nerve fibre loss and aberrant sciatic nerve morphology. Furthermore, VEGF-A165b inhibited a STZ-induced increase in Evans Blue extravasation in dorsal root ganglia (DRG), saphenous nerve and plantar skin of the hind paw. Increased transient receptor potential ankyrin 1 (TRPA1) channel activity is associated with the onset of diabetic neuropathy. VEGF-A165b also prevented hyperglycaemia-enhanced TRPA1 activity in an in vitro sensory neuronal cell line indicating a novel direct neuronal mechanism that could underlie the anti-nociceptive effect observed in vivo. These results demonstrate that in a model of Type I diabetes VEGF-A165b attenuates altered pain behaviour and prevents neuronal stress, possibly through an effect on TRPA1 activity.

CLINICAL PERSPECTIVE

  • Diabetic complications such as neuropathy affect up to 50% of all diabetics. The pain associated with diabetic neuropathy is a large unmet clinical need due the increasing numbers of diabetics worldwide. VEGF-A has been suggested to be a possible therapeutic for diabetic neuropathy; this is a family of proteins generated through alternative pre-mRNA splicing that can exert different physiological effects.

  • We show that human recombinant protein VEGF-A165b can ameliorate both diabetic neuronal injury and neuropathic pain in experimental rodents.

  • Therapeutics that alter the VEGF-A isoform complement in favour of VEGF-A165b may be novel treatments for diabetic neuropathy.

INTRODUCTION

Diabetes mellitus affects 6.7% of European and 10.5% of U.S. populations and results in a myriad of debilitating complications that constitute an increasing burden on healthcare systems. A major complication is diabetic neuropathy, affecting up to 50% of diabetic patients [1]. Hyperglycaemia affects sensory afferents and autonomic and motor nerves [2], influencing unmyelinated [3] and myelinated [4] fibre function, leading to autonomic and motor dysfunction and altered proprioception and nociception. Nociceptors are rendered more sensitive to activation as a result of changes in the local microenvironment, such as compromised vascular perfusion and/or direct actions upon neurons. Sensory neurons are particularly susceptible to hyperglycaemic damage, due to a lack of insulin-regulated glucose uptake [5] and are also affected by the generation of reactive metabolites such as oxidative or glycosylated by-products under hyperglycaemic conditions [6]. Exposure to persistently high glucose levels in humans and in rodent models leads to classical signs of diabetic neuropathy, axonal atrophy and demyelination, reductions in conduction velocity and epidermal innervation and symptoms of allodynia (pain in response to normally non-painful stimuli), hyperalgesia (more severe pain in response to a normally painful stimulus), burning or ongoing pain. Sensory neuronal damage and loss may ultimately lead to hypoalgesia or anaesthesia [7]. A number of mechanisms have been proposed to underlie diabetes-induced neuropathology [2,5], including activation of transient receptor potential ankyrin 1 (TRPA1) channels [8], blockade of which attenuates both diabetes-induced peripheral neuronal loss and neuropathic pain [9].

Neuronal complications are attributed both to microvascular damage and to reduction in neural trophic support, for example both nerve growth factor (NGF) [10] and vascular endothelial growth factor-A (VEGF-A) [11] are reduced. VEGF-A165a is the archetypal pro-angiogenic factor that also has neuroprotective capacity in experimental diabetes [12]. The products of the VEGFA gene consist of two isoform families, VEGF-Axxxa and VEGF-Axxxb (xxx relates to amino acid number), generated by alternative pre-mRNA splicing that differ only in their terminal six amino acid sequences. Whereas each family's predominant isoform, VEGF-A165a and VEGF-A165b respectively, is neuroprotective for sensory neurons [13,14], VEGF-A165a is also pro-nociceptive [15]. In contrast, VEGF-A165b is anti-angiogenic and anti-nociceptive yet also neuroprotective, in vivo and in vitro, through alternative VEGFR2 activation and signalling mechanisms [13,1618]. In diabetic neuropathy, intra-epidermal nerve fibre loss and the severity of pain are related to the degree of decrease in VEGF and VEGFR2 expression, suggesting a relationship between VEGF-A and diabetic neuronal damage [19]. Therefore, we tested the hypothesis that the anti-angiogenic, neuroprotective and anti-nociceptive VEGF-A isoform VEGF-A165b protects against neuronal damage and pain in a model of Type 1 diabetes. Our results are the first to demonstrate that in this diabetes model, systemic delivery of recombinant human VEGF-A165b can reverse pain-like behaviour and prevent peripheral neuropathy.

MATERIALS AND METHODS

Experimental animals and procedures

Thirty-seven female Sprague-Dawley rats (∼250 g) were used in the present study in accordance with the UK Home Office Animals (Scientific Procedures) Act 1986 and amendments (2012). Experiments were performed in laboratories at the Universities of Bristol (behaviour and subsequent tissue analysis, primary culture) and Nottingham (Evans Blue, tissue analysis) and were approved by the Universities of Bristol and Nottingham AWERBs (Animal Welfare and Ethical Review Boards). Diabetes was induced with intraperitoneal (i.p.) injection of streptozotocin (STZ; 50 mg/kg) and one-third of an insulin pellet (Linshin) was implanted in the scruff of the neck using the supplied trocar (Linshin) under isoflurane anaesthesia (2–3% in O2) [20]. For the first 24 h, animals had ad libitum access to saturated sucrose solution, animals had ad libitum access to standard chow throughout, were regularly weighed, and housing and drinking water were regularly checked over the 9-week study. Experimental groups were diabetic animals with or without saline (i.p., n=14) or recombinant human VEGF-A165b (20 ng/g body weight, i.p. twice weekly from week 1, n=12) [13] and untreated naive animals (n=11). This twice weekly i.p. dosing regime has proven successful in the amelioration of diabetic nephropathy [21], traumatic nerve injury [16] and in vivo tumour models [22].

Nociceptive behavioural testing

Behavioural assessments were carried out on a total of 22 rats. Rats were habituated to the behavioural testing environments for 2 weeks (enclosures with metal mesh or Perspex floors) and for 10 min before the start of each session. The operator was blind to treatment throughout.

Mechanical withdrawal thresholds: von Frey (vF) monofilaments of increasing forces were applied to the plantar surface of the hind paws. Each monofilament was applied five times to generate a force–response (withdrawal) relationship and the 50% withdrawal threshold was calculated.

Cold nociception

A single drop of acetone was applied to the plantar surface of the hind paw five times with quantification of licking/shaking nociceptive behaviours [23] (positive response=1, no response=0, maximum=5).

Heat nociception

Withdrawal to heat was determined using the Hargreaves test [24]. The intensity of the radiant heat source was set and remained constant for the duration of the study. Each hind paw was stimulated three times and mean withdrawal latency was calculated. Mechanical and thermal nociceptive behaviours were tested weekly throughout the study. Response to a noxious chemical stimulus was determined by subcutaneous injection (dorsal surface of hind paw) of 0.5% formalin prior to diabetes induction [20] and at 2 and 7 weeks after diabetes onset. This is a low concentration of formalin that induces only a slight and brief acute phase (0–15 min) and no second phase (15–60 min) response in normal rats (Figure 2D) and does not induce prolonged sensitization.

Immunofluorescence and Toluidine Blue histology

After 7 weeks, animals were terminally anaesthetized (sodium pentobarbital 60 mg/kg i.p.) and transcardially perfused with either PBS followed by 4% paraformaldehyde (PFA) in PBS (pH 7.4) for immunofluorescence (n=3/group) or PBS followed by 2.5% glutaraldehyde in cacodylate-buffered saline (CBS, pH 7.4) for sciatic nerve myelinated fibre counts (n=3/group). L4 and L5 DRG (dorsal root ganglia) and hind paw plantar skin were removed, post-fixed overnight (4% PFA at 4°C) and transferred to 30% sucrose overnight at 4°C. Tissue was embedded, frozen and stored at −80°C until sectioning and processing. Cryosections were cut at 8 μm (DRG) or 20 μm (plantar skin), mounted on slides (Superfrost plus, VWR International), washed (PBS) and incubated in blocking solution (5% BSA and 10% FBS) for, 2 h at room temperature, followed by primary antibodies (see below) in blocking solution overnight at 4°C. Sections were washed with PBS, incubated in secondary antibodies in PBS+0.2% Triton X-100 at room temperature for 2 h, coverslipped in Vectorshield mounting medium (Vector Laboratories; +DAPI). Sciatic nerve was removed, post-fixed in 2.5% glutaraldehyde and CBS (pH 7.4) at 4°C until processed by washing in CBS, incubation in 1% osmium tetroxide in 0.1 M CBS and CBS wash. Samples were dehydrated, embedded in resin and semi-thick sections (1 μm) were stained with Toluidine Blue [25].

Primary antibodies and dilutions were: polyclonal rabbit anti-activated caspase 3 (AC3; 1: 500 dilution, New England Biolabs [NEB]); monoclonal mouse anti-NeuN (1: 200 dilution, Millipore); isolectin B4-biotin conjugate (IB4; 1: 500 dilution, Sigma–Aldrich); monoclonal mouse anti-neurofilament 200 (NF200; 1: 1000 dilution, Sigma–Aldrich); polyclonal rabbit anti-PGP9.5 (protein gene protein 9.5; 1: 200 dilution, Ultraclone); polyclonal rabbit anti-NSE-1 (neuron-specific enolase-1; 1: 1000 dilution, VWR International); monoclonal rabbit anti-VEGF receptor 2 (VEGFR2; Cell Signaling Technology 55B11, 1: 200 dilution); goat polyclonal anti-actin (Santa Cruz Biotechnology SC-1616, 2 μg/ml). Secondary antibodies for immunofluorescence were Alexa Fluor 488-conjugated chicken anti-mouse, Alexa Fluor 555-conjugated donkey anti-rabbit and streptavidin-conjugated anti-Alexa Fluor 488 (all Invitrogen). Controls included replacement of primary antibody with species- and concentration-matched IgG or pre-incubation with appropriate blocking peptide (activated/cleaved caspase III; NEB). Secondary antibodies for Western blotting were Licor IRDye 680RD-conjugated donkey anti-rabbit, 1: 5000 dilution, and donkey anti-goat conjugated to Licor IRDye 800CW, 1: 5000 dilution.

