VPAC1 receptors play a dominant role in PACAP-induced vasorelaxation in female mice

Background PACAP and VIP are closely related neuropeptides with wide distribution and potent effect in the vasculature. We previously reported vasomotor activity in peripheral vasculature of male wild type (WT) and PACAP-deficient (KO) mice. However, female vascular responses are still unexplored. We hypothesized that PACAP-like activity is maintained in female PACAP KO mice and the mechanism through which it is regulated differs from that of male PACAP KO animals. Methods We investigated the vasomotor effects of VIP and PACAP isoforms and their selective blockers in WT and PACAP KO female mice in carotid and femoral arteries. The expression and level of different PACAP receptors in the vessels were measured by RT-PCR and Western blot. Results In both carotid and femoral arteries of WT mice, PACAP1-38, PACAP1-27 or VIP induced relaxation, without pronounced differences between them. Reduced relaxation was recorded only in the carotid arteries of KO mice as compared to their WT controls. The specific VPAC1R antagonist completely blocked the PACAP/VIP-induced relaxation in both arteries of all mice, while PAC1R antagonist affected relaxation only in their femoral arteries. Conclusion In female WT mice, VPAC1 receptors appear to play a dominant role in PACAP-induced vasorelaxation both in carotid and in femoral arteries. In the PACAP KO group PAC1R activation exerts vasorelaxation in the femoral arteries but in carotid arteries there is no significant effect of the activation of this receptor. In the background of this regional difference, decreased PAC1R and increased VPAC1R availability in the carotid arteries was found.


Methods
We investigated the vasomotor effects of VIP and PACAP isoforms and their selective blockers in WT and PACAP KO female mice in carotid and femoral arteries. The expression and level of different PACAP receptors in the vessels were measured by RT-PCR and Western blot.

Results
In both carotid and femoral arteries of WT mice, PACAP1-38, PACAP1-27 or VIP induced relaxation, without pronounced differences between them. Reduced relaxation was recorded only in the carotid arteries of KO mice as compared to their WT controls. The specific VPAC1R antagonist completely blocked the PACAP/VIP-induced relaxation in both PLOS
This action is mediated through all three PACAP receptors localized mainly on the surface of the smooth muscle in arteries and arterioles [14,25]. Although highly expressed, receptors are not detected equally across the vasculature. Receptors can be found in the small pulmonary arterioles and cerebral microvessels and also in large vessels like the aorta [7,8,14,23,26]. We have also confirmed the presence of PAC1R and VPAC1R in carotid and femoral artery of male mice [17].
Utilization of PACAP-deficient mouse model (knockout-KO) enables insight of physiological roles of PACAP both in vivo and in vitro. A thorough analysis of PACAP KO mice reported a number of different abnormalities resulting in several pathological alterations in different organs and higher mortality rate. Among others, PACAP deficiency leads to developmental alterations [27][28][29] and increased vulnerability to harmful stimuli as well as accelerated aging [30][31][32]. Our recent results show that PACAP deficiency leads to accelerated age-related systemic amyloidosis, with more severe and more generalized appearance of amyloid deposits in several organs in knockout mice [33].
The cardiovascular system is also affected by PACAP-deficiency. A recent study reported reduced dilatator ability of meningeal arteries upon nitroglycerin treatment in PACAP KO mice [18]. Another abnormality is related to the heart, where lack of PACAP leads to degeneration of the myocardium, increased fibrosis and reduced cardiac function [34,35]. Our group reported increased arterial relaxation to PACAP1-27 and VIP and absence of relaxation to PACAP1-38 in male PACAP KO mice [17]. Although previously we did investigate how PACAP-deficiency influences vasomotor responses in males, such studies have not been conducted in females.
With regard to gender differences in various PACAP effects, the literature reported certain clinically important phenomena that differ in male and female groups. PACAP deficiency leads to decreased tear secretion and dry eye symptoms, which is more pronounced in female mice with aging [31]. The PACAP-PAC1 receptor pathway has been demonstrated to play a role in the human psychological stress responses. This study also revealed that in women the perturbations in the PACAP-PAC1 pathway are involved in abnormal stress responses [36]. Another group reported that a variant in the gene encoding PAC1 is associated with post-traumatic stress disorders in females [37] and later the role of estrogen was confirmed in this gender-dependent association [38]. As another significant clinical aspect, PACAP receptor emerged as a target in the therapy of migraine [39][40][41]. Migraine is one of the most common neurological disorders that is thought to be elicited by cerebral and meningeal arterial vasodilation [39][40][41]. Therefore, a thorough assessment of the vascular responses also in females is an important question for safe clinical use of such new drugs.
Our main question was whether the vascular response of isolated arteries to PACAP isoforms will be altered in female mice. Therefore, we aimed to investigate the relaxation properties of the carotid and femoral arteries of PACAP wild-type (WT) and KO female mice to cumulative administration of PACAP1-38, PACAP1-27 and VIP and in the presence of selective receptor blockers. We also aimed to investigate the presence of the various PACAP receptors in carotid and femoral arteries of female WT and PACAP KO mice.

