prognosis of patients with acute lower ischemia still remains poor. The formation of new blood vessels in response to ischemia is an important adaptive response that preserves tissue integrity and is regulated by hypoxia and inflammation (8). New blood vessels grow postnatally by means of angiogenesis [i.e., capillary sprouting of resident endothelial cells (ECs)], arteriogenesis, and vasculogenesis (i.e., de novo vascularization from EC precursors) but also grow in adults (8). It is widely known that the neuronal system plays a fundamental role in the maturation of primitive embryonic vasculature. Mutations that disrupt peripheral sensory nerves or Schwann cells prevent proper arteriogenesis, whereas those that disorganize the nerves maintain the alignment of arteries with misrouted axons (24). Sensory neurons have been shown to modulate the expression of arterial markers on ECs via the secretion of vascular endothelial growth factor (VEGF) (23).

Primary afferent sensory neurons transmit sensory information from peripheral tissues to the spinal cord and brain. The cell bodies of the sensory nerve fibers that innervate the head and body are located in the trigeminal ganglia and dorsal root ganglia (DRG) and can be divided into two main categories, namely myelinated A-fibers and unmyelinated C-fibers. These sensory fibers are specialized sensory neurons known as nociceptors, which detect environmental stimuli. Nociceptors express a diverse repertoire of receptors and transduction molecules that can sense forms of noxious stimulation (i.e., thermal, mechanical, and chemical stimulation) with varying degrees of sensitivity (21). Nociceptor activation results in the release of neurotransmitters, such as calcitonin gene-related peptide (CGRP), endothelin, histamine, glutamate, and substance P (21).

CGRP is a 37-amino acid neuropeptide produced in DRG by tissue-specific alternative splicing of the primary transcript of the calcitonin/CGRP gene (3). CGRP is widely distributed in the central and peripheral neuronal systems and exhibits numerous biological activities, including responses to sensory stimuli, cardiovascular regulation, and vasodilation (3, 6, 15). Endogenous CGRP has been shown to protect the heart (16) against ischemic injury and damage. We have reported that CGRP plays critical roles in the maintenance of gastric mucosal integrity (4, 5, 28). Maintenance of gastric mucosal integrity is highly dependent on the alarm systems that can rapidly sense the harmful chemical or mechanical stimuli to which the mucosa is exposed. The gastrointestinal tract is known to be rich in neuronal systems, among which afferent neurons of extrinsic origin are reported to operate as an emergency protective system (15). The functions of these afferents sensitive to chemicals are reported to be mediated by CGRP released in the gastric mucosa (28).

It was reported that one of the sensory neuropeptides, substance P, has a proangiogenic activity (10); however, the details of the mechanism underlying enhanced angiogenesis were not precisely studied. It was reported that CGRP stimulates the proliferation of cultured human umbilical vein endothelial cells (HUVECs) (14). In addition, we recently found that CGRP has a proangiogenic activity. CGRP enhances tube formation activity in a coculture system using HUVECs and fibroblasts (29). Moreover, endogenous CGRP promotes tumor growth and tumor-associated angiogenesis (29). These observations suggest that CGRP would be involved in the regulation of angiogenesis; however, little is known about the role of CGRP in the revascularization in response to acute ischemia in vivo. Thus the present study was conducted to investigate the roles of CGRP in blood flow recovery and angiogenesis during unilateral hindlimb ischemia using CGRP gene-disrupted mice (26).

MATERIALS AND METHODS

Animals.

Male C57BL/6 mice (8 wk of age) were obtained from CLEA Japan (Tokyo, Japan). A strain of male CGRP knockout mice (CGRP^−/−, 8 wk of age) developed by us (26) and their wild-type littermates (CGRP^+/+) were used. All mice were maintained at a constant humidity (60 ± 5%) and temperature (25 ± 1°C) and kept continuously on a 12:12-h light-dark cycle. All procedures conducted on animals were performed in accordance with the guidelines for animal experimentation of Kitasato University School of Medicine. The study protocol was approved by the Animal Care and Use Committee (2076, 2010–125).