For DRG (n=3) and plantar skin (n=3) analyses, a minimum of five randomly selected non-serial images from at least five sections from each animal in each treatment group were used, at a magnification of ×20 (plantar skin) or ×10 (DRG). In skin, all PGP9.5-positive nerve fibres were counted per image area (35 mm2). In DRG, the mean number of neurons (identified by NeuN/neurofilament) counted per animal was ∼800. Myelinated axon counts were performed on a minimum of five random non-serial images per section, five sections per animal (a mean of ∼450 axons per animal).

Evans Blue extravasation and calculation of clearance

In vivo vascular dysfunction was evaluated using Evans Blue dye perfusion [26] in an additional 15 rats. Seven week diabetic+saline, diabetic+VEGF-A165b and age-matched controls (n=5 each) were terminally anaesthetized [ketamine/medetomidine T intravenous (i.v.) 50 mg/kg] and infused with Evans Blue dye i.v. (Sigma–Aldrich, 45 mg/kg. At 2 min post-infusion, 0.2 ml of arterial blood was withdrawn, followed by subsequent 0.1 ml of withdrawals every 15 min for 2 h. After 2 h of Evans Blue dye circulation, 0.2 ml of blood was withdrawn from the heart followed by cardiac perfusion of 50 ml of saline at 120 mmHg. DRG L3, L4 and L5, left saphenous nerve and left plantar skin were excised and weighed to determine their wet weight. Tissue was then dried at 70°C overnight and weighed to determine dry weight. To determine Evans Blue extravasation, dried tissue was incubated in 0.15 ml of formamide (Sigma–Aldrich) at 70°C overnight. Blood samples were centrifuged at 10625 g for 45 min at 4°C and the supernatant from tissue and blood samples were analysed at 620 nm. Evans Blue clearance was calculated as Evans Blue wet weight solute flux (μg/h/g) divided by time-averaged plasma Evans Blue concentration (μg/ml) multiplied by total time (h).

Primary dorsal root ganglion neuronal culture

Adult male Wistar rats were perfused transcardially with PBS under anaesthesia (sodium pentobarbital 60 mg/kg, i.p.) and T1-L6 DRG were extracted, enzymatically and mechanically dissociated and cultured on poly-L-lysine/laminin-coated glass coverslips in Ham's F12 plus 1× N2 supplement (Life Technologies), 0.3% BSA+1% penicillin/streptomycin. After cell attachment, mitotic cells were inhibited with 30 μg/ml 5-fluoro-2′-deoxyuridine. Cultures were pre-treated overnight (18 h) with 2.5 nM VEGF-A165b or vehicle as previously described [13], then incubated under either high glucose (final concentration 50 mM) or a basal neuronal culture glucose level (10 mM) [27] for 6 h. Proportions of AC3-positive DRG neurons (co-labelled NSE-1) were determined by immunofluorescence as previously described [13].

Epifluorescence imaging was performed on a Nikon Eclipse E400 microscope equipped with a ×40 objective lens and a Nikon DN100 camera. Confocal imaging was performed on Leica TCS SPE confocal microscope. Images were acquired using a ×63 oil immersion objective lens and Leica application suite. All imaging was performed at room temperature. Computer-aided analysis of images was performed using ImageJ (NIH; available at http://rsb.info.nih.gov/nih-image). In each instance, the analyser was blinded to treatment group.

Immortalized neuronal cell line (50B11) culture

TRPA1 channel activation was assayed in immortalized embryonic rat sensory neurons, 50B11, in line with the requirements of U.K. legislation and the 3Rs. 50B11 cells were maintained under previously published conditions [28]: Neurobasal medium (Invitrogen), supplemented with FBS (10%), glutamine (0.55 mM), B27 supplement and additional glucose (11 mM) making the total glucose concentration 36 mM (basal glucose). Incubation with lower, more physiological, glucose concentrations in vitro affects their neuronal phenotype, impairing normal neurite outgrowth (Bestall S.M., Hulse, R.P., Bates, D.O. and Donaldson, L.F., unpublished work). For these cells, this high glucose level (compared with in vivo) is a requirement for normal function and is therefore considered ‘basal’. 50B11 cells were plated into 96-well plates and incubated for 2 days, then differentiated with 75 μM forskolin maintained in the medium for at least 24 h, after which cells were incubated for 24 h with the following reagents: 2.5 nM VEGF-A165b, PBS vehicle, 200 nM PTK787 (VEGFR2 inhibitor, in 0.02% DMSO), 0.02% DMSO control, 8.05 μg/ml DC101 (rat monoclonal anti-mouse VEGFR2, BioXcell) or 8.05 μg/ml rat IgG control. In vitro high glucose conditions were maintained for 24 h, with addition of a further 30 mM glucose (final concentration 66 mM) to the medium or 30 mM mannitol to control for osmotic effects. After differentiation and incubation under experimental conditions, cells were loaded with Fluo-4 (Invitrogen) in Hanks balanced salt solution containing 20 mM HEPES and 2 mM CaCl2 for 1 h. The TRPA1 agonist allyl isothiocyanate (AITC, Sigma–Aldrich) in PBS was used to activate TRPA1 channels and calcium fluorescence was measured in a Wallac 1420 Victor 3 multi-plate reader (PerkinElmer) at 37°C. A TRPA1 antagonist, AP-18 (Tocris), was used to determine the specificity of AITC in this assay. Baseline readings were determined prior to stimulation by 100 μM AITC and after AITC application, sequential fluorescence emission values were recorded. The time of the first reading was ∼25 s after AITC application.

The effect of high glucose treatment or 300 μM AITC for 24 h (application twice at 12 h intervals) on AC3 in 50B11 cells was also determined. After 24 h of differentiation with forskolin, 50B11 cells were treated for 24 h, then fixed and stained for AC3, as detailed above. Total image AC3 fluorescence intensity was expressed over total DAPI fluorescence (total cell nuclei) intensity to account for differences in cell number.

Western blotting

The presence of VEGFR2 in 50B11 cells was confirmed with Western blotting. 50B11 cells grown in either normal (36 mM) or high (66 mM) glucose were lysed and protein extracted in the presence of PMSF and proteinase inhibitors. Protein lysate as electrophoresed on SDS/PAGE (10% gels), blotted on PVDF membrane by wet transfer and incubated for 24 h at 4°C with anti-VEGFR2 antibody in blocking buffer (Tris-buffered saline, 0.1% Tween 20 and 5% BSA). After secondary antibody incubation for 1 h at room temperature in blocking buffer, blots were imaged using a Licor imaging system. Protein loading differences between wells were controlled for by probing for VEGFR2 expression.

Data extraction and statistical analysis

Acquired data were processed and graphed using Microsoft Excel 2007 and/or GraphPad Prism v5.6. Data are shown as means ± S.E.M. unless otherwise stated. Three or more groups were compared using between subjects one- or two-way ANOVA with post-hoc Bonferroni tests and two groups were compared using Wilcoxon signed rank tests, as stated in the figure legends. Representative data values for representative outcome measures are given in Table 1. The effect sizes for treatment compared with diabetes alone shown in Table 1 were calculated from the means, S.D. and sample sizes given in the Table, using GPower (http://www.gpower.hhu.de/).

Table 1
Representative data for outcome measures shown in the text

*†P<0.05, **††P<0.01, ***†††P<0.001. *compared with control group, †compared with diabetic group. ns, not significant. Statistical significance levels shown are derived from analyses as described in the text.

