Identification of a Putative Microvascular Oxygen Sensor
The vascular response to changes in oxygen levels in the blood and tissue is a highly adaptive physiological response that functions to match tissue oxygen supply to metabolic demand. Defining the cellular mechanisms that can sense physiologically relevant changes in Po2 and adjust vascular diameter are vital to our understanding of this process. A cytochrome P450 (P450) enzyme of the 4A family of ω-hydroxylases was localized in renal microvessels, renal cortex, and a striated muscle microvascular bed (cremaster) of the rat. In the presence of molecular oxygen, this P450 enzyme catalyzes formation of 20-HETE from arachidonic acid (AA). Prior studies have shown that 20-HETE potently contracts renal and cerebral arteries and arterioles. The present study demonstrates that 20-HETE constricts striated muscle arterioles as well. In both intact renal microvessels and enriched renal cortical microsomal enzyme preparations, the formation of 20-HETE was linearly dependent on Po2 between 20 and 140 mm Hg. Homogenates of cremaster tissue produced 20-HETE when incubated with AA. They also expressed message for P450 4A enzyme, as determined by Southern and Western blots. Administration of 17-octadecynoic acid (17-ODYA), which is a P450 4A inhibitor, attenuated the constriction of third-order cremasteric arterioles in response to elevation of superfusion solution Po2 from ≈3 to 5 mm Hg to ≈35 mm Hg. 17-ODYA had no effect on basal vascular tone or response of cremaster arterioles to vasoactive compounds. These results demonstrate the existence of P450 ω-hydroxylase activity and 20-HETE formation in the vasculature and parenchyma of at least two microvascular beds. Our data suggest that a P450 enzyme of the 4A family has the potential to function as an oxygen sensor in mammalian microcirculatory beds and to regulate arteriolar caliber by generating 20-HETE in an oxygen-dependent manner.
The vasculature is sensitive to changes in blood and tissue Po2. Reducing blood Po2 or inducing tissue ischemia, with the notable exception of the lung, dilates arteries and arterioles, which increases blood flow and oxygen delivery.1 Conversely, increasing tissue Po2 delivery against a background of stable tissue oxygen demand elicits a vasoconstrictor response in most peripheral arterial beds.2 3 Despite a great deal of work in this area, the cellular or molecular sensor(s) responsible for mediating vascular responses to a changing Po2 remains poorly defined. The importance of identifying the cellular oxygen sensor/sensors lies in the fact that the vascular response to oxygen is a vital and highly adaptive physiological response that matches tissue oxygen delivery to metabolic demand. An inability of the vasculature to respond to elevations in metabolic oxygen demand, eg, during exercise, tachycardia or elevated neuronal activity, can lead to tissue ischemia and cell death. Conversely, overperfusion of a vascular bed leads to vasoconstriction and, over the long term, possible loss of microvessels (rarefaction) and hypertension.4
It is generally thought that both parenchymal tissue and the vascular wall can respond to a changing Po2. For example H+, adenosine, K+, prostaglandins, and other metabolites are released from nonvascular tissue during hypoxia, and all of these are capable of actively dilating the microvasculature to increase blood flow.5 Similarly, endothelial cells in the arterial wall release vasodilator products of the cyclooxygenase pathway in response to reductions in Po2.6 Most studies to date have used large reductions of Po2 to induce dilation of isolated vessels or intact vascular beds by the above mechanisms. An enzyme system that is capable of generating a vasoactive product and that possesses a Km for oxygen within the normal physiological range of blood and tissue Po2 has yet to be identified.
