Multiphasic response to HOCl and endothelium involvement

To determine the effect of HOCl on the PCA, different concentrations (100 and 500 μM HOCl) were added to the PCA rings pre-contracted with U46619. At 100 μM, HOCl induced a rapid and transient initial relaxation (62.45 ± 13.87%) which lasted 2–4 min and then returned to baseline after 5–6 min; after that, further relaxation was observed for the remaining 60 min (103.63 ± 3.94%) (n = 7) (Fig. 1A). A rapid and transient relaxation response (28.77 ± 11.48%) was also induced by 500 μM HOCl, lasting about 40 s, followed by contraction which peaked at between 6 and 10 min (52.39 ± 23.27%), and then slow relaxation was observed (97.80 ± 4.20%) (n = 7) (Fig. 1A). Figure 1B is a plot of the first six minutes to show more clearly the initial rapid and transient responses to 100 and 500 μM HOCl.

Fig. 1

Responses of porcine coronary arteries induced by HOCl at 100 and 500 μM and the effect of endothelium removal. (A) responses measured for 60 min at 1 min intervals; (B) expanded time course of the first 6 min measured at 10 s intervals (360 s) to show the initial rapid relaxation; (C–D) representative traces for: (C) 100 μM HOCl and (D) 500 μM HOCl. Porcine coronary artery responses evoked by (E) 100 μM HOCl and (F) 500 μM HOCl in endothelium-intact and denuded vessels (n = 6). (G–H) representative trace for: (G) 100 μM HOCl and (H) 500 μM HOCl in endothelium-denuded vessels. Representative traces C,D,G and H show the second of two contractions to KCl (60 mM) which was used to set the level of precontraction. Artery segments were pre-contracted with U46619. Data are means ± SEM. Control (n = 6), 100 μM HOCl (n = 7), and 500 μM HOCl (n = 7)

The involvement of the endothelium in the responses to HOCl was then investigated. In endothelium-denuded PCA, the transient relaxations observed in PCA with intact endothelium to 100 μM and 500 μM HOCl (70 ± 4.5%, n = 6 and 57.5 ± 5.5%, n = 6, respectively) were abolished (Fig. 1E-H). The contraction at 500 μM HOCl in endothelium-intact arteries (54.7 ± 8%, n = 6) was also abolished in endothelium-denuded arteries (Fig. 1F, H). The relaxation to 100 μM HOCl at 60 min was significantly larger in the PCA without endothelium (111 ± 3.2%, n = 6) compared to endothelium-intact PCA (97 ± 1.5%, n = 6) (p = 0.002) (Fig. 1E).

The levels of U46119 pre-contraction (% of KCl) achieved in the vessels without endothelium in the 100 and 500 µM HOCl groups were 55.4 ± 4.5% and 60 ± 2%, respectively (both n = 6). Moreover, with the endothelium intact, the contraction levels with U46619 were 60 ± 5% and 59 ± 3.3% for 100 and 500 µM HOCl, respectively (both n = 6). There was no significant difference in tone between the vessels with and without endothelium for either the 100 µM HOCl experimental group (p = 0.47) or the 500 µM HOCl experimental group (p = 0.47). The concentration of U46619 used to pre-contract the coronary arteries was significantly lower in the vessels without endothelium (4.3 ± 0.4 nM) than in those with intact endothelium (7.5 ± 0.9 nM, n = 6, and 6) (p < 0.007).

Purine receptor involvement in HOCl vasomotor responses

Comprehensive data for the effects of antagonists and inhibitors on the rapid and slow relaxation response to 100 μM HOCl described in this section are in Tables 2 and 3. Comprehensive data for the effects of antagonists and inhibitors on the rapid relaxation, contraction and slow relaxation to 500 μM HOCl are in Tables 4, 5, and 6.

Table 2 Effect of antagonists and inhibitors on rapid relaxation to 100 μM HOClTable 3 Effect of antagonist and inhibitors on slow relaxation to 100 μM HOClTable 4 Effect of antagonists and inhibitors on rapid relaxation to 500 μM HOClTable 5 Effect of antagonists and inhibitors on contraction to 500 μM HOClTable 6 Effect of antagonists and inhibitors on slow relaxation to 500 μM HOClEffect of 8-(p-sulphophenyl)theophylline (adenosine receptor antagonist) and suramin (non-selective P2 receptor antagonist)

The possible involvement of purine receptors in the response to HOCl was then investigated using 8-SPT, an adenosine/P1 receptor antagonist, and suramin, a non-selective P2 receptor antagonist which can have other effects including membrane channel inhibition [12, 51]. 8-SPT (100 μM) blocked the rapid relaxation response induced by 100 μM HOCl (72.6 ± 4.6%, n = 6) (Fig. 2A). At a concentration of 500 μM HOCl, the transient rapid relaxation that was observed in control PCA (42.7 ± 4.3%, n = 6) was also blocked when 8-SPT was added (Fig. 2B). The contraction to 500 μM HOCl (57.3 ± 5.9%, n = 6) was blocked too in the presence of 8-SPT (Fig. 2B). These results show an involvement of adenosine receptors in the rapid relaxation and contraction to HOCl.

