Jump to
Abstract
Abstract We have reported previously that several blood vessels of the rat and cow contain immunoreactive vasopressin and further suggested that this peptide might be produced locally. To provide additional support for this hypothesis, we conducted the present study to determine whether mRNA for arginine vasopressin is also present in blood vessels. Ribonuclease protection analysis of total RNA isolated from rat hypothalamus and aorta revealed the presence of arginine vasopressin message in both tissues but not in RNA isolated from liver, a tissue devoid of vasopressin. Subsequent comparison of the autoradiographic intensities of the signals in these two tissues indicated that vasopressin message was 100- to 1000-fold lower in aorta. Additional studies showed that RNA isolated from endothelium-denuded vessels contained levels of arginine vasopressin message similar to those in intact vessels, indicating that endothelium was not a major source of this message. These data were substantiated by further studies using a vasopressin radioimmunoassay, which showed that vasopressin peptide levels in intact and endothelium-denuded vessels did not differ. Thus, the present study showed that rat aorta contains arginine vasopressin mRNA as well as the vasopressin peptide and that both the message and the peptide are contained in nonendothelial structures. However, the data do not rule out endothelium as a possible source of vasopressin. These studies add further support to the hypothesis that blood vessels are capable of producing vasopressin.
Arginine vasopressin (AVP) is an antidiuretic hormone with potent vasoconstricting activity. This hormone is synthesized primarily in the supraoptic and paraventricular nuclei of the hypothalamus and then transported to the posterior pituitary where it is stored before secretion into the general circulation. Circulating AVP appears to be responsible for regulating antidiuretic activity within the kidney, but it remains unclear whether the levels of the circulating hormone in nonpathological states are sufficient to regulate vascular contractility.1
Over the last decade, a number of peripheral tissues in several species have been shown to contain and produce AVP. For example, AVP is present in significant quantities in the ovaries of rats, cows, and humans2 3 4 and is synthesized by isolated bovine ovarian cells.5 Immunoreactive AVP has also been found in both human and rat testis6 7 as well as adrenal6 8 and thymus.9 Several of these tissues have also been shown to contain AVP mRNA,10 11 thus providing additional support that these peptides are locally produced. Interestingly, Brattleboro rats, which have minimal amounts of central nervous system and circulating AVP, have normal levels of AVP in peripheral tissues,4 12 13 suggesting that peripheral and central AVP systems function independently.
We have recently reported the presence of immunoreactive AVP in various blood vessels of the rat and cow.14 This immunoreactive material appears indistinguishable from authentic AVP when analyzed by immunologic, biochemical, and physiochemical means. Burnstock and colleagues (Lincoln et al15 and Loesch et al16 17 ) have provided immunohistochemical data also indicating the presence of AVP in vascular endothelial cells of several species. These observations, plus the fact that blood vessels of Brattleboro rats and hypophysectomized Sprague-Dawley rats contain levels of immunoreactive AVP that are no different from those in normal animals,14 led us to suggest that the vascular peptide was not derived from the circulation but was produced locally.14 To provide further support for this hypothesis, we performed the present studies to determine whether blood vessels contained mRNA for AVP.
Methods
RNA Preparation
Male Sprague-Dawley rats (250 to 350 g; Harlan Sprague Dawley, Madison, Wis) were killed by CO2 asphyxiation following procedures approved by the Institutional Animal Care and Use Committee of the University of Iowa. Various tissues were quickly removed, frozen in liquid nitrogen, and stored at −70°C. RNA was then isolated from each tissue by the method of Glisin et al.18 Tissues were pulverized with a mortar and pestle, homogenized in 10 vol of 20 mmol/L Tris (pH 7.4), 7 mol/L guanidine HCl, 10 mL/L N-lauroylsarcosine, and 1 mmol/L Na2EDTA and then centrifuged at 3000g at 4°C for 15 minutes. The supernatant was layered into sterile polyallomer tubes (14×95 mm) containing 3 mL CsCl (5.7 mol/L CsCl, 10 mmol/L Na2EDTA) and centrifuged at 100 000g in a Sorvall SW-27 rotor for 16 hours at 18°C. After centrifugation, the upper solution was aspirated and the tube inverted to drain the CsCl. The tube was then cut 2 cm from the bottom and allowed to drain inverted. The RNA pellet was resuspended in H2O (two times), transferred to a 1.5-mL microfuge tube, and precipitated overnight at −20°C in 0.3 mol/L sodium acetate and 2.5 vol ethanol. Precipitated RNA was collected by centrifugation (12 000g, 10 minutes, 4°C), washed once with 70% ethanol, and resuspended in H2O. RNA was measured by absorbance at 260 nm (A260), and purity was determined by the A260/A280 ratio.