Model of acute hindlimb ischemia.

Hindlimb ischemia was induced as described elsewhere (2, 17). Under anesthesia with pentobarbital sodium (50 mg/kg ip), a midline incision was made in the abdominal skin, which permitted dissection to expose the external iliac artery and vein in the upper part of the left limb. The artery and vein were then ligated both proximally and distally using 6–0 silk suture, and the intervening 6-mm section was excised. Next, the incision was closed.

Continuous administration of a CGRP antagonist, CGRP8–37.

A CGRP antagonist, CGRP8–37 (Peptide Institute, Osaka, Japan), in the physiological saline was infused in the subcutaneous tissues at a rate of 50 nmol/day using osmotic pumps (Azlet; DURECT, Cupertino, CA). The delivery rate was 0.5 μl/h, and the mice received CGRP8–37 for 14 days. One day after the implantation of pumps, mice were subjected to hindlimb ischemia. Mice treated with vehicle (physiological saline) served as controls.

Laser doppler flow analysis.

Blood flow to the right and left hindlimbs was assessed by scanning the lower abdomen and limbs of the mice with a laser Doppler blood flowmeter (Laser Doppler Perfusion Imager System, moorLDI-Mark 2; Moor Instruments, Wilmington, DE), as previously reported (2, 17). The ratio of blood flow in the ischemic (left) to the nonischemic (right) limb was calculated by dividing the integrated blood flow in an area of the image that included the left foot pad by the integrated blood flow in an area of the same size that included the right foot pad.

Measurement of blood pressure and heart rate.

The mean arterial blood pressure (MAP) and heart rate (HR) in conscious mice were determined by a programmable sphygmomanometer (BP-98A; Softron, Tokyo, Japan) using the tail-cuff method as reported previously (17).

Immunohistochemistry.

Immunostaining was performed with the use of CD31 (1:100 dilution; Santa Cruz Biotechnology, Santa Cruz, CA) or CGRP (1:100 dilution; Enzo Life Sciences, Plymouth Meeting, PA), as reported previously (17). For quantification, the numbers of CD31-positive cells were counted in 10 randomly selected transverse sections in each animal. The results were averaged, and capillary density was expressed in terms of the number of CD31-positive cells per high-power field.

In vivo microscopy.

Seven days after surgery, animals were anesthetized with pentobarbital sodium (50 mg/kg ip) and were prepared for in vivo fluorescence microscopy as previously described (17, 19). The hindlimb microcirculation was observed using a fluorescence microscope (ECLIPSE E600, upright type; Nikon) with a 100-W mercury lamp for epi-illumination. The microscopic images were obtained with a long-working-distance objective lens (M plan 40/0.40 SLWD; Nikon) and a ×10 eyepiece lens. Images of the microcirculation were transmitted through a charge-coupled device camera (C7190; Hamamatsu Photonics; Hamamatsu) to a television monitor screen (PVM-144Q; Sony) and were recorded for subsequent off-line analysis on videotape with an S-VHS recorder (BR-S600; Victor). Plasma was labeled with 1 mg/ml iv of fluorescein isothiocyanate-dextran (Sigma) just before the observation. Microcirculation in the quadriceps muscles (muscular regions) or through the tissues adjacent to the femoral artery and vein (perifemoral regions) was observed.

Analysis of in vivo microscopy.

To evaluate the blood flow through the capillaries in the ischemic limb, the total length of perfused capillaries per observation area was measured and expressed as functional capillary density (mm/mm²). The functional capillary density was determined in 10 different perifemoral and muscular regions in each animal.

Isolation of lumbar DRG neurons.

Under ether anesthesia, the medulla spinalis was removed, and the lumbar DRG neurons of the L₁-L₅ levels were dissected from the spinal cord.

Real-time RT-PCR.