The present study describes a P450 enzyme in renal and striated microvascular beds that is capable of producing a potent endogenous constrictor substance and uses molecular oxygen within the normal physiological range of blood and tissue Po2. P450 enzymes oxidize a plethora of substrates, and >230 cDNAs have been identified.7 One of the endogenous substrates for the P450 4A family of enzymes is AA. P450 requires several cofactors, including molecular oxygen.8 9 P450 enzymes produce a series of vasoactive metabolites.8 One of these metabolites is an ω-hydroxylation product of AA, 20-HETE, which is produced by renal10 and cerebral arterioles11 and by skeletal muscle and renal parenchymal tissue (present study). We have recently reported that 20-HETE is a potent constrictor of renal10 and cerebral11 resistance arteries and arterioles and that the threshold concentration necessary to constrict these vessels is <10−10 mol/L.11
Materials and Methods
In the present study, we used both the renal and striated muscle (cremaster) vascular beds for several reasons. First, one of the aims of this study was to determine the oxygen-dependent nature of 20-HETE formation by P450 4A ω-hydroxylase enzymes. Since we have previously demonstrated that 20-HETE is avidly produced by renal microvessels and that these vessels are easy to isolate in bulk,10 we chose renal vessels to examine whether 20-HETE production in vascular tissue is dependent on oxygen concentration. Second, to rule out the possibility that the oxygen-dependent production of 20-HETE in intact microvessels is due to a diffusion limitation, we needed an enriched source of the enzyme in a broken cell preparation. Because of the amount of tissue needed, it was impractical to use microvessels to make a microsomal fraction in sufficient quantities for such studies. Since we have reported that the renal cortex of rats contains the same P450 4A isoforms as renal vessels, we used the kidney as an abundant source of tissue from which to make enzyme-enriched microsomes. Finally, to determine the physiological significance of P450 4A enzymes as oxygen sensors within a microvascular bed, we needed a preparation that responds to changes in Po2 and can be studied in vivo and whose oxygen responses are well documented. To fulfill this requirement, we used the rat cremaster microvascular preparation, which was documented to generate 20-HETE and to express P450 4A ω-hydroxylase message in the present study. All experimental protocols for the studies presented here were reviewed and approved by the Animal Care and Use Committee of the Medical College of Wisconsin.
Preparation of Renal Microvessels
Male Sprague-Dawley rats weighing between 250 and 280 g were used in the study. The rats were anesthetized with Inactin (10 mg/100 g body weight). Renal preglomerular microvessels were isolated using a modification of the method of Chaudhari and Kirschenbaum.12 Briefly, a midline abdominal incision was made, and the abdominal aorta was cannulated below the renal arteries. The aorta above the renal arteries was ligated, and the kidneys were first flushed with PSS and then filled with 15 mL of a 5% (wt/vol) suspension of iron oxide particles (5 to 10 μm, Aldrich Chemical) in PSS. After the kidneys were filled, they were removed, and the renal cortical tissue was dissected out. The microvessels were separated from the cortical tissue by pressing the tissue through a stainless steel screen (180-μm mesh) and rinsing with PSS. The retained tissue was passed several times through an 18-gauge needle attached to a 20-mL syringe to shear off any adhering glomeruli. A magnet (Bio Mag separator, Advanced Magnetics) was used to recover the vascular from nonvascular tissue. The microvessels that were obtained in this manner were then treated with collagenase type II (Worthington Biochemical) to free any remaining tubules attached to them. After the collagenase treatment, the microvessels were checked for purity by viewing them under a stereomicroscope (magnification, ×60) and by measuring alkaline phosphatase activity to rule out the presence of renal tubules. Renal tubules contain very high alkaline phosphatase, whereas microvessels do not. Microvessel preparations containing >5% renal tubules, resulting in detectable alkaline phosphatase activity, were not used for these studies.
Intact iron oxide–filled renal microvessels or microsomes prepared as described below from cremaster muscle (0.5 mg protein) were incubated with [1-14C]AA (0.5 μCi/mL, 10 μmol/L, DuPont; specific activity, 53 mCi/mmol) in 2 mL of a 100 mmol/L potassium phosphate buffer (pH 7.4) containing (mmol/L) MgCl2 5, EDTA 1, and NADPH 1 and an NADPH-regenerating system (10 mmol/L isocitrate and 0.4 U/mL isocitrate dehydrogenase) for 60 minutes at 37°C in the presence of various concentrations of oxygen. Humidified oxygen gas mixture was bubbled overnight directly into the assay solution from a weather balloon attached to a pump. In the morning, the concentration of oxygen in an aliquot of the medium was measured using a blood gas analyzer (ABL-2, Radiometer). When the desired oxygen concentration was established in the medium, renal microvessels and NADPH were added to the medium, and the reaction was started by the addition of [14C]AA. The reaction was terminated after 60 minutes by acidification to pH 4.0 with formic acid. The metabolites of AA were extracted with ethyl acetate and separated using a 2 mm×25 cm C18 reverse-phase analytical HPLC column (Supelco Corp) in a linear elution gradient from 100% solution A (acetonitrile/water/acetic acid, 50:50:0.1 [vol/vol/vol]) to 100% solution B (acetonitrile/acetic acid, 100:0.1 [vol/vol]) over 40 minutes, with a final 10 minutes in 100% solution B. The radioactive metabolites were monitored using a radioactive flow detector (model 120, Radiomatic Instrument) as previously described.11 13 14
Preparation of Microsomes
Renal cortical tissue and cremaster muscle were homogenized separately in 3 vol of 10 mmol/L potassium phosphate buffer at pH 7.7 containing (mmol/L) sucrose 250, EDTA 1, phenylmethylsulfonyl fluoride 0.1, and MgCl2 10 using a Wheaton overhead homogenizer. The homogenate was centrifuged at 3500g to remove tissue chunks. The supernatant was then centrifuged for 15 minutes at 9000g followed by 60 minutes at 100 000g to obtain a microsomal pellet. All centrifugations were carried out at 4°C. Microsomes were resuspended in 100 mmol/L potassium phosphate buffer at pH 7.25 containing (mmol/L) EDTA 1, dithiothreitol 1, and phenylmethylsulfonyl fluoride 0.1, along with 30% glycerol.