ParameterNaive mean ± S.E.M. (n)Diabetic mean ± S.E.M. (n)Diabetic+VEGF-A165b mean ± S.E.M. (n)Effect size (Cohen's d) (effect of diabetes+treatment compared with diabetes alone)Relevant Figure
Glycaemia (mmol/l) 6.29±0.21 (6) 19.7±6.82** (9) 20.5±7.29 (7) 0.04 1A 
Weight (g) 248±4.0 (6) 243±1.6 (9) 241±1.7 (7) 0.4 1B 
Mechanical withdrawal threshold (g) at 7 weeks 17.1±0.79 (6) 13.1±0.78 (9) 18.0±1.0 (7) 2.2 1C 
Cold allodynia (number of withdrawals) at 5 weeks 0.42±0.23 (6) 1.28±0.28** (9) 0.07±0.07*** (7) 1.5 1D 
Heat hyperalgesia (withdrawal latency in seconds) at 7 weeks 10.7±0.71 (6) 7.34±0.3*** (9) 10.6±0.9ns (7) 3.8 1E 
Duration of response to formalin (s) at 2 weeks 48.8±7.5(6) 107.6±19.6* (9) 65.8±15.3ns (7) 0.75 1F 
AC-positive neurons (L5) (%) 18.2±4.0 (3) 48.3±1.2* (3) 32.29±3.55ns (3) 4.3 2D 
Neurons < 400 μm (%) 18.7±2.7 (3) 37.9±1.1*** (3) 20.3±8.9 (3) 1.96 2E 
AC3/IB4-positive neurons 29.5±1.3 (3) 50.8±2.4* (3) 29.8±4.45 (3) 4.15 3B 
AC3/NF200-positive neurons 29.4±3.2 (3) 45.3±1.64ns (3) 27.1±3.58 (3) 4.6 3D 
Intra-epidermal nerve fibres 2.34±0.17 (3) 1.40±0.14***(3) 2.09±0.20 (3) 2.83 4B 
Langerhans cell number 1.88±0.25 (3) 4.42±0.35*** (3) 2.85±0.25†† (3) 3.69 4C 
Myelinated axon cross sectional area (μm245.6±2.7 (3) 28.9±1.7*** (3) 44.6±3.1 (3) 4.43 5B 
Aberrant myelination (%) 2.07±0.18 (3) 6.32±0.45*** (3) 3.44±0.28†† (3) 5.4 5E 
DRG Evans Blue extravasation (nl/h/g) 3.3±1.1 (5) 14.0±3.0** (5) 7.8±0.8ns (5) 1.4 6A 
Saphenous Evans Blue extravasation (nl/h/g) 1.1±0.7 (5) 14.0±2.2* (5) 1.9±2.1 (5) 2.8 6B 
Skin Evans Blue extravasation (nl/h/g) 4.7±1.2 (5) 15.0±5.2ns (5) 1.4±1.4 (5) 1.79 6C 
ParameterNaive mean ± S.E.M. (n)Diabetic mean ± S.E.M. (n)Diabetic+VEGF-A165b mean ± S.E.M. (n)Effect size (Cohen's d) (effect of diabetes+treatment compared with diabetes alone)Relevant Figure
Glycaemia (mmol/l) 6.29±0.21 (6) 19.7±6.82** (9) 20.5±7.29 (7) 0.04 1A 
Weight (g) 248±4.0 (6) 243±1.6 (9) 241±1.7 (7) 0.4 1B 
Mechanical withdrawal threshold (g) at 7 weeks 17.1±0.79 (6) 13.1±0.78 (9) 18.0±1.0 (7) 2.2 1C 
Cold allodynia (number of withdrawals) at 5 weeks 0.42±0.23 (6) 1.28±0.28** (9) 0.07±0.07*** (7) 1.5 1D 
Heat hyperalgesia (withdrawal latency in seconds) at 7 weeks 10.7±0.71 (6) 7.34±0.3*** (9) 10.6±0.9ns (7) 3.8 1E 
Duration of response to formalin (s) at 2 weeks 48.8±7.5(6) 107.6±19.6* (9) 65.8±15.3ns (7) 0.75 1F 
AC-positive neurons (L5) (%) 18.2±4.0 (3) 48.3±1.2* (3) 32.29±3.55ns (3) 4.3 2D 
Neurons < 400 μm (%) 18.7±2.7 (3) 37.9±1.1*** (3) 20.3±8.9 (3) 1.96 2E 
AC3/IB4-positive neurons 29.5±1.3 (3) 50.8±2.4* (3) 29.8±4.45 (3) 4.15 3B 
AC3/NF200-positive neurons 29.4±3.2 (3) 45.3±1.64ns (3) 27.1±3.58 (3) 4.6 3D 
Intra-epidermal nerve fibres 2.34±0.17 (3) 1.40±0.14***(3) 2.09±0.20 (3) 2.83 4B 
Langerhans cell number 1.88±0.25 (3) 4.42±0.35*** (3) 2.85±0.25†† (3) 3.69 4C 
Myelinated axon cross sectional area (μm245.6±2.7 (3) 28.9±1.7*** (3) 44.6±3.1 (3) 4.43 5B 
Aberrant myelination (%) 2.07±0.18 (3) 6.32±0.45*** (3) 3.44±0.28†† (3) 5.4 5E 
DRG Evans Blue extravasation (nl/h/g) 3.3±1.1 (5) 14.0±3.0** (5) 7.8±0.8ns (5) 1.4 6A 
Saphenous Evans Blue extravasation (nl/h/g) 1.1±0.7 (5) 14.0±2.2* (5) 1.9±2.1 (5) 2.8 6B 
Skin Evans Blue extravasation (nl/h/g) 4.7±1.2 (5) 15.0±5.2ns (5) 1.4±1.4 (5) 1.79 6C 

RESULTS

Recombinant human VEGF-A165b prevented diabetic neuropathic pain in vivo

As previously reported [20], STZ-injection with low-dose insulin supplementation resulted in a maintained hyperglycaemia (Figure 1A) and body weight (Figure 1B) over the duration of the study. Also, as previously reported [20], STZ treatment with low-dose insulin supplementation resulted in a rapid (1–2 weeks) and significant lowering of mechanical-stimulus withdrawal thresholds (mechanical allodynia), maintained for the duration of the study (Figure 1C). In addition, diabetic animals developed significant cold allodynia (Figure 1D) and heat hyperalgesia (Figure 1E). Biweekly, systemic VEGF-A165b treatment significantly reversed the initial changes in mechanical allodynia (Figure 1C), prevented the onset of cold allodynia (Figure 1D) and reversed the heat hyperalgesia (Figure 1E). Chemical nociception in response to a low dose of formalin (0.5%) was enhanced in diabetic animals, measured at week 2 (Figure 1F). This enhanced acute chemo-nociception at week 2 was ameliorated by VEGF-A165b treatment (Figure 1F). The significant increase in chemo-nociceptive behaviour in diabetic rats was lost by week 7 (Figure 1F). No significant second phase behaviour (indicative of central sensitization) was observed in response to formalin in any animals at any time point.

VEGF-A165b reverses diabetic neuropathic pain behaviour in rats

Figure 1
VEGF-A165b reverses diabetic neuropathic pain behaviour in rats

(A) Blood glucose levels measured in the STZ-treated animals (STZ injection on day 0, open circles) compared with age matched naive (vehicle injected, closed circles) animals. (B) The body weights of the groups. Two experimental groups of diabetic (STZ) animals were treated with either vehicle or VEGF-A165b during week 1 (arrowhead) and treatment was continued biweekly for the duration of the experiment. STZ treatment led to neuropathic pain phenotypes in the diabetic+vehicle group, demonstrated in (C). Mechanical allodynia measured by vF hair withdrawal threshold. (D) Cold allodynia measured by withdrawal to acetone. (E) Heat hyperalgesia measured by latency of withdrawal. (F) Acute response to 0.5% formalin (0–15 min) in diabetic rats after 2 weeks. VEGF-A165b treatment attenuated diabetic neuropathic pain (CF) (two-way ANOVA+Bonferroni multiple comparisons test *P<0.05, **P<0.01, ***P<0.001 compared with diabetic group, naive n=6, diabetic n=9, diabetic+VEGF-A165b n=7). Note that the same symbols are used in all graphs for the different groups of animals.

Figure 1
VEGF-A165b reverses diabetic neuropathic pain behaviour in rats

(A) Blood glucose levels measured in the STZ-treated animals (STZ injection on day 0, open circles) compared with age matched naive (vehicle injected, closed circles) animals. (B) The body weights of the groups. Two experimental groups of diabetic (STZ) animals were treated with either vehicle or VEGF-A165b during week 1 (arrowhead) and treatment was continued biweekly for the duration of the experiment. STZ treatment led to neuropathic pain phenotypes in the diabetic+vehicle group, demonstrated in (C). Mechanical allodynia measured by vF hair withdrawal threshold. (D) Cold allodynia measured by withdrawal to acetone. (E) Heat hyperalgesia measured by latency of withdrawal. (F) Acute response to 0.5% formalin (0–15 min) in diabetic rats after 2 weeks. VEGF-A165b treatment attenuated diabetic neuropathic pain (CF) (two-way ANOVA+Bonferroni multiple comparisons test *P<0.05, **P<0.01, ***P<0.001 compared with diabetic group, naive n=6, diabetic n=9, diabetic+VEGF-A165b n=7). Note that the same symbols are used in all graphs for the different groups of animals.

Recombinant human VEGF-A165b protected sensory neurons against hyperglycaemic damage in vivo and high glucose treatment in vitro

Incubation of adult rat primary DRG neurons under increased glucose conditions for 6 h resulted in an increase of the percentage of neurons expressing AC3 (Figures 2A and 2B; basal glucose=26.3±3.5%, high glucose=53.0±6.1%), indicating neuronal stress under these conditions, which was blocked by recombinant human VEGF-A165b (Figure 2B; high glucose ± VEGF-A165b=35.0±2.3%). VEGF-A165b alone had no effect on expression of AC3. We then determined whether VEGF-A165b could protect sensory neurons against damage in vivo in the rodent model of Type I diabetic neuropathy in which VEGF-A165b had inhibited pain responses (Figure 1). Figures 2(C) and 2(D) show that the number of AC3-expressing L5 (Figures 2C–2F) and L4 (results not shown) DRG neurons was increased in diabetic rats in vivo compared with naive and this was reduced by VEGF-A165b treatment, predominantly in the small diameter population (Figure 2E). Furthermore, there was an increase in the total number of smaller DRG neurons (<400 μm2 PGP9.5-positive DRG neurons) within the diabetic animals compared with age-matched controls (Figure 2F). VEGF-A165b treatment prevented this increase in the number of smaller DRG neurons. Co-staining demonstrated that there was increased AC3 staining in nociceptive small IB4-reactive neurons (Figures 3A and 3B) and also in myelinated neurons (NF200-positive, Figures 3C and 3D). VEGF-A165b treatment reversed the increased AC3 expression in both neuronal groups (Figures 3B and 3D). A reduction in epidermal innervation is a diagnostic hallmark of diabetic neuropathy [29]. Seven weeks after STZ injection, there were fewer PGP9.5-positive nerve terminals at the dermal/epidermal border and entering the epidermis in hind paw plantar skin (Figure 4A, arrows), which was reversed with VEGF-A165b treatment at week 7 (Figure 4B). Langerhans cells, inflammatory cells reported to increase in number in diabetic patients [30], were also increased in the epidermis of diabetic rats at week 7 (Figure 4A, arrowheads) and this was also significantly reduced by VEGF-A165b treatment (Figure 4C).