Transcripts encoding CD31, VEGF-A, basic fibroblast growth factor (bFGF), transforming growth factor-β (TGF-β), pro-CGRP, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were quantified by real-time RT-PCR analysis as described previously (17). Total RNA was extracted from the ischemic muscle tissues or DRG using TRIzol (GIBCO), and single-stranded cDNA was generated from 1 μg of total RNA via reverse transcription using ReverTra Ace-α (TOYOBO). Quantitative PCR amplification was performed using SYBR Premix Ex Taq (Takara Bio) and the following gene-specific primers: for CD31, 5′-CCCTTCATTGACCTCAACTACAATGGT-3′(sense) and 5′-GAGGGGCCATCCACAGTCTTCTG-3′ (antisense); for VEGF, 5′-CCCCAGAATGAAGGTTACACA-3′ (sense) and 5′-TGTCAAAGGGAGAAGGGTTTT-3′ (antisense); for bFGF, 5′-CCCCAGAAAATGAAGGTTACACA- 3′ (sense) and 5′-TGTCAAACCGAAGGAGAAGGGTTTT-3′ (antisense); for TGF-β, 5′-CCCCAGAATGAAGGGTTACACA- 3′ (sense) and 5′-TGTCAAAGGGAGAAAGGGTTTT-3′ (antisense); for pro-CGRP, 5′-CCCCAGAATGAAGGTTACACA-3′ (sense) and 5′-TGTCAAAGGGAGAAGGGTTTT-3′ (antisense); and, for GAPDH, 5′-CCCTTCATTGACCTCCAACTACAATGGT-3′ (sense) and 5′-GAGGGGCCATCCACACGTCTTCTG-3′ (antisense).

RT-PCR analysis in HUVECs and fibroblasts.

HUVECs purchased from Kurabo (Tokyo, Japan) were cultured in 10% FBS and endothelial cell growth supplement (EGM-2 MV; Cambrex Bioscience, Wekerville, MD). Confluent HUVEC (incubated in serum-free media) were treated with CGRP (3 and 30 μM) and PBS for 6 h. Next, HUVECs were collected and homogenized with Trizol (Invitrogen, Carlsbad, CA). Murine fibroblasts (L929) were obtained from Cell Bank, RIKEN Bioresource Center (Ibaraki, Japan). They were cultured in DMEM supplemented with 10% (vol/vol) FBS and 100 U/500 ml penicillin, all of which were obtained from GIBCO-BRL, Life Technologies (Rockville, MD), at 37°C in 5% humidified CO₂ as reported elsewhere (18). L929 fibroblasts (3 × 10⁵ cells/well) were incubated for 6 h with CGRP (3 and 30 μM) and PBS. We tested the mRNA expression of VEGF-A and CD31 in HUVECs and fibroblasts by real-time RT-PCR.

We also tested the mRNA expression of the following CGRP receptors: calcitonin receptor-like receptor (CRLR) (a subunit of CGRP1), calcitonin receptor (CTR) a (a subunit of CGRP2), and receptor activity-modifying protein (RAMP)−1 (a common subunit of CGRP receptor) in HUVECs and fibroblasts by real-time RT-PCR. The primers used for HUVECs were as follows: for CRLR, 5′-AACAACCAGGCCTTAGTAGCC-3′ (sense) and 5′-CCTTCACAGAGCATCCAAAAG-3′ (antisense); for CTRa, 5′-GCTTGGCACTGTTTCTTCTTC-3′ (sense) and 5′-CATCCATCATCTTCTTTCGTC-3′ (antisense); for RAMP-1, 5′-TCACCTCTTCATGACCACACTGC-3′ (sense) and 5′-TCCCTGTAGCTCCTGATGGTC-3′ (antisense); and for GAPDH, 5′-CAACTTTGGTATCGTGGAAGG-3′ (sense) and 5′-AGAGGCAGGGATGATGTTCTG-3′ (antisense). The primers used for L929 fibroblasts were as follows: for CRLR, 5′-CTTCTGGATGCTCTGTGAAGG-3′ (sense) and 5′-CCCAGCCGAGAAAATAATACC-3′ (antisense); for CTRa, 5′-GAACTGTCACCACCCTTACCC-3′ (sense) and 5′-TCGCAGAGCATCCAGAAGTAG-3′ (antisense); for RAMP-1, 5′-CCATCTCTTCATGGTCACTGC-3′ (sense) and 5′-AGCGTCTTCCCAATAGTCTCC-3′ (antisense); and for GAPDH, 5′-ACATCAAGAAGGTGGTGAAGC-3′ (sense) and 5′-AAGGTGGAAGAGTGGGAGTTG-3′ (antisense).