Western Blot Analysis
Western blot analysis was performed on renal cortical microsomes and homogenized renal microvessels using a polyclonal antibody raised against a synthetic peptide corresponding to a sequence in the rat P450 4A ω-hydroxylase enzyme. This antibody cross-reacts with P450 4A1, 4A2, and 4A3 isoforms in renal tissue of the rat. The renal microsomes and the renal microvessel homogenates were separated by electrophoresis on a 7.5% SDS-polyacrylamide gel and transferred to a nitrocellulose membrane. The transfer was carried out at room temperature at 100 V for 1 hour at pH 8.3 in a solution containing 25 mmol/L tris(hydroxymethyl)aminomethane, 192 mmol/L glycine, and 20% (vol/vol) methanol. After transfer, nonspecific binding was blocked by immersing the membranes in 5% blocking reagent (Bio-Rad) in TBS-T (50 mmol/L Tris HCl, 0.2 mol/L NaCl, and 0.1% Tween 20, pH 7.5) followed by incubation with the polyclonal P450 4A rabbit antiserum (1:3000 dilution) for 2 hours at room temperature. After incubation with the P450 4A antiserum, the membrane was washed three times with TBS-T before incubating with a 1:1000 dilution of horseradish peroxidase–labeled secondary antibody (Bio-Rad). Immunoblots were visualized using an enhanced chemiluminescence kit (Amersham), and the image was captured on x-ray film.
RT-PCR and Southern Blot
Total RNA was isolated from cremaster muscle using Trizol reagent (GIBCO BRL). Oligo dT (5 μg)–primed RNA was reverse-transcribed for 1 hour at 37°C with M-MuLV RT (Pharmacia) in a volume of 33 μL. Aliquots of RT reactions (5 μL) were used for PCR amplification in a volume of 100 μL with 250 μmol/L dNTP, 1.5 mmol/L Mg2+, 100 pmol of each primer, and 2.5 U Taq DNA polymerase (Pharmacia). Specific primers for cytochrome P450 4A1, 4A2, and 4A3 were as follows: 4A1 forward, 5′-GTATCCAAGTCACACTCTCCA-3′; 4A1 reverse, 5′-CAGGACACTGGACACTTTATTG-3′; 4A2 forward, 5′-CTGTACCTTCTGTGAGTCGAG-3′; 4A2 re-verse 5′-GCTGGGAAGGTGTCTGGAGT-3′; 4A3 forward, 5′-CAGTACCTTCTGTGAGTCGAGA-3′; and 4A3 reverse, 5′-CTCTCTACTGTTCTGTATCAGA-3′.