VEGF-A165b is neuroprotective under high glucose conditions

Figure 2
VEGF-A165b is neuroprotective under high glucose conditions

(A) Primary adult DRG neurons stained for NSE-1 (green) and AC3 (red) under different experimental conditions. (B) The percentage of AC3-expressing cells was increased by high glucose treatment and blocked by 18 h of 2.5 nM VEGF-A165b treatment (n=4, scale bars=25 μm). (C) Images of L5 DRG expressing AC3 from naive, diabetic (D and E, scale bar=50 μm), diabetic+VEGF-A165b and negative controls. (D) AC3 expression significantly increased in diabetic animals (n=3), but not in VEGF-A165b-treated diabetic animals. (E) AC3 expression by neuronal size. AC3 expression was increased in small/medium L5 DRG neuronal cell bodies (<400 μm2, 401–800 μm2) in diabetic animals (n=3) and was partially ameliorated by VEGF-A165b. (F) The percentage of total neuronal cell bodies by size with an area less than 400 μm2 was increased in diabetic animals compared with naive. VEGF-A165b treatment prevented this effect of diabetes on DRG neuronal cell body cross-sectional area (two-way ANOVA+Bonferroni multiple comparisons *P<0.05, **P<0.01, ***P<0.001).

Figure 2
VEGF-A165b is neuroprotective under high glucose conditions

(A) Primary adult DRG neurons stained for NSE-1 (green) and AC3 (red) under different experimental conditions. (B) The percentage of AC3-expressing cells was increased by high glucose treatment and blocked by 18 h of 2.5 nM VEGF-A165b treatment (n=4, scale bars=25 μm). (C) Images of L5 DRG expressing AC3 from naive, diabetic (D and E, scale bar=50 μm), diabetic+VEGF-A165b and negative controls. (D) AC3 expression significantly increased in diabetic animals (n=3), but not in VEGF-A165b-treated diabetic animals. (E) AC3 expression by neuronal size. AC3 expression was increased in small/medium L5 DRG neuronal cell bodies (<400 μm2, 401–800 μm2) in diabetic animals (n=3) and was partially ameliorated by VEGF-A165b. (F) The percentage of total neuronal cell bodies by size with an area less than 400 μm2 was increased in diabetic animals compared with naive. VEGF-A165b treatment prevented this effect of diabetes on DRG neuronal cell body cross-sectional area (two-way ANOVA+Bonferroni multiple comparisons *P<0.05, **P<0.01, ***P<0.001).

VEGF-A165b reduces markers of cell stress in sub-populations of DRG neurons

Figure 3
VEGF-A165b reduces markers of cell stress in sub-populations of DRG neurons

(A) AC3 expression in IB4-positive neurons. (B) The number of IB4-positive neurons expressing AC3 was increased in diabetic and reduced in VEGF-A165b-treated rats. (C) AC3 expression in neurofilament (NF200)-positive DRG subsets. (D) This was increased in diabetic compared with naive and was reduced in diabetic+VEGF-A165b (n=3/group, Kruskal–Wallis test *P<0.05).

Figure 3
VEGF-A165b reduces markers of cell stress in sub-populations of DRG neurons

(A) AC3 expression in IB4-positive neurons. (B) The number of IB4-positive neurons expressing AC3 was increased in diabetic and reduced in VEGF-A165b-treated rats. (C) AC3 expression in neurofilament (NF200)-positive DRG subsets. (D) This was increased in diabetic compared with naive and was reduced in diabetic+VEGF-A165b (n=3/group, Kruskal–Wallis test *P<0.05).

Systemic VEGF-A165b treatment protects against intra-epidermal neuronal loss in diabetes

Figure 4
Systemic VEGF-A165b treatment protects against intra-epidermal neuronal loss in diabetes

(A) Representative images of epidermal/dermal (dashed line) nerve innervation (arrows) and Langerhans cells (arrowheads) stained with PGP9.5 in plantar skin from naive, diabetic and VEGF-A165b-treated diabetic rats. Nerves and Langerhans cells were quantified in the same sections. (B). There was a lower dermal/epidermal border innervation density in diabetic animals that was reversed by VEGF-A165b treatment. (C) Diabetic animals had increased intra-epidermal Langerhans cell numbers, which was prevented by VEGF-A165b treatment. (n=3, Kruskal–Wallis test *P<0.05, **P<0.01, ***P<0.001. Scale bar=50 μm.)

Figure 4
Systemic VEGF-A165b treatment protects against intra-epidermal neuronal loss in diabetes

(A) Representative images of epidermal/dermal (dashed line) nerve innervation (arrows) and Langerhans cells (arrowheads) stained with PGP9.5 in plantar skin from naive, diabetic and VEGF-A165b-treated diabetic rats. Nerves and Langerhans cells were quantified in the same sections. (B). There was a lower dermal/epidermal border innervation density in diabetic animals that was reversed by VEGF-A165b treatment. (C) Diabetic animals had increased intra-epidermal Langerhans cell numbers, which was prevented by VEGF-A165b treatment. (n=3, Kruskal–Wallis test *P<0.05, **P<0.01, ***P<0.001. Scale bar=50 μm.)

Morphological changes, including axon diameter and myelin structure of sciatic nerve, have been reported in SOD1−/− and db/db mice [25] and are thought to result from hyperglycaemic damage of the neuronal myelin through either structural or compositional alterations [25]. In diabetic rats, we found similar changes (Figure 5A), in that there was an overall reduction in myelinated axon cross-sectional area in the sciatic nerve (Figure 5B) compared with naïve animals and this was reversed by VEGF-A165b. The reduction in the number of ‘medium-sized’ fibres with cross-sectional areas <90 μm2 but >30 μm2 [and an increase in the number of small myelinated fibres with areas <30 μm2 (Figure 5C)], may reflect a reduction in larger axon size resulting in an apparent increase in the number of smaller myelinated fibres. Again, as previously observed [25], there was an increase in the number of axons with gross indications of aberrant myelin morphology in diabetic animals (Figures 5D and 5E; arrows in Figure 5A), which was reversed by VEGF-A165b (Figure 5E). Furthermore, NF200-positive neuronal cell bodies in L5 DRG had significantly smaller cross-sectional areas in the diabetic group compared with the naive age-matched control group (Figures 5F and 5G). VEGF-A165b treatment prevented this reduction in size of the neurofilament sensory neuronal population.

Systemic VEGF-A165b treatment protects myelinated sensory afferents against damage in diabetes

Figure 5
Systemic VEGF-A165b treatment protects myelinated sensory afferents against damage in diabetes

(AC) Sciatic nerves from naive, diabetic and VEGF-A165b-treated animals were stained with Toluidine Blue. (B) Myelinated fibre cross-sectional area was reduced in diabetes, which was prevented by long-term systemic administration of VEGF-A165b (n=3/group, scale bar=50 μm). (C) Myelinated nerve fibre numbers broken down by cross-sectional area. (D) Higher power image of axons demonstrating aberrant myelin morphology (* in A). (E) The percentage of aberrant axons was increased in diabetic animals and inhibited by VEGF-A165b treatment. Scale bars=20 μm, 10 μm. (F) NF200 staining of DRG neurons in normal, diabetic and VEGF-A165b-treated animals. (G) The cross-sectional area profile of NF200-positive DRG neuronal cell bodies was significantly different in L5 DRG of diabetic animals compared with naive, a difference that was not observed with VEGF-A165b treatment (Kruskal–Wallis test and two-way ANOVA with Bonferroni multiple comparisons test were performed. **P<0.01, ***P<0.001).

Figure 5
Systemic VEGF-A165b treatment protects myelinated sensory afferents against damage in diabetes

(AC) Sciatic nerves from naive, diabetic and VEGF-A165b-treated animals were stained with Toluidine Blue. (B) Myelinated fibre cross-sectional area was reduced in diabetes, which was prevented by long-term systemic administration of VEGF-A165b (n=3/group, scale bar=50 μm). (C) Myelinated nerve fibre numbers broken down by cross-sectional area. (D) Higher power image of axons demonstrating aberrant myelin morphology (* in A). (E) The percentage of aberrant axons was increased in diabetic animals and inhibited by VEGF-A165b treatment. Scale bars=20 μm, 10 μm. (F) NF200 staining of DRG neurons in normal, diabetic and VEGF-A165b-treated animals. (G) The cross-sectional area profile of NF200-positive DRG neuronal cell bodies was significantly different in L5 DRG of diabetic animals compared with naive, a difference that was not observed with VEGF-A165b treatment (Kruskal–Wallis test and two-way ANOVA with Bonferroni multiple comparisons test were performed. **P<0.01, ***P<0.001).

VEGF-A165b prevented STZ-induced increase in Evans Blue clearance in DRG, the saphenous nerve and hind paw plantar skin

There was a significant increase in Evans Blue clearance in DRG (Figure 6A) and saphenous nerve (Figure 6B) in diabetic animals and with a trend for an increase in the plantar hind paw skin (Figure 6C) compared with naive animals. Biweekly treatment with VEGF-A165b significantly reduced the clearance in these tissues compared with vehicle-treated animals (Figures 6A–6C).

Evans Blue clearance was increased in diabetic nerves and reduced by VEGF-A165b

Figure 6
Evans Blue clearance was increased in diabetic nerves and reduced by VEGF-A165b

(A) Animals were perfused with Evans Blue for 2 h and the clearance (tissue dye as a proportion of plasma dye concentration, per hour per gram of tissue) quantified by spectroscopy. STZ increased Evans Blue clearance (n=5) that increased in diabetes and reduced by long-term systemic VEGF-A165b treatment in (A) L3, L4 and L5 DRG, (B) saphenous nerve and (C) plantar skin (n=5, Kruskal–Wallis test *P<0.05 **P<0.01).

Figure 6
Evans Blue clearance was increased in diabetic nerves and reduced by VEGF-A165b

(A) Animals were perfused with Evans Blue for 2 h and the clearance (tissue dye as a proportion of plasma dye concentration, per hour per gram of tissue) quantified by spectroscopy. STZ increased Evans Blue clearance (n=5) that increased in diabetes and reduced by long-term systemic VEGF-A165b treatment in (A) L3, L4 and L5 DRG, (B) saphenous nerve and (C) plantar skin (n=5, Kruskal–Wallis test *P<0.05 **P<0.01).