Statistics.

All data were expressed as means ± SE. Multiple comparisons were performed using one-way ANOVA with a post hoc Fisher's test. The statistical difference between the two groups was examined using Student's unpaired t-test after confirming that the variance of data was not heterogeneous. Differences were considered to be significant for P values of <0.05.

RESULTS

Impaired angiogenesis during hindlimb ischemia in CGRP knockout mice.

To examine the role of CGRP in neovascularization, blood flow recovery was assessed by laser Doppler flowmetry in ischemic and nonischemic limbs after femoral ligation in CGRP^−/− and CGRP^+/+. Figure 1A shows representative images of hindlimb blood flow recorded before and after surgical induction of ischemia. The ratio of blood flow in ischemic to nonischemic hindlimbs was similar in the two groups of mice before surgery (Fig. 1B). Immediately after surgery, hindlimb blood flow was severely reduced in both strains, to the same extent, indicating that the severity of ischemia was comparable. In CGRP^+/+, hindlimb blood flow gradually recovered to ∼77% of the levels of the nonischemic limb over the next 14 days. In contrast, the blood flow recovery was impaired in CGRP^−/− compared with CGRP^+/+ (Fig. 1B). Ischemic/nonischemic limb blood flow ratios in CGRP^−/− were 28% (day 3), 40% (day 7), and 14% (day 14) less than in CGRP^+/+. At 21 days after surgery, the levels of blood flow recovery in CGRP^+/+ and CGRP^−/− were restored to 78 and 74%, respectively, compared with baseline levels. There was no difference in blood flow levels between the two groups. Figure 1C demonstrates photomicrographs of representative ischemic limb sections stained with an antibody against CD31 at 7 days after surgery. Quantitative analysis revealed that capillary density expressed in terms of the number of CD31-positive cells per field in CGRP^−/− 7 days after surgery was significantly reduced by 46% compared with that in CGRP^+/+ (Fig. 1D). At 14 days after surgery, capillary density in CGRP^−/− was still lower (25% reduction) than that in CGRP^+/+ (Fig. 1D), although the difference between the two groups at 14 days was smaller than that at 7 days. We also measured MAP and HR in conscious mice. MAPs in CGRP^−/− were higher than those in WT, although there was no significant difference in HRs between the two groups throughout the experiments (Fig. 1E).

We observed the microcirculation through the tissue adjacent to the femoral artery and vein (perifemoral regions) or in the quadriceps muscles (muscular regions) 7 days after surgery (Fig. 2). Higher-magnification images of in vivo fluorescence microscopy demonstrated that revascularization in the perifemoral and in the muscular regions was suppressed in CGRP^−/− (Fig. 3A). Quantitative analysis revealed that the functional capillary density in the perifemoral and muscular regions was reduced (by 46 and 44%, respectively) in CGRP^−/− compared with that in CGRP^+/+ (Fig. 3B).

These results suggested that endogenous CGRP plays a role in blood flow recovery and angiogenesis during hindlimb ischemia.

Increased pro-CGRP expression in the lumbar DRG during hindlimb ischemia.

To estimate the increase in CGRP release during ischemia-induced angiogenesis, we determined the mRNA levels of pro-CGRP, a precursor of CGRP, in DRG (L₁-L₅). Hindlimb ischemia caused increased pro-CGRP expression in the ipsilateral lumbar DRG 1, 3, and 7 days after surgery by 19, 32, and 15%, respectively compared with sham-operated mice (Fig. 4A). Expression of pro-CGRP returned to normal levels by 14 days after surgery. These results suggested that hindlimb ischemia upregulated pro-CGRP in the DRG, which innervated the area of ischemia.

