OUP user menu

A reactive oxygen species-mediated component in neurogenic vasodilatation

Anna Starr, Rabea Graepel, Julie Keeble, Sabine Schmidhuber, Natalie Clark, Andrew Grant, Ajay M. Shah, Susan D. Brain
DOI: http://dx.doi.org/10.1093/cvr/cvn012 139-147 First published online: 18 January 2008


Aims Activation of the transient receptor potential vanilloid receptor 1 (TRPV1) leads to release of potent microvascular vasodilator neuropeptides. This study was designed to investigate in vivo mechanisms involved in TRPV1-mediated peripheral vasodilatation.

Methods and results Wildtype (WT) and TRPV1 knockout (KO) mice were investigated in a model of peripheral vasodilatation. Blood flow was measured by laser Doppler flowmetry under anaesthesia and following local application of the TRPV1 agonist capsaicin. A sustained (60 min) increase in blood flow was observed in WT but not TRPV1 KO mouse ears. This response was resistant to blockers of classic vasodilators but inhibited in pharmacogenetic experiments that targeted blockade of the substance P (SP) and calcitonin gene-related peptide (CGRP) pathways. The TRPV1-mediated vasodilatation was also attenuated by treatment with superoxide dismutase and the hydrogen peroxide scavenger catalase, but not by deactivated enzymes, supporting a novel role for reactive oxygen species (ROS) generation. Furthermore, neurogenic vasodilatation was observed neither in the presence of the selective NADPH inhibitor apocynin, nor in gp91phox KO mice, under conditions where prostaglandin E1-induced vasodilatation occurred. Finally, a role of neuropeptides in initiating a ROS-dependent component was verified as superoxide dismutase, catalase, and apocynin inhibited SP and CGRP vasodilatation.

Conclusion These studies provide in vivo evidence that ROS are involved in mediating TRPV1- and neuropeptide-dependent neurogenic vasodilatation. An essential role of NADPH oxidase-dependent ROS is revealed that may be of fundamental importance to the neurogenic vasodilator component involved in circulatory homeostasis and the pathophysiology of certain cardiovascular diseases.

  • Reactive oxygen species
  • Neurogenic vasodilatation
  • TRPV1
  • Capsaicin
  • Substance P
  • CGRP
  • NADPH oxidase

1. Introduction

It has been established, through a series of classic studies performed over the last 150 years, that a population of capsaicin-sensitive sensory fibres are widely distributed in the cardiovascular system and mediate neurogenic vasodilatation.1 Two vasodilator neuropeptides, released from the sensory nerves, are considered to primarily mediate the responses, namely, substance P (SP) and calcitonin gene-related peptide (CGRP),2,3 with CGRP often described as the major vasodilator neurotransmitter.4,5 Sensory nerves are suggested to have a physiologically important role in maintaining vascular homeostasis, but their exact role is debated.6,7 There is evidence that the reduction of sensory nerves, as occurs in ageing and diabetes, may be sufficient to accelerate the onset of salt-induced hypertension.7 The potent vasodilator activity of CGRP, in particular, has led to the suggestion that neurogenic vasodilatation is essential for maintaining vascular homeostasis in peripheral organs and a lack of neurogenic vasodilatation is involved in the defective blood flow observed in Raynaud's disease, diabetic neuropathy, and poor wound healing.5 However, despite numerous studies on the activities of the neuropeptides, the understanding of mechanisms that relate to activation of capsaicin-sensitive sensory nerves and release of the neuropeptides is limited.