Components of reactions excluding Taq polymerase were mixed, overlaid with mineral oil, and heated to 95°C for 3 minutes; then they were cooled to 80°C for the addition of Taq polymerase and cycled for 40 cycles at 94°C for 1 minute, 56°C for 1 minute, and 72°C for 2 minutes, ending with a 7-minute extension at 72°C. To verify the identity of an appropriately sized RT-PCR product, a Southern blot hybridization was performed with 4A1-, 4A2-, and 4A3-specific oligonucleotide probes. Forty microliters of each of the RT-PCR reaction products was electrophoresed on a 1.5% agarose gel and visualized by ethidium bromide staining under UV light. Gels were then denatured in 0.5 mol/L NaOH and vacuum-blotted onto Nytran-plus nylon membranes in 10× SSC for 90 minutes using a vacuum blotter (Bio-Rad). After transfer, the membrane was air-dried, and DNA was immobilized by UV cross-linking (Stratalinker, Stratagene). The membrane was hybridized for 12 hours at 42°C with fluorescein-11-dUTP 3′ end-labeled oligonucleotide probes (Amersham) specific for internal regions of 4A1, 4A2, and 4A3 PCR products. Probe sequences were as follows: 4A1, 5′-GCCACAATCACCTTCATCTCACTC-3′ (1412 to 1389); 4A2, 5′-GATTGTCCCAAGACTCTGAGAACTG-3′ (1876 to 1900); and 4A3, 5′-CATAGCCATGCTTATCTGCCATTC-3′ (1330 to 1353). The blots were washed with 5× SSC and 0.1% SDS twice for 5 minutes at room temperature, followed by two washes with 1× SSC and 0.1% SDS for 15 minutes at 50°C. Blots were then incubated with anti-fluorescein horseradish peroxidase–conjugated monoclonal antibody (Amersham), washed, treated with an enhanced chemiluminescence solution (Amersham), and exposed to x-ray film for ≈30 seconds. Positive controls consisted of full-length 4A1, 4A2, and 4A3 cDNA clones, which were isolated from the kidney of rats and inserted into a pCRII vector (Invitrogen). These templates were amplified with the respective 4A primers as above. Negative control reactions for each primer pair consisted of (1) RT with water and PCR with water and (2) PCR with non–reverse-transcribed total RNA. DNA molecular weight markers (Pharmacia 100-bp ladder) were used to estimate product sizes on agarose gels and were also used as a negative control for Southern blots.
Microcirculatory Studies in the Rat Cremaster Muscle
For the microcirculatory studies, male Sprague-Dawley or Wistar-Kyoto rats were anesthetized with sodium pentobarbital (45 to 50 mg/kg IP). The trachea was cannulated to ensure a patent airway, and a carotid artery and femoral vein were cannulated for the measurement of arterial pressure and for the administration of supplemental anesthesia, respectively. After the initial surgery was complete, the right cremaster muscle was prepared for television microscopy, without interrupting the deferential feed vessels.15
The rat was placed on the stage of a Leitz Laborlux microscope, and the cremaster muscle was continuously superfused with PSS equilibrated with a 0% O2/5% CO2/95% N2 gas mixture to ensure that oxygen delivery to the tissue was from the microcirculation and not from the superfusion solution. The PSS used in these experiments had the following ionic composition (mmol/L): NaCl 130, CaCl2 1.6, NaH2PO4 1.18, MgSO4 1.17, NaHCO3 14.9, and disodium EDTA 0.026. Succinylcholine chloride (0.1 mmol/L) was also added to the superfusion solution to prevent spontaneous contractions of the cremaster muscle.
Internal diameters of third-order arterioles were measured by television microscopy16 using a Leitz ×20/0.32 long-working-distance objective, an RCA model TC2011 television camera, a Setchell Carlson model 17M922 monitor (AVONIX), and a model IV-550 videomicrometer (For-A Instruments). The final magnification on the face of the television monitor was ×700, and the accuracy of the system was ±1 μm.
Resting tone in the arterioles was assessed by measuring the increase in diameter occurring during maximal dilation of the vessel in response to topical application of 10−4 mol/L adenosine. Endothelium-dependent vasodilator responses were assessed by determining the effect of 1 μmol/L acetylcholine on resting diameter of the arterioles. Arteriolar responses to increased oxygen availability were assessed by measuring vessel diameters before and after increasing the oxygen concentration of the superfusion solution from the control gas equilibration mixture (0% O2/5% CO2/95% N2) to one containing 5% O2/5% CO2/90% N2. This procedure increases oxygen availability to the tissue independent of a change in Po2 in the lumen of the vessels and provides an indication of the oxygen-dependent autoregulatory mechanisms in the tissue. The response of the vessels to the increase in oxygen availability during 5% O2 superfusion was tested before and after exposure of the cremaster muscle to 17-ODYA, a suicide substrate inhibitor of the P450A ω-hydroxylase enzyme and the formation of 20-HETE in vascular tissue.17
To determine the magnitude of the increase in Po2 occurring during elevation of superfusion solution oxygen concentration, tissue Po2 was measured in another series of experiments using Whalen-type oxygen microelectrodes (tip diameter, 2 to 6 μm). The electrodes were polarized at −0.7 V and calibrated in saline equilibrated with 0% O2 and 21% O2 immediately before and after (respectively) measurements were obtained in the tissue. Measurements were discarded if any significant change occurred in the calibration currents. Tissue Po2 was measured between capillaries in order to provide an estimate of Po2 in the least oxygenated parts of the tissue.18
One series of control experiments was performed to verify that any reduction in the constrictor response to 5% O2 superfusion in the presence of 17-ODYA was not due to a nonspecific inhibitory effect of the compound on the ability of the arterioles to contract in response to other activating stimuli. In that series of experiments, the ability of the arterioles to constrict in response to topical application of 1 μmol/L norepinephrine was tested before and after a 30-minute treatment of the cremaster muscle with 17-ODYA.
