Phenylephrine induces relaxation of longitudinal strips from small arteries of goat legs

Alpha adrenergic stimulation is known to produce vasoconstriction. We have earlier shown that, in spiral strips of small arteries Phenylephrine (PE) caused vasorelaxation under high nitric oxide (NO) environment. However, on further experimentation it was realized that the PE-induced vasorelaxant response occurred only with longitudinal strips of small arteries even under normal NO environment while circular strips showed contraction with PE even under high NO environment. Such PE-induced vasorelaxation of longitudinal strips was blocked by Phentolamine, an alpha-adrenergic receptor blocker. On delineation of specific receptor subtype, PE-induced relaxation was found to be mediated through alpha 1D receptor. However, this phenomenon is specific to small artery, as longitudinal smooth muscle of aorta showed only contractile response to adrenergic stimulation. There is no prior report of longitudinal smooth muscle in small artery up to our knowledge. The results of this study and histological examination of vessel sections suggest the presence of longitudinal smooth muscle in small artery and their relaxant response to alpha adrenergic stimulation is a novel phenomenon.


Introduction
Phenylephrine (PE) is an alpha-adrenergic agonist which is known to cause contraction of vascular smooth muscle. We had recently reported that Phenylephrine caused relaxation of spiral strips made from small arteries from goat legs in a high nitric oxide environment, while causing contraction in normal circumstances. However, with further experimentation we have realized that longitudinal strips made from small arteries always relaxed with Phenylephrine, even in normal nitric oxide environment. On the other hand, PE caused contraction of circular strips of small arteries even in a high nitric oxide environment. Spiral strips used in earlier experiments have led to erroneous inferences, as the method of making the strip is prone to experimenter bias. In this paper, we compare the responses of transverse (or circular) and longitudinal preparations made from aorta and small arteries from goats. While PE increases vascular tension in both transverse and longitudinal strips made from aorta, as do circular strips of small arteries, longitudinal strips from small arteries show reduction in vascular tension in the presence of PE.
An important determinant of arterial pressure is peripheral vascular resistance, which in turn is primarily determined by the diameter of small arteries and arterioles which are referred to as resistance vessels [1]. The walls of the resistance vessels have smooth muscle cells arranged in a concentric fashion in the tunica media which is referred to as circular smooth muscle [2]. Contraction of circular smooth muscle reduces lumen diameter and therefore increases vascular resistance [1,3]. The balance between vasoconstrictor and vasodilator signals acting on the circular smooth muscle determines the diameter and therefore the tone of resistance vessels.
While muscular arteries (medium-sized) are documented to have concentrically arranged circular smooth muscle in the tunica media, veins have longitudinal smooth muscle just next to the intima [4]. There is evidence for the presence of longitudinal smooth muscle in the walls of large arteries and aorta too [5]. Function of longitudinal smooth muscle in the large arteries has not been clearly understood.
There is, however, no prior report of longitudinal smooth muscle in small arteries, up to our knowledge. The results of this study suggest the presence of longitudinal smooth muscle in small arteries, and their relaxant response to alpha adrenergic stimulation is a novel phenomenon.

Materials & methods
This study was approved by Institutional Review Board of Christian medical College, Vellore, (IRB number 10959 dated 7 th November 2017).
10 mmol/L stock solution was prepared for phenylephrine and L-Arginine in distilled water. Appropriate amount of the drug was added both to the organ bath and to the drug reservoir to obtain the final concentration.

Isolation of small sized artery
Fresh goat leg (Capra hircus) from a registered slaughterhouse, Rafi's slaughterhouse, accreditation number: 309/13 (Co-ordinates 12.93 o N 79.13 o E) was procured on the day of experiment, washed, and the skin removed. A vascular bundle close to the muscle below the knee was identified. The artery in the bundle was identified by patency of its lumen. Considerable length of the artery was dissected and transferred to a petri dish containing cold ECF solution. Adventitious tissue attached to the artery was removed and the artery was cut into segments of 1.5-2 cm length. It was ensured that the arterial segments were devoid of side branches. Two different arterial preparations were made: i. Transverse cylinder ii. Longitudinal strip Transverse cylinder preparation: Two fine threads were inserted through the lumen of the arterial segment and loops were made on opposite sides. One loop was connected to a force transducer and the other loop was attached to a hook in the organ bath such that the tissue was suspended transversely. (Fig 1) Longitudinal strip preparation: An arterial segment of length 1.5-2 cm was selected and cut open longitudinally using an iris scissors thereby exposing the endothelium. One end of the longitudinal axis was connected to the force transducer while the other end was connected to the base of the organ bath. (Fig 1)