Ribonuclease Protection Assay
RNA was analyzed for AVP mRNA by a ribonuclease (RNase) protection assay following the method of Liu et al.19 RNA samples (0.1 to 180 μg) dissolved in 200 μL H2O were combined with 5×105 cpm 32P-labeled cRNA probe (prepared as described below) and precipitated with 0.3 mol/L sodium acetate and 2.5 vol ethanol for 15 minutes at −70°C. Precipitated RNA and probe were collected by centrifugation (12 000g at 4°C, 10 minutes), washed once with 70% ethanol, and dried for 15 minutes in a vacuum oven at 55°C. Samples were resuspended in 30 μL hybridization solution (800 mL/L deionized formamide, 400 mmol/L NaCl, 1 mmol/L EDTA, and 40 mmol/L piperazine-N-N′-bis[2-ethanesulfonic acid] [PIPES], pH 6.4), heated at 85°C for 10 minutes, and incubated overnight at 55°C. Single-stranded RNA was digested by addition of 270 μL digestion buffer (300 mmol/L NaCl, 5 mmol/L EDTA, 10 mmol/L Tris-HCl, pH 7.4) containing RNase T1 (500 U/mL, GIBCO-BRL) and RNase A (10 mg/mL, Sigma Chemical Co) and then incubated for 1 hour at 22°C. Next, 20 μL sodium dodecyl sulfate (100 g/L) and 10 μL proteinase K (10 mg/mL, Sigma) were added and the samples incubated at 37°C for 30 minutes. After digestion, nucleic acids were extracted once with phenol/chloroform and once with chloroform, precipitated with ethanol at −70°C for 15 minutes, and pelleted by centrifugation as above. Pellets were washed with 70% ethanol, dried, and resuspended in 5 μL H2O and 10 μL stop buffer (Promega). The protected RNA fragments were separated by electrophoresis on an acrylamide (60 g/L)/urea (7.5 mol/L) gel. Gels were then fixed for 30 minutes in 10% methanol/12% acetic acid, dried, and exposed to Kodak XAR-5 film for 2 to 10 days. Probe fragment sizes were determined by including a lane containing a 100-bp DNA ladder (GIBCO-BRL) that was end-labeled with [α-32P]ATP (3000 Ci/mmol, DuPont-NEN) using T4 polynucleotide kinase (GIBCO-BRL),20 and input probe size was determined by including a lane containing 5×103 cpm of 32P-labeled cRNA probe.
Probe Preparation
A plasmid containing 622 bp of the AVP sequence (designated AVP8c and provided by Dr Thomas Sherman, University of Pittsburgh) was used to generate an AVP-specific cRNA probe. The plasmid was linearized by digestion with HindIII, and a cRNA probe was transcribed using DNA-dependent SP6 RNA polymerase and [α-32P]CTP (800 Ci/mmol, DuPont-NEN).20 This procedure produced a 674-nucleotide cRNA probe specific for AVP mRNA. Typically, specific activities were 108 cpm/μg RNA.
AVP Radioimmunoassay
Tissues were extracted and assayed as previously described.14 Briefly, tissues were removed, rinsed free of blood, and homogenized in 10 vol of 1.0 mol/L HCl. Homogenates were centrifuged to remove precipitated protein, and the neutralized supernatants were assayed for AVP immunoreactivity in a specific AVP radioimmunoassay. Cross-reactivity of the assay was 100% for lysine vasopressin, 0.7% for arginine vasotocin, and less than 0.1% for oxytocin, angiotensin II, and endothelin-1. The intra-assay and interassay coefficients of variation were 20% and 5%, respectively. The sensitivity of this short, overnight assay was 7 pg per tube.