TRPA1-mediated calcium fluorescence was affected by VEGF-A165b in vitro

Diabetic neuronal terminal loss and pain are suggested to occur, at least in part, through sustained activation of C-fibre nociceptors [3], particularly those expressing the non-selective cation channel TRPA1 [9]. Immortalized rat sensory neurons (50B11 [28]) express nociceptive channels such as functional TRPV1 (transient receptor potential vanilloid 1), P2X [28], TRPA1 (Figure 7A) and also receptors for VEGF (Figure 7B). VEGFR2 expression was not changed under high glucose conditions (results not shown). The TRPA1 agonist AITC evoked a concentration-dependent increase in intracellular calcium levels in 50B11 cells maintained under basal glucose concentrations [28] (Figure 7A). AITC has been shown to activate TRPV1 as well as TRPA1 [31]. The TRPA1 antagonist AP-18 significantly blocked the calcium response to 100 μM AITC (Figures 7B and 7C), indicating that in this assay AITC stimulated exclusively TRPA1 to increase intra cellular calcium. The response to 100 μM AITC was significantly inhibited by pre-treatment with VEGF-A165b for 24 h (Figures 7D and 7E). AITC treatment (300 μM, 24 h) also caused a significant increase in AC3 fluorescence intensity measured in 50B11 cells (Figures 7F and 7G).

Cultured differentiated 50B11 immortalized DRG neurons were loaded with the calcium indicator dye Fluo4, treated with the TRPA1 agonist AITC and fluorescence intensity measured and normalized to baseline

Figure 7
Cultured differentiated 50B11 immortalized DRG neurons were loaded with the calcium indicator dye Fluo4, treated with the TRPA1 agonist AITC and fluorescence intensity measured and normalized to baseline

(A) AITC evoked a concentration-dependent increase in intracellular calcium (n=5). Cells were pre-treated with the TRPA1 antagonist AP-18 (100 μM). (B) 50B11 cells express VEGFR2, which is unaltered by glucose concentration. (C and D) The response to 100 μM AITC was blocked by AP-18. (E and F). Cells were pre-treated with 2.5 nM VEGF-A165b for 24 h and the calcium response to 100 μM AITC measured. The TRPA1 calcium increase was significantly reduced by overnight pre-treatment with VEGF-A165b (n=8). (G) 50B11 cells were treated for 24 h with 300 μM AITC and stained for AC3. (H) AC3 fluorescence intensity was increased by AITC. (A and CF) Two-way ANOVA+Bonferroni multiple comparisons test, Kruskal–Wallis test and (F) Mann–Whitney test *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

Figure 7
Cultured differentiated 50B11 immortalized DRG neurons were loaded with the calcium indicator dye Fluo4, treated with the TRPA1 agonist AITC and fluorescence intensity measured and normalized to baseline

(A) AITC evoked a concentration-dependent increase in intracellular calcium (n=5). Cells were pre-treated with the TRPA1 antagonist AP-18 (100 μM). (B) 50B11 cells express VEGFR2, which is unaltered by glucose concentration. (C and D) The response to 100 μM AITC was blocked by AP-18. (E and F). Cells were pre-treated with 2.5 nM VEGF-A165b for 24 h and the calcium response to 100 μM AITC measured. The TRPA1 calcium increase was significantly reduced by overnight pre-treatment with VEGF-A165b (n=8). (G) 50B11 cells were treated for 24 h with 300 μM AITC and stained for AC3. (H) AC3 fluorescence intensity was increased by AITC. (A and CF) Two-way ANOVA+Bonferroni multiple comparisons test, Kruskal–Wallis test and (F) Mann–Whitney test *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

Exposure of 50B11 cells to high glucose treatment caused a significant increase in AC3 fluorescence intensity (Figures 8A and 8B). High glucose also enhanced the level of intracellular calcium in response to 100 μM AITC stimulation indicating either altered channel kinetics or altered calcium extrusion under high glucose treatment, an effect that was not attributable to an osmotic effect of the additional glucose (Figures 8C and 8D). Treatment with VEGF-A165b significantly inhibited AITC-evoked calcium fluorescence under high glucose conditions (Figures 8E and 8F). Under high glucose conditions, the effect of VEGF-A165b on the calcium response was blocked by a VEGFR2 tyrosine kinase inhibitor (PTK787, 200 nM; Figure 8G) and a VEGFR2 receptor-neutralizing antibody (DC101; Figure 8H).

VEGF-A165b reverses TRPA1 sensitisation in immortalised sensory neurons caused by high glucose

Figure 8
VEGF-A165b reverses TRPA1 sensitisation in immortalised sensory neurons caused by high glucose

(A) Cultured, differentiated 50B11 immortalized DRG neurons were treated with high glucose for 24 h and AC3 expression determined by immunofluorescence. (B) High glucose increased AC3 expression. (C) Cells were treated with AITC after incubation in basal glucose medium or high glucose and calcium measured by Fluo4 fluorescence intensity. (D) High glucose enhanced the sustained increase in intracellular calcium in the presence of AITC, compared with basal glucose, which was not attributable to an osmotic effect, as mannitol had no effect (n=10). (E) Cells were pre-treated with high glucose, with or without VEGF-A165b or inhibitors and calcium measured in response to AITC. (F) VEGF-A165b blocked the increase in calcium induced by AITC under high glucose conditions (n=9) and this was inhibited by (G) the VEGF receptor antagonist PTK787 at 200 nM (n=13) and (H) the specific VEGFR2-neutralizing antibody DC101, 8.05 μg/ml (n=7). All experimental culture conditions are displayed on the x-axes (two-way ANOVA+Bonferroni multiple comparisons test, Kruskal–Wallis test and Mann–Whitney tests *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001).

Figure 8
VEGF-A165b reverses TRPA1 sensitisation in immortalised sensory neurons caused by high glucose

(A) Cultured, differentiated 50B11 immortalized DRG neurons were treated with high glucose for 24 h and AC3 expression determined by immunofluorescence. (B) High glucose increased AC3 expression. (C) Cells were treated with AITC after incubation in basal glucose medium or high glucose and calcium measured by Fluo4 fluorescence intensity. (D) High glucose enhanced the sustained increase in intracellular calcium in the presence of AITC, compared with basal glucose, which was not attributable to an osmotic effect, as mannitol had no effect (n=10). (E) Cells were pre-treated with high glucose, with or without VEGF-A165b or inhibitors and calcium measured in response to AITC. (F) VEGF-A165b blocked the increase in calcium induced by AITC under high glucose conditions (n=9) and this was inhibited by (G) the VEGF receptor antagonist PTK787 at 200 nM (n=13) and (H) the specific VEGFR2-neutralizing antibody DC101, 8.05 μg/ml (n=7). All experimental culture conditions are displayed on the x-axes (two-way ANOVA+Bonferroni multiple comparisons test, Kruskal–Wallis test and Mann–Whitney tests *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001).

DISCUSSION

Diabetic patients complain of abnormal sensations, including allodynia and hyperalgesia. Diabetic neuropathy is also associated with loss of sensation as peripheral sensory nerve terminals are lost from the dermis and epidermis. Classically these symptoms occur in the extremities, e.g. toes, affecting more proximal regions as the disease progresses. Thus diabetic neuropathy is classically characterized by peripheral nerve fibre degeneration in both myelinated and unmyelinated sensory nerves with abnormal pain perception. STZ-induced hyperglycaemia leads to a pronounced neuropathic pain phenotype in the rat [20] and high levels of blood glucose lead to the development of mechanical and cold allodynia and heat hyperalgesia. Heightened pain responses may arise as a result of damage to the peripheral nervous system from direct actions of glucose [5], metabolite production [9], inflammation [32] or microvascular damage and other associated micro-environmental changes [33]. The observed increase in small myelinated fibre profiles in sciatic nerve trunk may represent large fibre neuropathy, with demyelination of larger fibres [25], and hence smaller apparent cross-sectional area both in the nerve fibre and somata. Myelin damage could affect sensory fibres or motor axons, although hyperglycaemia has been shown to affect sensory fibres to a greater extent than motor axons [34]. AC3 co-expression with NF200 in DRG neurons is also consistent with this interpretation. Concurrent small fibre neuropathy in these animals is also indicated by the increased epidermal recruitment of Langerhans cells [30] and AC3 expression in the smaller-diameter neurons. Langerhans cells are dendritic cells that increase their expression of PGP9.5 following denervation and inflammation [35]. Changes in Langerhans cell number could be attributable to peripheral denervation and/or altered capillary permeability and plasma/albumin extravasation resulting from local inflammatory responses.

Both endoneurial capillary permeability and oedema can be seen in models of STZ-induced diabetes [36]. After 7 weeks of hyperglycaemia, the significantly increased Evans Blue clearance in sciatic nerve and DRG are suggestive of an increase in neural perfusion and/or capillary permeability that could be attributable to hyperglycaemic microvascular damage. Although sensory neurons are particularly susceptible to hyperglycaemic injury [37] as they lack insulin-regulated glucose uptake [5], local hypoxia as a consequence of ischaemia resulting from damage to the vasa nervosa may also contribute to peripheral neuropathy [38].

A number of diabetic complications are associated with increased (e.g. nephropathy, retinopathy) or impaired angiogenesis (e.g. neuropathy, wound healing) [39]. Total VEGF plasma levels are raised in symptomatic diabetic neuropathy [40,41]. In contrast, neuronal expression of total-VEGF-A is reduced in sensory neurons, in diabetic rats [11] and in epidermis in diabetic patients, where it is associated with more severe neuropathy [19]. The principal approach used to increase neuronal access to VEGF-A in different neuropathies has been to enhance local VEGF-A isoform transcription/production through gene therapy approaches, such as intramuscular injection of plasmid or viral vector containing VEGF-A cDNA, approaches that would increase either VEGF-A165a alone or all isoforms of VEGF. This has been effective in animal models of diabetic [11,4245] neuropathy and has had some success in the small number of early clinical trials to date [46,47]. Increasing total VEGF-A transcription with a zinc-finger protein transcription factor showed early preclinical promise [11,44], but unfortunately failed in Phase II clinical trial [48], suggesting that a more targeted approach may be necessary. The use of recombinant human VEGF-A, although also successful in animal models, had limited success in humans, largely as a result of short half-life and off-target effects at high doses [49]. Alternative therapeutic strategies that might better control the balance of isoforms, such as modulation of alternative splicing, are under consideration, as current strategies where only VEGF-Axxxa is replaced could result in increased angiogenesis and pain [16].

