A Heme Oxygenase Product, Presumably Carbon Monoxide, Mediates a Vasodepressor Function in Rats
Abstract Heme oxygenase is a mammalian enzyme that converts heme to biliverdin and carbon monoxide. Carbon monoxide activates soluble guanylate cyclase and relaxes vascular smooth muscle, and it has been implicated as a potential neuromessenger. The regulatory functions of endogenous carbon monoxide on hemodynamics are not known. Zinc deuteroporphyrin 2,4-bis glycol (ZnDPBG) inhibits heme oxygenase in rats and thus permits assessment of the hemodynamic response to inhibition of endogenous carbon monoxide synthesis. In chronically instrumented, awake male Sprague-Dawley rats, ZnDPBG (45 μmol/kg IP) increased mean arterial pressure (19±2%, P<.05) and total peripheral resistance (47±4%, P<.05), decreased cardiac output (−16±2%, P<.05), but did not affect heart rate. Another heme oxygenase inhibitor, zinc protoporphyrin IX (45 μmol/kg IP), also increased arterial pressure (17±5%, P<.05), with no effect on heart rate. In contrast, neither the nonmetallic deuteroporphyrin 2,4-bis glycol (45 μmol/kg IP) nor biliverdin (45 μmol/kg IP) had any effect on blood pressure or heart rate. These findings suggest that ZnDPBG and zinc protoporphyrin IX increase arterial pressure by inhibiting heme oxygenase activity. After pretreatment with chlorisondamine (5 mg/kg IP) or prazosin (5 mg/kg IP) to inhibit autonomic ganglionic or α1-adrenoceptor functions, respectively, ZnDPBG did not affect arterial pressure or heart rate. This suggests that ZnDPBG-induced increases in blood pressure rely on autonomic nervous function. We conclude that the pressor response to heme oxygenase inhibitors results from withdrawal of the inhibitory influence of endogenous carbon monoxide on a pressor mechanism mediated by the autonomic nervous system.
Heme oxygenase is a widely distributed enzyme that daily converts almost 1% of the heme in blood to biliverdin and carbon monoxide.1 2 3 4 The biliverdin by-product, in mammals, is rapidly transformed into bile pigments for excretion, whereas the endogenously generated carbon monoxide binds avidly to a variety of heme- and sulfydryl-bearing structures until eliminated by ventilation.3 Carbon monoxide is an activator of soluble guanylate cyclase5 and relaxes vascular smooth muscle via a cGMP-dependent mechanism.6 A recent study has provided evidence that a subset of glutamate receptors, involved in the function of the afferent arm of the baroreceptor reflex, may be coupled with heme oxygenase–mediated production of carbon monoxide.7 There has been speculation that heme oxygenase–generated carbon monoxide production might play a physiological role in blood pressure (BP) regulation,3 8 but no such role has been established. Zinc deuteroporphyrin 2,4-bis glycol (ZnDPBG) is an inhibitor of heme oxygenase activity.9 10 11 ZnDPBG has been shown to inhibit endogenous carbon monoxide production in rats12 and thus permits the physiological actions of heme oxygenase activity to be studied in vivo.
We designed the present study to assess the contribution of heme oxygenase activity to resting BP in the awake rat. To accomplish this, we contrasted hemodynamic measurements before and after short-term administration of ZnDPBG. We conducted additional experiments to assess potential involvement of the autonomic nervous system in ZnDPBG-induced BP effects.
Porphyrins and biliverdin were obtained from Porphyrin Products and chlorisondamine (Ecolid) from Ciba Pharmaceutical Products. All other drugs were purchased from Sigma Chemical Co. Porphyrins were prepared in 50 mmol/L Na2CO3 (15 μmol of drug per milliliter) immediately before use. Other drugs were dissolved in normal saline on the day of the experiment.
Sixty-five male Sprague-Dawley rats (Charles River, Wilmington, Mass) ranging in weight from 300 to 375 g were used in these studies. Rats were individually housed in a controlled temperature of 27°C, with automatic lighting that provided a 12-hour on-off cycle. Rats had free access to commercial rat chow (Ralston Purina) and tap water.