In a final series of experiments, we verified that 20-HETE causes constriction of cremasteric arterioles, as previously demonstrated for small arteries of the cerebral11 and renal19 vascular beds. In those experiments, the production of endogenous 20-HETE and the conversion of exogenous 20-HETE to other products of AA metabolism were blocked by superfusing the cremaster muscle with PSS containing 20 μmol/L miconazole to block the P450 system, 10 μmol/L baicalein to block the lipoxygenase pathway, and 1 μmol/L indomethacin to block the cyclooxygenase pathway. After a 20-minute superfusion with the blocking agents, arteriolar diameter was measured before and during superfusion of the bed with 1 nmol/L 20-HETE.
Data were summarized as mean±SEM. Differences between two means were assessed using Student's t test. ANOVA was used to compare multiple means, and when significant differences were detected, a Newman-Keuls post hoc test was used to determine which means were different. A probability value of P≤.05 was considered to be statistically significant for all tests.
Fig 1A⇓ shows a Western blot using a P450 4A polyclonal antibody that demonstrates that the same P450 4A isoforms are expressed in the renal cortex and renal microvessels of the rat. Fig 1B⇓ is a reverse-phase HPLC chromatogram indicating that the major P450 metabolite of AA that is generated by both renal cortical microsomes and by renal microvessels coelutes with authentic 20-HETE standards. In Fig 1B⇓, the peaks labeled as 1, 2, and 3 coelute with 14,15-, 11,12-, and 8,9-diHETEs, respectively, which are hydrolysis products of 14,15-, 11,12-, and 8,9-EETs.
In these studies, rat renal microvessels were used as a model system for the vascular production of 20-HETE, since they can be readily harvested in sufficient quantities for biochemical and molecular analysis. These vessels were incubated with radiolabeled AA, and the conversion to 20-HETE was determined under conditions of varying Po2. Fig 2⇓ shows that the formation of 20-HETE by renal microvessels depends on the Po2 of the incubation medium. Half-maximal generation of 20-HETE occurs at a Po2 of ≈60 mm Hg. We also measured the formation of EETs by renal microvessels as a function of Po2. As can be seen in Fig 2⇓, the inhibition of EET formation during reductions in Po2 was markedly less than that of 20-HETE and was not significantly reduced from control values until Po2 was lowered to <30 mm Hg.
At this point, we felt that it was important to determine whether the effect of lowering Po2 on 20-HETE formation was due to a direct effect on the P450 enzyme responsible for 20-HETE formation or whether it was simply due to diffusion limitation for oxygen across the vascular wall. To answer this question, we needed to prepare an enriched source of the enzyme. Because it was impractical to obtain a sufficient quantity of renal microvessels for such analysis (eg, one experiment would require vessels from 25 to 50 rats), we used renal cortical microsomes as an enriched source of P450 4A enzymes. As seen in Fig 1⇑, the rat renal cortex expresses the same P450 4A isozymes that are found in renal microvessels and generates 20-HETE when incubated with AA. Fig 3⇓ demonstrates that the production of 20-HETE by renal cortical microsomes is also highly dependent on Po2 in the range of oxygen tensions between 30 and 140 mm Hg. In these experiments, the half-maximal inhibition of the production of 20-HETE occurred at a Po2 between 60 and 70 mm Hg.
The difference between Figs 2 and 3⇑⇑ (ie, the vessel segments in Fig 2⇑ show no change in 20-HETE production between 140 and 90 mm Hg, whereas the enriched microsomal fractions in Fig 3⇑ do) was most likely due to (1) diffusion limitation of oxygen in the vessel segments versus a lack of diffusional limitation in microsomes and (2) the more linear response to changes in substrate (oxygen) concentration that would be expected in the enzyme-enriched microsomes at both higher and lower substrate levels.
