Altered Serotonin Receptor Subtypes Mediate Coronary Microvascular Hyperreactivityin Pigs With Chronic Inhibitionof Nitric Oxide Synthesis

Animal Preparation

Twenty-seven male domestic pigs (Nihon Crea Inc) weighing from 26 to 32 kg were used. Sixteen pigs were fed a regular diet supplemented with L-NAME (30 mg·kg⁻¹·d⁻¹) for 4 weeks, and 11 pigs were fed a regular diet alone for the same period.

The pigs were sedated with ketamine hydrochloride (12.5 mg/kg IM) and anesthetized with sodium pentobarbital (20 mg/kg IV). They were then endotracheally intubated and mechanically ventilated with a respirator. Under aseptic conditions, a left thoracotomy was performed, and an ultrasonic transit-time flow probe (Transonic Systems Inc) was placed in the midportion of the left anterior descending coronary artery. The chest was then closed, and the animals were allowed to recover from the surgery for at least 1 week. Postoperative diets were the same as preoperative diets.

This study was reviewed by the Committee on Ethics in Animal Experiments of the Faculty of Medicine, Kyushu University, and was carried out according to the Guidelines for Animal Experimentation of the Faculty of Medicine, Kyushu University, and the Law (No. 105) and Notification (No. 6) of the Japanese Government.

Experimental Protocols

The animals were again sedated, anesthetized, and intubated as described above. Arterial pH, Po₂, and Pco₂ were kept within normal ranges. Then, a 6F pigtail catheter was introduced into a femoral artery and advanced into the left ventricle. A 7F Kifa catheter was introduced into a carotid artery and advanced to the orifice of the left coronary artery. Various vasoactive agents were administered into the left coronary artery through the Kifa catheter while CBF and other hemodynamic parameters were measured.

Measurements

CBF was measured by connecting the flow probe to an ultrasonic transit-time flowmeter (Transonic T201D, Transonic System Inc). Peak responses of CBF to drugs were used for analysis. Aortic pressure and LV pressure were measured with pressure transducers (Nihon-Kohden Inc) connected to the Kifa and pigtail catheters, respectively. The LV dP/dt was obtained by electronic differentiation. Heart rate was measured by means of a cardiotachometer triggered by a pulse wave of LV pressure (AT 600G, Nihon-Kohden Inc). These hemodynamic variables and ECGs (V₁ and V₆ leads) were monitored continuously and recorded with a polygraph system (Polygraph 360, NEC San-Ei Inc).

Selective coronary angiography was performed in a left anterior oblique projection with a cineangiography system (DG-15GB, Toshiba Medical Inc) as previously described.^{25 26} The angle of the projection, the posture of the animal, the position of the x-ray focus, and the image intensifier were carefully kept constant during each experiment. Coronary angiograms were recorded on 35-mm cinefilm (Varicath I, Hirata Sangyo Inc) at 48 frames per second. Cineangiograms were projected on a screen with a cineprojector (ELK-35CB, Nishimoto Sangyo Inc), and the end-diastolic images were selected and printed. The luminal diameters of the coronary arteries were measured with calipers. The angiographic catheter was used to calibrate the diameter. The readily identifiable coronary branch points were used as reference points to assess serial changes in the diameter at the same arterial site. With this technique, excellent correlations between repeated measurements (r=.99) and between different observers (r=.98) were obtained in the range of coronary artery diameters from 0.9 to 5.0 mm. In each animal, three to five proximal segments of the left anterior descending and circumflex coronary arteries were measured, and the average change in diameter in response to the various drugs was used for analysis. The percent change in the luminal diameter is reported.