Each animal was anesthetized with sodium pentobarbital (Anpro Pharmaceuticals, 60 mg/kg IP), and a chronic arterial catheter was implanted for BP and heart rate (HR) determinations. Each arterial catheter (PE-50) was filled with heparinized normal saline, introduced through a femoral artery, and advanced into the lower abdominal aorta. Each was tunneled subcutaneously to an exit point at the nape of the neck and sealed with a steel pin until use.
In some experiments, animals were also chronically instrumented with aortic flow probes (model 2.5S, Transonic Systems) for measurement of cardiac output. Each of these animals was temporarily intubated for mechanical ventilation (Rodent Ventilator, Harvard Bioscience) while the flow probe was placed around the ascending aorta. The probe electrical leads were guided through the right second intercostal space, tunneled subcutaneously, and exited at the nape of the neck. All animals received ampicillin (30 mg/kg per 12 hours SC) for at least 3 days after surgery. A postsurgical recovery period of at least 4 days was allowed before experiments.
All experiments were conducted in awake, unrestrained rats. Femoral arterial catheters were connected to pressure transducers (model P23XL, Statham) coupled to a polygraph (model 7D, Grass Instrument Co) for continuous arterial pressure measurements. The aortic flow probes were connected to a flowmeter (Transonic model T208) for cardiac output determinations.
In protocol 1, arterial pressure, HR, and cardiac output were measured before and after injection of 50 mmol/L Na2CO3 (3 mL/kg IP) vehicle or ZnDPBG (45 μmol/kg IP) to inhibit heme oxygenase activity.9 10 11 12 Another control group received nonmetallic DPBG (45 μmol/kg IP), which does not affect heme oxygenase activity (unpublished observations). Total peripheral resistance was calculated as the ratio of mean arterial pressure (MAP) to cardiac output and was expressed as millimeters of mercury per milliliter per minute. In complementary experiments, rats were injected with ZnDPBG (45 μmol/kg IP, n=3) or DPBG (45 μmol/kg IP, n=3). Thirty minutes later, rats were given a sodium pentobarbital overdose (150 mg/kg IP), and the brains were immediately removed for assay of microsomal heme oxygenase activity as previously described.9 10 The microsomal heme oxygenase activity was not detectable in animals injected with ZnDPBG, whereas it averaged 0.160±0.025 nmol bilirubin/mg protein per hour in those injected with DPBG. In protocol 2, arterial pressure and HR were measured before and after an injection of 50 mmol/L Na2CO3 vehicle alone (3 mL/kg IP); zinc protoporphyrin IX (ZnPP, 45 μmol/kg IP), which inhibits heme oxygenase activity; or biliverdin (45 μmol/kg IP), which is a heme oxygenase product.1 2 4 In protocol 3, rats were pretreated for 15 minutes with either chlorisondamine (5 mg/kg IP) or prazosin (5 mg/kg IP) to block autonomic ganglionic or α1-adrenoceptor functions,13 respectively. Arterial pressure and HR were then measured before and after injection of the vehicle alone (3 mL/kg IP) or ZnDPBG (45 μmol/kg IP) to inhibit heme oxygenase activity. In some experiments, the effect of ZnDPBG on the BP of rats pretreated with chlorisondamine was examined in animals receiving phenylephrine at a rate (4±1 μg/kg per minute IV) sufficient to offset the vasodepressor effect of the ganglionic blocker. In protocol 4, rings of descending thoracic aorta from untreated rats were prepared for recording of isometric tension in organ baths filled with Krebs’ bicarbonate buffer, according to published procedures.13 In these rings we examined the effect of ZnDPBG (50 μmol/L) on resting tension as well as on the tension development induced by the cumulative addition of phenylephrine (10−9 to 10−6 mol/L) to the bath.
Results are expressed as mean±SEM. Data were analyzed by ANOVA, with a value of P<.05 being significant. This was followed by orthogonal contrasts (α=0.05) with Bonferroni correction.