VEGF-Axxxb is a potent neuroprotective factor for neurons including DRG [13], as is the alternatively spliced variant VEGF-Axxxa, in chemotherapeutic, diabetic and traumatic neuropathies [12,14]. VEGF-Axxxa and VEGF-Axxxb signal through different downstream pathways on VEGFR2 activation, although both isoform families can act through ERK1/2 (extracellular signal-related kinase 1/2) to exert neuroprotective actions [13]. A reduction in expression of either VEGF-A isoform family will therefore affect the potential for neuroprotection and/or regeneration. Systemic treatment with VEGF-A165b protein in vivo was able to prevent both neuropathic pain and the neuronal changes seen in diabetes that lead to neuropathy possibly by restoring some of the lost trophic support.

TRPA1 is a non-specific cation channel that has been implicated in the neuronal loss seen in diabetic [9] and chemotherapy-induced neuropathy [50]. Blockade of TRPA1 activity attenuates both diabetic neuropathic pain and the associated neuronal terminal loss [9]. TRPA1 is activated by the low formalin concentrations used in the present study [51], responses that are sensitized at week 2 and attenuated by VEGF-A165b treatment. Importantly, TRPA1 has been localized to the IB4 nociceptor subset, neurons that are readily sensitized, are responsible for ‘hyperalgesic priming’ [52] and, as we show, are affected by diabetic neuropathy. Furthermore, TRPA1 sensitization/activation in sensory nerve terminals can result in neurogenic inflammation and increased cutaneous plasma extravasation [53], thus suppression of sensory neuronal sensitization through TRPA1 by VEGF-A165b, acting through VEGFR2 [16], may also underlie the observed effect on plasma extravasation in the skin. VEGF-A165b can antagonize the vasodilatatory effects of VEGF-Axxxa [54] so the observed effect of VEGF-A165b on Evans Blue extravasation may also be due, at least in part, to direct actions on the vascular tone.

To explore whether VEGF-A165b could exert some of its actions through TRPA1 we used an in vitro assay of TRPA1 sensitization. Sensory neurons compromised by a peripheral nerve injury [indicated by expression neuronal stress/death/degeneration markers (e.g. AC3)], undergo changes in excitability that underpin neuropathic pain [55]. TRP channels including TRPA1 contribute to peripheral C-fibre nociceptor sensitization in a range of neuropathic [50] and inflammatory conditions [56]. Although the exact function of TRPA1 is still debated, it can be considered as a ‘universal sensitizer’ channel in peripheral nociceptors, contributing to mechanical [56], cold [57] and chemical pain, especially that generated by endogenous metabolites generated in hyperglycaemic conditions [9,58].

STZ-induced hyperglycaemia causes an increase in endoneural levels of the toxic glucose metabolite methylglyoxal and methylglyoxal-glycated proteins (advanced glycated end products) [6]. Methylglyoxal is known to induce sustained TRPA1 activity under hyperglycaemic conditions [9,58] and its production may produce the observed sustained TRPA1 activation following high glucose treatment. VEGF-A165b-induced blockade of TRPA1-mediated calcium entry/handling in these high glucose conditions in vitro could therefore reduce both neuronal damage resulting from enhanced TRPA1-mediated calcium entry and the mechanical and thermal pain resulting from TRPA1 sensitization of peripheral nerve terminals.

The present study demonstrates that exogenous recombinant VEGF-A165b protein delivered systemically can reverse signs of diabetic neuropathy in the rat. VEGF-A165b exerted neuroprotective and anti-nociceptive actions on peripheral sensory neurons, possibly through a TRPA1-mediated mechanism. It is therefore important to determine whether control of VEGF-A165b levels, either through exogenous protein therapy or control of alternative RNA splicing mechanisms [59] represents a possible novel therapeutic strategy for the treatment of multiple diabetic complications, including diabetic neuropathy.

AUTHOR CONTRIBUTION

Richard Hulse, Nicholas Beazley-Long, Nikita Ved, Samuel Bestall, Hamza Riaz, Priya Singhal, and Kurt Ballmer Hofer performed research. Richard Hulse, Nicholas Beazley-Long, Nikita Ved, Steve Harper, David Bates and Lucy Donaldson designed the research and analysed data. Richard Hulse, Nicholas Beazley-Long, David Bates and Lucy Donaldson wrote the manuscript with contributions from Steve Harper and final approval from all authors.

We thank Ahmet Hoke, John's Hopkins University, Baltimore, MD, U.S.A., and Damon Lowes, University of Aberdeen, Aberdeen, U.K., for the gift of the 50B11 cell line.

FUNDING

This work was supported by Diabetes UK [grant number 11/0004192 (to L.F.D. and D.O.B.)]; the Wellcome Trust [grant number 079736 (to D.O.B.)]; the Swiss National Science Foundation [grant number 31003A (to K.B-H.)]; the Oncosuisse [grant number OC201200-08-2007 (to K.B-H)]; and The Richard Bright VEGF Research Trust [grant number 1095785 (to D.O.B.)].

DECLARATIONS OF INTEREST

Lucy F. Donaldson, David O. Bates and Steve J. Harper are co-inventors on patents protecting VEGF-A165b and alternative RNA splicing control for therapeutic application in a number of different conditions. Lucy F. Donaldson, David O. Bates and Steve J. Harper are founder equity holders in, and David O. Bates and Steve J. Harper are directors of, Exonate Ltd, a company with a focus on development of alternative RNA splicing control for therapeutic application in a number of different conditions, including diabetic complications.

Abbreviations

     
  • AC3

    activated caspase 3

  •  
  • AITC

    allyl isothiocyanate

  •  
  • CBS

    cacodylate-buffered saline

  •  
  • DRG

    dorsal root ganglia

  •  
  • IB4

    isolectin

  •  
  • i.p.

    intraperitoneal

  •  
  • i.v.

    intravenous

  •  
  • NF200

    neurofilament 200

  •  
  • NSE-1

    neuron-specific enolase-1

  •  
  • PFA

    paraformaldehyde

  •  
  • PGP9.5

    protein gene protein 9.5

  •  
  • STZ

    streptozotocin

  •  
  • TRPA1

    transient receptor potential ankyrin 1

  •  
  • VEGF

    vascular endothelial growth factor

  •  
  • VEGFR2

    VEGF receptor 2

  •  
  • vF

    von Frey

References

References
1
Obrosova
 
I.G.
 
Diabetic painful and insensate neuropathy: pathogenesis and potential treatments
Neurotherapeutics
2009
, vol. 
6
 (pg. 
638
-
647
)
[PubMed]
2
Tomlinson
 
D.R.
Gardiner
 
N.J.
 
Diabetic neuropathies: components of etiology
J. Peripher. Nerv. Syst.
2008
, vol. 
13
 (pg. 
112
-
121
)
[PubMed]
3
Chen
 
X.
Levine
 
J.D.
 
Altered temporal pattern of mechanically evoked C-fiber activity in a model of diabetic neuropathy in the rat
Neuroscience
2003
, vol. 
121
 (pg. 
1007
-
1015
)
[PubMed]
4
Christianson
 
J.A.
Ryals
 
J.M.
Johnson
 
M.S.
Dobrowsky
 
R.T.
Wright
 
D.E.
 
Neurotrophic modulation of myelinated cutaneous innervation and mechanical sensory loss in diabetic mice
Neuroscience
2007
, vol. 
145
 (pg. 
303
-
313
)
[PubMed]
5
Tomlinson
 
D.R.
Gardiner
 
N.J.
 
Glucose neurotoxicity
Nat. Rev. Neurosci.
2008
, vol. 
9
 (pg. 
36
-
45
)
[PubMed]
6
Duran-Jimenez
 
B.
Dobler
 
D.
Moffatt
 
S.
Rabbani
 
N.
Streuli
 
C.H.
Thornalley
 
P.J.
Tomlinson
 
D.R.
Gardiner
 
N.J.
 
Advanced glycation end products in extracellular matrix proteins contribute to the failure of sensory nerve regeneration in diabetes
Diabetes
2009
, vol. 
58
 (pg. 
2893
-
2903
)
[PubMed]
7
Calcutt
 
N.A.
Jolivalt
 
C.G.
Fernyhough
 
P.
 
Growth factors as therapeutics for diabetic neuropathy
Curr. Drug Targets
2008
, vol. 
9
 (pg. 
47
-
59
)
[PubMed]
8
Wei
 
H.
Hamalainen
 
M.M.
Saarnilehto
 
M.
Koivisto
 
A.
Pertovaara
 
A.
 
Attenuation of mechanical hypersensitivity by an antagonist of the TRPA1 ion channel in diabetic animals
Anesthesiology
2009
, vol. 
111
 (pg. 
147
-
154
)
[PubMed]
9
Koivisto
 
A.
Hukkanen
 
M.
Saarnilehto
 
M.
Chapman
 
H.
Kuokkanen
 
K.
Wei
 
H.
Viisanen
 
H.
Akerman
 
K.E.
Lindstedt
 
K.
Pertovaara
 
A.
 
Inhibiting TRPA1 ion channel reduces loss of cutaneous nerve fiber function in diabetic animals: sustained activation of the TRPA1 channel contributes to the pathogenesis of peripheral diabetic neuropathy
Pharmacol. Res.
2012
, vol. 
65
 (pg. 
149
-
158
)
[PubMed]
10
Kamiya
 
H.
Murakawa
 
Y.
Zhang
 
W.
Sima
 
A.A.
 