Histopathological Study

After the in vivo experiments, the pigs were killed with a lethal dose of sodium pentobarbital. After the hearts had been excised, the left coronary artery was flushed with 20 mL saline containing 200 μg nitroglycerin and 900 μg adenosine for 3 minutes and then perfused with 6% formaldehyde at a pressure of 120 mm Hg for 30 minutes. The hearts were fixed in 6% formaldehyde for several days, then were cut transversely from apex to base serially at 1-cm intervals. The tissue was embedded in paraffin, sectioned 5 μm thick, and stained with Masson's trichrome, van Gieson's elastic solution, and hematoxylin-eosin. All stained sections were carefully scanned, and photographs of small intramyocardial arteries and epicardial coronary arteries (50 to 1800 μm in ID) were taken at ×40 to ×400 magnification with a Nikon light microscope equipped with a two-dimensional analysis system (SONY Inc). The inner border of the vascular intima and the outer border of the vascular media were traced, and the area between the tracings was calculated automatically. Nonround vascular profiles resulting from oblique sectioning or branching were excluded from the measurement, and only circular vascular profiles were measured.

To evaluate narrowing of the vessels, the wall-to-lumen ratio was calculated by the formula (O−I)/I, where O and I are the OD and ID of the vessels, respectively. In each heart, 20 to 30 microvessels (ID, <300 μm) and 3 to 5 large epicardial vessels were measured, and the averaged values were used for analysis.

Drugs

The following drugs were used: L-NAME, 5-HT, bradykinin (Sigma Chemical Co), nitroglycerin (Nihon-Kayaku Pharmaceutical Co), PGF_2α (Ono Pharmaceutical Co), ketanserin (Kyowa Hakko Co), methiothepin, and 5-CT (ICN Pharmaceuticals Inc). All drugs were dissolved in physiological saline immediately before use.

Statistical Analysis

Data are expressed as mean±SEM. Serial changes in CBF and other hemodynamic parameters of the two groups were compared by a two-way ANOVA followed by Bonferroni's test for multiple comparisons. A probability value of P<.05 was considered statistically significant.

Effects of Acute L-NAME Administration on Microvascular Hyperreactivity to 5-HT (Protocol 1)

There were no significant differences in basal CBF, heart rate, or LV dP/dt between the control and L-NAME pigs (Tables 1⇓ and 2). Basal aortic pressure and LV end-diastolic pressure were higher (P<.01) in the L-NAME pigs than in the control pigs. Aortic pressure, heart rate, LV dP/dt, and LV end-diastolic pressure did not change in response to intracoronary 5-HT, PGF_2α, bradykinin, or nitroglycerin.

Physiological saline injection had no effect on the hemodynamic variables in either group. In the control pigs, 5-HT at 3 μg/kg increased (P<.01) CBF and 5-HT at 30 μg/kg decreased (P<.01) CBF, whereas in the L-NAME pigs, 5-HT at the same doses only decreased (P<.01) CBF. The percent decrease in CBF in response to 30 μg/kg 5-HT was significantly greater in the L-NAME pigs than in the control pigs (Figs 1⇓, 2, and 3). The percent decrease in CBF in response to 0.5 μg/kg PGF_2α was similar (P=NS) between the two groups (Fig 3⇓). The increase in CBF in response to bradykinin was significantly less in the L-NAME pigs than in the controls, but the response to nitroglycerin was similar (P=NS) between the two groups (Fig 4⇓).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Real-time tracings of CBF in a control pig. Changes in CBF to 5-HT (serotonin) 30 μg/kg before and after ketanserin are presented.

Acute L-NAME administration significantly increased (P<.05) aortic pressure but did not affect the other variables in either group (Tables 1 and 2⇑⇓). The degree of the acute L-NAME–induced increase in aortic pressure was less (P<.05) in the L-NAME pigs than in the controls. In the control pigs, acute L-NAME administration attenuated (P<.01) the percent increase in CBF in response to 5-HT at 3 μg/kg and to bradykinin but did not affect the response to nitroglycerin (Figs 3 and 4⇓⇓). In the L-NAME pigs, acute L-NAME administration did not affect the percent decreases in CBF induced by 5-HT at 3 or 30 μg/kg or the percent increases in CBF induced by bradykinin and nitroglycerin (Figs 3 and 4⇓⇓).

Baseline coronary artery diameter was similar between the control pigs (3.3±0.2 mm) and L-NAME pigs (3.4±0.2 mm, P=NS versus control). 5-HT and PGF_2α constricted while bradykinin and nitroglycerin dilated the large epicardial coronary arteries. The percent change in the diameters in response to these drugs was similar (P=NS) between the two groups (Table 3⇓). Acute L-NAME administration did not alter the changes in diameter induced by these drugs (Table 3⇓).

