The GPCR modulator protein RAMP2 is essential for angiogenesis and vascular integrity

Adrenomedullin (AM) is a peptide involved both in the pathogenesis of cardiovascular diseases and in circulatory homeostasis. The high-affinity AM receptor is composed of receptor activity–modifying protein 2 or 3 (RAMP2 or -3) and the GPCR calcitonin receptor–like receptor. Testing our hypothesis that RAMP2 is a key determinant of the effects of AM on the vasculature, we generated and analyzed mice lacking RAMP2. Similar to AM^–/– embryos, RAMP2^–/– embryos died in utero at midgestation due to vascular fragility that led to severe edema and hemorrhage. Vascular ECs in RAMP2^–/– embryos were severely deformed and detached from the basement membrane. In addition, the abnormally thin arterial walls of these mice had a severe disruption of their typically multilayer structure. Expression of tight junction, adherence junction, and basement membrane molecules by ECs was diminished in RAMP2^–/– embryos, leading to paracellular leakage and likely contributing to the severe edema observed. In adult RAMP2^+/– mice, reduced RAMP2 expression led to vascular hyperpermeability and impaired neovascularization. Conversely, ECs overexpressing RAMP2 had enhanced capillary formation, firmer tight junctions, and reduced vascular permeability. Our findings in human cells and in mice demonstrate that RAMP2 is a key determinant of the effects of AM on the vasculature and is essential for angiogenesis and vascular integrity.

Many vasoactive substances play critical roles in the maintenance of cardiovascular homeostasis; moreover, imbalances among these substances have been implicated in the pathogenesis of various cardiovascular diseases. Among these, adrenomedullin (AM) was originally identified as a vasodilating peptide isolated from human pheochromocytoma (1) and, based on its structural homology and similar vasodilatory effects, has been classified as a member of the peptide family that also includes calcitonin gene–related peptide (CGRP). Although produced by a variety of tissues and cell types, AM is primarily secreted by vascular cells and functions as a local autocrine or paracrine (2) mediator, as well as a circulating hormone (3). Apart from its vasodilatory effect, AM also exerts diuretic (4) and cardiotonic (5) effects and is involved in the regulation of hormone release (6, 7), inflammation (8), and oxidative stress (9, 10) as well as the proliferation, migration, and differentiation of various cell types (11–13). Thus, AM is now recognized to be a pleiotropic vasoactive molecule. To better understand the in vivo functions of AM, we established and analyzed genetically engineered AM-deficient mice (14–25). Homozygous AM KO (AM^–/–) mice die in utero at around midgestation from systemic hemorrhage and edema resulting from the fragility of their vasculature (14). In addition to mediating vascularization during development, we found that AM also enhances revascularization in adult tissues subjected to ischemia (25). The potential clinical applications of AM implied by these findings have attracted much attention, with particular attention being paid to AM’s ability to stimulate vascular regeneration in ischemic tissue (26, 27).

AM signaling is regulated by a unique control system (28–31). The AM receptor is a 7–transmembrane domain GPCR named calcitonin receptor–like receptor (CRLR). CRLR associates with an accessory protein, receptor activity–modifying protein (RAMP), which is composed of about 160 amino acids and includes a single membrane-spanning domain. So far, 3 RAMP subtypes have been identified. By interacting with RAMP1, CRLR acquires a high affinity for CGRP, whereas by interacting with either RAMP2 or RAMP3, CRLR acquires a high affinity for AM. This novel system enables CRLR to transduce the signals of multiple ligands, although the precise mechanism remains largely unknown.

We hypothesized that not only the receptor-ligand specificity, but also the diversity, of AM’s biological activities reflects its novel regulation by RAMPs. To test this idea, we generated RAMP2 KO mice, which were then used to analyze the physiological functions of the AM-RAMP2 system.

Generation of RAMP2-null mice. We initially analyzed the expression of AM and its related genes during midgestational development (E11.5–E14.5), the stage at which AM^–/– embryos typically die. We found that in WT mice, AM, CRLR, RAMP2, and RAMP3 all continued to be expressed at midgestation (Figure 1A), which is consistent with previous findings (32, 33). Using in situ hybridization, we detected AM expression in the vascular system (data not shown) and found that, among the RAMPs, only RAMP2 was specifically expressed in the vasculature at that stage (Figure 1B). We therefore speculated that it is RAMP2 that determines AM’s function during vascular development and proceeded to generate RAMP2-specific KO mice to directly assess the functions of the AM-RAMP2 system in vivo.

Figure 1

RAMP2 expression during development and generation of RAMP2 knockout mice. (A) Real-time PCR analysis of gene expression during E11.5–E14.5 in WT embryos. Expression is shown relative to that at E11.5. n = 5 per time point. AM, CRLR and RAMPs were expressed during midgestation. (B) In situ hybridization of RAMP2 in WT embryos. Sections of umbilical artery, aortic arch, and lung from E12.5 WT embryos were used. Intense RAMP2 expression was detected in the vascular ECs. Scale bars: 20 μm. (C) Targeted disruption of mouse RAMP2. The genomic locus and predicted targeted locus are shown. Boxes denote exons 1–4 of RAMP2; ScaI and NheI restriction sites and loxP sites are indicated. Probes for Southern blot analysis are shown. (D–F) Southern blot analysis of mouse genomic DNA. (D) DNA was digested with NheI and probed with the 5′ probe. The 7.9-kb and 18.3-kb fragments denote floxed and WT alleles, respectively. (E) DNA was digested with ScaI and probed with the Neo probe. The 8.1-kb fragment denotes flox. (F) DNA was digested with ScaI and probed with the 3′ probe. The 8.1-kb ScaI fragment denotes flox; the 12.4-kb fragment denotes WT. The 4.9-kb fragment denotes the KO allele, generated by deletion of the loxP site using Cre recombinase.