The recognition of the capsaicin receptor as the transient receptor potential vanilloid 1 (TRPV1) receptor, also known as the VR1 receptor,8 has enabled advances in understanding to be made. TRPV1 is a membrane-associated cation channel that is primarily localized on a subset of C and Aδ sensory fibres. Agonists include vanilloids (e.g. the chilli extract, capsaicin).9 Endogenous agonists include noxious heat and protons,10,11 as well as an increasing range of proposed chemical mediators,12 although their relative importance in vivo is not yet known. Recently, low pH has been shown to release CGRP, proposed to be protective in ischaemic conditions, in wildtype (WT) but not TRPV1 KO mice hearts.13 Use of these mice has also revealed a protective role of TRPV1 in myocardial ischaemia, although the role of CGRP was debated.14 From a therapeutic viewpoint, the use of capsaicin skin patches leads to improved ischaemic thresholds and exercise time in patients with stable coronary disease.15

The above studies point to a pivotal role of the TRPV1 receptor and the resulting neuropeptide-mediated neurogenic vasodilatation in cardiovascular regulation in biology and disease, as well as a potential target for novel therapeutic approaches. However, limited knowledge exists of the critical mechanisms involved in the vasodilatation observed following stimulation of TRPV1. The aim of these studies was to investigate fundamental mechanisms involved in TRPV1-dependent vasodilatation in vivo. We have utilized a pharmacogenetic approach to determine the roles of classic vasodilators when compared with the potent vasoactive neuropeptides CGRP and SP. Having revealed a hitherto unknown involvement of reactive oxygen species (ROS) in the TRPV1-dependent vasodilatation, we have proceeded to investigate the possibility that the Nox2 (gp91phox) isoform of NADPH oxidase, a ROS generator, is the enzymatic source.

2. Methods

2.1 Preparation of animals

Experiments were carried out in accordance with the UK Home Office Animals (Scientific Procedures) Act, 1986, and local Ethics Committee guidelines. These conform with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996). C57BL6/129SVJ mice (25–35 g), either genetically unaltered WT or lacking the gene for the TRPV1 receptor (TRPV1 KO), were bred from WT and TRPV1 KO breeding colonies16 that were donated by Dr S. Boyce, MSD, Terlings Park, UK. Tachykinin NK1 receptor knockout (NK1 KO) Sv129 + C57BL/617 mice from Prof. N. Gerard, Perlmutter Laboratory, Children's Hospital, Boston, MA, USA; αCGRP KO C57BL/6 mice18 from Dr A.-M. Salmon, Institut Pasteur, Paris, France, and gp91phox KO mice (The Jackson Laboratory, Maine)19 or in each case their respective wildtypes were also used. All strains displayed normal growth and behaviour and were paired according to sex and weight for all procedures. Female CD1 mice (25 g) were also used. All non-recovery procedures were carried out under urethane anaesthesia (2.5 mg/g, ip). All substances were obtained from Sigma (Dorset, UK), unless otherwise stated.

2.2 Measurement of vasoactive responses

Blood flow was continuously measured simultaneously in both ears of anaesthetized mice using a non-invasive two channel laser Doppler flowmeter (Moor Instruments, UK) connected to a PowerLab (ADInstuments, UK).20 Measurements of flux changes in the ears in response to topical application of ethanol containing capsaicin (200 µg) or PGE1 (100 nmol) were made over 60 min. Contralateral ears were treated with ethanol. Responses were also measured after SP (200 pmol) and CGRP (6.6 pmol) in Tyrode's solution (137 mM NaCl, 2.7 mM KCl, 0.42 mM NaH2PO4, 11.9 mM NaHCO3, 1.05 mM MgCl2, and 5.55 mM glucose) were intradermally injected into the base of the ear using BD microfine insulin syringes. Contralateral ears were injected with Tyrode alone and blood flow was recorded for 50 min. Blood flow data were recorded in arbitrary flux units, which are proportional to blood flow through the vessel. Results are expressed as the integral of the recorded flux vs. time trace (×103 flux units).

Images were also obtained from a Laser Doppler blood flow Imager (LDI, Moor Instruments) that works on the same principle as the probe except the laser beam scans in a raster fashion using a moving mirror. This enables blood flow to be mapped and colour coded images prior to and at intervals of 2 min after capsaicin application to be displayed alongside a corresponding photo image and regions of interest to be analysed.21

In some experiments, plasma extravasation was measured concomitantly with blood flow by the extravascular accumulation of intravenously injected [125I]-labelled BSA over 60 min.22 Plasma and tissue radioactivity was assessed (1260 Multigamma II, EG&G Wallac, UK) and plasma extravasation in the ears was expressed as μL plasma/g tissue.