Unmyelinated fiber sensory neuropathy differs in type 1 and type 2 diabetes
Diabetes Metab. Res. Rev.
2005
, vol. 
21
 (pg. 
448
-
458
)
[PubMed]
11
Pawson
 
E.J.
Duran-Jimenez
 
B.
Surosky
 
R.
Brooke
 
H.E.
Spratt
 
S.K.
Tomlinson
 
D.R.
Gardiner
 
N.J.
 
Engineered zinc finger protein-mediated VEGF-a activation restores deficient VEGF-a in sensory neurons in experimental diabetes
Diabetes
2010
, vol. 
59
 (pg. 
509
-
518
)
[PubMed]
12
Verheyen
 
A.
Peeraer
 
E.
Lambrechts
 
D.
Poesen
 
K.
Carmeliet
 
P.
Shibuya
 
M.
Pintelon
 
I.
Timmermans
 
J.P.
Nuydens
 
R.
Meert
 
T.
 
Therapeutic potential of VEGF and VEGF-derived peptide in peripheral neuropathies
Neurosci
2013
, vol. 
244
 (pg. 
77
-
89
)
[PubMed]
13
Beazley-Long
 
N.
Hua
 
J.
Jehle
 
T.
Hulse
 
R.P.
Dersch
 
R.
Lehrling
 
C.
Bevan
 
H.
Qiu
 
Y.
Lagreze
 
W.A.
Wynick
 
D.
, et al 
VEGF-A165b Is an endogenous neuroprotective splice isoform of vascular endothelial growth factor A in vivo and in vitro
Am. J. Pathol
2013
, vol. 
183
 (pg. 
918
-
929
)
[PubMed]
14
Sondell
 
M.
Lundborg
 
G.
Kanje
 
M.
 
Vascular endothelial growth factor has neurotrophic activity and stimulates axonal outgrowth, enhancing cell survival and Schwann cell proliferation in the peripheral nervous system
J. Neurosci.
1999
, vol. 
19
 (pg. 
5731
-
5740
)
[PubMed]
15
Lin
 
J.
Li
 
G.
Den
 
X.
Xu
 
C.
Liu
 
S.
Gao
 
Y.
Liu
 
H.
Zhang
 
J.
Li
 
X.
Liang
 
S.
 
VEGF and its receptor-2 involved in neuropathic pain transmission mediated by P2X(2/3) receptor of primary sensory neurons
Brain Res. Bull.
2010
, vol. 
83
 (pg. 
284
-
291
)
[PubMed]
16
Hulse
 
R.P.
Beazley-Long
 
N.
Hua
 
J.
Kennedy
 
H.
Prager
 
J.
Bevan
 
H.
Qiu
 
Y.
Fernandes
 
E.S.
Gammons
 
M.V.
Ballmer-Hofer
 
K.
, et al 
Regulation of alternative VEGF-A mRNA splicing is a therapeutic target for analgesia
Neurobiol. Dis.
2014
, vol. 
71
 (pg. 
245
-
259
)
[PubMed]
17
Kawamura
 
H.
Li
 
X.
Goishi
 
K.
van Meeteren
 
L.A.
Jakobsson
 
L.
Cebe-Suarez
 
S.
Shimizu
 
A.
Edholm
 
D.
Ballmer-Hofer
 
K.
Kjellen
 
L.
, et al 
Neuropilin-1 in regulation of VEGF-induced activation of p38MAPK and endothelial cell organization
Blood
2008
, vol. 
112
 (pg. 
3638
-
3649
)
[PubMed]
18
Woolard
 
J.
Wang
 
W.Y.
Bevan
 
H.S.
Qiu
 
Y.
Morbidelli
 
L.
Pritchard-Jones
 
R.O.
Cui
 
T.G.
Sugiono
 
M.
Waine
 
E.
Perrin
 
R.
 
VEGF165b, an inhibitory vascular endothelial growth factor splice variant: mechanism of action, in vivo effect on angiogenesis and endogenous protein expression
Cancer Res.
2004
, vol. 
64
 (pg. 
7822
-
7835
)
[PubMed]
19
Quattrini
 
C.
Jeziorska
 
M.
Boulton
 
A.J.
Malik
 
R.A.
 
Reduced vascular endothelial growth factor expression and intra-epidermal nerve fiber loss in human diabetic neuropathy
Diabetes Care
2008
, vol. 
31
 (pg. 
140
-
145
)
[PubMed]
20
Calcutt
 
N.A.
 
Luo
 
Z.D.
 
Modeling diabetic sensory neuropathy in rats
Pain Research: Methods and Protocols
2004
Totowa
Humana Press
(pg. 
55
-
65
)
21
Oltean
 
S.
Qiu
 
Y.
Ferguson
 
J.K.
Stevens
 
M.
Neal
 
C.
Russell
 
A.
Kaura
 
A.
Arkill
 
K.P.
Harris
 
K.
Symonds
 
C.
, et al 
Vascular endothelial growth factor-A165b is protective and restores endothelial glycocalyx in diabetic nephropathy
J. Am. Soc. Nephrol.
2014
 
doi:10.1681/ASN.2014040350
[PubMed]
22
Rennel
 
E.S.
Hamdollah-Zadeh
 
M.A.
Wheatley
 
E.R.
Magnussen
 
A.
Schuler
 
Y.
Kelly
 
S.P.
Finucane
 
C.
Ellison
 
D.
Cebe-Suarez
 
S.
Ballmer-Hofer
 
K.
, et al 
Recombinant human VEGF165b protein is an effective anti-cancer agent in mice
Eur. J. Cancer
2008
, vol. 
44
 (pg. 
1883
-
1894
)
[PubMed]
23
Drake
 
R.A.
Hulse
 
R.P.
Lumb
 
B.M.
Donaldson
 
L.F.
 
The degree of acute descending control of spinal nociception in an area of primary hyperalgesia is dependent on the peripheral domain of afferent input
J. Physiol.
2014
, vol. 
592
 (pg. 
3611
-
3624
)
[PubMed]
24
Hargreaves
 
K.
Dubner
 
R.
Brown
 
F.
Flores
 
C.
Joris
 
J.
 
A new and sensitive method for measuring thermal nociception in cutaneous hyperalgesia
Pain
1988
, vol. 
32
 (pg. 
77
-
88
)
[PubMed]
25
Hamilton
 
R.T.
Bhattacharya
 
A.
Walsh
 
M.E.
Shi
 
Y.
Wei
 
R.
Zhang
 
Y.
Rodriguez
 
K.A.
Buffenstein
 
R.
Chaudhuri
 
A.R.
Van Remmen
 
H.
 
Elevated protein carbonylation, and misfolding in sciatic nerve from db/db and Sod1(-/-) mice: plausible link between oxidative stress and demyelination
PLoS One
2013
, vol. 
8
 pg. 
e65725
 
[PubMed]
26
Xu
 
Q.
Qaum
 
T.
Adamis
 
A.P.
 
Sensitive blood-retinal barrier breakdown quantitation using Evans blue
Invest. Ophthalmol. Vis. Sci.
2001
, vol. 
42
 (pg. 
789
-
794
)
[PubMed]
27
Russell
 
J.W.
Golovoy
 
D.
Vincent
 
A.M.
Mahendru
 
P.
Olzmann
 
J.A.
Mentzer
 
A.
Feldman
 
E.L.
 
High glucose-induced oxidative stress and mitochondrial dysfunction in neurons
FASEB J.
2002
, vol. 
16
 (pg. 
1738
-
1748
)
[PubMed]
28
Chen
 
W.
Mi
 
R.
Haughey
 
N.
Oz
 
M.
Hoke
 
A.
 
Immortalization and characterization of a nociceptive dorsal root ganglion sensory neuronal line
J. Peripher. Nerv. Syst.
2007
, vol. 
12
 (pg. 
121
-
130
)
[PubMed]
29
Arimura
 
A.
Deguchi
 
T.
Sugimoto
 
K.
Uto
 
T.
Nakamura
 
T.
Arimura
 
Y.
Arimura
 
K.
Yagihashi
 
S.
Nishio
 
Y.
Takashima
 
H.
 
Intraepidermal nerve fiber density and nerve conduction study parameters correlate with clinical staging of diabetic polyneuropathy
Diabetes Res. Clin. Pract.
2013
, vol. 
99
 (pg. 
24
-
29
)
[PubMed]
30
Casanova-Molla
 
J.
Morales
 
M.
Planas-Rigol
 
E.
Bosch
 
A.
Calvo
 
M.
Grau-Junyent
 
J.M.
Valls-Sole
 
J.
 
Epidermal Langerhans cells in small fiber neuropathies
Pain
2012
, vol. 
153
 (pg. 
982
-
989
)
[PubMed]
31
Ohta
 
T.
Imagawa
 
T.
Ito
 
S.
 
Novel agonistic action of mustard oil on recombinant and endogenous porcine transient receptor potential V1 (pTRPV1) channels
Biochem. Pharmacol.
2007
, vol. 
73
 (pg. 
1646
-
1656
)
[PubMed]
32
Galloway
 
C.
Chattopadhyay
 
M.
 
Increases in inflammatory mediators in DRG implicate in the pathogenesis of painful neuropathy in type 2 diabetes
Cytokine
2013
, vol. 
63
 (pg. 
1
-
5
)
[PubMed]
33
Ragavendran
 
J.V.
Laferriere
 
A.
Xiao
 
W.H.
Bennett
 
G.J.
Padi
 
S.S.
Zhang
 
J.
Coderre
 
T.J.
 
Topical combinations aimed at treating microvascular dysfunction reduce allodynia in rat models of CRPS-I and neuropathic pain
J. Pain
2013
, vol. 
14
 (pg. 
66
-
78
)
[PubMed]
34
Habib
 
A.A.
Brannagan
 
T.H
 
Therapeutic strategies for diabetic neuropathy
Curr. Neurol. Neurosci. Rep.
2010
, vol. 
10
 (pg. 
92
-
100
)
[PubMed]
35
Hamzeh
 
H.
Gaudillere
 
A.
Sabido
 
O.
Tchou
 
I.
Lambert
 
C.
Schmitt
 
D.
Genin
 
C.
Misery
 
L.
 