The targeting strategy and analysis of homologous recombination are shown in Figure 1, C–F. RAMP2 heterozygous KO mice (RAMP2^+/–) were apparently normal, although the number of live births was markedly diminished when RAMP2^+/– mice were intercrossed. No RAMP2 homozygous KO (RAMP2^–/–) newborns were obtained, and analysis of the embryos from timed RAMP2^+/– intercrosses showed that the RAMP2^–/– genotype was lethal at midgestation. The mortality rate among RAMP2^–/– embryos was 13% at E13.5, 92% at E14.5, and 100% at E15.5 (Table 1). The most lethal stage (E13.5–E14.5) was nearly identical to that of the AM^–/– genotype.

Table 1

Genotype of embryos from RAMP2^+/– male and female mouse intercrosses

Real-time PCR analyses (Figure 2A) showed there to be no expression of RAMP2 in RAMP2^–/– embryos, confirming that the RAMP2 gene was successfully destroyed. By contrast, expression of RAMP3 did not differ in RAMP2^–/– and WT mice, which indicates that no functional redundancy exists between RAMP2 and RAMP3 during development. Moreover, the expression of AM was upregulated by more than 5-fold in RAMP2^–/– mice, presumably as a compensatory response to the absence of a functional AM receptor.

Figure 2

Quantitative real-time PCR analysis, macroscopic analysis, and histology of RAMP2^–/– embryos. (A) Gene expression of AM, CRLR, and RAMPs in E13.5 WT and RAMP2^–/– embryos, assessed by real-time PCR of total RNA. No RAMP2 expression was detected in RAMP2^–/– mice, confirming RAMP2 was successfully destroyed. Conversely, RAMP3 expression did not differ between RAMP2^–/– and WT mice, showing that the absence of RAMP2 did not induce compensatory upregulation of RAMP3 during development. AM expression was upregulated more than 5-fold in RAMP2^–/– mice. n = 6 per group. **P < 0.01 vs. WT. (B–L) Development of blood vessels in E13.5 WT and RAMP2^–/– mice. Appearance of the yolk sac (B) and vitelline arteries (C and D). (E and F) CD31 immunostaining of sections of yolk sacs. Arrows indicate sections of vitelline arteries. (G and H) Whole-mount immunofluorescence staining of CD31 in yolk sacs. In C–H, vitelline arteries were well developed on the yolk sacs of WT mice but poorly developed on those of RAMP2^–/– mice. (I–L) TUNEL staining of sections of vitelline artery (I and J) and umbilical vessel (K and L) in E13.5 WT and RAMP2^–/– embryos. Apoptosis was visualized in green fluorescence. Arrows indicate vessel lumens. Some ECs in RAMP2^–/– mice were TUNEL positive. (M and N) Severe systemic edema observed in RAMP2^–/–. Front (M) and side (N) views of WT and RAMP2^–/– embryos at midgestation. Some RAMP2^–/– embryos showed severe systemic edema. (O–R) Pericardial effusion in RAMP2^–/– mice. (O and P) Magnified side view of embryos at midgestation revealing the appearance of the pericardial space in RAMP2^–/– embryos. (Q and R) Sagittal sections showing the pericardial space in embryos at midgestation. The pericardial space was larger in RAMP2^–/– than WT embryos and showed the accumulation of pericardial effusion. (S–U) Severe hemorrhagic changes in RAMP2^–/– mice. (S) Side view of WT and RAMP2^–/– embryos at midgestation. (T and U) Sections of the liver at the same stage. Some RAMP2^–/– embryos showed severe hemorrhagic changes that were apparent on their surface and within the liver. Scale bars: 20 μm (E and F); 50 μm (I–L, T, and U); 200 μm (Q and R).

Macroscopic and histological observation of RAMP2^–/– embryos. At E13.5, well-developed vitelline arteries were detected on the yolk sacs of WT embryos, whereas RAMP2^–/– embryos had only poorly developed vitelline arteries (Figure 2, B–D). Histological examination revealed that the vitelline arteries from RAMP2^–/– embryos were smaller than those from WT embryos and appeared disorganized (Figure 2, E–H). That these phenotypes resembled those of AM^–/– embryos (14) showed that deletion of RAMP2 was sufficient to reproduce the major phenotypes of the vascular abnormality seen in AM^–/– mice. In RAMP2^–/– embryos, moreover, some of the ECs in the vitelline (Figure 2, I and J) and umbilical (Figure 2, K and L) arteries appeared apoptotic.