Animals
Experiments were performed on 8-12 week-old female PACAP-KO mice on a CD-1 background and their WT littermates (total n = 54). Our study has been approved by the University of Pecs Ethical Committee for the Protection of Animals in Research (BA02/2000-15024/ 2011). All procedures were in accordance with the main directives of the National Ethical Council for Animal Research and those of the European Communities Council (86/609/EEC, Directive 2010/63/EU of the European Parliament and of the Council).

Surgery
The common carotid arteries and the proximal part of the femoral arteries were isolated using an Olympus surgical microscope (model SZX7; Olympus Inc., Japan) under anesthesia induced by intraperitoneal injection of ketamine (Gedeon Richter Plc., Budapest, Hungary) and xylazine (Eurovet Animal Health B.V., Bladel, Netherland) mixture (81.7 and 9.3 mg/kg, respectively). The proximal and distal ends of the isolated segment were ligated; the vessels were excised between the ligations, and then transferred to refrigerated Krebs solution. Carotid and femoral arteries of both sides were used. After the removal of the arteries, the animal was euthanized with an intraperitoneal injection of pentobarbital (100 mg/kg, Ceva Sante Animale, Libourna, France). Before every surgery, the female estrus cycle was tested by vaginal cytology [42].

Measurement of isometric force of isolated arteries
Measurements of the isometric force of isolated arteries were according to Mulvany's protocol and our previous experiments [17,43]. After removal of the arteries, they were quickly transferred into cold oxygenated (95% O 2 / 5% CO 2 ) (Linde, Repcelak, Hungary) physiological Krebs solution (NaCl: 119 mM, KCl: 4.7 mM, KH 2 PO 4 : 1.2 mM, NaHCO 3 : 25 mM, Mg 2 SO 4 : 1.2 mM, CaCl 2 × 2H 2 O: 1.6 mM, EDTA: 0.026 mM, glucose: 11.1 mM). NaCl and KCl were purchased from VWR International (Radnor, PA, USA). All other chemicals and drugs were obtained from Sigma-Aldrich (St. Louis, MO, USA). The arteries were dissected into 2 mm long rings. Each ring was positioned between two tungsten wires (diameter of the wire was 0.04 mm for CA and 0.02 mm for FA) in the organ chamber of the myograph in 5 ml Krebs bath solution. The bath solution was continuously oxygenated with a gas mixture of 95% O 2 and 5% CO 2 , and kept at 36.9±0.1˚C.
Isometric contraction forces were measured with a DMT 610M Wire Myograph (Danish Myo Technology, Aarhus, Denmark). Normalization was performed according to Mulvany and Harpern [43]. LabChart 8 (AD Instruments, Dunedin, New Zealand) and Myodaq 2.01 (Danish Myotechnologies, Denmark) software were used for data acquisition and display. After normalization, vessels were allowed to stabilize for 60 minutes.