Isolation of aorta
Fresh goat heart procured from a registered slaughterhouse (details as before) was transported to lab in ice cold mammalian ECF solution. The heart was washed, pericardial sac and fat removed, and ascending aorta was identified by the patency of its lumen. The aorta was then dissected from its point of origin to the point before it branches. The dissected aorta was transferred to a petri dish containing cold ECF solution and the adventitious tissue removed. Two different preparations were made.
i. Aortic circular strip ii. Aortic longitudinal strip Aortic circular strip preparation: A ring of aorta was cut, with thickness of 5 mm. One end of the ring was attached to the force transducer and the other end to the base of the organ bath.
Aortic Longitudinal strip Preparation: A length of aortic cylinder was cut open in longitudinal orientation with surgical scissors exposing the endothelium. Opposite ends of the longitudinal strip were connected to a force transducer and the base of the organ bath.
Data acquisition and analysis. The force transducer was connected to Powerlab data acquisition system for recording vessel tension. The thread was kept taut with an optimum preload and the resting tension was maintained between 0.2 to 0.6g. Change in tension was recorded when drugs were added into the organ bath. The data were acquired at 1 KHz and was processed using MATLAB R2018a software. Change in vascular tension with interventions is expressed as scatter plots. The percentage change in tension is also shown as bar diagrams showing mean and SD.

Statistical analysis
Statistical analysis was done using SPSS software ver. 16.0. Comparison of vascular tension before and after intervention in same group was done using Wilcoxon signed rank (WSR) test. Changes in vascular tension between different groups were done using Mann-Whitney U (MWU) test. P value � 0.05 was considered statistically significant.

Histological examination of vessel preparations
Transverse and longitudinal preparations of small artery and aorta were immersed in 10% buffered formalin for 2-3 days for fixation. Ascending grades of alcohol were used to dehydrate the tissues which were finally cleared with xylene. Then paraffin blocks of the tissues were made by impregnating with liquid paraffin. 5μm thick sections were made using microtome. Tissue sections were subjected to dewaxination and were hydrated with descending grades of alcohol and were washed in distilled water. Staining was done using hematoxylin and eosin and sections were observed under light microscope.

Results
A summary of the major results obtained is presented in Table 1 and the details are provided:

PE induces contraction of circular smooth muscle of aorta under normal and high NO environments
In circular strips of aorta, PE 100 μmol/L increased tension from 0.37 to 0.51g. (Median, n = 5, P value = 0.043 with WSR test comparing vascular tension before and after PE), Fig 2, (i). The concentration of PE for the experiments on aortic as well as small artery preparations was chosen as 100μmol/L because the small artery preparation did not show a response to 10μmol/L PE.
L-Arginine 100μmol/L was added to the bath to increase NO levels in the circular strip of aorta. No change in tension was observed with L-Arginine. Subsequent addition of PE 100 μmol/L increased tension from 0.31g to 0.65g. (Median, n = 6, P value = 0.028 with WSR test comparing tension before and after PE), Fig 2, (ii).
Addition of sodium nitroprusside (SNP) 400μmol/L to the bath to create high NO environment, did not change vascular tension in the circular aortic strip. Subsequent addition of PE increased tension from 0.35g to 0.49g. (Median, n = 5, P value = 0.043 with WSR when tension before and after PE in presence of SNP were compared), Fig 2, (iii). There was no significant difference when percentage change in aortic vascular tension due to PE with and without L-Arginine (P value = 0.273 with MWU) were compared and also when PE with and without SNP (P value = 0.754 with MWU) were compared, showing that PE-induced vasoconstriction in aortic circular muscle was not affected by high levels of NO. There was no significant difference when the percentage change in vascular tension due to PE with and without L-Arginine were compared (P value = 0.076 with Mann-Whitney U test, MWU) and also when percentage changes in vascular tension due to PE with and without SNP were compared (P value = 0.251 with MWU) showing that PE-induced vasoconstriction in aortic longitudinal muscle was not affected by high NO conditions.