As for the embryos, the most apparent finding in RAMP2^–/– mice was severe systemic edema (Figure 2, M and N). They also showed accumulation of pericardial effusion suggestive of cardiac failure (Figure 2, O–R), and some had bleeding that was observable under the skin (Figure 2S) and within organs (Figure 2U). These phenotypes were also observed in AM^–/– embryos (14), although RAMP2^–/– mice showed systemic edema much more severe than that in AM^–/– mice.

Vascular abnormalities and gene expression in RAMP2^–/– mice. To determine whether the developmental abnormalities described above were the cause of the vascular fragility seen in RAMP2^–/– embryos, we analyzed in detail the vascular structure at E12.5. Electron microscopic observation of the ECs of the vitelline arteries in RAMP2^–/– embryos revealed deformity and detachment from the basement membrane (Figure 3B). Similar EC detachment was also detected in the liver (Figure 3D). In addition, there was abnormal thinning of the arterial walls in RAMP2^–/– embryos (WT, 1.75 ± 0.12 μm; RAMP2^–/–, 1.36 ± 0.05 μm; P < 0.05, n = 5 per group; Figure 3F), and the multiple layers of smooth muscle cells and basement membrane that normally comprise arterial walls were severely disrupted in RAMP2^–/– embryos (Figure 3H). We also found that expression of molecules involved in cell adhesion was altered in arteries from RAMP2^–/– mice. In particular, expression of vascular endothelial cadherin (VE-cadherin), claudin 5 (CDN5), and type IV collagen was diminished compared with WT mice (Figure 3I). These molecules are all important for the composition of tight junctions, adherence junctions, and the basement membrane of vascular ECs, and abnormalities involving them lead to paracellular leakage from the vascular lumen, which likely explains the severe edema seen in RAMP2^–/– mice.

Figure 3

Abnormalities of vascular structure and gene expression in RAMP2^–/– embryos.

Mechanisms underlying the angiogenesis and vascular stability mediated by the AM-RAMP2 system. To analyze the mechanisms underlying the angiogenesis and vascular stability mediated by RAMP2, we next generated EC lines that stably overexpressed RAMP2 (RAMP2O/E cells), utilizing EAhy926 ECs (Figure 4, A and B; see Methods for details). EAhy926 ECs are known to form capillary-like tubes on Matrigel (34, 35). We observed that RAMP2O/E cells showed much greater capillary formation than did control cells in Matrigel assays (Figure 4, C–E), clearly demonstrating that upregulation of the AM-RAMP2 system enhances angiogenesis.

Figure 4

Establishment and functional analysis of the RAMP2O/E line. (A) Plasmid vector used to overexpress RAMP2 (see Methods). (B) Western blot analysis of the membrane protein fraction from RAMP2O/E cells showing expression of the transfected gene. (C–E) Capillary formation by EAhy926 cells on Matrigel. RAMP2O/E cells or control ECs were cultured in 24-well culture plates coated with Matrigel in medium containing 10^–7 M AM, and capillary formation was monitored microscopically. (C) Capillary area relative to day-1 cell surface area. RAMP2O/E cells exhibited greater angiogenesis than control. n = 8 per group. (D and E) Representative photomicrographs of RAMP2O/E and control cells. (F) In vitro vascular permeability assay (see Methods). The permeability of the monolayer, assessed using a fluorescence microplate reader, is expressed relative to control at 5 min. RAMP2O/E cells showed significantly lower permeability than control ECs. n = 10 per group. *P < 0.05, **P < 0.01 vs. control. (G–J) Immunostaining of ZO-1. ECs were cultured until confluent on chamber slides in the presence of 10^-7 M AM. Two hours after treatment with 0.5 mM H₂O₂, the cells were immunostained using anti–ZO-1 antibody and Hoechst 33342. (K) Comparison of the tight junctions illustrated by the immunostaining in G–J. Tight junctions were better preserved after H₂O₂ treatment in RAMP2O/E cells than control ECs. **P < 0.01 vs. H₂O₂-treated control; comparison in 4 microscopic fields each from 3 independent experiments. (L) Quantitative real-time PCR analysis of gene expression in ECs cultured on Matrigel. Values are relative to control ECs treated with 10^–7 M AM. RAMP2O/E cells showed stronger expression of VEGF, eNOS, and CDN5 than control cells; this effect was blocked by LY294002 (10^-6 M) or a PKA inhibitor (10^-6 M). n = 6 per group. ^##P < 0.01 and ^#P < 0.05 vs. AM-treated control. **P < 0.01 and *P < 0.05 vs. AM-treated RAMP2O/E. Scale bars: 50 μm (D and E); 10 μm (G–J).

We then assessed endothelial barrier function by assaying vascular permeability in vitro. Cells were seeded onto semipermeable membranes in permeability chambers, after which the passage of FITC-dextran through confluent EC monolayers was monitored. Monolayers of RAMP2O/E cells were significantly less permeable than those of control cells, which suggests that upregulation of the AM-RAMP2 system also enhances vascular barrier function and reduces permeability (Figure 4F). We hypothesized that the reduced permeability reflected the firmer structure of the tight junctions formed by RAMP2O/E cells. To test this idea, we treated RAMP2O/E and control cells with H₂O₂ (0.5 mM), which leads to formation of intercellular gaps and reduced tight junction formation between ECs. Subsequent immunostaining of the tight junction marker zona occludens–1 (ZO-1) revealed that the structure was better preserved in RAMP2O/E than control cells (Figure 4, G–K). In addition, MTT and TUNEL assays revealed that RAMP2O/E cells showed significantly better viability after H₂O₂-induced damage than did control cells (data not shown).