Pharmacological agents
At the beginning of the experiments, the functional integrity of the vessels was verified with viability tests. To study the endothelium-dependent [acetylcholine (Ach)-induced] and -independent [sodium nitroprusside (SNP)-induced] relaxation ability of the vessels first the vessels were pre-contracted with 60 mM KCl. When the contraction reached the plateau phase, 50 μl of increasing doses of either Ach or SNP were administered directly into the bath solution to reach final concentrations of 10 −9 M to 10 −5 M. After this viability test, the chambers were washed out with Krebs solution. Vascular responses of the vessels were measured in response to increasing doses of different vasoactive substances in a similar fashion. First, 60 mM KCl was administered to establish a tone [21,44]. Once the vessel reached plateau phase, the pharmacological agent was applied directly into the vessel chamber to reach final concentration. After the administration of each dose of a specific substance, the isometric force was registered. We used PACAP1-38, PACAP1-27 and VIP with final concentrations of 10 -9 M to 10 -6 M (Bachem, Bubendorf, Switzerland). PACAP1-38 and PACAP1-27 were synthesized as previously described [45]. Selective agonists for PAC1R (maxadilan; Tocris, Bioscience, Bristol, UK), VPAC1R (Ala 11,22,28 VIP; Bachem) and VPAC2R (Bay55-9837; Bachem) were used (10 -10 M to 10 -7 M). In addition, antagonists for PAC1R/VPAC1R/VPAC2R (PACAP6-38, 10 -7 M), PAC1R (M65, 10 -7 M) and VPAC1R (VIP6-28, 10 -7 M) were also used (Bachem).
All drugs were dissolved in distilled water. When only the solvent (distilled water) was administrated, there was no change in force. Changes in the vasomotor activity were measured by the difference compared to maximal contraction induced by 60 mM KCl (in graphs marked as a baseline, 10 -0 M), for each administered drug, artery and mice genotype.

RT-PCR analysis
Tissues were cryo-ground in liquid nitrogen and dissolved in Trizol (Applied Biosystems, Foster City, CA, USA), and after the addition of 20% RNase free chloroform samples were centrifuged at 4˚C at 10,000×g for 15 min. Samples were incubated in 500 μL of RNase-free isopropanol at -20˚C for 1 h then total RNA was harvested in RNase free water and stored at -  Table 1 for 1 min; extension, 72˚C, 90 sec) and then 72˚C, 10 min. PCR products were analyzed by electrophoresis in 1.2% agarose gel containing ethidium bromide. Actin was used as internal control. Signals were developed with gel documentary system (Fluorchem E, ProteinSimple, CA, USA). The optical density of signals was measured by using ImageJ 1.40g freeware and results were normalized to the optical density of control tissue.

Statistical analysis
All data represented here were collected in series, as single-point measurement, compared across genotypes and dose points by two-way ANOVA (post hoc-Tukey). For Western Blot analysis, the student t-test was used. Statistical analyses were performed using Sigma Plot 12.5 (Systat, Chicago, IL, USA). Significate difference value was set at p <0.05. The data are reported as mean ± SEM.

Administration of PACAP1-38, PACAP1-27 and VIP leads to relaxation of carotid and femoral arteries in female mice
Original records (Fig 1) and summary data (Fig 2A-2C (Fig 2D-2F). The female estrus cycle did not influence these vasomotor responses of the arteries (S4 Fig).
As compared with females, PACAP-and VIP-induced arterial relaxations were significantly stronger both in male WT and in male PACAP KO mice as shown by S1 Fig.

Selective PAC1R antagonist M65 blocked PACAP1-38-, PACAP1-27-and VIP-induced relaxation in femoral but not in carotid arteries in female mice
Summary data show the effects of cumulative doses of PACAP1-38, PACAP1-27 and VIP with or without the presence of PAC1R antagonist M65 on the vasomotor responses in the carotid (Fig 3A-3C) and femoral arteries (Fig 3D-3F) of WT and PACAP KO mice. In the carotid arteries of both WT and PACAP KO mice, the presence of M65 had no significant impact on PACAP-and VIP-induced responses (Fig 3A-3C). However, in femoral arteries, the administration of M65 significantly reduced PACAP-and VIP-induced relaxation in both WT and PACAP KO mice at 10 −7 -10 -6 M (for PACAP1-27 and VIP in WT mice already at 10 -8 M) (Fig  3D and 3E).  (Fig 4A-4C) and femoral arteries (Fig  4D-4F) of WT and PACAP KO mice. In the arteries of both WT and KO mice, the presence of VIP6-28 resulted in blockade of PACAP-and VIP-induced vasomotor responses as PACAP-induced vasorelaxation in female mice compared to their controls (Fig 4A-4F). The difference was always significant at 10 -6 M in WT and KO mice in both arteries for all examined substances. For additional significant differences see