Expression of PGP9.5 on Langerhans' cells and their precursors
Acta Derm. Venereol.
2000
, vol. 
80
 (pg. 
14
-
16
)
[PubMed]
36
Simard
 
B.
Gabra
 
B.H.
Sirois
 
P.
 
Inhibitory effect of a novel bradykinin B1 receptor antagonist, R-954, on enhanced vascular permeability in type 1 diabetic mice
Can. J. Physiol. Pharmacol.
2002
, vol. 
80
 (pg. 
1203
-
1207
)
[PubMed]
37
Zochodne
 
D.W.
 
Is early diabetic neuropathy a disorder of the dorsal root ganglion? A hypothesis and critique of some current ideas on the etiology of diabetic neuropathy
J. Peripher. Nerv. Syst.
1996
, vol. 
1
 (pg. 
119
-
130
)
[PubMed]
38
Ibrahim
 
S.
Harris
 
N.D.
Radatz
 
M.
Selmi
 
F.
Rajbhandari
 
S.
Brady
 
L.
Jakubowski
 
J.
Ward
 
J.D.
 
A new minimally invasive technique to show nerve ischaemia in diabetic neuropathy
Diabetologia
1999
, vol. 
42
 (pg. 
737
-
742
)
[PubMed]
39
Costa
 
P.Z.
Soares
 
R.
 
Neovascularization in diabetes and its complications. Unraveling the angiogenic paradox
Life Sci.
2013
, vol. 
92
 (pg. 
1037
-
1045
)
[PubMed]
40
Deguchi
 
T.
Hashiguchi
 
T.
Horinouchi
 
S.
Uto
 
T.
Oku
 
H.
Kimura
 
K.
Makisumi
 
K.
Arimura
 
K.
 
Serum VEGF increases in diabetic polyneuropathy, particularly in the neurologically active symptomatic stage
Diabet. Med.
2009
, vol. 
26
 (pg. 
247
-
252
)
[PubMed]
41
Motawi
 
T.K.
Rizk
 
S.M.
Ibrahim
 
I.A.
El-Emady
 
Y.F.
 
Alterations in circulating angiogenic and anti-angiogenic factors in type 2 diabetic patients with neuropathy
Cell Biochem. Funct.
2014
, vol. 
32
 (pg. 
155
-
163
)
[PubMed]
42
Chattopadhyay
 
M.
Krisky
 
D.
Wolfe
 
D.
Glorioso
 
J.C.
Mata
 
M.
Fink
 
D.J.
 
HSV-mediated gene transfer of vascular endothelial growth factor to dorsal root ganglia prevents diabetic neuropathy
Gene Ther.
2005
, vol. 
12
 (pg. 
1377
-
1384
)
[PubMed]
43
Murakami
 
T.
Arai
 
M.
Sunada
 
Y.
Nakamura
 
A.
 
VEGF 164 gene transfer by electroporation improves diabetic sensory neuropathy in mice
J. Gene Med.
2006
, vol. 
8
 (pg. 
773
-
781
)
[PubMed]
44
Price
 
S.A.
Dent
 
C.
Duran-Jimenez
 
B.
Liang
 
Y.
Zhang
 
L.
Rebar
 
E.J.
Case
 
C.C.
Gregory
 
P.D.
Martin
 
T.J.
Spratt
 
S.K.
Tomlinson
 
D.R.
 
Gene transfer of an engineered transcription factor promoting expression of VEGF-A protects against experimental diabetic neuropathy
Diabetes
2006
, vol. 
55
 (pg. 
1847
-
1854
)
[PubMed]
45
Schratzberger
 
P.
Walter
 
D.H.
Rittig
 
K.
Bahlmann
 
F.H.
Pola
 
R.
Curry
 
C.
Silver
 
M.
Krainin
 
J.G.
Weinberg
 
D.H.
Ropper
 
A.H.
Isner
 
J.M.
 
Reversal of experimental diabetic neuropathy by VEGF gene transfer
J. Clin. Invest.
2001
, vol. 
107
 (pg. 
1083
-
1092
)
[PubMed]
46
Ropper
 
A.H.
Gorson
 
K.C.
Gooch
 
C.L.
Weinberg
 
D.H.
Pieczek
 
A.
Ware
 
J.H.
Kershen
 
J.
Rogers
 
A.
Simovic
 
D.
Schratzberger
 
P.
, et al 
Vascular endothelial growth factor gene transfer for diabetic polyneuropathy: a randomized, double-blinded trial
Ann. Neurol.
2009
, vol. 
65
 (pg. 
386
-
393
)
[PubMed]
47
Simovic
 
D.
Isner
 
J.M.
Ropper
 
A.H.
Pieczek
 
A.
Weinberg
 
D.H.
 
Improvement in chronic ischemic neuropathy after intramuscular phVEGF165 gene transfer in patients with critical limb ischemia
Arch. Neurol.
2001
, vol. 
58
 (pg. 
761
-
768
)
[PubMed]
48
Eisenstein
 
M.
 
Sangamo's lead zinc-finger therapy flops in diabetic neuropathy
Nat. Biotechnol.
2012
, vol. 
30
 (pg. 
121
-
123
)
[PubMed]
49
Simon-Yarza
 
T.
Formiga
 
F.R.
Tamayo
 
E.
Pelacho
 
B.
Prosper
 
F.
Blanco-Prieto
 
M.J.
 
Vascular endothelial growth factor-delivery systems for cardiac repair: an overview
Theranostics
2012
, vol. 
2
 (pg. 
541
-
552
)
[PubMed]
50
Materazzi
 
S.
Fusi
 
C.
Benemei
 
S.
Pedretti
 
P.
Patacchini
 
R.
Nilius
 
B.
Prenen
 
J.
Creminon
 
C.
Geppetti
 
P.
Nassini
 
R.
 
TRPA1 and TRPV4 mediate paclitaxel-induced peripheral neuropathy in mice via a glutathione-sensitive mechanism
Pflugers Arch.
2012
, vol. 
463
 (pg. 
561
-
569
)
[PubMed]
51
McNamara
 
C.R.
Mandel-Brehm
 
J.
Bautista
 
D.M.
Siemens
 
J.
Deranian
 
K.L.
Zhao
 
M.
Hayward
 
N.J.
Chong
 
J.A.
Julius
 
D.
Moran
 
M.M.
Fanger
 
C.M.
 
TRPA1 mediates formalin-induced pain
Proc. Natl. Acad. Sci. U.S.A.
2007
, vol. 
104
 (pg. 
13525
-
13530
)
[PubMed]
52
Ferrari
 
L.F.
Bogen
 
O.
Levine
 
J.D.
 
Nociceptor subpopulations involved in hyperalgesic priming
Neuroscience
2010
, vol. 
165
 (pg. 
896
-
901
)
[PubMed]
53
Meseguer
 
V.
Alpizar
 
Y.A.
Luis
 
E.
Tajada
 
S.
Denlinger
 
B.
Fajardo
 
O.
Manenschijn
 
J.A.
Fernandez-Pena
 
C.
Talavera
 
A.
Kichko
 
T.
, et al 
TRPA1 channels mediate acute neurogenic inflammation and pain produced by bacterial endotoxins
Nat. Commun.
2014
, vol. 
5
 pg. 
3125
 
[PubMed]
54
Bates
 
D.O.
Hillman
 
N.J.
Williams
 
B.
Neal
 
C.R.
Pocock
 
T.M.
 
Regulation of microvascular permeability by vascular endothelial growth factors
J. Anat.
2002
, vol. 
200
 (pg. 
581
-
597
)
[PubMed]
55
Tsantoulas
 
C.
Zhu
 
L.
Shaifta
 
Y.
Grist
 
J.
Ward
 
J.P.
Raouf
 
R.
Michael
 
G.J.
McMahon
 
S.B.
 
Sensory neuron downregulation of the Kv9.1 potassium channel subunit mediates neuropathic pain following nerve injury
J. Neurosci.
2012
, vol. 
32
 (pg. 
17502
-
17513
)
[PubMed]
56
Dunham
 
J.P.
Kelly
 
S.
Donaldson
 
L.F.
 
Inflammation reduces mechanical thresholds in a population of transient receptor potential channel A1-expressing nociceptors in the rat
Eur. J. Neurosci.
2008
, vol. 
27
 (pg. 
3151
-
3160
)
[PubMed]
57
Barriere
 
D.A.
Rieusset
 
J.
Chanteranne
 
D.
Busserolles
 
J.
Chauvin
 
M.A.
Chapuis
 
L.
Salles
 
J.
Dubray
 
C.
Morio
 
B.
 
Paclitaxel therapy potentiates cold hyperalgesia in streptozotocin-induced diabetic rats through enhanced mitochondrial reactive oxygen species production and TRPA1 sensitization
Pain
2012
, vol. 
153
 (pg. 
553
-
561
)
[PubMed]
58
Eberhardt
 
M.J.
Filipovic
 
M.R.
Leffler
 
A.
de la Roche
 
J.
Kistner
 
K.
Fischer
 
M.J.
Fleming
 
T.
Zimmermann
 
K.
Ivanovic-Burmazovic
 
I.
Nawroth
 
P.P.
, et al 
Methylglyoxal activates nociceptors through transient receptor potential channel A1 (TRPA1): a possible mechanism of metabolic neuropathies
J. Biol. Chem.
2012
, vol. 
287
 (pg. 
28291
-
28306
)
[PubMed]
59
Nowak
 
D.G.
Woolard
 
J.
Amin
 
E.M.
Konopatskaya
 
O.
Saleem
 
M.A.
Churchill
 
A.J.
Ladomery
 
M.R.
Harper
 
S.J.
Bates
 
D.O.
 
Expression of pro- and anti-angiogenic isoforms of VEGF is differentially regulated by splicing and growth factors
J. Cell Sci.
2008
, vol. 
121
 (pg. 
3487
-
3495
)
[PubMed]