We also found that expression of eNOS, VEGF, and CDN5 was upregulated in the RAMP2O/E cells and that treatment with the PI3K inhibitor LY294002 or a PKA inhibitor blocked those effects (Figure 4L). Thus, signaling via a PI3K- and PKA-dependent pathway appears to play a key role in AM-RAMP2 mediated angiogenesis and vascular stability. By contrast, RAMP3O/E cells did not show either enhanced angiogenesis or changes in permeability (data not shown). Apparently, the vascular functions of AM are exclusively regulated by RAMP2.

Reduced responses to angiogenic stimuli in adult RAMP2^+/– mice. Unlike their homozygous KO littermates, RAMP2^+/– mice survived until adulthood and were fertile, though aortic expression of RAMP2 was reduced to about half that seen in WT mice (Figure 5A), and they had higher BP than did WT mice (systolic BP, WT, 102.8 ± 2.2 mmHg; RAMP2^+/–, 112.9 ± 2.2 mmHg; P < 0.01, n = 9 per group). With acute infusion of AM (10 nmol/kg), RAMP2^+/– mice showed a smaller BP response than WT mice (maximum percent change in systolic BP, WT, 18.4 ± 1.4%; RAMP2^+/–, 13.3 ± 1.3%; P < 0.05, n = 6 per group). In contrast, CGRP-induced depressor effects did not differ in RAMP2^+/– and WT mice (maximum percent change in systolic BP, WT, 15.9 ± 1.2%; RAMP2^+/–, 14.4 ± 1.3%; NS, n = 6 per group). Interestingly, AM expression was significantly upregulated in the aortas of RAMP2^+/– mice, which suggests that reducing the number of functional AM receptors caused a compensatory upregulation of AM expression and implies that the AM-RAMP2 system is also important in the vascular function of adults. This prompted us to analyze the angiogenic properties of the AM-RAMP2 system in adult mice. We found that aortic ring explants cultured in collagen gel sprouted microvessels when stimulated with VEGF and that this angiogenic effect was greatly diminished in explants from RAMP2^+/– mice (Figure 5, B–E). Similarly, in Matrigel plug assays, RAMP2^+/– mice showed reduced neovascularization in response to stimulation with bFGF (Figure 5, F and G). We also cultured tissue from the aorta-gonad-mesonephros (AGM) regions of E10.5 embryos, which were plated on mouse OP9 stromal cells to promote angiogenesis. We found that there was substantially less development of a vascular network in tissue from RAMP2^–/– than WT mice (Figure 5, H and I).

Figure 5

Reduced responses to angiogenic stimuli in adult RAMP2^+/– mice. (A) Gene expression in aortas from 8-week-old WT and RAMP2^+/– mice. Quantitative real-time PCR analysis of total RNA extracted. In RAMP2^+/– mice, RAMP2 expression was about half that in WT mice, while AM expression was significantly upregulated. n = 6 per group. (B–E) Aortic ring assay. Representative photomicrographs of 7-day collagen gel cultures of aortas from 8-week-old mice (see Methods). Aortic explants from WT and RAMP2^+/– mice were cultured in the absence or presence of VEGF (50 ng/ml), and capillaries sprouting from the edges of the rings were analyzed. The aortic explants from RAMP2^+/– mice showed diminished angiogenesis. n = 6 per group. (F and G) Matrigel angiogenesis assay (see Methods). Representative photomicrographs showing angiogenesis in response to injected Matrigel. Capillary formation toward the Matrigel was greatly reduced in RAMP2^+/– (G) compared with WT (F). n = 6 per group. (H and I) In vitro culture of AGM explant (see Methods). Representative photomicrographs showing the vascular network formation from the tissue cultured AGM region of E10.5 WT (H) and RAMP2^–/– embryos (I). Vascular network formation was diminished in RAMP2^–/– mice. n = 4 per group.

Enhanced vascular permeability in adult RAMP2^+/– mice. Based on the findings presented thus far, we hypothesized that the AM-RAMP2 system regulates vascular stability in adults as well as during development, which we tested by analyzing vascular permeability in adult RAMP2^+/– mice. We initially generated a footpad edema model by subcutaneously injecting λ-carrageenan, a sulfated high-MW polygalactan. As expected, RAMP2^+/–mice showed greater swelling than did WT mice (Figure 6A). We then directly measured vascular permeability using a skin edema model. Mice were injected with FITC-BSA via the tail vein as a tracer of vascular permeability, after which serum exudation caused by subcutaneous injection of histamine was measured using a fluorescence microplate reader. As shown in Figure 6B, RAMP2^+/– mice exhibited greater vascular permeability than did WT mice.