2.3 Measurement of blood pressure

Intravascular blood pressure was recorded in anaesthetized mice using a fluid-filled transducer and PowerLab data acquisition system (PowerLab ADInstruments, Disposable BP Transducer). Alternatively, blood pressure was determined in pre-trained conscious mice by the tail-cuff method (CODA 6 system, Kent Scientific, USA).

2.4 Proposed antagonists and inhibitors

The NK1 receptor antagonist SR140333 ((S)1-(2-[3-(3,4-dichlorophnyl)-1-(3-isopropoxyphenylacetyl)piperidin-3-yl]ethyl)-4-phenyl-1-azoniabicyclo[2.2.2]octane chloride), a gift from Dr X. Emonds-Alt, Sanofi, Toulouse, France, was administered at a dose of 0.32 mg/kg.23 The non-peptide CGRP receptor antagonist BIBN4096BS (BIBN) was obtained from Dr M. Schindler, Boehringer-Ingelheim, Germany, and was administered at 0.3 mg/kg.24,25 The nitric oxide synthase inhibitor L-NAME (LG-nitro-l-arginine methyl ester) and cyclo-oxygenase inhibitor indomethacin were administered at 15 and 20 mg/kg, respectively.26 The non-selective calcium operated potassium channel blocker tetraethylammonium chloride (TEA) was administered at 6 mg/kg to give a plasma concentration of ∼0.5 mM that was previously shown to inhibit forearm blood flow in response to SP in humans.27 These drugs were given iv 5 min prior to blood flow recording. The H2O2 scavenger catalase and superoxide dismutase (SOD) or denatured enzymes were administered ip at a dose of 25 000 U/kg (based on preliminary dose response), 5 min prior to capsaicin application. Denatured (95°C for 20 min) catalase and SOD enzymes were used as vehicle control treatment. Plasma levels of SOD were significantly raised 15 min after ip SOD administration from non-detectable levels in vehicle-treated mice to 0.55 ± 0.12 U/5 µL (P < 0.05, n = 4). Corresponding SOD levels were measured in the peritoneal lavage of vehicle- and SOD-treated mice (0.63 ± 0.14 U/5 µL vs. 1.14 ± 0.12 U/5 µL, P < 0.05, n = 5). The iron chelator (blocks formation of hydroxyl radicals), deferoxamine (mesylate salt), and the non-selective NADPH oxidase inhibitor apocynin were given at 25 and 20 mg/kg, respectively.28,29

2.5 Measurement of superoxide dismutase levels

Uptake of SOD into the blood stream following ip administration was determined. Plasma and peritoneal lavage samples were taken 15 min after ip injections of enzyme or vehicle. Blood samples were taken by cardiac puncture and centrifuged at 2800 g for 10 min at 4°C for plasma. The peritoneal cavity was washed with 500 µL of phosphate-buffered saline to obtain the peritoneal lavage fluid. SOD levels were measured using a modified method by Elstner and Heupel, as described by Sharma et al.30

2.6 Analysis of results and statistical evaluation

Results are expressed as the mean ± SEM. Statistical analysis was carried out using ANOVA analysis followed by Bonferroni's comparison test or Dunnett's comparison test or unpaired t-tests where appropriate. P-values <0.05 were considered significant.