Effects of the 5-HT Receptor Antagonists on Microvascular Hyperreactivity to 5-HT (Protocol 1)

In the control pigs, the percent decreases in CBF in response to 5-HT at 3 and 30 μg/kg after acute L-NAME administration were nearly abolished after treatment with ketanserin at 1 mg/kg (Figs 1 and 3⇑⇓, Table 1⇑). In contrast, in the L-NAME pigs, the augmented decreases in CBF in response to 5-HT at 3 and 30 μg/kg were partially but significantly attenuated by ketanserin at 1 and 3 mg/kg (Fig 2⇓, Table 2⇑). The degree of inhibition by 1 mg/kg ketanserin of 5-HT responses was similar to that by 3 mg/kg ketanserin in the L-NAME pigs (Table 2⇑). The decrease in CBF due to 5-HT after ketanserin was significantly greater in the L-NAME pigs than in the controls. The augmented decrease in CBF due to 5-HT in the L-NAME pigs was markedly (P<.01) inhibited by additional treatment with 1 mg/kg methiothepin (Figs 2 and 3⇓⇓). The percent decrease in CBF in response to PGF_2α was not significantly altered by these 5-HT antagonists (Fig 3⇓).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Real-time tracings of CBF in an L-NAME pig. Changes in CBF to 5-HT (serotonin) 30 μg/kg at baseline, after ketanserin, and after ketanserin plus methiothepin are presented.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Effects of acute L-NAME and 5-HT receptor antagonists on the CBF response to 5-HT and PGF_2α. Open bars indicate no treatment; dark hatched bars, after acute L-NAME (1 mg/kg) treatment; light stippled bars, after acute L-NAME plus ketanserin (1 mg/kg) treatment; and solid bars, after acute L-NAME plus ketanserin plus methiothepin (1 mg/kg) treatment. *P<.01 vs control group; †P<.01 vs acute L-NAME plus ketanserin treatment; and ¶P<.01 vs control group (no treatment) by two-way ANOVA followed by Bonferroni's multiple comparison test.

The percent decrease in coronary artery diameter in response to 5-HT was significantly attenuated by ketanserin in both groups, but the changes in diameter due to PGF_2α, bradykinin, and nitroglycerin were not affected by ketanserin in either group (Table 3⇑). The changes in diameter due to these drugs were not affected by additional treatment with methiothepin in the L-NAME–treated pigs (Table 3⇑).

Reproducibility of 5-HT–Induced Response (Protocol 2)

In the time-control experiments, changes in CBF due to 5-HT, PGF_2α, bradykinin, and nitroglycerin were reproducible in the control and L-NAME pigs (data not shown).

Effects of Acute L-NAME or 5-HT Antagonists on the Changes in CBF due to 5-CT (Protocols 3 and 4)

In the control pigs, intracoronary 5-CT increased CBF (P<.01). This increase was blocked completely by acute administration of L-NAME (protocol 3) (Fig 4⇓) and ketanserin and methiothepin treatment (Fig 5⇓) but not by ketanserin alone (protocol 4) (Fig 5⇓). The increase in CBF due to nitroglycerin and the decrease in CBF due to PGF_2α were not affected by these 5-HT receptor antagonists (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Effects of acute L-NAME administration on the CBF response to 5-CT 100 ng/kg, bradykinin 100 ng/kg, and nitroglycerin 10 mg/kg. Open bars indicate before acute L-NAME; shaded bars, after acute L-NAME. *P<.01 vs before acute L-NAME; †P<.01 vs control group by two-way ANOVA followed by Bonferroni's multiple comparison test.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Effects of the 5-HT receptor antagonists on the CBF responses to 5-CT in the control and L-NAME pigs. Open bars indicate no treatment; shaded bars, after ketanserin; and solid bars, after ketanserin plus methiothepin. *P<.01 vs no treatment by two-way ANOVA followed by Bonferroni's multiple comparison test.

