 |
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
|
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
BASIC SCIENCE |
*Renovascular Pharmacology and Physiology (INSERM), University Louis Pasteur School of Medicine, Strasbourg, France; University Louis Pasteur Hospital Institute of Pathology, Strasbourg, France; and Institute of Pharmacology, University Louis Pasteur Medical School, Strasbourg, France
Correspondence to Dr. Jean-Jacques Helwig, Pharmacologie & Physiologie Rénovasculaires (INSERM-ULP 0015), 11 rue Humann, Bâtiment 4, 1er étage, F67085 Strasbourg Cedex, France. Phone: 333-90-24-34-54; Fax: 333-90-24-34-59; E-mail: jean-jacques.helwig{at}pharmaco-ulp.u-strasbg.fr
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
PTH-related protein (PTHrP), produced by most cells and organs throughout the normal body, is recognized to be a locally active polyhormone (7,8). PTHrP is posttranslationally processed to a family of daughter peptides, each having its own activity and likely its own receptor. PTHrP(1-36), the only PTHrP isoform structurally and functionally related to PTH, is equipotent to PTH(1-34) in binding and activating the so-called type 1 PTH/PTHrP receptor (PTH1R) in most organs (9,10), which is the only receptor for PTHrP that has been identified so far. Therefore, it seems reasonable to assume that the physiologically relevant ligand of the PTH1R responsible for inducing renin release or vasodilation is probably the locally produced PTHrP, and not PTH.
PTHrP(1-36), like PTH(1-34), stimulates the release of renin activity through direct interaction with renin-secreting juxtaglomerular cells in a dose-dependent manner (11). In these studies, PTHrP(1-36) was shown to stimulate the release of renin activity from nonfiltering kidneys, perfused under stable pressure and flow conditions, as well as from isolated perfused cells enriched in isolated juxtaglomerular cells (12). However, whether the endogenous PTHrP/PTH1R system controls the various stages of the RAS, including hepatic angiotensinogen (AGT) formation, renal renin release, and plasma renin activity (PRA), under physiologic conditions remains unclear. That angiotensin II (Ang II), the most effective end product of the RAS, has been established to induce rapidly and transiently PTHrP mRNA in vessels (13), including the renal vasculature in vivo (14), further suggests an interaction between endogenous PTHrP and the RAS.
In terms of vasoactivity, the PTH-related peptide system has been acknowledged to be expressed and to exhibit PTH1R-mediated vasodilatory properties in all vascular beds tested so far (15), including the renal vascular bed. For instance, in the rat, infusion of PTHrP(1-36) in the isolated perfused kidney (16,17) or in vivo directly into the renal artery of anesthetized animals (4) strongly decreases renal tone through a PTH1R-mediated pathway. In vivo, PTHrP(1-36) increases renal blood flow, GFR, and diuresis. In transgenic mice overexpressing the PTH1R or PTHrP, using the smooth muscle–specific SMP-8 
-actin promoter, BP was shown to be reduced, which correlates nicely with the vasodilatory properties of this peptide system (18,19). In PTH1R-overexpressing mice, in addition to a reduction in BP (19), the renal vasodilation induced by saline volume expansion has been shown to be exaggerated, whereas the vasoconstriction induced by Ang II was blunted (20). Importantly, smooth muscle–targeted overexpression of PTHrP (18) also resulted in desensitization of the vasculature to PTHrP, as well as to acetylcholine and probably also to other vasodilatory factors. These findings strongly support the involvement of locally produced PTHrP in the regulation of renal and systemic hemodynamics.
Nevertheless, the interpretation of the results obtained with transgenic models are difficult to consolidate. This is mostly due to the numerous compensatory mechanisms that would be expected to take place during fetal and postnatal life, which obviously increases the complexity of the phenotypic alterations associated with these transgenic models. This is valid for null-mice models, too, although the lack of PTHrP or PTH1R in knockout mice is lethal, at birth or in utero, because of a multitude of bone deformities (21). Ideally, a mouse strain in which the PTHrP or the PTH1R gene could be excised in a particular vascular cell type at a time of the postnatal life would allow the exploration of vascular biology in the full absence of this peptide system. The technology for generating such a conditional knockout could be based on the Cre/LoxP site-specific recombination system (22). Although mice with floxed PTHrP alleles have recently been generated (23), mouse strains that specifically express Cre-recombinase in vascular cells seem difficult to generate and are not yet available.
An alternative way to investigate the physiologic actions of the endogenous PTH/PTHrP/PTH1R system in the postnatal state is the intravascular injection of naked plasmid DNA (24). It is now well recognized that this simple method of transgenesis results in gene transfer and expression, predominantly in liver (25,26), as well as in most, if not all, vascular-accessible cells (24,27,28). Systemic administration of DNA solution also results in gene transfer in a number of extravascular organs, including lung (29,30), heart (30,31), spleen (30), kidneys (30), and skin (32). Overall, intravascular administration of DNA appears as an easy tool not only to analyze the regulation of gene expression in normal situations or to test the beneficial effect of replenishing the expression of downregulated genes in physiopathologic situations but also to uncover the functions of gene product in normal animals, with limited long-term compensatory mechanisms that take place during development in transgenic models. We recently reported in spontaneously hypertensive rats in which the PTH1R is downregulated that the intravascular administration of a DNA plasmid encoding the human PTH1R (hPTH1R) alters renal tone, as well as PRA (33). These findings strongly suggest, at least in this model of genetic hypertension, that locally produced PTHrP interacts with the RAS.