3. Results

3.1 Transient receptor potential vanilloid 1-mediated vasodilatation and importance of neuropeptides

Conscious pre-trained WT and TRPV1 KO mice exhibited similar basal blood pressure when measured using a tail-cuff system (123.0 ± 8.0 mmHg in WT vs. 124.2 ± 8.2 mmHg in TRPV1 KO mice, n = 12). Local application of the TRPV1 agonist capsaicin in ethanol caused a sustained increase in blood flow that remained high until the experiment was terminated after 60 min. This was preceded by an initial transient decrease in measured blood flow flux, possibly related to cooling of the skin and influence of alcohol on light transmission/probe attachment as this occurred with ethanol vehicle treatment also. A typical response obtained by the laser Doppler probe system is shown in response to ethanol (Figure 1A) and capsaicin (Figure 1B) in WT mice. Figure 1C shows a photograph of the ears alongside a laser scan image, obtained at 60 min following application of either capsaicin or ethanol to WT and TRPV1 KO mice. Blood flow was assessed using the integral of the flux response curve over 60 min. Group data demonstrate a significant increase in blood flow in WT mice and a clear absent response in TRPV1 KO mice (Figure 1D).

Figure 1

TRPV1-mediated blood flow. A representative trace of blood flow assessed laser Doppler flux vs. time trace after topical application of (A) ethanol or (B) capsaicin (20 µL of 10 mg/mL) to the WT mouse ear is shown. An upward deflection in the trace is proportional to an increase in blood flow. (C) The response as observed by the laser Doppler imager is shown alongside the laser Doppler grey/black scale ‘photo’ image. (D) Group data for increased blood flow in response to capsaicin are shown for WT and TRPV1 KO mice (n = 9). ***P < 0.001 compared with ethanol-treated contralateral ears. ###P < 0.001 compared with capsaicin-treated ears of vehicle-treated animals.

The reliance of TRPV1-mediated increased blood flow on mechanisms involving both the major vasodilator neuropeptides CGRP and SP was shown through investigation of mice lacking αCGRP or lacking the SP, NK1 receptor. CGRP WT mice vasodilatory responses to capsaicin were high in the presence of vehicle and reduced, albeit not significantly, by an NK1 receptor antagonist (Figure 2A). Whereas SR140333 abolished the increased blood flow seen in CGRP KO mice treated with vehicle (Figure 2B). In a complementary manner, the CGRP antagonist, BIBN, only reduced vasodilatory responses observed in NK1 KO mice but not in NK1 WT mice (Figure 2C and D). This component of the study extends studies carried out in this laboratory previously that suggested that a combination of SP and CGRP mediate neurogenic vasodilatation in the mouse.20

Figure 2

Neuropeptide involvement in neurogenic vasodilation. (A) αCGRP WT and (B) αCGRP KO mice in the presence of the NK1 antagonist SR140333 (0.32 mg/kg), and (C) NK1 WT and (D) NK1 KO mice in the presence of the CGRP antagonist BIBN4096BS (0.3 mg/kg), (n = 7). *P < 0.05 and ***P < 0.01 compared with ethanol-treated contralateral ears. ##P < 0.01 compared with capsaicin-treated ears of vehicle-treated animals.

3.2 Vasodilator mechanisms

Effective doses of L-NAME, indomethacin and TEA, blockers of classic endothelial-dependent vasodilators nitric oxide, prostaglandins, and K+ channels, respectively,26,27 had no significant effect on the neurogenic vasodilatation induced by capsaicin in CD1 mice (Table 1). Doses were chosen that had inhibited vasodilatation in vivo.

View this table:
Table 1

Effect of classic vasodilator inhibitors on capsaicin-induced vasodilatation

Treatment (ip)nTopical application ethanol (20 µL)capsaicin (20 µL of 10 mg/mL)
Vehicle735.0 ± 9.0101.8 ± 18.7*
L-NAME611.6 ± 2.3136.3 ± 34.7**
Indo441.6 ± 2.5156.6 ± 9.3*
TEA629.8 ± 5.487.6 ± 5.6*
Vehicle841.3 ± 7.0100.1 ± 13.7*
L-NAME + indo551.7 ± 22.3129.2 ± 43.0*
Indo + TEA425.9 ± 3.1083.6 ± 3.11*
L-NAME + indo + TEA539.6 ± 10.4192.4 ± 47.2*
  • Blood flow was measured in response to topical application of capsaicin (20 µL of 10 mg/mL) and ethanol (20 µL)-treated ears in each case. Capsaicin induced a significant increase in blood flow in all treated animals, compared with ethanol (*P < 0.05, **P < 0.01), but no significant difference in capsaicin-induced blood flow between inhibitor treatments or vehicle was observed (ANOVA + Dunnett's post test).