PE induces contraction of circular smooth muscle of small artery under normal and high NO environments
10μmol/L concentration of PE did not alter the tension in small-artery preparation, but at 100 μmol/L concentration, PE increased the tension of circular smooth muscle of small artery from 0.09g to 0.54g. (Median, n = 5, P value = 0.028 with Wilcoxon signed rank test when tension before and after PE was compared, Fig 4, (i).
No change in basal tension was observed when high NO environment was created by adding L-Arginine 100μmol/L. Subsequent addition of PE 100μmol/L increased the tension of circular strip of small artery from 0.07g to 0.79g (Median, n = 5, P value = 0.043 with WSR test, Fig 4, (ii).
There was no significant difference in percentage change in vascular tension due to PE with or without L-Arginine (P value = 0.273 with MWU test).

L-NNA inhibited PE-induced contraction of circular smooth muscle in small artery
Addition of L-NNA, a competitive inhibitor of Nitric oxide synthase, caused a transient increase in tension which returned to baseline. Subsequent addition of PE 100μmol/L failed to produce contraction of circular smooth muscle. The vessel tension before and after addition of PE in the presence of L-NNA were 0.09 and 0.08g (median, n = 5, P value = 0.08 with WSR test), Fig 4, (iii).
The lack of response to PE in the presence of L-NNA was statistically significant as compared to the contractile response to PE alone (Median, n = 5, P value = 0.008, with MWU test).

PE induces relaxation of longitudinal smooth muscle of small artery both under normal and high NO environments
PE 100μmol/L reduced the tension of longitudinal smooth muscle from 0.25g to 0.09g in small arteries. (Median, n = 5, P value = 0.043, WSR), Fig 5, (i)).
While L-Arginine 100μmol/L did not reduce the basal tension by itself, subsequent addition of PE 100μmol/L reduced the tension from 0.28 to 0.07g (Median, n = 5, P value = 0.043, WSR), Fig 5, (ii).  The lack of response to PE in the presence of L-NNA was statistically significant as compared to the relaxant response to PE alone (Median, n = 5, P value = 0.008, with MWU test).

PE-induced relaxation of longitudinal smooth muscle is alpha adrenoceptor-mediated
PE failed to produce relaxation of longitudinal smooth muscle in the presence of phentolamine 50μmol/L, an alpha adrenoceptor blocker. The vascular tension before and after the addition of PE (in the presence of phentolamine 50μmol/L) were 0.24g and 0.24g respectively (median, n = 5, P value = 0.08, WSR test). Fig 5, (iv).
The lack of relaxant response to PE in the presence of phentolamine was statistically significant as compared to the relaxant response to PE alone (Median, n = 5, P value = 0.008, with MWU test).

PE-induced relaxation of longitudinal smooth muscle is mediated through alpha-1D receptor subtype
Delineation of the specific subtype of alpha-1 adrenergic receptor responsible for PE-induced vasorelaxation was done by testing the effect of PE in the presence of different blockers of alpha-adrenergic receptor subtypes. Alpha-1A receptor blocker, RS17053 (10μmol/L) and alpha-1B receptor blocker, CEC (10μmol/L) were not able to prevent the PE-induced reduction in tension in longitudinal strip. However, an alpha-1D receptor blocker, BMY 7378 (10μmol/L) prevented PE-induced reduction in tension in longitudinal strip.
The vascular tension before addition of PE (in the presence of 10μmol/L RS 17053) was 0.32g and after addition of PE was 0.24g (median, n = 5, p value = 0.043 with WSR test), Fig 7, (i).
A statistically significant difference was not observed (P value = 0.347, MWU test) when percentage changes of vascular tension due to PE in the presence and absence of presence of RS 17053 10μmol/L were compared.
Tension recordings in the longitudinal strip before and after addition of PE (in the presence of CEC 10μmol/L) were 0.30g and 0.18g respectively (median, n = 5, P value = 0.043 with WSR test), Fig 7, (ii).

PLOS ONE
Phenyl ephrine relaxes longitudinal strips of small arteries There was a 37% reduction of baseline tension by PE in the absence of CEC 10μmol/L, (Median, n = 5) whereas the reduction of baseline tension by PE in the presence of CEC 10μmol/L was about 56% (Median, n = 5). However, for this sample size there was no significant difference when percentage changes in tension due to PE with and without CEC were compared (P value = 0.056 with MWU test).
The tension in longitudinal strip before addition of PE (in the presence of BMY 7378 10μmol/L) was 0.26g and after addition of PE was 0.24g (median, n = 5, P value = 0.08 with WSR test), Fig 7, (iii).
There was a significant difference between the control group (PE alone) and BMY 7378/PE group when the percentage changes in tension were compared (P value = 0.008 with MWU test).