We assessed the immunoreactivity with CGRP in the DRG during hindlimb ischemia (Fig. 4B). The enhanced CGRP-expressing DRG neurons together with numerous CGPR-immunoreactive nerve fibers were shown 3 days after the induction of ischemia compared with sham operation. Next, the immunoreactivity with CGRP was attenuated at 7 days. In addition, enhanced CGRP immunoreactivity in CGRP^+/+ was demonstrated on peripheral nerves distributed in the ischemic muscles (Fig. 4C, top) of the ischemic limbs 7 days after ischemia compared with CGRP^−/− (Fig. 4C, bottom).

Downregulated expression of proangiogenic factors in CGRP^−/− mice.

Because angiogenesis is affected by proangiogenic factors, we determined the levels of mRNA expression of angiogenic factors in the ischemic tissues 3 and 7 days after surgery, by real-time RT-PCR analysis. The mRNA expression of proangiogenic factors, including CD31 (60 and 40%) (Fig. 5A), VEGF-A (28 and 44%) (Fig. 5B), bFGF (38 and 32%) (Fig. 5C), and TGF-β (64 and 39%) (Fig. 5D), in the ischemic tissues of CGRP^−/− was significantly reduced compared with that of CGRP^+/+.

Upregulated expression of proangiogneic factors in HUVECs and fibroblasts treated with CGRP.

To elucidate the contribution of CGRP to angiogenesis, HUVECs and fibroblasts were cultured in the presence of CGRP. We found that CGRP at final concentrations of 30 μM enhanced the mRNA expression of CD31 (Fig. 6A) and VEGF-A (Fig. 6B) in HUVECs. In addition, VEGF-A mRNA expression in L929 fibroblasts was upregulated (Fig. 6C).

Furthermore, we tested mRNA expression of the CGRP receptors CRLR, CTRa, and RAMP-1 in HUVECs and fibroblasts (Fig. 7). Real-time RT-PCR analysis on CGRP receptors revealed that transcripts encoding CRLR, a subunit of CGRP1, were detected on both HUVECs and fibroblasts. By contrast, CTRa, a subunit of CGRP2, was not detected in either HUVECs or fibroblasts. The mRNA expression of RAMP-1, a common subunit of CGRP receptors, was present, but their levels were weak.

Blockade of the CGRP receptor impaired ischemia-induced revascularization.

We also investigated the effects of continuous infusion of a CGRP antagonist, CGRP8–37, using miniosmotic pumps, on ischemia-induced revascularization (Fig. 8). The ratio of ischemic to nonischemic flow 7 and 14 days after surgery was 36 and 12% less, respectively, in mice treated with CGRP8–37 compared with that of mice treated with vehicle (Fig. 8A). Capillary density in CGRP8–37-treated mice was lower (41% decrease) than that in vehicle-treated mice (Fig. 8B). In vivo microscopic studies demonstrated that the functional capillary density in the perifemoral and muscular regions 7 days after the induction of ischemia was reduced (by 46 and 44%, respectively) (Fig. 8C). These findings indicate that the results from CGRP^−/− are essentially the same as those from mice treated with the CGRP antagonist.

Downregulated expression of growth factors in CGRP8–37-treated mice.

We determined the mRNA expression levels of proangiogenic factors by real-time RT-PCR. Three days after the hindlimb ischemia, the levels of expression for CD31, VEGF-A, bFGF, and TGF-β in CGRP8–37-treated mice were decreased by 50, 67, 46, and 53%, respectively, compared with vehicle-treated mice (Fig. 9).