Catalase, an H2O2 scavenger, significantly reduced neurogenic vasodilatation induced by capsaicin when compared with responses in the presence of the deactivated enzyme. Furthermore, co-administration of SOD with catalase abolished neurogenic vasodilatation, whereas treatment with SOD alone had no inhibitory effect. The membrane permeable SOD mimetic tempol also had no effect. Results as follows, for flux over 60 min: vehicle-treated mice, ethanol 42.4 ± 7.7 vs. capsaicin 120.6 ± 24.0 × 103 flux units, n = 5; tempol-treated mice (30 mg/kg iv), ethanol 45.6 ± 12.4 vs. capsaicin 117.6 ± 21.4 × 103 flux units, n = 8. The effectiveness of the combined SOD and catalase treatment was supported by the findings that the denatured enzymes did not inhibit (Figure 3) and that, in separate studies, systemic blood pressure was similar in anaesthetized mice treated with vehicle (55.8 ± 2.7 mmHg, n = 5) or with SOD and catalase (58.9 ± 4.4 mmHg, n = 5), indicating that inhibition was not secondary to an effect on blood pressure. Furthermore in support H2O2 (20 µL of 2% w/v solution) induced a transient but significant increase in blood flow (8.4 ± 2.7 × 103 flux units compared with ethanol vehicle 3.3 ± 1.3 × 103 flux units, n = 5) over 10–20 min after topical application. Interestingly, the oedema component of the neurogenic response remained intact in the presence of SOD and catalase (vehicle-treated mice, ethanol 31.4 ± 5.0, capsaicin 101.7 ± 6.7; SOD + catalase-treated mice, ethanol, 26.9 ± 4.0, capsaicin 109.9 ± 10.1 µL/g of tissue, n = 5), indicating that the ROS inhibitors do not markedly inhibit neuropeptide release.

Figure 3

Effect of ROS inhibitors on TRPV1-mediated vasodilatation. (A) Responses in CD1 mice treated with catalase (25 000 U/kg) and SOD (25 000 U/kg, n = 5). (B) Representative laser Doppler images, depicting the position of the mouse head and flux images taken every 6 min, after treatment with denatured enzymes (vehicle) or catalase and SOD. A star denotes increased blood flow. Statistical significance is shown by ##P < 0.01 and ###P < 0.001 compared with capsaicin-treated ears of vehicle-treated animals.

The possibility that OH radicals may also be involved was investigated using deferoxamine, but it had no effect on vasodilatation (vehicle, ethanol 46.4 ± 25.6 vs. capsaicin-treated ears 145.8 ± 28.8 × 103 flux units compared with deferoxamine, 32.7 ± 10.5 vs. 152.2 ± 40.3 × 103 flux units, mean ± SEM, n = 6) at doses known to substantially inhibit OH radical production in response to radiation-induced oedema responses in mouse skin.29

3.3 Role of NADPH oxidase in neurogenic vasodilatation

The possible source of ROS was then investigated. Experiments were carried out following administration of apocynin, a selective inhibitor of NADPH oxidase. Increased blood flow in untreated CD1 mice induced by capsaicin was inhibited by apocynin (Figure 4C), but increased blood flow induced by the sensory-nerve-independent vasodilator PGE1 was not affected (Figure 4D). The gp91phox component of NADPH oxidase is the major subunit responsible for Nox2-dependent enzyme activity. The possible involvement of this subunit was investigated by determining TRPV1-mediated vasodilatation in WT and gp91phox KO mice. Capsaicin-induced blood flow was not observed in gp91phox KO mice compared with their WT counterparts (Figure 4E). By comparison, PGE1-induced vasodilatation was similar in both WT and gp91phox KO mice ears (Figure 4F).