Fig. 2

Inhibitory effects of 8-SPT and suramin on responses to HOCl. Porcine coronary artery responses produced by (A) 100 μM HOCl (n = 6) and (B) 500 μM HOCl (n = 6) in the absence or presence of 8-SPT (100 μM). Responses produced by (C) 100 μM HOCl (n = 6) and (D) 500 μM HOCl (n = 7) in the absence or presence of suramin (100 μM) (n = 6). Artery segments were pre-contracted with U46619. Data are means ± SEM

Suramin (100 µM) blocked the rapid relaxation observed in response to 100 μM HOCl (Fig. 2C). In contrast, the rapid relaxation observed in response to 500 μM HOCl (50.5 ± 5.9%, n = 7) was evident in the presence of suramin (50.2 ± 8.2%, n = 7) at 40 s (p = 0.97) (Fig. 2D). The contraction observed in control tissues in response to 500 μM HOCl (64.3 ± 9.1%, n = 7) was blocked by suramin (Fig. 2D). These data suggest a possible involvement of P2 receptors in the response to HOCl so the next experiments used other selective and non-selective P2 receptor antagonists with an aim of identifying the P2 receptor subtypes involved.

Effect of ZM 241385 (A2A receptor antagonist) and DPCPX (A1 receptor antagonist) on HOCl response

The transient relaxation of PCA segments evoked by 100 μM HOCl was not different in the presence and absence of ZM 241385 (1 µM) (Table 2). Additionally, the slow relaxation at 60 min was similar in PCA with ZM 241385 compared to the control (Table 3). At 500 μM HOCl, the transient relaxation, contraction, and slow relaxation in the presence of ZM 241385 were similar to responses produced in control arteries (Tables 4, 5, and 6).

At 100 μM HOCl, DPCPX (100 nM) did not alter the transient relaxation response of HOCl compared to that in the absence of DPCPX (Table 2). Likewise, the slow relaxation response at 60 min in the absence or presence of DPCPX was similar (Table 3). With 500 μM HOCl, the rapid relaxation, subsequent contraction and the later relaxation response were also similar in the absence and presence of DPCPX (Tables 4, 5, and 6).

Effect of PPADS (P2 receptor antagonist), MRS2179 (P2Y1 receptor antagonist), reactive blue 2 (non-selective P2Y receptor antagonist), NF449 (P2X1 receptor antagonist), α,β-methylene ATP (P2X1 receptor desensitising agent), and BX430 (P2X4 receptor antagonist)

PPADS (10 μM), a non-selective P2 receptor antagonist, did not alter the effect of 100 μM HOCl (Fig. 3A). In addition, there was no difference between responses to 500 μM HOCl in the absence and presence of PPADS (Fig. 3B).

Fig. 3

Effect of PPADS (P2 antagonist), MRS2179 (P2Y1 antagonist), reactive blue 2 (P2Y antagonist), NF449 (P2X1 antagonist), α,β-methylene ATP (P2X1 desensitizer) and BX430 (P2X4 antagonist) on HOCl responses. Porcine coronary artery responses produced by 100 μM HOCl in the absence or presence of: (A) PPADS (10 μM) (n = 7) (C) MRS 2179 (10 μM) (n = 9), (E) reactive blue 2 (30 μM) (n = 6), (G) NF449 (10 μM) (n = 6), (I) α,β-methylene ATP (10 μM) (n = 6) and (K) BX430 (10 μM) (n = 6). Responses produced by 500 μM HOCl in the absence or presence of: (B) PPADS (10 μM) (n = 6), (D) MRS 2179 (10 μM) (n = 6), (F) reactive blue 2 (30 μM) (n = 6), (H) NF449 (10 μM) (n = 8), (J) α,β-methylene ATP (10 μM) (n = 6) and (L) BX430 (10 μM) (n = 6). Data are means ± SEM

The effect of 100 μM HOCl on the PCA segments was similar in the presence and absence of MRS 2179 (10 µM), a P2Y1 receptor antagonist (Fig. 3C). At 500 μM HOCl, the transient relaxation, contraction, and slow relaxation in the presence of MRS 2179 were similar in control arteries (Fig. 3D).