Selective PAC1R agonist maxadilan, VPAC1R agonist Ala 11,22,28 VIP, but not VPAC2R agonist Bay55-9837 induced relaxation in female mice
Summary data show the effects of cumulative doses of receptor agonists in the carotid (Fig 5A-5C) and femoral arteries (Fig 5D-5F) of WT and PACAP KO mice with or without their respective selective blockers. Selective PAC1R agonist maxadilan induced significant relaxations of carotid and femoral arteries only at the highest concentration (10 -7 M) (Fig 5A and 5D). PAC1R effect was remarkable stronger on femoral arteries of PACAP KO mice at this concentration compared to relaxation effect of WT mice (Fig 5D). However, selective PAC1R antagonist M65 had no effect on the strong femoral relaxation of PACAP KO mice (Fig 5A and 5D). Selective VPAC1 agonist Ala 11,22,28 VIP elicited significant relaxation in femoral arteries of both WT and PACAP KO mice at 10 −8 -10 -7 M dose (on carotid arteries of PACAP KO mice only at 10 -7 M) (Fig 5B and 5E). There was a significant difference between WT and PACAP KO mice in the femoral arteries (10 −8 -10 -7 M), where the selective VPAC1 agonist Ala 11,22,28-VIP induced greater relaxations in the PACAP KO mice compared to WT mice. The selective VPAC1R antagonist VIP6-28 completely inhibited all the above mentioned VPAC1R agonist Ala 11,22,28 VIP-induced relaxations.
On the other hand, selective VPAC2R agonist Bay55-9837 had no significant effect on vasomotor responses (Fig 5C and 5F). Fig 6 shows the mRNA expressions of PAC1R, VPAC1R, and VPAC2R detected by RT-PCR in carotid and femoral arteries of WT and PACAP KO mice. The analysis showed reduced expression of PAC1R and increased expression of VPAC1R in the arteries of PACAP KO mice as compared to WT controls (t-test, � p < 0.05 vs. control) (Fig 6). No VPAC2R expression was detected in either sample.

Arteries of female PACAP KO mice express increased amount of VPAC1R and decreased amount of PAC1R compared to those of WT mice
Protein levels analyzed by Western blot were in accord with the expression findings (t-test, � p < 0.05 vs. control). No signal for PAC1R protein was detected in carotid arteries of PACAP KO mice (S2 Fig).

Arterial viability tests
In both types of arteries of PACAP KO mice, a 60mM KCl elicited significantly greater contractions as compared to their WT controls (S3A Fig).
The magnitudes of ACh-(endothelium dependent) and SNP-induced (endothelium independent) arterial relaxations were similar in WT and PACAP KO mice. The SNP-induced responses in femoral arteries were stronger compared to those of carotid arteries (S3B and  S3C Fig).

Discussion
Our results showed that arteries of female mice reacted to administration of PACAP/VIP with vasorelaxation, similar to our previous observations in male mice [17]. However, vessels of female mice show weaker relaxations compared to males. Females also seem to be less sensitive to the lack of PACAP than males. The main novel finding of the current study is, that in female mice, VPAC1 receptors appear to play a dominant role in PACAP-induced vasorelaxation  both in carotid and in femoral arteries. While in the femoral arteries PAC1R activation exerts vasorelaxation, in carotid arteries of PACAP KO mice, there is no significant effect of the activation of this receptor.
These findings are supported by multiple lines of evidence: 1) functional experiments show, that a selective VPAC1R antagonist is able to decrease the vasomotor effect of all PACAP isoforms both in carotid and in femoral arteries in both WT and PACAP KO mice; 2) both RT-PCR and the Western blot verify the VPAC1R upregulation together with the reduction of PAC1R mRNA and protein levels in PACAP KO mice.