Figure 6

In vivo vascular permeability assay. (A) Footpad edema model. λ-Carrageenan was injected into the footpad of 8-week-old RAMP2^+/– and WT mice to induce edema for the evaluation of vascular permeability in adult mice; swelling of the footpad was measured hourly using a thickness gauge. RAMP2^+/– mice showed significantly greater swelling than WT mice. n = 12 per group. **P < 0.01, *P < 0.05 vs. WT. (B) Skin edema model (see Methods). Fluorescence intensity was measured using a fluorescence microplate reader. Permeability levels are presented relative to WT. RAMP2^+/– mice (n = 8) showed significantly greater vascular permeability than WT mice (n = 13). **P < 0.01 vs. WT. (C) Brain edema model (see Methods). Vascular permeability in RAMP2^+/– mice (n = 12) is presented relative to that in WT mice (n = 10). RAMP2^+/– mice showed significantly greater vascular permeability than WT mice. *P < 0.05 vs. WT.

Finally, to prove that the capacity of the AM-RAMP2 system to regulate vascular permeability could in fact make it a useful therapeutic target, we generated a brain edema model in which edema was caused by injuring the brain using a liquid nitrogen–cooled copper probe. Twenty-four hours after the injury, RAMP2^+/– mice showed greater vascular permeability than did WT mice (Figure 6C), which suggests that the AM-RAMP2 system is important for maintenance of the blood-brain barrier.

The mechanism by which a stable and functional vascular network is generated and regulated by humoral factors is still not fully understood. For example, VEGF alone is insufficient for stable vessel formation (36) and presents major disadvantages for therapeutic angiogenesis, in that it increases vascular permeability and may exacerbate arteriosclerosis.

Studies of gene-targeted mice have led to the identification of angiogenic factors that had not previously been recognized for their angiogenic properties. AM was originally identified as a vasodilator, although it is now known to possess a variety of biological activities. Indicative of AM’s novel angiogenic properties is our previous finding that AM^–/– embryos die in utero due to hemorrhage and edema resulting from abnormalities of vascular development (14). We also showed previously that exogenous administration of AM enhances angiogenesis in ischemic tissues in adults, and therapeutic application of AM is much anticipated (25). Gene-targeted mice also provide information about the fundamental roles played by AM during the multistep process of angiogenesis. It is noteworthy, for instance, that AM^–/– embryos die at a relatively late stage of development (E13.5–E14.5) compared with KO mice lacking other substances classified as angiogenic factors. This suggests that the vasculature does develop in AM^–/– mice, but its fragile structure is likely disrupted after the start of circulation. It also clearly shows that AM is essential not only for angiogenesis, but also for vascular integrity.

As with other growth factors, the clinical applicability of AM has 2 serious limitations: AM is a peptide with a short half-life in the bloodstream, and the cost of the recombinant protein makes its use in the treatment of chronic diseases impractical. This prompted us to focus on AM’s receptor system. McLatchie et al. showed that AM signaling is regulated by a unique control system (28). The main body of the AM receptor is thought to be CRLR, a 7–transmembrane domain GPCR. CRLR associates with 1 of 3 subtypes of RAMP, which determines the affinity of CRLR for its ligands. By generating RAMP2-specific KO mice, we have been able to demonstrate that RAMP2 is the key determinant of AM’s function during vascular development. In our RAMP2^–/– mice, CRLR and the other RAMPs were preserved; nevertheless, deletion of RAMP2 was sufficient to reproduce the phenotypes of the AM^–/– genotype. Our finding that AM expression was upregulated in RAMP2^–/– embryos just before their death further confirms that RAMP2 is essential for AM signaling during vascular development.

In both AM^–/– and RAMP2^–/– embryos, vascular fragility ultimately leads to hemorrhage and edema. Notably, however, the systemic edema was much more severe in RAMP2^–/– mice. Edema was sometimes detected in AM^–/– mice, but its severity varied. Moreover, administration of recombinant AM to crossbred AM^+/– females increased the survival rate of AM^–/– embryos at E14.5 (14), suggesting that maternally supplied AM partially compensates for the lack of embryonic AM expression. By contrast, RAMP2^–/– mice cannot express a functional AM receptor in their vasculature and thus can not respond to maternal AM.

We found that neovascularization was diminished and vascular permeability was increased in adult RAMP2^+/– mice, which showed reduced expression of RAMP2. We also found that RAMP2^+/– mice had higher BP than did their WT littermates, which confirms that RAMP2 continues to be a crucial determinant of vascular function in the adult. Interestingly, we also found that the edema developed by RAMP2^+/– mice in various disease models was more severe than that in WT mice, suggesting the AM-RAMP2 system could be an attractive therapeutic target for treating the edema often associated with vascular regenerative therapies, brain trauma, and infarction. In that regard, it is noteworthy that we were able to modulate the vascular function of AM by modulating RAMP2. Using RAMP2O/E cells, we clearly showed that by upregulating RAMP2 signaling, we could enhance capillary formation, firm up tight junctions, and reduce vascular permeability. RAMP2O/E cells were also resistant to apoptosis. Thus, RAMP2 could be a therapeutic target by which to manipulate the vascular functions of AM.