Fig 1⇓ displays hemodynamic data before and after ZnDPBG or DPBG administration. ZnDPBG administration, to inhibit heme oxygenase activity, increased MAP (P<.05) and total peripheral resistance (P<.05). ZnDPBG decreased cardiac output (P<.05) but did not affect HR (P>.05). DPBG administration, which does not affect heme oxygenase activity, did not affect MAP, total peripheral resistance, cardiac output, or HR (P>.05 each). Also, vehicle administration did not affect (n=5, P>.05) MAP (104±1 versus 101±1 mm Hg), total peripheral resistance (1.40±0.02 versus 1.41±0.02 mm Hg/[mL/min]), cardiac output (76±1 versus 74±1 mL/min), or HR (437±1 versus 436±2 beats per minute).
Fig 2⇓ shows the data on HR and MAP before and after administration of ZnPP, biliverdin, or vehicle. ZnPP administration, to inhibit heme oxygenase activity, increased MAP (P<.05) but did not affect HR (P>.05). Biliverdin administration, which is a heme oxygenase product, did not affect either MAP or HR (P>.05 each). Also, vehicle administration did not affect arterial pressure or HR (P>.05 each).
Fig 3⇓ shows the data of MAP and HR before and after vehicle or ZnDPBG administration in rats pretreated with chlorisondamine to block ganglionic function. Chlorisondamine pretreatment lowered BP and decreased HR (P<.05 each). In chlorisondamine-pretreated animals, ZnDPBG administration had no effect on BP (P>.05) but did decrease HR (P<.05). Vehicle administration had no effect on BP or HR (P>.05). ZnDPBG did not affect the BP (105±5 versus 97±3 mm Hg, n=5) of rats in which the vasodepressor response to pretreatment with chlorisondamine was offset by infusion of phenylephrine.
Fig 4⇓ shows the data of MAP and HR before and after vehicle and ZnDPBG administration in rats pretreated with prazosin to inhibit α1-adrenoceptor function. Prazosin pretreatment reduced BP (P<.05) and increased HR (P<.05). In prazosin-pretreated animals, neither vehicle nor ZnDPBG administration had any affect on BP or HR (P>.05, each).
The resting tone of rings of descending thoracic aorta bathed in Krebs’ bicarbonate buffer did not change with the addition of ZnDPBG (50 μmol/L) to the bath. Also, aortic constrictor responses to phenylephrine (1 μmol/L) were similar in the presence and absence of ZnDPBG (87±8% versus 83±5% of contractions induced by 120 mmol/L KCl).
Since ZnDPBG and ZnPP inhibit heme oxygenase–mediated conversion of heme to carbon monoxide and biliverdin,9 10 11 12 the effects of these inhibitors on hemodynamics may provide insights into the physiological influence of heme oxygenase products on BP regulation. ZnDPBG and ZnPP administration produces pressor responses that are maximal within 5 minutes and maintained for approximately 1 hour. ZnDPBG-induced increases in arterial pressure are paralleled by increases in total peripheral resistance, establishing systemic vasoconstriction as the major hemodynamic determinant of the increase in BP. Importantly, nonmetallic DPBG, which does not inhibit heme oxygenase activity, does not affect hemodynamics. This strengthens the argument that the pressor effect of ZnDPBG is a consequence of decreased heme oxygenase activity. If so, these findings imply that heme oxygenase activity is subserving a vasodepressor function by promoting vasodilation.
Elevation of BP after pretreatment with an inhibitor of heme oxygenase may be a consequence of increased cellular levels of heme or diminished formation of biliverdin or carbon monoxide products. As heme turnover rate is slow,3 it is unlikely that the ZnDPBG- and ZnPP-induced pressor responses, which have a rapid onset, are a consequence of heme accumulation, secondary to decreased heme degradation. Although biliverdin is an antioxidant and may potentially lower BP by inhibiting lipid peroxidation and/or by prolonging the half-life of nitric oxide,5 14 in our studies biliverdin does not decrease BP. Therefore, ZnDPBG- and ZnPP-induced increases in arterial pressure are more likely to be the consequence of a diminished carbon monoxide production than of heme accumulation or decreased biliverdin production.