Results
Detection of AVP mRNA in Rat Vascular Tissue
Initial experiments using Northern blot analysis of total RNA isolated from rat hypothalamus, aorta, or adrenal consistently identified a single AVP mRNA transcript of approximately 750 nucleotides in RNA isolated from rat hypothalamus but not in RNA isolated from rat adrenal, a tissue previously shown to contain low levels of AVP mRNA,8 or from rat aorta (data not shown). Consequently, we performed subsequent studies using the more sensitive RNase protection assay, which also allows the use of greater quantities of RNA. For each experiment, hypothalamus, aorta, adrenal, cerebellum, and liver were obtained from 20 rats and extracted for total RNA, and the extracts were examined for the presence of AVP mRNA by RNase protection analysis. Fig 1⇓ shows the results from a typical experiment that was replicated four times. A 622-bp protected cRNA probe fragment was identified in incubations containing RNA from hypothalamus or aorta but not in those containing RNA from liver or cerebellum; also present was a small amount of undigested input probe (674 bases). Although not visible in the figure, longer autoradiographic exposures revealed a protected probe fragment in incubations using RNA isolated from adrenal. Interestingly, a smaller fragment of approximately 350 bp was consistently detected in incubations using RNA isolated from cerebellum. This fragment may represent a portion of the AVP message previously identified in this tissue by Ivell et al11 and Rehbein et al.21 The autoradiographic intensities of the signals from rat aortic RNA in the four experiments were equivalent to those obtained with from 0.1 to 1 μg of hypothalamic RNA. These data suggest that the level of AVP mRNA in rat aorta is approximately 100- to 1000-fold lower than that in hypothalamus.
Blots show RNase protection analysis of RNA from rat tissues for arginine vasopressin (AVP) mRNA. Total cellular RNA was isolated from rat hypothalamus (lane 2, 10 μg; lane 3, 3 μg; lane 4, 1 μg), adrenal (lane 5, 150 μg), aorta (lane 6, 115 μg), cerebellum (lane 7, 130 μg), and liver (lane 8, 140 μg) and then analyzed for AVP mRNA by RNase protection assay. Nondigested input probe (674 bases, 5×103 cpm) is shown in lane 1. The AVP protected probe fragment (622 bp) is present in lanes 2, 3, 4, and 6. Probe sizes were determined by including a 100-bp DNA ladder that had been end-labeled with 32P (not shown). Data are from a typical experiment that was reproduced four times.
Vascular Localization of AVP and AVP mRNA
Burnstock and colleagues15 16 17 have previously demonstrated that between 10% and 50% of endothelial cells from rat vessels, including the aorta, mesenteric, and pulmonary arteries, stain positive for AVP by immunohistochemistry. However, they were unable to detect AVP immunoreactivity in any other vascular structures, including smooth muscle and adventitia. From their findings, we reasoned that if AVP were produced in endothelial cells, endothelium removal should deplete the vessels of AVP mRNA. To test this hypothesis, isolated rat aortas were first mechanically denuded of endothelium by dissecting the vessels longitudinally and scraping the endothelial surface with a blunt spatula. The effectiveness of this procedure was assessed by comparing scanning electron micrographs of the luminal surfaces of the intact and endothelium-denuded vessels. As shown in Fig 2⇓, the smooth appearance of the intact aorta (A) indicates the presence of vascular endothelium, whereas the rough, textured appearance of the denuded aorta (B) confirms the removal of endothelium. Subsequently, an experiment was performed in which aortas were collected from 40 rats; half were denuded of endothelium and half were left intact. The two groups were then extracted for RNA and examined for AVP mRNA by RNase protection assay. Fig 3⇓ shows results from this experiment, which was replicated three separate times and revealed 622-bp protected probe fragments when RNA from either endothelium-intact or endothelium-denuded vessels was used. Again, the autoradiographic intensities of the protected probe fragments from both intact and denuded vessels were approximately equivalent to the intensity obtained with 0.1 μg hypothalamic RNA. Since these studies indicated that endothelium denudation had little if any effect on the level of AVP mRNA present in the vessels, we next compared the level of AVP peptide in intact and endothelium-denuded vessels using a specific radioimmunoassay for AVP. These studies (n=3) showed that the level of AVP immunoreactivity in endothelium-denuded aorta (4.9±1.0 ng/g tissue) was not significantly different from that in intact aorta (4.5±1.7 ng/g tissue).
Scanning electron photomicrographs show intact (A) and endothelium-denuded (B) rat aorta. The lumina of intact and endothelium-denuded rat aorta were examined by scanning electron microscopy to determine the extent of endothelium removal after the mechanical denudation procedure outlined in “Results.” The lumen of normal, intact vessels is smooth in appearance, whereas that of the scraped, denuded vessels is rough in appearance and devoid of endothelial cells. White scale bars represent 100 μm. Original magnification ×356.