Figure 4

Role of NADPH oxidase generated ROS in TRPV1-mediated vasodilatation. (A) Effect of the NADPH oxidase inhibitor apocynin (20 mg/kg) on increased blood flow induced by topical capsaicin (n = 6–8) and (B) topical PGE1 (n = 6). (C) Blood flow induced by capsaicin in gp91phox WT and KO mice (n = 6–8) and (D) PGE1 (n = 6). Statistical significance is shown by ***P < 0.001, **P < 0.01, and *P < 0.05 compared with ethanol-treated contralateral ears. ##P < 0.05 and #P < 0.05 compared with capsaicin-treated ears of vehicle-treated mice.

3.4 Effect of reactive oxygen species inhibitors on neuropeptide responses

Co-administration of catalase and SOD was also able to significantly attenuate the neurogenic vasodilatory responses observed in αCGRP WT, αCGRP KO, NK1 WT, and NK1 KO mice (Figure 5A–D), demonstrating that ROS inhibitors can influence the activity of either of the major neuropeptides.

Figure 5

Effect of ROS inhibitors on neurogenic vasodilatation. Effects of catalase and SOD (25 000 U/kg each) or vehicle on vasodilatation induced by capsaicin in (A) αCGRP WT (n = 7) and (B) αCGRP KO mice (n = 7), and (C) NK1 WT (n = 6) and (D) NK1 KO mice (n = 6). Statistical significance is shown as *P < 0.05, **P < 0.01, and ***P < 0.001 compared with ethanol-treated ears. #P < 0.05, ##P < 0.01, and ###P < 0.001 compared with ear of vehicle-treated mouse.

To mimic neurogenic vasodilatation, the vasodilatory effect of SP and CGRP was induced by direct exogenous injection into the base of the ear. A range of SP and CGRP doses were tested and a co-injection of SP (200 pmol) and CGRP (6.6 pmol) induced a significant increase in blood flow (Figure 6A). Vasodilatation induced by a co-injection of SP and CGRP was inhibited by SOD and catalase and apocynin (Figure 6B and C), thereby verifying a role for ROS in mediating neuropeptide responses. Furthermore, oedema formation was unaffected by treatment with SOD and catalase (vehicle-treated mice, ethanol 135.6 ± 42.0, SP + CGRP 296.0 ± 20.9; SOD + catalase-treated mice, ethanol, 144.1 ± 21.7, SP + CGRP 279.2 ± 22.0µL/g of tissue, n = 7).

Figure 6

Effect of ROS inhibitors on exogenous SP- and CGRP-induced vasodilatation. (A) Time-dependent effects of exogenous SP (200 pmol) and CGRP (6.5 pmol) on blood flow injected intradermally into ear, over 50 min (n = 21). Effects of (B) deactivated enzymes (vehicle) and of SOD and catalase (n = 7) and (C) vehicle and apocynin on id SP + CGRP stimulated blood flow (n = 6). *P < 0.05, **P < 0.01, and ***P < 0.001 compared with Tyrode-treated ears. #P < 0.05 compared with ear of vehicle-treated mouse.

4. Discussion

This study has revealed a pivotal role for ROS in the vasodilator response to neuropeptides released following TRPV1 receptor activation. Initial studies confirmed that topical capsaicin application to the mouse ear causes increased blood flow that is TRPV1 dependent, as this response is not observed in TRPV1 KO mice. The classic neurogenic nature of the response was confirmed as vasodilatation was not observed when the principal sensory dilator neuropeptide pathways of CGRP and SP were blocked. Intriguingly, capsaicin-induced vasodilatation was also inhibited by ROS inhibitors. Moreover, selective inhibition of TRPV1-dependent vasodilatation by the NADPH oxidase inhibitor apocynin and a significant inhibition of TRPV1-dependent vasodilatation in mice lacking the NADPH oxidase subunit gp91phox (Nox2) reveal the importance of NADPH oxidase in this response. To our knowledge, this is the first time ROS have been shown to mediate the vasodilator response to TRPV1 activation by capsaicin. To determine whether the ROS production was the consequence of an intrinsic mechanism between the TRPV1 receptor and NADPH oxidase or the consequence of neuropeptide release, exogenous neuropeptides were used to show that ROS production is indeed due to mechanisms downstream of neuropeptide release from sensory nerves.