By contrast, RAMP3O/E cell lines did not show either enhanced angiogenesis or improved vascular stability, although RAMP3 has previously been shown to work with CRLR to function as another AM receptor (37). Furthermore, our finding that RAMP3 was expressed at WT levels in RAMP2^–/– mice confirmed that RAMP3 cannot compensate for the absence of RAMP2 during vascular development. Consistent with the distinctly different physiological roles played by RAMP2 and RAMP3, RAMP3^–/– mice live apparently normally until old age (38). In addition, whereas RAMP2 and CRLR are downregulated in an endotoxemia model, RAMP3 is markedly upregulated (39), and it has been suggested that RAMP3 may be involved in post-endocytic receptor trafficking, as it presents a PDZ type I domain (40).

Signal transduction via GPCRs and the regulation of their function has long attracted the interest of many researchers. Indeed, about 40% of the drugs in clinical use today target GPCRs. We suggest that RAMP2 is an alternative therapeutic target by which to affect CRLR function. Because RAMP2 is a low-MW protein, structural analysis and the synthesis of specific agonists or antagonists are much more realistic for RAMP2 than for 7–transmembrane domain GPCRs, which has proven difficult. Moreover, because RAMP2 determines the vascular functions of AM, it would be expected that greater specificity would be achieved by targeting RAMP2 than by targeting CRLR, which can also function as a receptor for other ligands. In that context, our findings provide a clear basis for the development of drugs to modulate RAMP2 and, thereby, the vascular effects of AM.

We thank C.J. Edgell for EAhy926 cells. This study was supported by a grant from the Takeda Medical Research Foundation; a Japan Heart Foundation Research Grant; a Grant for Research on Cardiovascular Disease from the Tanabe Medical Conference; a grant from the Novartis Foundation for Gerontological Research; an Astra Zeneca Research Grant; grants from the Mitsui Life Social Welfare Foundation, the Ichiro Kanehara Foundation, the Uehara Memorial Foundation, the Sankyo Foundation of Life Science, the Naito Foundation, the Mitsubishi Pharma Research Foundation, and the Salt Science Research Foundation; Research Grant for Cardiovascular Disease 19C-7 from the Ministry of Heath, Labour and Welfare; and a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan.

Address correspondence to: Takayuki Shindo, Department of Organ Regeneration, Shinshu University Graduate School of Medicine, 3-1-1 Asahi, Matsumoto 390-8621, Japan. Phone: 81-263-37-3192; Fax: 81-263-37-3437; E-mail: t-shindo@sch.md.shinshu-u.ac.jp.

Nonstandard abbreviations used: AGM, aorta-gonad-mesonephros; AM, adrenomedullin; CDN5, claudin 5; CGRP, calcitonin gene–related peptide; CRLR, calcitonin receptor–like receptor; RAMP, receptor activity–modifying protein; RAMP2O/E, RAMP2-overexpressing (cell); ZO-1, zona occludens–1.

Conflict of interest: The authors have declared that no conflict of interest exists.

Reference information: J. Clin. Invest.118:29–39 (2008). doi:10.1172/JCI33022

See the related articles at Blood is thicker than lymph, Adrenomedullin signaling is necessary for murine lymphatic vascular development, and Adrenomedullin signaling is necessary for murine lymphatic vascular development.