At 100 μM HOCl, reactive blue 2 (30 μM), a non-selective P2Y receptor antagonist, did not alter the response of HOCl (Fig. 3E). With 500 μM HOCl, the rapid relaxation and subsequent contraction were not different to those observed in the presence of reactive blue 2 compared to the control, in addition, the later relaxation response was similar in the control artery and in the presence of reactive blue 2 (Fig. 3F).

NF449 (10 μM), a P2X1 receptor antagonist, neither altered the transient relaxation nor slow relaxation to 100 μM HOCl compared to the control (Fig. 3G). In addition, there was no difference between responses in the presence of NF449 and the control at 500 μM HOCl (Fig. 3H).

At 100 μM HOCl, α,β-methylene ATP (10 μM), a P2X1 receptor desensitizing agent, did not alter the response in the PCA; both the rapid relaxation and the later relaxation response were similar in the absence and presence of α,β-methylene ATP (Fig. 3I). With 500 μM HOCl, the rapid relaxation, subsequent contraction, and the later relaxation response were not different in the presence of α,β-methylene ATP compared to the control (Fig. 3J).

BX430 (10 μM), a P2X4 receptor antagonist, did not affect the rapid relaxation and the later relaxation evoked by 100 µM HOCl (Fig. 3K). At 500 µM HOCl, the rapid relaxation, subsequent contraction, and later relaxation were not significantly different in the absence and presence of BX430 (Fig. 3L).

These data show that P2X1, P2X4 and P2Y1 receptors are not involved in the response to HOCl. Neither PPADS nor reactive blue affected the HOCl responses.

Effect of apyrase (hydrolyses nucleotides) and ARL67156 (ecto-ATPase inhibitor)

Apyrase, which hydrolyses nucleotides, was used to investigate the possible involvement of endogenous ATP in the response to HOCl. At 100 μM HOCl, apyrase (100 units/ml) completely blocked the HOCl response compared to the control. The rapid relaxation and slow relaxation in the PCA control group were 58.7 ± 5% and 104.5 ± 9.7%, respectively (n = 6), compared to the apyrase group at 4.6 ± 1.3% (p < 0.0001) and 10.8 ± 3.4% (p < 0.0001) (n = 6) (Fig. 4A). Apyrase also abolished the rapid relaxation observed at 500 μM HOCl (0.7 ± 0.5%, n = 6) compared to the control group (48.4 ± 5.1%, n = 6) (p < 0.0001). Moreover, apyrase blocked the contraction produced by 500 µM HOCl (24.9 ± 4.6%, n = 6) compared to that in the presence of apyrase (10.7 ± 2.8%, n = 6) (p = 0.02). In addition, the later relaxation at 60 min to HOCl (96.5 ± 6.2%, n = 6) was eliminated in the presence of apyrase (7 ± 9.6%, n = 6) (p < 0.0001) (Fig. 4B). Apyrase alone elicited vasoconstriction which can be seen in the representative trace in Fig. 7B. A lower concentration of apyrase (10 units/mL) had no significant effect on the relaxation response to HOCl.

Fig. 4

Effect of apyrase (hydrolyses nucleotides) and ARL67156 (ecto-ATPase inhibitor) on responses to HOCl. Responses of porcine coronary arteries to: (A) 100 μM HOCl (n = 6) and (B) 500 μM HOCl (n = 6) in the absence or presence of apyrase (100 units/ml). Responses to: (C) 100 μM HOCl (n = 6) and (D) 500 μM HOCl (n = 6) in the absence or presence of ARL67156 (100 μM). Artery segments were pre-contracted with U46619. Data are means ± SEM

ARL67156, an ecto-ATPase inhibitor, was also used to determine whether ATP is involved in any of the components of the HOCl response. ARL67156 (100 μM) did not alter transient or slow relaxation at 60 min of 100 μM HOCl (p = 0.48) (Fig. 4C). Moreover, there was no difference between responses to 500 μM HOCl in the presence and absence of ARL67156; not in the transient relaxation, the subsequent contraction, and in the later relaxation (Fig. 4D).

Effect of ATP on PCA and endothelium involvement

The effect of ATP was investigated as a control. ATP was added as a single concentration (100 μM) (rather than as cumulative concentrations) to match the additions of HOCl. ATP caused relaxation in U46619 pre-contracted PCA rings (Fig. 5A). A rapid relaxation response to ATP occurred at 1 min, with a maximum of 50.2 ± 10.7% (n = 6). This was followed by a sustained relaxation response with a maximum (at about 10 min) of 88.3 ± 7.1% (n = 6). At 60 min, the observed relaxation was 84.4 ± 8.3% (n = 6) (Fig. 5A).