4.1 Neurogenic vasodilatation: relevance to established vasodilator mechanisms

The historical understanding that TRPV1-dependent neurogenic vasodilatation is mediated by the sensory-nerve-derived neuropeptide vasodilators CGRP, via the CGRP receptor, and SP, via the NK1 receptor, is supported in this study and extends previous studies in this model.20 The results demonstrate a relationship between the peptides that involves functional redundancy, such that inhibition of one neuropeptide vasodilator will not compromise dilator tone in the periphery. However, the concomitant blockade of both CGRP- and SP-induced vasodilation by use of either antagonists or KO mice abolished neurogenic vasodilatation. These results are in keeping with studies in humans where neither a selective CGRP antagonist31 nor a selective SP NK1 antagonist32 affected baseline cardiovascular parameters, when given alone. However, there is a dense perivascular innervation by TRPV1-containing sensory nerves of distal arteriolar vessels,33,34 of relevance to the potential for sensory nerves to contribute to the control of local vascular tone (Figure 7). In addition, there is evidence that CGRP especially has potent vasodilator effects in the coronary, cutaneous, and cerebral microvasculature.5

Figure 7

Proposed role of ROS in TRPV1-mediated vasodilatation. A blood vessel with attached sensory nerve is shown. Capsaicin activates the TRPV1 receptor to release SP and CGRP. These then act on vascular receptors to stimulate NADPH oxidase to mediate ROS-dependent relaxation.

As SP and CGRP mediate their actions through distinct receptors and intracellular pathways, mechanisms of vascular relaxation in this model were elucidated further. Nitric oxide, prostaglandins, and endothelium-derived hyperpolarizing factor are endothelial-dependent vasodilators involved in the regulation of vascular tone, however inhibition of these mediators had no effect on neurogenic vasodilatation induced by capsaicin, when administered alone or in combination. As classical vasodilators played little role in our model, other putative vasodilators were therefore considered.

4.2 Role of reactive oxygen species in neurogenic vasodilatation

ROS, especially H2O2, have previously been proposed to mediate vascular relaxation,35 and a weak vasodilator response is observed following the topical application of H2O2, but a regulatory role in neurogenic vasodilatation has not been previously shown. Here, mice were pretreated with SOD and catalase to determine a possible role for ROS in the capsaicin response. The results demonstrate an involvement of ROS in the vasodilator response following TRPV1 activation by capsaicin as the combined treatment of the antioxidants blocked the vasodilatation in response to capsaicin and to exogenous SP and CGRP. Perhaps surprisingly neither SOD nor the cell-permeable tempol when given alone modulated neurogenic vasodilatation. One might expect SOD to play a role in causing relaxation by converting vasoconstrictor superoxide into vasodilator H2O2 or by preventing the normal O2 scavenging of vasodilator NO.36 However, if this occurred, it did not result in increased vasodilatation. One possibility is that peroxynitrite is involved, produced by interaction of O2 and NO, which has been shown to possess vasoactive properties under some situations. However, if this were so, one would have not expected catalase to block vasodilatation. The effect of catalase, when in combination with SOD, but not of the deactivated enzymes to abolish neurogenic vasodilatation points to a role of H2O2. It is possible that OH radicals may be formed. This possibility was investigated using deferoxamine, at a dose that substantially inhibited ROS-dependent murine inflammation but no inhibition of vasodilatation was observed, thus a role of the OH radical cannot be suggested. We attempted to measure H2O2 in homogenized tissue samples using two different assays, but could not obtain reliable results, presumably due to the low amounts of vascular tissue within the ear.