DISCUSSION

CGRP is a 37-amino acid neuropeptide produced by tissue-specific alternative splicing of the primary transcript of the calcitonin/CGRP gene (3). Although cardiovascular phenotypes were studied (13, 20), little attention was focused on the contribution of CGRP to angiogenesis in response to ischemia in CGRP null mice. The present study showed that CGRP accelerated vascular regeneration in the murine ischemic model by enhancing the expression of proangiogenic growth factors (Figs. 1, 3, and 5). Impaired blood flow recovery from ischemia and reduced capillary density in CGRP knockout mice provide evidence that CGRP upregulates ischemia-induced angiogenesis. In vivo microscopic observation revealed that revascularization was observed in the ischemic tissues and that that in CGRP knockout mice was suppressed. The finding that infusion of a CGRP antagonist, CGRP8–37, suppressed angiogenesis supports the results from CGRP knockout mice (Fig. 8). These results suggest that CGRP, which is synthesized in the neuronal systems, enhances angiogenesis in this model. Thus, the blood flow recovery from ischemia that is dependent on angiogenesis may be facilitated by endogenous CGRP.

Within the nervous system, CGRP has been detected in spinal cord motor neurons, DRG, and motor nerve endings (30). Sensory nerves can be activated by a variety of physical and chemical stimuli (6). Upon induction of inflammation, CGRP is upregulated, and pro-CGRP mRNA levels in the DRG that innervate the sites of inflammation rise (31). In the current study, pro-CGRP mRNA levels in the DRG (L₁-L₅) were upregulated during the evolution of revascularization in response to ischemia (Fig. 4). The pro-CGRP expression in the DRG was correlated with the expression of CGRP in the DRG as well as in the nerves of the ischemic limbs. These findings indicate that hindlimb ischemia enhances the expression of pro-CGRP in the DRG that innervated the area affected by ischemia and that CGRP is synthesized in the neuronal systems and is delivered to the periphery of the nerves innervating the sites of revascularization. The effects of a CGRP antagonist and of CGRP gene disruption on revascularization suggest that CGRP release from the peripheral nerve endings during hindlimb ischemia may stimulate the proangiogenic activities. It has been shown that myocardial ischemia-reperfusion results in the release of CGRP from sensory nerve terminals in rodents (22). The tissues exposed to ischemia might produce various metabolites, including bradykinin, reactive oxygen species, protons, and arachidonate metabolites, some of which may activate sensory C-fibers to release CGRP (11). Moreover, immunoelectron microscopic studies reveal that CGRP-immunoreactive gold particles are localized on the ECs (7, 9). Thus, in addition to the DRG, ECs themselves could synthesize CGPR during hindlimb ischemia.

It was reported that one of the sensory neuropeptides, substance P, has a proangiogenic activity (10); however, the details of the mechanism underlying enhanced angiogenesis were not precisely studied. In the present experiment, another neuropeptide, CGRP, was found to facilitate ischemia-induced angiogenesis by enhancing the expression of proangiogenic factors, including CD31, VEGF-A, bFGF, and TGF-β, in the ischemic tissues where angiogenesis was predominantly observed (Fig. 5).

The current study also demonstrates that CGRP exhibits proangiogenic activity by enhancing the expression of VEGF and CD31 in HUVECs and the expression of VEGF in fibroblasts (Fig. 6). Consistent with this result, our recent report (29) demonstrated that tumor growth and tumor-associated angiogenesis are inhibited in CGRP^−/− mice implanted with Lewis lung carcinoma cells, which is associated with reduced VEGF expression in tumor tissues. In addition, we have shown that CGRP enhances tube formation activity in a coculture system using HUVECs and fibroblasts (29). The capillary-like tube formation elicited by CGRP is blunted by VEGF antibody (29). Moreover, we reported that VEGF is involved in hindlimb ischemia-induced angiogenesis because an antibody against VEGF inhibited blood flow recovery from hindlimb ischemia (2), suggesting that VEGF may work as a downstream molecule. The other proangiogenic factors, such as bFGF and TGF-β, may also participate in CGRP-mediated angiogenesis during hindlimb ischemia. Because CGRP is a potent vasodilator, its vasodilatory activity in addition to its proangiogenic activity would increase blood flow during hindlimb ischemia.