Fig. 5

Response to exogenous ATP and effect of endothelium removal. (A) Porcine coronary artery (PCA) responses produced by 100 μM ATP (n = 8). Artery segments were pre-contracted with U46619. (B) Original trace showing the response of the PCA to 100 μM ATP in vessels pre-contracted with U46619; shows the second of two contractions to KCl (60 mM) which was used to set the level of precontraction. (C) Responses evoked by 100 μM ATP in endothelium-intact and -denuded PCA (n = 8). Data are means ± SEM

The rapid relaxation response of the PCA with intact endothelium to ATP (100 μM) (measured at 1 min) was 68.9 ± 4.3% (n = 8). It was blocked after removing the endothelium: 16.5 ± 5.3%, n = 8 (p < 0.0001). The maximum relaxation response of the PCA with intact endothelium to ATP (100 μM) (measured at ~ 15 min) was 82.2 ± 8.9% (n = 6). After removing the endothelium, the relaxation response was larger, being 97.1 ± 6.0% (n = 6), but this difference was not statistically significant (p = 0.19). At 60 min, the relaxation response in endothelium-denuded PCA was 95.2 ± 7.0% (n = 8), which was bigger but not significantly different compared to endothelium-intact PCA, which showed a response of 72.4 ± 11.5% (n = 8) (p = 0.11) (Fig. 5C).

Effect of 8-SPT and suramin on ATP response

The effects of 8-SPT and suramin on the response to exogenous ATP were investigated to characterise the purine receptors involved. 8-SPT (100 μM) reduced the rapid relaxation response caused by ATP (100 μM) (33.2 ± 7.43%, n = 6) compared to the relaxation response observed under normal conditions (64.3 ± 8.4%, n = 6) (p = 0.02) (Fig. 6A). It also reduced the maximum relaxation response caused by ATP (100 μM) (36.6 ± 7.8%, n = 6) compared to the relaxation response observed under normal conditions (84.1 ± 6.1%, n = 6) (p = 0.0009) Furthermore, after 60 min, the relaxation response to ATP in the presence of 8-SPT (25.7 ± 10.3%, n = 6) was significantly lower than that in the control group (68.8 ± 12.6%, n = 8) (p = 0.03) (Fig. 6A).

Fig. 6

Effect of 8-SPT (adenosine receptor antagonist) and suramin (P2 receptor antagonist and membrane channel inhibitor) on responses to ATP. Porcine coronary artery responses produced by 100 μM ATP in the absence or presence of (A) 8-SPT (100 μM) (n = 6), and (B) suramin (100 μM) (n = 6). Artery segments were pre-contracted with U46619. Data are means ± SEM

The presence of suramin (100 μM) did not alter the response of PCA to ATP; the maximum relaxation was 87.8 ± 7.3% (n = 6) in the absence of suramin compared to 101.3 ± 1.3% (n = 6) in the presence of suramin (p = 0.09). Likewise, the relaxation responses at 60 min in the absence or presence of suramin were similar at 84.4 ± 8.3% and 102.6 ± 1.6% (n = 6) (p = 0.05), respectively (Fig. 6B).

Thus, ATP causes relaxation via P1 receptors, but not via suramin-sensitive P2 receptors.

Effect of apyrase on ATP and adenosine responses

The effect of apyrase, which hydrolyses nucleotides, on the ATP response was investigated. Apyrase (100 units/mL) completely inhibited the maximum ATP response when compared to the control group, with relaxation of 8.8 ± 2.1% and 92.2 ± 1.1% respectively (p = 0.0005) (n = 6). At 60 min, the control group exhibited a relaxation of 78.5 ± 11.0% (n = 6), whereas the apyrase group showed a significantly lower relaxation of 4.4 ± 4.9% (p = 0.0001) (Fig. 7A, B). Apyrase (100 units/mL) also blocked the relaxation response to adenosine (30 µM). The adenosine relaxation at 15 min, 74 ± 12.1% (n = 6), was completely blocked by 100 units/ml of apyrase, 19.0 ± 4.9% (n = 6) (p = 001). The relaxation at 60 min in the control group was 70.9 ± 13.1% (n = 6) compared to the apyrase group at 7.5 ± 4.9% (n = 6) (p = 0.001) (Fig. 7C, D). Apyrase (100 units/mL) had no significant effect on relaxation responses to bradykinin: the bradykinin pEC50 of 8.5 ± 0.4 and Rmax = 64.5 ± 4.5% (n = 6) were not changed in arteries with added apyrase, with a pEC50 of 8.4 ± 0.5 and Rmax = 61.3 ± 4.8% (n = 6). Apyrase alone evoked a vasocontractile effect on the PCA (Fig. 7B, D).