1. Kitamura, K., et al. 1993. Adrenomedullin: a novel hypotensive peptide isolated from human pheochromocytoma. Biochem. Biophys. Res. Commun. 192:553-560.
   View this article via: CrossRef PubMed Google Scholar
2. Michibata, H., et al. 1998. Autocrine/paracrine role of adrenomedullin in cultured endothelial and mesangial cells. Kidney Int. 53:979-985.
   View this article via: PubMed Google Scholar
3. Kitamura, K., et al. 1994. Immunoreactive adrenomedullin in human plasma. FEBS Lett. 341:288-290.
   View this article via: CrossRef PubMed Google Scholar
4. Jougasaki, M., et al. 1995. Renal localization and actions of adrenomedullin: a natriuretic peptide. Am. J. Physiol. 268:F657-F663.
   View this article via: PubMed Google Scholar
5. Nishikimi, T., Matsuoka, H. 2005. Cardiac adrenomedullin: its role in cardiac hypertrophy and heart failure. Curr. Med. Chem. Cardiovasc. Hematol. Agents. 3:231-242.
   View this article via: CrossRef PubMed Google Scholar
6. Samson, W.K., Murphy, T., Schell, D.A. 1995. A novel vasoactive peptide, adrenomedullin, inhibits pituitary adrenocorticotropin release. Endocrinology. 136:2349-2352.
   View this article via: CrossRef PubMed Google Scholar
7. Petrie, M.C., Hillier, C., Morton, J.J., McMurray, J.J. 2000. Adrenomedullin selectively inhibits angiotensin II-induced aldosterone secretion in humans. J. Hypertens. 18:61-64.
   View this article via: PubMed Google Scholar
8. Isumi, Y., Kubo, A., Katafuchi, T., Kangawa, K., Minamino, N. 1999. Adrenomedullin suppresses interleukin-1beta-induced tumor necrosis factor-alpha production in Swiss 3T3 cells. FEBS Lett. 463:110-114.
   View this article via: CrossRef PubMed Google Scholar
9. Shimosawa, T., et al. 2002. Adrenomedullin, an endogenous peptide, counteracts cardiovascular damage. Circulation. 105:106-111.
   View this article via: CrossRef PubMed Google Scholar
10. Shimosawa, T., et al. 2003. Deficiency of adrenomedullin induces insulin resistance by increasing oxidative stress. Hypertension. 41:1080-1085.
    View this article via: CrossRef PubMed Google Scholar
11. Kano, H., et al. 1996. Adrenomedullin as a novel antiproliferative factor of vascular smooth muscle cells. J. Hypertens. 14:209-213.
    View this article via: CrossRef PubMed Google Scholar
12. Miyashita, K., et al. 2003. Adrenomedullin promotes proliferation and migration of cultured endothelial cells. Hypertens. Res. 26(Suppl.):S93-S98.
13. Iwasaki, H., Eguchi, S., Shichiri, M., Marumo, F., Hirata, Y. 1998. Adrenomedullin as a novel growth-promoting factor for cultured vascular smooth muscle cells: role of tyrosine kinase-mediated mitogen-activated protein kinase activation. Endocrinology. 139:3432-3441.
    View this article via: CrossRef PubMed Google Scholar
14. Shindo, T., et al. 2001. Vascular abnormalities and elevated blood pressure in mice lacking adrenomedullin gene. Circulation. 104:1964-1971.
    View this article via: CrossRef PubMed Google Scholar
15. Shindo, T., et al. 2000. Hypotension and resistance to lipopolysaccharide-induced shock in transgenic mice overexpressing adrenomedullin in their vasculature. Circulation. 101:2309-2316.
    View this article via: PubMed Google Scholar
16. Oh-hashi, Y., et al. 2001. Elevated sympathetic nervous activity in mice deficient in alphaCGRP. Circ. Res. 89:983-990.
    View this article via: CrossRef PubMed Google Scholar
17. Niu, P., et al. 2004. Protective effects of endogenous adrenomedullin on cardiac hypertrophy, fibrosis, and renal damage. Circulation. 109:1789-1794.
    View this article via: CrossRef PubMed Google Scholar
18. Niu, P., et al. 2003. Accelerated cardiac hypertrophy and renal damage induced by angiotensin II in adrenomedullin knockout mice. Hypertens. Res. 26:731-736.
    View this article via: CrossRef PubMed Google Scholar
19. Nishimatsu, H., et al. 2002. Role of endogenous adrenomedullin in the regulation of vascular tone and ischemic renal injury: studies on transgenic/knockout mice of adrenomedullin gene. Circ. Res. 90:657-663.
    View this article via: CrossRef PubMed Google Scholar
20. Imai, Y., et al. 2002. Resistance to neointimal hyperplasia and fatty streak formation in mice with adrenomedullin overexpression. Arterioscler. Thromb. Vasc. Biol. 22:1310-1315.
    View this article via: CrossRef PubMed Google Scholar
21. Yamamoto, H., et al. 2007. Adrenomedullin insufficiency increases allergen-induced airway hyperresponsiveness in mice. J. Appl. Physiol. 102:2361-2368.
    View this article via: CrossRef PubMed Google Scholar
22. Kurihara, H., Shindo, T., Oh-Hashi, Y., Kurihar, Y., Kuwaki, T. 2003. Targeted disruption of adrenomedullin and alphaCGRP genes reveals their distinct biological roles. Hypertens. Res. 26(Suppl.):S105-S108.
23. Nishimatsu, H., et al. 2003. Endothelial responses of the aorta from adrenomedullin transgenic mice and knockout mice. Hypertens. Res. 26(Suppl.):S79-S84.
24. Aoki-Nagase, T., et al. 2002. Attenuation of antigen-induced airway hyperresponsiveness in CGRP-deficient mice. Am. J. Physiol. Lung Cell Mol. Physiol. 283:L963-L970.
    View this article via: PubMed Google Scholar
25. Iimuro, S., et al. 2004. Angiogenic effects of adrenomedullin in ischemia and tumor growth. Circ. Res. 95:415-423.
    View this article via: CrossRef PubMed Google Scholar
26. Fujii, T., et al. 2005. Adrenomedullin enhances therapeutic potency of bone marrow transplantation for myocardial infarction in rats. Am. J. Physiol. Heart Circ. Physiol. 288:H1444-H1450.