According to the present study, inhibitors of heme oxygenase increase BP via mechanisms that rely on autonomic nervous function. This conclusion is based on the observation that the ZnDPBG-induced rise in BP is prevented by pretreatment with chlorisondamine to block ganglionic function. Also, the ZnDPBG-induced increase in arterial pressure is prevented by pretreatment with prazosin, to block α1-adrenoceptors, implying that the pressor effect of the heme oxygenase inhibitor depends on α1-adrenoceptor function. That ZnDPBG does not affect constrictor responses of aortic smooth muscle to phenylephrine, or increase the BP of chlorisondamine-pretreated rats undergoing infusion of phenylephrine, argues against the possibility that the inhibitor of heme oxygenase acts postsynaptically to increase sympathetically mediated vasoconstrictor tone. Rather, the results of the present study are consistent with the notion that inhibitors of heme oxygenase affect presynaptic events leading to augmentation of sympathetic activity and BP. Of note, an inhibitor of heme oxygenase was recently shown to block the effect of metabotropic glutamate receptor activation in the nucleus tractus solitarii of rats,7 implicating carbon monoxide in the function of the afferent arm of the baroreceptor reflex and, consequently, in the neurogenic control of BP.
To the extent that carbon monoxide arising from heme oxygenase activity subserves a vasodepressor function, interventions that increase heme oxygenase–catalyzed conversion of heme to carbon monoxide may be expected to lower BP. A previous study15 has shown that increased heme oxygenase activity produced by tin chloride treatment is associated with a lowering of BP in spontaneously hypertensive rats. Another study16 demonstrated that heme arginate treatment, which also increased heme oxygenase activity, similarly lowered BP in that rat strain.
In summary, ZnDPBG and ZnPP, known to inhibit heme oxygenase–mediated conversion of heme to carbon monoxide, increase arterial pressure. Pretreatment with chlorisondamine or prazosin prevents ZnDPBG from increasing arterial pressure, implying that the pressor effect of this agent relies on autonomic nervous function. We conclude that the pressor response to heme oxygenase inhibitors results from withdrawal of the inhibitory influence of endogenous carbon monoxide on a pressor mechanism mediated by the autonomic nervous system. A corollary of this conclusion is that carbon monoxide arising from heme via metabolism by heme oxygenase exerts a tonic counterregulatory influence on autonomic mechanisms that promote elevation of BP.
This research was supported by grants 5PO1 HL-34300, HL-36670, and HL-18579 from the National Institutes of Health, Bethesda, Md. We would like to thank Jennifer Brown for assistance in preparing the manuscript.
- Received June 16, 1994.
- Revision received August 16, 1994.
- Accepted October 25, 1994.
Maines MD. Heme oxygenase: function, multiplicity, regulatory mechanisms, and clinical applications. FASEB J. 1988;2:2557-2568.
Moore MR, McColl KEL, Rimington C, Goldberg A. Porphyrins and enzymes of the heme biosynthetic pathway. In: Wintrobe MM, ed. Disorders of Porphyrin Metabolism. New York, NY: Plenum Medical Book Co; 1987:21-44.
Glaum SR, Miller RJ. Zinc protoporphyrin-IX blocks the effects of metabotropic glutamate receptor activation in the rat nucleus tractus solitarii. Mol Pharmacol. 1993;43:965-969.
Verma A, Hirsch DJ, Glatt CE, Ronnett GV, Snyder SH. Carbon monoxide: a putative neural messenger. Science. 1993;259:381-384.
Pucci ML, Lin L, Nasjletti A. Pressor and renal vasoconstrictor effects of NG-nitro-L-arginine as affected by blockade of pressor mechanisms mediated by the sympathetic nervous system, angiotensin, prostanoids and vasopressin. J Pharmacol Exp Ther. 1992;261:240-245.
Sacerdoti D, Escalante B, Abraham NG, McGiff JC, Levere RD, Schwartzman ML. Treatment with tin prevents the development of hypertension in spontaneously hypertensive rats. Science. 1989;243:388-390.
Levere RD, Martasek P, Escalante B, Schwartzman ML, Abraham NG. Effect of heme arginate administration on blood pressure in spontaneously hypertensive rats. J Clin Invest. 1990;86:213-219.