Considering the profound role of ROS in vasodilatation, it was interesting that in the present study, ROS inhibition had no effect on the neurogenic oedema response. For many years, neurogenic inflammation was considered to consist of oedema formation and vasodilatation. However, it is now realized that neurogenic oedema mediated by the SP NK1 receptor22,37 is of minimal relevance in humans since NK1 antagonists were found to be ineffective in a range of clinical trials of diseases with an inflammatory component.38,39 Furthermore, we have shown in the model used here that the topical application of capsaicin to skin is associated with neurogenic vasodilatation, at a lower concentration than necessary for oedema formation.20 The lack of effect of the ROS inhibitors in inhibiting neurogenic oedema is in contrast to studies in lung where direct treatment with ROS increased trans-endothelial permeability.40 Thus, while evidence suggests that ROS are important in modulating neurogenic vasodilatation, results show that ROS are not markedly involved in mediating neuropeptide-induced oedema formation and, as a consequence, are unlikely to act by inhibiting SP release from sensory nerves. As CGRP and SP are co-localized in sensory nerves, this in turn suggests that ROS do not play a critical role in mediating neuropeptide release per se.

4.3 Reactive oxygen species production

In view of this novel role for ROS in the TRPV1-induced vasodilator response, it was important to determine the enzymatic source of the ROS. It has been recently shown that NADPH oxidase-dependent ROS generation may be important as a physiological vasodilatory mechanism, in the cerebral circulation35,41 when compared with other areas of the body and, of direct relevance to this study, there are increased levels of NADPH oxidase activity in the cerebral when compared with systemic vasculature.42 The sensory nerves that innervate the ear originate from the trigeminal ganglion that also innervates the cerebral circulation. Hence, it is conceivable that ROS may well be acting as vasodilatory mediators in this vasculature. It was then questioned whether ROS could be generated when TRPV1 was intact but neurogenic vasodilatation blocked, through study of NK1 KO mice in the presence of a CGRP antagonist. Under these conditions and in CGRP KO mice treated with an NK1 antagonist, blood flow was not detected above basal levels. This provides further positive evidence for a direct link between ROS and neuropeptide-mediated activity, and evidence for the distinct possibility that ROS are released downstream of SP and CGRP at the vascular level, following their release from sensory nerves.

To our knowledge, this is the first evidence that neuropeptide-mediated ROS production is an important modulatory component in neurogenic vasodilatation, although links exist with ROS production under inflammatory conditions. In order to clarify the relationship between neuropeptides and ROS production, exogenous SP and CGRP responses were studied. Subsequent inhibition of the response first by catalase and SOD and secondly by apocynin provides additional evidence that ROS derived from NADPH oxidase is produced in response to neuropeptides.

In conclusion, the present study provides an important and novel insight into the mechanisms of vascular relaxation that involve sensory nerves. While supporting the traditional concept that neuropeptides mediate neurogenic vasodilatation, this study also introduces the hypothesis that ROS play a pivotal role in neurogenic vasodilatation and that ROS generation is specifically linked to activation of a Nox2-containing NADPH oxidase in the cerebral vasculature. It is established that Nox2 is a major enzymatic source of damaging ROS in endothelial cells in cardiovascular disease.19,43 However, TRPV1-dependent vasodilatation involving ROS may be important from a protective view, as there is a large range of physiological and pathophysiological situations that the ‘molecular sensing’ TRPV1 may respond to44 and ischaemic damage has already been shown to be enhanced in TRPV1 KO mice.14 Thus, we propose neuropeptide-mediated ROS production following TRPV1 activation to be a noteworthy pathophysiological target in the modulation of blood flow and tissue perfusion.


This work was supported by the Biotechnology and Biological Sciences Research Council and the British Heart Foundation.

Conflict of interest: none declared.


View Abstract