ransformation did not render the samples Gaussian, non-parametric tests were used, in which case multiple groups were compared with KruskalWallis or Friedman’s tests followed by post-hoc Dunn’s tests. Two group tests were 2 tailed Student’s t-tests with Welch’s correction where necessary for unequal variance, or Mann Whitney U tests for non-parametric data. Numbers of DRG neurons with TRPV1-activated currents were compared using Fisher’s exact test. 250 R.P. Hulse et al. / Neurobiology of Disease 71 245259 Results VEGF-A splice isoforms differentially affect pain behaviors, through direct VEGFR2-mediated effects on primary sensory nociceptive neurons Systemic delivery of anti-mouse VEGF antibody acutely sensitized animals to mechanical and thermal stimulation. Neutralization of VEGF-A165b, and inhibition of VEGF receptor-2 by selective and specific inhibitors PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/19841886 also produced sensitization. Systemic recombinant human VEGF-A165b had no effect on mechanical, or thermal nociceptive behavior, whereas rhVEGF-A165a sensitized to mechanical but not thermal stimuli. rhVEGF-A121a, which has the same C-terminal six amino acid sequence as the VEGF-Axxxa family but reduced affinity for neuropilin-1 also resulted in mechanical sensitization. rVEGF-A159, which lacks the six C terminal amino acids, had no effect on pain, showing that the mechanism through which VEGF-A165a and VEGF-A121a enhance pain is C-terminal sequence dependent . We then determined the neuronal mechanism through which systemic rhVEGF-A165a might alter nociceptive behavior. VEGFR2 protein was detected in DRG neurons as previously described by immunofluorescence in proportions of both TrkA- and isolectin B4, nociceptive neurons, with increased expression 
following traumatic nerve injury. Inhibition of VEGFR2 directly sensitized nociceptors to mechanical stimulation. As endogenous VEGF-A isoforms can exert potent vascular effects, we also determined whether VEGFR2 inhibitors PTK787 and ZM323881 overtly affected local blood flow. Neither receptor blocker resulted in any reduction in local blood flow as measured by laser Doppler flowmetry, or in skin temperature, in contrast to local adrenaline R 115777 web injection used as a positive control. To determine the roles of VEGF-A isoforms on sensory afferents, effects on nociceptors were determined before and after injection of vehicle, VEGF-A165a or VEGF-A165b. Injection of rhVEGF-A165a, but not rhVEGF-A165b into individual characterized sensory neuronal receptive fields resulted in the initiation of spontaneous ongoing firing in 56% of mechano-sensitive primary afferent nociceptors tested, indicating expression of functional VEGF receptors in a large proportion of the sampled afferents. The proportion of neurons responding to VEGF-A was significantly higher than the proportion of VEGFR2 positive IB4/TrkA + ve neurons — posited to be nociceptors. There are a number of explanations for this including: unconscious bias in our search strategy for afferents leading to an overrepresentation of afferents expressing VEGFR2; detection of protein by immunofluorescence underestimating the degree to which functional VEGFR are found on sensory neurons; or a higher proportion of VEGF sensitive nociceptors in the paw than in other regions through which Fig. 2. VEGF-A isoforms differentially affect pain depending on VEGFR2 activation. A. Intraperitoneal injection of 6 g/g anti-VEGF-A antibody induced significant mechanical allodynia in mice. B. Systemic injection of