To explore the possible physiological relevance of oxygen-dependent changes in vascular 20-HETE production, we studied a vascular bed (the rat cremaster muscle), which consistently responds to changes in local Po2 in the physiological range. When incubated with [14C]AA, we observed generation of 20-HETE upon HPLC (Fig 4⇓). Furthermore, a Southern blot of RT-PCR products from the rat cremaster demonstrates the presence of both of the P450 4A ω-hydroxylase isoforms found in renal tissue (Fig 5⇓).
Fig 6⇓ summarizes the response of in situ third-order arterioles of the cremaster muscle to a step elevation in superfusion oxygen concentration from 0% O2 (superfusate Po2, 3 to 5 mm Hg) to 5% O2 (superfusate Po2, ≈35 mm Hg). In a separate series of experiments (eight sites in four rats), we demonstrated that our procedure for elevation of superfusion solution oxygen concentration results in a significant increase (P<.05) in tissue Po2 from 21±3.0 mm Hg during 0% O2 superfusion to 38±4.6 mm Hg during 5% O2 superfusion. In the present experiments, cremasteric microvessels exhibited a 6-μm (Fig 6A⇓) or 25% (Fig 6B⇓) reduction in diameter in response to the elevation of superfusion solution oxygen concentration from 0% O2 to 5% O2. After treatment with 17-ODYA, the P450 suicide substrate inhibitor of P450 ω-hydroxylase,13 14 the constriction of the arterioles in response to increased superfusion solution Po2 was significantly inhibited, independent of any vehicle effect (Fig 6⇓).
After the observations summarized in Fig 6⇑ were made, we conducted a series of control studies to rule out any effect of 17-ODYA on basal tone that might interfere with the ability of the arterioles to respond to changes in Po2. Fig 7⇓ is a reverse-phase HPLC chromatogram of AA metabolites demonstrating that 10 μmol/L 17-ODYA, which blocked the response of the cremaster arterioles to elevation of superfusion solution Po2, also blocks the formation of 20-HETE in the rat renal microsomes. 17-ODYA (10 μmol/L) had no effect on resting tone in the arterioles, nor did it enter the systemic circulation to affect arterial pressure (Fig 8⇓). To rule out any effect of 17-ODYA on the ability of the arterioles to respond to other endogenous vasoactive substances, we conducted a separate series of experiments to determine the effect of 17-ODYA on the response of cremasteric arterioles to acetylcholine, adenosine, and norepinephrine. 17-ODYA (10 μmol/L) did not affect acetylcholine- or adenosine-induced dilation (Fig 9⇓) or norepinephrine-induced constriction (Fig 10⇓) of the arterioles. To determine the direct effect of 20-HETE on third-order cremasteric arterioles, we superfused the preparation with 1 nmol/L 20-HETE as described in “Materials and Methods.” In these experiments, the arterioles exhibited significant constriction from a control diameter of 27±2.4 to a final diameter of 15±4.3 μm.
The present results demonstrate that the generation of the P450 4A metabolite 20-HETE by vascular tissue is directly dependent on the concentration of oxygen within the normal physiological range of blood and tissue Po2. Although all P450 enzymes require molecular oxygen, the majority of them (such as those found in the liver) require only very low Po2 levels for normal activity. The unique characteristic of the extrahepatic P450 enzymes responsible for 20-HETE formation is the high Po2 required for the catalytic activity of these enzymes. 20-HETE is generated through ω-hydroxylation of AA catalyzed by enzymes of the P450 4A gene family.13 We have used RT-PCR and Western blot analysis to identify that P450 4A1 and 4A2 isoforms are expressed in renal cortical tissues, renal and cerebral microvessels,10 11 19 20 21 22 and now in the cremaster muscle of the rat.