PE-induced relaxation of longitudinal smooth muscle is not mediated through alpha-2 receptor
To identify if PE-induced vasorelaxation in longitudinal strip of small artery is mediated through alpha-2 adrenergic receptor, yohimbine 50μmol/L, an alpha-2 antagonist was used. Tension in the longitudinal strip prior to the addition of PE (in the presence of Yohimbine 50μmol/L) was 0.39g and after addition of PE was 0.21g (median, n = 5, P value = 0.043 with WSR test), Fig 7, (iv).
There was no statistically significant difference in percent change of vascular tension between control group (PE alone) and PE in the presence of yohimbine 50μmol/L (P value = 0.117, MWU test)

PE-induced relaxation of longitudinal smooth muscle is not mediated through cAMP
To delineate the role of cAMP in PE-induced vasorelaxation in longitudinal strips of small artery, MDL 12330A, an adenylyl cyclase inhibitor was used. Tension in the longitudinal strip prior to the addition of PE (in the presence of MDL 50μmol/L) was 0.22g and after addition of PE was 0.05g (median, n = 5, P value = 0.043 with WSR test), Fig 8, (i).
There was no significant difference when the percent changes in tension were compared between the control group (PE alone) and MDL 12330A/PE group P value = 1.00, with MWU test).
Forskolin (FSK), an adenylyl cyclase stimulator failed to bring about reduction in tension in longitudinal strips of small artery by itself. Tension in the longitudinal strip prior to the addition of FSK 100μmol/L was 0.24g and after addition of FSK was 0.23g (median, n = 5, P value = 0.08, with WSR test), Fig 8, (ii). Fig 9 shows the summary of percent change in tension from baseline tension due to PE in the presence various blockers from in longitudinal strip preparation of small artery.

Histological examination
To look for anatomical evidence for presence of longitudinal smooth muscle in small artery, histological examination was done. Light microscopic images of transverse cylinder preparation showed elliptical nucleus of circular smooth muscle predominantly while a few nuclei were seen circular representing a different arrangement of muscle fibres (Fig 10A and 10B). In light microscopic images of longitudinal strip, nuclei of circular smooth muscle were seen circular and a few were seen elliptical representing the longitudinal smooth muscle fibres ( Fig  10C and 10D).

PLOS ONE
Phenyl ephrine relaxes longitudinal strips of small arteries Similar picture is seen in aorta in which the existence of longitudinal smooth muscle is well known (Fig 11A, 11B, 11C & 11D).
Vasoconstriction by alpha adrenergic stimulation occurs in two ways-either stimulation of myosin light chain kinase (MLCK) by calcium/Calmodulin complex [12] or inhibition of myosin light chain phosphatase by (MLCP) by phosphorylated CPI-17 [13,14] It has been shown earlier that PE could cause alpha-adrenoceptor mediated vasorelaxation in spiral strips of small arteries in a high nitric oxide (NO) environment [15]. However, it was realized that whether a spiral strip will relax or constrict depended not on NO, but in the manner in which it was prepared. There was PE-induced relaxation with strips prepared with fewer spirals (in hindsight this would include more longitudinal component) and PE-induced contraction even in a high NO environment with strips prepared in a manner to include more concentric muscle. To clarify this paradox this study focuses on the effect of PE on two different preparations-transverse cylindrical arterial strip (circular smooth muscle) and longitudinal strip (longitudinal muscle) respectively. Both these types of preparations were made with small arteries and aorta. It is shown here that there is a component of longitudinal muscle in small arteries which relaxes on alpha adrenergic stimulation. However, circular strips of small arteries demonstrated the well-known vasoconstrictor response with alpha adrenergic stimulation (with PE). The relaxant response of longitudinal smooth muscle to PE that was observed in small arteries did not occur in the aorta. Both circular and longitudinal preparations of aorta showed a vasoconstrictor response with PE.
Phenylephrine, a selective alpha agonist [16] is a known vasoconstrictor, the effects of which are mediated via alpha-1 and alpha-2 adrenergic receptors. The action of Phenylephrine on alpha 1 receptor is mediated through phospholipase-C on the cell membrane [17]. Phospholipase-C causes breakdown of phosphatidylinositol-4,5 bisphosphate (PIP2) on cell membrane to inositol 1,4,5-triphosphate (IP 3 ) and 1,2-diacylglycerol (DAG). IP 3 then enters the cytoplasm, acts on IP3 receptors on sarcoplasmic reticulum and causes release of calcium. The released calcium then binds to calmodulin, a regulatory protein forming calcium/calmodulin complex [18]. Calcium/Calmodulin complex causes stimulation of MLCK which causes phosphorylation of myosin light chain thereby causing contraction [12]. Calcium released from sarcoplasmic reticulum combines with DAG and stimulates protein kinase C (PKC). PKC causes phosphorylation of a smooth muscle phosphoprotein, CPI-17. CPI-17 inhibits MLCP, which prevents dephosphorylation of myosin light chain thereby supporting contraction [13,14]. Activation of alpha-2 adrenoceptor stimulates G i type of G-protein, which in turn inhibits adenylyl cyclase and decreases cAMP, again supporting contraction [18].
Raj et al., [15] have demonstrated in spiral strip preparation of goat arteries, scenarios in which PE could cause vasorelaxation from basal tone, one being high NO environment. It was