Blots show RNase protection analysis of RNA from intact and endothelium-denuded rat aorta for arginine vasopressin (AVP) mRNA. Total RNA was isolated from rat hypothalamus (lane 2, 0.3 μg; lane 3, 0.1 μg), intact aorta (lane 4, 120 μg), and endothelium-denuded aorta (lane 5, 100 μg) and then analyzed for AVP mRNA by RNase protection assay. Nondigested input probe (674 bases, 5×103 cpm) is shown in lane 1. The AVP protected fragment (622 bp) is present in lanes 2, 3, 4, and 5. Probe sizes were determined by including a 100-bp DNA ladder that had been end-labeled with 32P (not shown). Data are from a typical experiment that was reproduced three times.
Discussion
These studies have demonstrated the presence of AVP mRNA in rat aorta by use of the RNase protection assay. Addition of RNA from either hypothalamus or aorta to this assay resulted in the appearance of a 622-nucleotide protected probe fragment, the size predicted for samples containing AVP message. Because this probe contains the entire AVP gene sequence, this observation suggests that the AVP message in aorta is identical in structure to that in hypothalamus, although it is possible that the lengths of the poly(A) tails, which are not examined in this assay, may vary. In addition, the autoradiographic intensities of the protected probe fragment using slightly more than 100 μg rat aortic RNA were approximately equivalent to those obtained using 0.1 to 1 μg hypothalamic RNA, thus indicating that the AVP mRNA concentration in the aorta is approximately 1000- to 100-fold lower than in the hypothalamus. Interestingly, this ratio of aortic to hypothalamic mRNA is comparable to that previously found for AVP peptide levels in these same two tissues.14 In light of these discrepant levels, it is not surprising that Northern blot analysis was unable to detect AVP message in RNA isolated from rat aorta but was able to detect it in RNA isolated from rat hypothalamus. Other investigators have obtained similar results when attempting to identify low-level mRNAs in vascular tissue. For example, Holycross et al22 were unable to identify renin mRNA in rat aortic smooth muscle using Northern blot analysis, but Samani et al23 were able to detect this message by RNase protection analysis using RNA isolated from intact aorta.
Previous immunocytochemical studies performed by Burnstock’s group15 16 17 detected the presence of AVP only in vascular endothelium. When attempting to verify these findings in our studies, we were surprised to find no difference in the levels of AVP mRNA or AVP immunoreactivity between intact and endothelium-denuded aorta. These data clearly indicate that subendothelial structures such as vascular smooth muscle or perivascular adventitia possess both AVP message and peptide and may thus represent significant sites of AVP production. However, because endothelial cells constitute such a small proportion of total aortic cell mass (probably <1%), it is possible, and perhaps likely, that the decrease in total AVP peptide and message levels produced by the removal of these cells is insufficient to be detected in our assays. Thus, the present data do not address the issue of whether endothelial cells contain AVP message. However, the data of Loesch et al,16 showing that primary cultures of rabbit aortic endothelial cells contain AVP, support the possibility that endothelial cells do in fact contain, and probably produce, AVP. If so, endothelial cells may in fact contain AVP message. The present observations are not the first to indicate local vascular production of a vasoactive peptide. Several groups have previously demonstrated that endothelial cells synthesize a number of vasoactive peptides, including endothelin,24 angiotensin II,25 and C-type natriuretic peptide.26 In fact, endothelin was originally characterized, sequenced, and cloned from cultured porcine aortic endothelial cells. In addition to these vasoactive peptides, vascular cells have also been shown to contain other components of the angiotensin II–producing system, including angiotensin-converting enzyme,27 renin,22 23 and angiotensinogen.28
In conclusion, these studies have demonstrated the presence of AVP mRNA in rat aorta. This observation, coupled with the previous identification and characterization of AVP immunoreactivity in aorta (and other blood vessels) as well as the presence of AVP in the vasculature of AVP-deficient Brattleboro and hypophysectomized rats, provides compelling evidence that blood vessels are capable of synthesizing AVP. Although the functions of this locally produced peptide are not yet clear, it may have important regulatory functions within the vasculature, including regulation of local vascular tone through its vasoconstricting effects as well as regulation of vascular regrowth and remodeling through its growth-promoting effects.29 30
Acknowledgments
This work was supported by a grant from the National Institutes of Health (HL-44546), Bethesda, Md, and the University of Iowa Diabetes and Endocrine Research Center (DK 25295).
Footnotes
-
Reprint requests to Barry G. Kasson, PhD, Department of Pharmacology, College of Medicine, University of Iowa, Iowa City, IA 52242.
- Received May 30, 1994.
- Revision received September 21, 1994.
- Accepted January 3, 1995.
References
This Issue
Jump to
Article Tools
Share this Article
- Identification of Vasopressin mRNA in Rat Aorta