Recent reports from our laboratory and others strongly support the hypothesis that 20-HETE is a mediator of physiologically important functions. For example, inhibition of 20-HETE formation blocks autoregulation of renal blood flow in rats.23 20-HETE is endogenously produced in smooth muscle cells isolated from renal or cerebral arterioles. 20-HETE is an extremely potent activator of arterial muscle, constricting isolated cerebral and renal arteries at nanomolar concentrations,11 19 similar to their vasoactive properties on rat cremaster muscle arterioles in the present study. Therefore, conditions that enhance the endogenous activity of P450 ω-hydroxylase activity and 20-HETE formation in the vasculature would be expected to lead to vasoconstriction. Reduction of P450 ω-hydroxylase activity, on the other hand, might be expected to inhibit physiological responses mediated by 20-HETE. The in vivo studies described in the present study were designed to explore the possibility that 20-HETE may participate in the vasoconstrictor response to elevations in peripheral Po2. Toward this end, we demonstrated that a concentration of 17-ODYA that is sufficient to block the formation of 20-HETE by renal arterioles10 inhibits the response of arterioles of the rat cremaster muscle to elevations in superfusion Po2, but it did not affect resting vascular tone (which might reduce the ability of oxygen to reduce arteriolar caliber). Although the absolute specificity of any pharmacological inhibitor should always be questioned, we have, to the best of our abilities, shown that 17-ODYA blocks formation of P450 metabolites while leaving other products of AA metabolism intact. Although it is still possible that 17-ODYA might exert some nonspecific action on arterial muscle that influences the vascular response to a changing Po2, it does not do so by altering resting tone in the arterioles.
In the present study, we used multiple tissue types to determine the oxygen sensitivity of the P450 4A ω-hydroxylase and in vivo response to P450 inhibition for several reasons. First, it is impossible to obtain sufficient numbers of microvessels to prepare microsomes. Therefore, we used renal cortical microsomes as an enriched source of the enzyme, since renal microvessels and renal cortex both produce 20-HETE and express the same isoforms of P450 4A enzymes. Second, the response of renal microvessels to changes in Po2 is not well documented, and it would be difficult to study oxygen-mediated responses of this vascular bed in vivo. As a result, we wanted to use an established in vivo preparation that is routinely used as a model of normal microvascular responses and in which the response to changes in Po2 are well established. In the present study, we demonstrated that 20-HETE was made by the rat cremaster muscle and that the constriction of rat cremasteric arterioles in response to increased oxygen availability is inhibited by 17-ODYA, which inhibits 20-HETE formation by the P450 4A ω-hydroxylase enzyme.
Prior studies from our laboratory have used gas chromatography/mass spectroscopy to confirm that 20-HETE is the major P450 metabolite of AA that is endogenously formed by microvessels.11 One of the principal cellular mechanisms of action of 20-HETE is to inhibit the activity of the large-conductance KCa channel.11 Thus, changes in 20-HETE concentration within cells may alter the activity of the KCa channel, depending on intracellular Ca2+ concentration and pH. If increases in Po2 lead to increases in 20-HETE production and reducing Po2 inhibits formation of 20-HETE in the vascular smooth muscle cells or parenchymal tissue, it would be reasonable to expect that higher concentrations of 20-HETE during increased oxygen availability would lead to decreases in KCa channel activity, causing depolarization and contraction of the vessels. Conversely, decreased formation of 20-HETE during exposure to reduced Po2 could lead to increases in KCa channel activity, hyperpolarizing the vascular smooth muscle and leading to arteriolar dilation. Although it is not established that this sequence of events occurs in response to changes in Po2 in all blood vessels, hypoxia does increase the activity of the KCa channel in isolated cat cerebral arterial muscle cells.20
In summary, the present experiments and previous studies in the literature show that (1) 20-HETE is a potent vasoactive agent generated by a P450 ω-hydroxylase enzyme of the 4A family of genes, (2) the rat renal and cremaster muscle vasculature and surrounding parenchymal tissue express P450 4A protein and generate 20-HETE, and (3) the enzyme responsible for 20-HETE formation is sensitive to oxygen in the normal physiological range of tissue Po2. Although these conclusions do not imply that a P450 4A enzyme is an oxygen sensor in all microvascular beds or that it is the only enzyme whose activity is sensitive to physiological levels of Po2, the present findings provide a tenable hypothesis upon which further studies on this fundamentally important topic can be designed.
Selected Abbreviations and Acronyms
|KCa channel||=||Ca2+-activated K+ channel|
|M-MuLV||=||Moloney murine leukemia virus|
|PCR||=||polymerase chain reaction|
|PSS||=||physiological salt solution|
This study was supported in part by grants HL-33833, HL-37374, HL-36279, and NS-32321 from the National Institutes of Health and by a grant from the Veterans Administration. Andrew Lange is a predoctoral candidate supported by the Medical Scientist training program of the Medical College of Wisconsin. The authors would like to thank Dr William Jackson for his valuable suggestions and Nicole Gruber for secretarial assistance.