In contrast, in the L-NAME pigs, 5-CT decreased CBF (P<.01). This decrease was blocked completely by ketanserin and methiothepin treatment but not by ketanserin alone or acute L-NAME administration (Figs 4 and 5⇑⇑). 5-CT did not affect the diameters of large epicardial coronary arteries or other hemodynamic variables in either group (data not shown).

Histopathology

Histopathology demonstrated thickening of the vascular media in coronary microvessels but not in large epicardial coronary vessels in the L-NAME pigs.¹⁶ The wall-to-lumen ratios (an index of medial thickening) in microvessels were significantly greater (P<.01) in the L-NAME pigs (0.77±0.09) than in controls (0.16±0.02). However, the ratios in the large epicardial vessels were similar between the two groups (0.14±0.01 versus 0.16±0.02, P=NS).

Contribution of Defective Endothelium-Dependent Dilation to Microvascular Hyperreactivity due to 5-HT

Endothelium-dependent increases in CBF by bradykinin and 5-HT₁ receptor stimulation with 5-HT (3 μg/kg) and 5-CT (100 ng/kg) were markedly less in the L-NAME pigs than in the controls. Acute L-NAME administration markedly inhibited the increase in CBF evoked by 5-HT₁ receptor stimulation and bradykinin in the controls but did not alter the increase in CBF by bradykinin in the L-NAME pigs. These observations indicate that endothelium-derived NO-mediated dilation of the coronary microcirculation due to activation of endothelial bradykinin and 5-HT₁ receptors was defective in the L-NAME pigs. Although the rise in aortic pressure after acute administration of L-NAME was less in the L-NAME pigs than in controls, acute L-NAME administration produced a significant increase in aortic pressure in the L-NAME pigs. The latter findings suggest that NO-generating capacity at baseline may also be reduced but still present despite chronic administration of L-NAME. Therefore, the activities of NO synthase in the L-NAME pigs remain to be investigated.

This study also demonstrated that acute L-NAME administration did not affect the 5-HT–induced decrease in CBF in either the L-NAME pigs or the controls. Importantly, the percent decreases in CBF due to 5-HT or 5-CT after acute L-NAME were greater in the L-NAME pigs than in the controls. These observations suggest that the augmented decreases in CBF in the L-NAME pigs were mediated largely by hyperconstriction of the coronary microvascular smooth muscle but did not result from defective NO synthesis per se.

It is reported that 5-HT₁ receptor stimulation or bradykinin leads to endothelium-dependent dilation of excised large epicardial coronary arteries in vitro.^{29 30} However, intracoronary infusion of 5-HT or 5-CT did not dilate the large epicardial coronary arteries in the L-NAME or control pigs in vivo in this study. Bradykinin-induced dilation of the large epicardial coronary arteries was comparable between the two groups. Vasomotion of the large epicardial coronary arteries induced by bradykinin, 5-HT, and 5-CT was not affected by acute L-NAME administration. These observations suggest that NO may not be involved in bradykinin-induced or 5-HT–induced vasomotion of large epicardial coronary arteries in pigs in vivo.

Relative Contributions of 5-HT₁ and 5-HT₂ Receptor Subtypes to Microvascular Hyperreactivity due to 5-HT

The vascular effects of 5-HT are mediated by two distinct receptor subtypes: 5-HT₁ and 5-HT₂.^{24 31 32 33} Pretreatment with ketanserin (a selective 5-HT₂ antagonist) completely blocked the 5-HT–induced decrease in CBF in the controls, thus indicating that the 5-HT–induced microvascular constriction was totally mediated by 5-HT₂ receptors in normal pigs. In contrast, in the L-NAME pigs, the augmented decrease in CBF evoked by 5-HT was not affected by ketanserin alone but was blocked completely by pretreatment with ketanserin and methiothepin (a nonselective 5-HT₁/5-HT₂ antagonist). The latter findings suggest that the 5-HT–induced microvascular hyperreactivity was mediated largely by 5-HT₁ receptors in the L-NAME pigs. It is unlikely that failure of inhibition of the 5-HT–induced microvascular hyperreactivity by ketanserin in the L-NAME pigs was due to incomplete blockade of 5-HT₂ receptors, because the augmented decrease in CBF by 5-HT was not changed by an increase in the dose of ketanserin from 1 to 3 mg/kg. It is also unlikely that the effects of the 5-HT receptor antagonists were due to nonspecific changes, because the CBF responses to 5-HT were reproducible in the time-control experiments. Also, the decrease in CBF induced by PGF_2α was not affected by the 5-HT receptor antagonists.