PLOS ONE
shown that PE/NO induced vasorelaxation is cGMP independent and requires alpha adrenergic activation. However, further experiments showed that contractile or relaxant response depends on the manner of spiral strip preparation rather than on the level of NO.
The results show that PE induces only contraction of circular smooth muscle both under normal NO and high NO environment as known earlier. However, the effect of PE on longitudinal component of smooth muscle is only relaxation both under normal NO and high NO environment. L-Arginine is reported to produce relaxant response by increasing NO levels [19]. However, we observed that L-Arginine by itself did not induce relaxation in both small artery and aorta. It is not as if the preparations were not capable of relaxation because in the case of longitudinal strip of small artery at least, PE could induce vasorelaxation. Similarly, SNP, a NO donor also failed to produce relaxation in both circular and longitudinal strips of aorta. These effects of both L-Arginine and SNP need further evaluation. The earlier observations of L-Arginine-induced relaxation were all observed in pre-constricted strips and not in unconstructed strips [20]. In the current study, we show that L-Arginine and SNP have no relaxant effect from baseline tension in unconstricted strips. We have also demonstrated structural integrity of the endothelium histologically. However, it is a limitation of the study that functional integrity of the endothelium was not assessed.
It was also found that such PE-induced vasorelaxation in longitudinal smooth muscle is alpha-adrenergic dependent, as phentolamine, an alpha adrenoceptor antagonist prevented such response. In order to delineate the receptor subtype involved in PE-induced vasorelaxation, four alpha adrenoceptor subtype blockers (alpha-1A, alpha-1B, alpha-1D and alpha-2) were used. Among them only alpha-1D blocker, BMY 7378 was successful in preventing PE induced vasorelaxation. BMY can block alpha-2C adrenoceptors when used at a concentration above 0.3μmol/L [21]. Though the concentration that we have used is 10μmol/L, its effect of blocking vasorelaxation in longitudinal strips is unlikely to be due to alpha 2C blockade, as Yohimbine, which is a non-selective antagonist or alpha-2 subtypes did not prevent the vasorelaxation induced by PE. Hence, prevention of PE-induced vasorelaxation by BMY 10μmol/L is through blockade of alpha-1D receptor.
The role of the second messenger cAMP, in PE-induced vasorelaxation was also studied. To decipher this, forskolin an adenylyl cyclase activator, and MDL 12330A, an adenylyl cyclase inhibitor were used. Forskolin by itself failed to produce vasorelaxation in longitudinal strips of small artery and MDL 12330A failed to prevent PE-induced vasorelaxation showing that PE-induced vasorelaxation in longitudinal strip is not mediated through cAMP.
It is also shown that L-NNA, a competitive inhibitor of NO synthase inhibited both PEinduced contraction and relaxation in small artery. The probability of L-NNA being a blocker of alpha-adrenergic receptors too, rather than being just a NO synthase blocker has to be evaluated further. Also, L-NNA caused a transient increase in tension in transverse cylinder preparation whereas it produced a sustained increase in tension in longitudinal preparation of small artery. This differential behaviour of L-NNA also needs further evaluation.
Further, anatomical evidence for the existence of longitudinal smooth muscle in small arteries has been demonstrated through histology.
Presence of longitudinal smooth muscle in aorta is well-known. However, the effect of PE on both circular and longitudinal smooth muscle of aorta is only vasoconstriction, unlike in small arteries. This is the first time within our knowledge, the existence of longitudinal smooth muscle in small arteries is being reported. It is also shown here that the longitudinal smooth muscle of small artery relaxes in response to α 1D adrenergic stimulation and such relaxant response is independent of cAMP. However, the physiological significance for the existence of such longitudinal component of smooth muscle in small arteries and their relaxant response to alpha adrenergic stimulation needs to be ascertained.