The status of CGRP receptor is critical to understand the pathophysiological roles of endogenous CGRP. The results of RT-PCR of CGRP receptors revealed the detection of the expression of CRLR mRNA, a subunit of CGRP1, in cultured HUVECs and fibroblasts. By contrast, CTRa, a subunit of CGRP2, was not detected. RAMP-1 mRNA, a common subunit of CGRP receptors, was present, but their level of expression was weak. These results suggest that CGRP1 in ECs and fibroblasts has a significant role in facilitating the revascularization during hindlimb ischemia. In fact, in the sponge model (29) as well as in the current model, the angiogenic reaction was inhibited by CGRP-(8–37), a CGRP1 antagonist. The mechanisms underlying that CGRP exerts its angiogenic activity through CGRP1 receptor signaling on ECs and fibroblasts are uncertain. However, CGRP1 signaling is known to activate adenylate cyclase to increase intracellular cAMP levels (6). We reported that increased cAMP through activation of adenylate cyclase/protein kinase A results in VEGF expression and facilitates angiogenesis in vivo (1). These results indicate that CGRP enhances the expression of VEGF through cAMP elevation followed by CGRP1 receptor stimulation. CGRP also is capable of activating phosphatidylinositol-3 kinase/protein kinase B and mitogen-activated protein kinase (MAPK) pathways in cultured human embryonic kidney cells via CGRP1 (27) and in cultured ECs via CRLR (25). Moreover, CGRP upregulates the expression of VEGF in human HaCaT keratinocytes by activation of extracellular signal-regulated kinase 1/2 MAPK (32). Further studies are necessary to clarify the molecular mechanisms by which CGRP regulates the angiogenic factors through activation of intrasignaling pathways.

During the prenatal development, the expression of CGRP in the lumber DRG first appears on embryonic day 16 in mice (12). On embryonic day 18, abundant DRG neurons expressing CGRP are observed. These findings indicate that CGRP develops well in the late stages of gestation. Therefore, it is plausible that CGPR may not be critical for vascular formation during embryogenesis. Indeed, CGRP^−/− mice have normal general development and display no defective vascular formation (26), indicating that CGRP plays a minimal role in prenatal angiogenesis and in postnatal or physiological vessel growth. Consistent with this, the current study showed that blood flow levels were same between CGRP^−/− and CGRP^+/+ before the induction of hindlimb ischemia.

Although CGRP plays a critical role in revascularization in response to hindlimb ischemia, the proangiogenic activity as indicated by growth factor expression and capillary density mainly exhibited within 7 days after induction of hindlimb ischemia. The blood perfusion in CGRP^−/− mice was restored to the similar levels of WT mice 3 wk after hindlimb ischemia. The short-term duration of its effect on proangiogenic activity is supported by the finding that expression of precursor-CGRP is transiently upregulated during revascularization in response to hindlimb ischemia. This suggests that CGRP is involved in the stimulation of revascularization during the early phase of ischemia-induced angiogenesis, even though the CGRP gene is disrupted. These results also suggest that other neuropeptides, including substance P and adrenomedulin, could be involved in revascularization after hindlimb ischemia.

In conclusion, we demonstrated that CGRP has a role in enhancing blood flow recovery during hindlimb ischemia. CGRP8–37 can block the CGRP-dependent enhancement of angiogenesis. Upregulation of the expression of proangiogenic growth factors was associated with enhanced blood flow recovery and revascularization. These results indicate that endogenous CGRP, which is possibly released from the neuronal systems, has a significant role in ischemia-associated angiogenesis. Targeting of CGRP may provide a promising method for controlling pathophysiological angiogenesis.

GRANTS

This work was supported by grants from the Ministry of Education, Culture, Sports, Science, and Technology (Research Grants 20659037 and 21591761, High-Tech Research Center Grant, Academic Frontier Project Grant, and The 21st Century COE Program Grant), the Integrative Research Program of the Graduate School of Medical Science, Kitasato University, and the Parents' Association Grant of Kitasato University School of Medicine.

DISCLOSURES

No conflicts of interest are declared by the authors.