Fig. 7

Effect of apyrase on response to ATP and adenosine. (A) Responses of porcine coronary arteries (PCA) mediated by 100 μM ATP in the absence or presence of apyrase (100 units/ml) (n = 6). (B) Original trace showing the responses of PCA to 100 μM ATP in a PCA pre-contracted with U46619 in the presence of apyrase (100 units/ml); shows the second of two contractions to KCl (60 mM) which was used to set the level of precontraction. (C) Responses to 30 μM adenosine in the absence and presence of apyrase (100 units/ml). (D) Representative traces showing responses to 30 μM adenosine in the absence and presence of apyrase (100 units/ml). Data are means ± SEM

Effect of probenecid (pannexin 1 channel blocker) and carbenoxolone (pannexin 1 and connexin channels blocker)

Suramin can inhibit connexin and pannexin channels [51]. For that reason the pannexin1 and connexin channels blockers probenecid and carbenoxolone were used to investigate whether these channels might be involved in the inhibitory effects of suramin on HOCl responses. Probenecid (1 mM) had no effect on the initial relaxation response to 100 µM HOCl (78.5 ± 2.1%, n = 7) compared to the control (69.7 ± 7.1%, n = 7) (p = 0.25). The later relaxation after 60 min was significantly increased in the presence of probenecid (105.8 ± 1.0%, n = 7) compared to the control group (94.7 ± 3.9%, n = 7) (p = 0.01) (Fig. 8A).

Fig. 8

Effect of probenecid and carbenoxelone on responses to HOCl. Porcine coronary artery responses produced by (A) 100 μM HOCl (n = 7) and (B) 500 μM HOCl (n = 6) in the absence or presence of probenecid (1 mM), and (C) 100 μM HOCl (n = 6) and (D) 500 μM HOCl (n = 6) in the absence or presence of carbenoxolone (100 μM). Artery segments were pre-contracted with U46619. Data are means ± SEM

In the presence of probenecid, the initial rapid relaxation to 500 µM HOCl was 47.1 ± 8.1% (n = 6) and in the control group it was similar at 52.3 ± 10.8% (n = 6) (p = 0.7). The subsequent contraction to 500 µM HOCl was 29.8 ± 12.1% (n = 6) which was significantly reduced in the presence of probenecid 11.1 ± 13.1% (n = 6) (p = 0.04). The later slow relaxation after 60 min was significantly increased in the presence of probenecid (104.5 ± 1.7%, n = 6) compared to HOCl alone (82.1 ± 5.4%, n = 6) (p = 0.002) (Fig. 8B).

Carbenoxolone (100 μM) failed to block the rapid relaxation evoked by 100 µM HOCl, as is shown in Fig. 8C. On the other hand, the response to HOCl in the presence of carbenoxolone failed to return to baseline (50.2 ± 8.9%, n = 6) compared to controls (22.2 ± 7.7%, n = 6) (p = 0.03). Moreover, HOCl produced a larger relaxation at 60 min in the presence of carbenoxolone compared to HOCl alone (114.1 ± 2.1%, n = 6; 98.8 ± 1.6%, n = 6, respectively) (p = 0.0002) (Fig. 8C).

At 500 µM HOCl alone, the early relaxation was 51.6 ± 5.2% (n = 6) in the control, which was not significantly different in the presence of carbenoxolone (41.2 ± 6.1%, n = 6) (p = 0.22). However, subsequent contraction was inhibited in the presence of carbenoxolone (1.7 ± 7.4%, n = 6) compared to the control (25.6 ± 5.5%, n = 6) (p = 0.03). The later relaxation after 60 min significantly increased in the presence of carbenoxolone (104.4 ± 1.3%, n = 6) compared to the control group (89.3 ± 2.3%, n = 6) (p = 0.0002) (Fig. 8D).

HOCl releases ATP from cultured human coronary artery endothelial cells

The ATP level in the supernatant of cells with added HOCl (100 µM) at 1 min was 0.18 ± 0.07 nmol/mg protein, which was significantly higher than that in the control group at 0.12 ± 0.06 nmol/mg protein (n = 9) (p = 0.02) (Fig. 9). The ATP level was not significantly different from the control at 5 min after HOCl addition.

Fig. 9

ATP release from cultured human coronary artery endothelial cells (HCAECs) by 100 µM HOCl. The culture medium bathing the cells was assayed for ATP at 1 and 5 min after the addition of HOCl (n = 9). *p < 0.05 (two-way ANOVA)