Vascular effect of PACAP isoforms and VIP in arteries isolated from female PACAP WT and KO mice
Potent vasomotor effect of PACAP has been described by many previous studies both in vivo and in vitro [7,15,16,19,35,46,47]. Various studies demonstrated relaxation responses to both PACAP isoforms and VIP in vessels of different origin, such as pulmonary [22], coronary PACAP-induced vasorelaxation in female mice [48], carotid [17,21], cerebral and intracerebral [19,48] of various species. In accordance with these studies, we have also observed arterial relaxation induced by PACAP/VIP administration in female mice. There was no difference between the magnitude of relaxation triggered by PACAP isoforms or VIP, in carotid and femoral arteries of WT mice. However, Cheng et al. reported the strongest vasomotor effect of PACAP1-27 in pulmonary arteries [22], whereas other studies reported the strongest vasodilator capacity of PACAP1-38 in rat mesenteric [49] and mice carotid and femoral arteries [17], suggesting isoform-and region-specific PACAP responses. Further studies confirmed such region-specific vascular effects of PACAP. In these studies, the same stimulus induced different magnitude of response in different organs [21], or even reacted differently within different regions of the same organ, e.g. brain vessels [14], probably reflecting different requirements of blood supply. The reason for this could be due to the different sensitivity of arteries to PACAP isoforms [50], and potency of receptors [7,14].
Introduction of the PACAP KO model allows us to investigate these mechanisms further. While relaxations of carotid arteries to PACAP isoforms were reduced in female PACAP KO mice compared to WT controls to some extent, femoral arteries did not show such a difference.

Involvement of PAC1, VPAC1 and VPAC2 receptors in vascular responses of female mice
As described earlier, PACAP/VIP receptor distribution can highly differ across and within organs [7]. Vasculature of animals and humans are innervated with PACAP-containing nerve fibers and as a result, they contain a high density of binding sites [7,14,20,24,51]. Our previous study also reported the presence of PAC1R and VPAC1R in peripheral arteries of mice [17], however, the presence of PAC1, VPAC1 and VPAC2 receptors in peripheral arteries of female mice have not been investigated yet.
According to our results, in the vasculature of female mice, presence and function of PAC1R was verified with the selective PAC1R agonist maxadilan. Interestingly, although PAC1 protein is either absent (in carotid arteries) or reduced (in femoral arteries) in PACAP KO mice, we observed relaxation in all arteries of both WT and PACAP KO mice after maxadilan treatment. We found similar observations in male mice [17]. PAC1R may be present in reduced amount in the vasculature of PACAP KO mice, which is not detectable with Western blot method. Selective antagonist of PAC1R had no effect on maxadilan-induced vasodilation, most likely due to the massive potency of maxadilan as compared with that of M65 or the existence of PAC1R splice variants [7] resulting in different affinity for maxadilan and M65. The idea that maxadilan is not selective enough should not be ignored, especially because it is not structurally related to PACAP/VIP. Interestingly, administration of M65 reduced PACAP/ VIP-induced relaxation in femoral arteries, but not in carotid arteries. We propose that this difference in the effectivity of the pharmacological agents could be due to different downstream signaling response, a mechanism shown by Hoover et al. in cardiac neurons [52]. However, these suggested mechanisms still cannot fully explain the surprising fact that M65, as a maxadilan derivative fails to block the vasomotor effect of maxadilan.
Involvement of VPAC1R in arterial relaxation in female mice was confirmed with selective receptor agonist VIP6-28. VPAC1R is functional in both arteries of WT and PACAP KO female mice. This is consistent with other previous findings [7,53] and with our previous work [17]. Our molecular analysis was in accord with the functional vasomotor responses of the arteries. VPAC1R agonist resulted in higher relaxation of femoral arteries in PACAP KO than in the WT mice. A possible explanation could be that endogenous PACAP suppresses receptor activity in benefit of PAC1R, and in the absence of PACAP, the signal transduction could shift toward VPAC1R. This is supported by the fact that PACAP upregulates the receptor expression [54], which can be altered by lack of PACAP. In addition, there is a possibility that in the absence of PACAP, the VPAC1R may undergo conformational changes allowing the binding of VPAC1R agonist Ala 11,22,28 VIP more effectively, despite a similar receptor expression. It could be the reason for an increased response to a selective agonist in the femoral arteries of KO mice. In contrast, there was no difference in the vasomotor responses of the carotid arteries of the WT and PACAP KO mice, which again may indicate the importance of region-specific vasomotor responses as mentioned above. Importantly, selective VPAC1R antagonist VIP6-28 abolished all PACAP-induced relaxations that suggests a crucial role of VPAC1R in the vasculature of female mice.
Observations with PACAP isoforms and VIP in the presence of selective PAC1R (M65) and VPAC1R (Ala 11,22,28 VIP) antagonists indicates that receptor signaling is different in carotid and femoral arteries. It seems that in carotid arteries, only VPAC1R is functional, while in femoral arteries synergistical activation of both PAC1R and VPAC1R allows a vascular response (Fig 7). Although molecular findings show some inconsistency with functional findings (effect of PAC1R agonist in carotid arteries), nevertheless, these findings indicate the importance of VPAC1R in the modulation of response not only in female WT [14] but also in PACAP KO mice.