Our observation that the 5-CT–induced increase in CBF in the control pigs was inhibited completely by pretreatment with both ketanserin and methiothepin but not by ketanserin alone suggests that activation of 5-HT₁ receptors stimulates the release of NO from the endothelium and does not cause vasoconstriction in normal pigs. In contrast, 5-CT markedly decreased CBF in the L-NAME pigs. This decrease was blocked by pretreatment with both ketanserin and methiothepin but not by ketanserin alone. Therefore, the 5-HT₁ receptors in microvascular smooth muscle cells play an important role in mediating the 5-HT–induced microvascular hyperreactivity in our animal model. These observations suggest that coronary microvascular hyperreactivity due to 5-HT in our animal model is mediated by relative changes in affinity for 5-HT receptor subtypes or de novo expression of 5-HT₁ receptor subtypes in microvascular smooth muscle cells.

In contrast to the altered microvascular response to 5-HT, constriction of the large epicardial coronary arteries caused by 5-HT was comparable between the L-NAME and control pigs; 5-CT did not cause significant vasoconstriction in either group. Ketanserin significantly inhibited the 5-HT–induced constriction of the large epicardial coronary arteries in the two groups. These observations suggest that 5-HT–induced constriction of large epicardial coronary arteries is mediated by the 5-HT₂ receptor subtype in the two groups. Also, altered 5-HT receptor subtypes occurred specifically in the coronary microvessels but not in the large epicardial vessels.

Golino et al³⁴ recently demonstrated that intracoronary infusion of 5-HT (10 μg·kg⁻¹·min⁻¹) significantly increased CBF and dilated large epicardial coronary arteries in patients with normal coronary arteries. In patients with coronary atherosclerosis, 5-HT significantly decreased CBF and constricted large epicardial coronary arteries, which was prevented by pretreatment with ketanserin. On the basis of these observations, Golino et al concluded that 5-HT–induced coronary vasoconstriction was mediated by the 5-HT₂ receptor subtype. However, it is uncertain whether the 5-HT–induced decrease in CBF was due to constriction of resistance coronary arteries in their study. The decrease in CBF due to 5-HT in their report was associated with the concomitant vasoconstriction (64% reduction in cross-sectional area) of large epicardial coronary arteries.

Chilian et al¹⁰ investigated the vasomotor responses of the coronary microcirculation to 5-HT in atherosclerotic monkeys and demonstrated that the enhanced vasoconstriction to 5-HT occurred not only in large atherosclerotic arteries but also in resistance arteries that were free of atherosclerotic lesions. They speculated that the enhanced constriction of resistance coronary arteries to 5-HT had been related to endothelial dysfunction of resistance arteries downstream of the atherosclerotic lesions. In the present study, constricting responses of resistance coronary arteries to 5-HT were augmented under conditions of chronic inhibition of NO synthesis. However, the present animal model differs from the animal models of atherosclerosis with regard to the structural changes in coronary microvessels.

Clinical Implications

Previous reports in animals and humans^{18 19 20 21 22} have suggested that 5-HT₁ receptors in vascular smooth muscle are involved in the altered contractile response of atherosclerotic arteries. The present study demonstrates that changes in the relative contributions of 5-HT₁ and 5-HT₂ receptor subtypes in microvascular smooth muscle cells may mediate coronary microvascular hyperreactivity due to 5-HT in our animal model. These data provide new insight into the pathophysiology of microvascular disease. Because 5-HT₁ receptor subtypes are heterogeneous, further investigations are needed to elucidate the cellular mechanism of microvascular hyperreactivity to 5-HT.