    View this article via: CrossRef PubMed Google Scholar
27. Nagaya, N., et al. 2005. Adrenomedullin: angiogenesis and gene therapy. Am. J. Physiol. Regul. Integr. Comp. Physiol. 288:R1432-R1437.
    View this article via: PubMed Google Scholar
28. McLatchie, L.M., et al. 1998. RAMPs regulate the transport and ligand specificity of the calcitonin-receptor-like receptor. Nature. 393:333-339.
    View this article via: CrossRef PubMed Google Scholar
29. Kuwasako, K., Cao, Y.N., Nagoshi, Y., Kitamura, K., Eto, T. 2004. Adrenomedullin receptors: pharmacological features and possible pathophysiological roles. Peptides. 25:2003-2012.
    View this article via: CrossRef PubMed Google Scholar
30. Parameswaran, N., Spielman, W.S. 2006. RAMPs: The past, present and future. Trends Biochem. Sci. 31:631-638.
    View this article via: CrossRef PubMed Google Scholar
31. Morfis, M., Christopoulos, A., Sexton, P.M. 2003. RAMPs: 5 years on, where to now? Trends Pharmacol. Sci. 24:596-601.
    View this article via: CrossRef PubMed Google Scholar
32. Montuenga, L.M., Martinez, A., Miller, M.J., Unsworth, E.J., Cuttitta, F. 1997. Expression of adrenomedullin and its receptor during embryogenesis suggests autocrine or paracrine modes of action. Endocrinology. 138:440-451.
    View this article via: CrossRef PubMed Google Scholar
33. Montuenga, L.M., Mariano, J.M., Prentice, M.A., Cuttitta, F., Jakowlew, S.B. 1998. Coordinate expression of transforming growth factor-beta1 and adrenomedullin in rodent embryogenesis. Endocrinology. 139:3946-3957.
    View this article via: CrossRef PubMed Google Scholar
34. Bauer, J., et al. 1992. In vitro model of angiogenesis using a human endothelium-derived permanent cell line: contributions of induced gene expression, G-proteins, and integrins. J. Cell Physiol. 153:437-449.
    View this article via: CrossRef PubMed Google Scholar
35. Edgell, C.J., Curiel, D.T., Hu, P.C., Marr, H.S. 1998. Efficient gene transfer to human endothelial cells using DNA complexed to adenovirus particles. Biotechniques. 25:264- 268 , 270–272 .
36. Lee, R.J., et al. 2000. VEGF gene delivery to myocardium: deleterious effects of unregulated expression. Circulation. 102:898-901.
    View this article via: PubMed Google Scholar
37. Sexton, P.M., Albiston, A., Morfis, M., Tilakaratne, N. 2001. Receptor activity modifying proteins. Cell Signal. 13:73-83.
    View this article via: CrossRef PubMed Google Scholar
38. Dackor, R.T., Fritz-Six, K.L., Smithies, O., Caron, K.M. 2007. Receptor activity modifying proteins 2 and 3 have distinct physiological functions from embryogenesis to old age. J. Biol. Chem. 282:18094-18099.
    View this article via: CrossRef PubMed Google Scholar
39. Ono, Y., Okano, I., Kojima, M., Okada, K., Kangawa, K. 2000. Decreased gene expression of adrenomedullin receptor in mouse lungs during sepsis. Biochem. Biophys. Res. Commun. 271:197-202.
    View this article via: CrossRef PubMed Google Scholar
40. Bomberger, J.M., Parameswaran, N., Hall, C.S., Aiyar, N., Spielman, W.S. 2005. Novel function for receptor activity-modifying proteins (RAMPs) in post-endocytic receptor trafficking. J. Biol. Chem. 280:9297-9307.
    View this article via: PubMed Google Scholar
41. Shindo, T., et al. 2000. ADAMTS-1: a metalloproteinase-disintegrin essential for normal growth, fertility, and organ morphology and function. J. Clin. Invest. 105:1345-1352.
    View this article via: JCI PubMed Google Scholar
42. Shindo, T., et al. 2002. Kruppel-like zinc-finger transcription factor KLF5/BTEB2 is a target for angiotensin II signaling and an essential regulator of cardiovascular remodeling. Nat. Med. 8:856-863.
    View this article via: PubMed Google Scholar
43. Lyons, G.E., Schiaffino, S., Sassoon, D., Barton, P., Buckingham, M. 1990. Developmental regulation of myosin gene expression in mouse cardiac muscle. J. Cell Biol. 111:2427-2436.
    View this article via: CrossRef PubMed Google Scholar
44. Zhu, W.H., Iurlaro, M., MacIntyre, A., Fogel, E., Nicosia, R.F. 2003. The mouse aorta model: influence of genetic background and aging on bFGF- and VEGF-induced angiogenic sprouting. Angiogenesis. 6:193-199.
    View this article via: CrossRef PubMed Google Scholar
45. Hamaguchi, I., et al. 1999. In vitro hematopoietic and endothelial cell development from cells expressing TEK receptor in murine aorta-gonad-mesonephros region. Blood. 93:1549-1556.
    View this article via: PubMed Google Scholar
46. Sakurai, K., et al. 1997. Anti-inflammatory activity of superoxide dismutase conjugated with sodium hyaluronate. Glycoconj. J. 14:723-728.
    View this article via: CrossRef PubMed Google Scholar
47. Yamaki, K., Takano-Ishikawa, Y., Goto, M., Kobori, M., Tsushida, T. 2002. An improved method for measuring vascular permeability in rat and mouse skin. J. Pharmacol. Toxicol. Methods. 48:81-86.
    View this article via: CrossRef PubMed Google Scholar
48. Hortobagyi, T., et al. 2000. A novel brain trauma model in the mouse: effects of dexamethasone treatment. Pflugers Arch. 441:409-415.
    View this article via: CrossRef PubMed Google Scholar
49. Schoch, H.J., Fischer, S., Marti, H.H. 2002. Hypoxia-induced vascular endothelial growth factor expression causes vascular leakage in the brain. Brain. 125:2549-2557.
    View this article via: PubMed CrossRef Google Scholar