Gender difference in PACAP-and VIP-induced relaxation in arteries of mice
Vessels of female mice show weaker vascular relaxations upon PACAP isoforms and VIP administration as compared to those of males [17]. Females also seem to be less sensitive to the lack of PACAP than males. Other peptides, like endothelin-1 and bradykinin (femoral and brachial arteries) [55] or neuropeptide Y (hind limb) [56] also show gender-dependent effects in the vasculature experimental animals. Different receptor modulation and downstream signaling in females could play a role in the observed gender-differences of vascular responses, as we can see from our experiments with selective antagonists.
The question arises, whether sex steroids affect PACAP/VIP-induced relaxation. However, there were no significant differences in PACAP-and VIP-induced vascular responses during the estrus cycle. Other observations reported by Dalsgaard and coworkers also showed that treatment with sex steroids induced no changes in the vascular effects of PACAP or VIP in ovariectomized rabbits [48]. Gender-difference was also reported with regard to other effects of PACAP or VIP. Kiss and coworkers demonstrated a marked increase in PACAP levels in the hypothalamus of male rats after food deprivation compared to the lower increase in females [57], while Lam and coworkers reported higher levels of VIP were found in the pituitaries of male as compared to females. Although these studies demonstrate PACAP/VIP related gender-differences [57,58], relatively little is known about differences in either expression of receptors or effect of PACAP in the macrovasculature. Further investigations in ovariectomized female mice would help clarify the underlying mechanisms.

Conclusions
This study is the first to show the relaxation effects of exogenous PACAP isoforms in peripheral arteries isolated from female WT and PACAP-deficient mice. These responses are more moderate than those of male mice. In female PACAP KO mice, carotid vasomotor responses to PACAP isoforms and VIP were reduced, but no differences were detected in the femoral arteries. In the background of this regional difference, decreased PAC1R and increased VPAC1R mediation of the carotid arteries were found. Our results also show, that in female mice, PACAP-induced vasorelaxation responses are mainly mediated by VPAC1R.  PACAP and VIP bind to PAC1R/VPAC1R G protein-coupled receptors, and stimulate the cAMP/PKA, promoting vasodilatation. PACAP isoforms bind to both receptors, while VIP binds only to VPAC1R (and PAC1R at >500nM [7]). PACAP isoforms and VIP induce similar relaxations in arteries of WT and KO mice. Lower panel: in carotid arteries of PACAP-deficient mice, PACAP-and VIP-induced relaxations were reduced, whereas in femoral arteries there is no difference in response as compared to WT mice. Relaxation of carotid arteries is modulated primarily via VPAC1R and relaxation of femoral arteries via both PAC1R and VPAC1R. Endogenous VIP can bind not only to VPAC1R, but also to PAC1R. Unchanged, reduced or absent relaxations to polypeptides are marked with thicker, thinner or dashed arrows, respectively (as compared to the PACAP wild type mice). "+" indicates polypeptides presented in the system and added exogenously. "-"indicates exogenously added polypeptides (in PACAP-deficient mice only).