- Received February 5, 1996.
- Accepted March 27, 1996.
Jackson WF. Arteriolar oxygen reactivity is inhibited by leukotriene antagonists. Am J Physiol. 1989;257:H1565-H1572.
Duling BR. Microvascular responses to alterations in oxygen tension. Circ Res. 1972;31:481-489.
Lombard JH, Hinojosa-Laborde C, Cowley AW. Hemodynamics and microcirculatory alterations in reduced renal mass hypertension. Hypertension. 1989;13:128-138.
Haddy FJ, Scott JB. Metabolically linked vasoactive chemicals in local regulation of blood flow. Physiol Rev. 1968;48:688-707.
Shawl PW, Campbell WB, Farrar MA, Magness RR. Oxygen modulates prostacyclin synthesis in ovine fetal pulmonary arteries by an effect on cyclooxygenase. J Clin Invest. 1992;90:2147-2155.
Nelson DR, Kamataki T, Waxman DJ, Guengerich FP, Estabrook RW, Feyereisen R, Gonzalez FJ, Coon, MJ, Cunsalus IC, Gothoh O, Okuda K, Nebert DW. The P450 superfamily: update on new sequences, gene mapping, accession numbers, early trivial names of enzymes, and nomenclature. DNA Cell Biol. 1993;12:1-51.
Ortiz de Montellano PR. Cytochrome P450 Structure, Mechanism, and Biochemistry. New York, NY: Plenum Publishing Corp; 1986:123-138.
Oliw EH, Guenerich FP, Oats JA. Oxygenation of arachidonic acid by hepatic monooxygenases: isolation and metabolism of four epoxide intermediates. J Biol Chem. 1982;257:3771-3781.
Zou AP, Imig JD, Ortiz de Montellano PR, Sui Z, Roman RJ. Inhibition of renal vascular 20-HETE production impairs autoregulation of renal blood flow. Am J Physiol. 1994;266:F275-F282.
Harder DR, Gebremedhin D, Narayanan J, Jefcoat C, Falck JR, Campbell WB, Roman RJ. Formation and action of a P450 4A metabolite of arachidonic acid in cat cerebral microvessels. Am J Physiol. 1994;266:H2098-H2107.
Chaudhari A, Kirschenbaum MA. A rapid method for isolating rabbit renal microvessels. Am J Physiol. 1988;254:F291-F296.
Zou AP, Ma YH, Ortiz de Montellano PR, Clark JE, Masters BS, Roman RJ. Effects of 17-ODYA, a suicide substrate inhibitor of P450 fatty acid ω-hydroxylase, on renal function in rats. J Pharmacol Exp Ther. 1994;268:474-481.
Lombard, JH, Hess ME, Stekiel WJ. Neural and local control of arterioles in SHR. Hypertension. 1984;6:530-535.
Lombard JH, Fleming JT, Liard JF, Roman RJ. Blockade of arachidonic acid metabolism by cytochrome P450 inhibits constriction of rat cremasteric arterioles in response to elevated PO2. Microcirculation. 1995;2:67. Abstract.
Lombard JH, Duling BR. Relative importance of tissue oxygenation and vascular smooth muscle hypoxia in determining arteriolar responses to occlusion in the hamster cheek pouch. Circ Res. 1977;41:546-551.
Imig JD, Zou AP, Stec DE, Harder DH, Falck JR, Roman RJ. Formation and actions of 20-hydroxyeicosatetraenoic acid in rat renal arterioles. Am J Physiol. 1996;270:R217-R227.
Kauser K, Clark JE, Masters BS, Ortiz de Montellano PR, Ma Y-H, Harder DR, Roman RJ. Inhibitors of cytochrome P-450 attenuate the myogenic response of dog renal arcuate arteries. Circ Res. 1991;68:1154-1163.
Ma Y-H, Gebremedhin D, Schwartzman ML, Falck JR, Clark JE, Master BS, Harder DR, Roman RJ. 20-Hydroxyeicosatetraenoic acid is an endogenous vasoconstrictor of canine renal arcuate arteries. Circ Res. 1993;72:126-136.
Zou AP, Imig JD, Ortiz de Montellano PR, Sui Z, Falck JR, Roman RJ. Effect of P450 ω-hydroxylase metabolites of arachidonic acid on tubuloglomerular feedback. Am J Physiol. 1994;266:F934-F941.