With this background, the present studies were designed to gain insight into the physiologic roles of the endogenous PTH/PTHrP/PTH1R system in the cardiovascular and the renovascular systems of normal adult rats, with particular emphasis on its role in regulating the key steps in the renin-angiotensin cascade. To address these questions, we explored the acute effect of direct intravenous administration of a DNA plasmid encoding the hPTH1R. This allowed a detailed analysis of the possible interactions of the PTH/PTHrP/PTH1R not only with systemic and renal vascular tone as documented before in SHR (33) but also with the RAS.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Human PTH1R Plasmid DNA Preparations and Delivery
The isolation of the hPTH1R cDNA (HKrk) from phage libraries and subcloning of the 1.9-kb insert, containing the entire coding sequence, into the 4-kb mammalian expression vector pcDNA1.1 has been described previously in detail (34). In previous studies, the amplification efficiency of pcDNA1.1 plasmid seemed low (33). The entire 1.9-kb hPTH1R gene therefore was removed from pcDNA1.1 and re-ligated into the 5.4-kb mammalian expression vector pcDNA3 (Invitrogen, Life Technologies, Carlsbad, CA). For that purpose, hPTH1R-pcDNA1.1 was digested using the restriction enzymes HindIII and XBaI, as indicated by the manufacturer (MBI Fermentas, Vilnius, Lithuania), to remove the 1.9 kb, corresponding to the hPTH1R cDNA. The digestion products were electrophoresed on 1.5% agarose gels, and the hPTH1R cDNA fragment was subcloned into the XBaI/HindIII restriction sites of the pcDNA3 vector, downstream of the cytomegalovirus (CMV) promoter. The plasmids pcDNA3-hPTH1R and empty pcDNA3 were transformed into the DH5 Escherichia coli host strain. Colonies were picked, amplified, and purified in a giga-preparation using a plasmid preparation kit (Qiagen, Chatsworth, CA), according to the manufacturer’s instructions. Nine-week-old Wistar rats received a single injection of 500 µg per animal of either hPTH1R-pcDNA3 or empty pcDNA3 (controls) in 700 µl of PBS-glucose 5% into the tail vein.
A human wild-type PTH1R fused at its C-terminal end to emerald green fluorescence protein of Aequorea Victoria (EGFP) inserted in pcDNA3 was also used. To this end, the receptor-EGFP conjugate in EGFP-N2 vector (provided by Dr. C. Silve, INSERM U.426, Paris, France) (35) was digested using the restriction enzymes NotI and HindIII (MBI Fermentas) under conditions suggested by the manufacturer. The digestion products were electrophoresed on 1.5% agarose gels, and the full-length hPTH1R-GFP cDNA (2.7 kb) was subcloned into the HindIII/NotI restriction sites of the multiple cloning site of pcDNA3 plasmid. The ligation mixture (pcDNA3-hPTH1R-GFP and empty pcDNA 3.1) was transformed into the DH5 E. coli host strain. Colonies were picked and amplified and plasmid DNA was prepared and injected using the same standard techniques as described above.
Two groups of two 9-wk-old rats received 500 µg per animal of either pcDNA3-hPTH1R-GFP plasmid cDNA or empty pcDNA3 plasmid (controls) in 700 µl of PBS-glucose 5% buffer into the tail vein. Three weeks after plasmid DNA administration, the rats were anesthetized by intraperitoneal injection of sodium pentobarbital (45 mg/kg). Kidneys, liver, and mesentery were fixed by intra-aortic perfusion of neutral-buffered paraformaldehyde (10%) and removed. Brain, heart, and aorta were fixed by an intracardiac perfusion of neutral-buffered paraformaldehyde (10%) and removed. After 2 d of buffered 10% paraformaldehyde fixation, the pieces were rinsed off of fixative, dehydrated in an ascending series of alcohols, passed through xylene, sectioned at 3 µm, and embedded in paraffin. For the GFP staining visualization, paraffin was removed by xylene baths and the sections were rehydrated in a decreased series of alcohol. Sections from empty pcDNA3-treated rat served as negative controls. Fluorescence was visualized using a Nikon epifluorescence microscope (TE200) equipped with a GFP(R)LP epi-fluorescence filter cube (excitation at 480 nm, emission at 510 nm).
Isolation of Rat Intrarenal Small Arteries
Twelve-week-old pcDNA3- and pcDNA3-hPTH1R–treated Wistar rats (n = 3 each) were anesthetized by intraperitoneal injection of sodium pentobarbital (45 mg/kg), and kidneys were removed, rinsed, and placed in ice-cold PBS. Renal arterial trees that contained mainly arcuate and interlobular arteries were isolated by a sieving method exactly as we have described previously (36). Isolated vessels were frozen in liquid nitrogen and ground before processing for mRNA, as described below.
Reverse Transcriptase–PCR Analysis for hPTH1R, Renin, and AGT Transcripts
Total RNA was isolated with TRIzol reagent (Life Technologies) from aorta, heart, lung, liver, brain, isolated renal vessels, and renal cortex according to the protocol of the manufacturer (Sigma-Aldrich, St. Quentin Fallavier, France) and quantified by its absorbance at 260 nm. Reverse transcription (RT) was performed on 5 or 15 µg of denatured RNA, using Moloney murine leukemia virus reverse transcriptase (Perkin Elmer, Roissy, France) and nonspecific P(dT)15 primer (2 µM) at 37°C for 1 h. PCR of the cDNA solutions was performed in the presence of specific sense and antisense primers of human PTH1R (33), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (33), renin, and AGT. The sense and antisense of the AGT primers were 5'-TTG TTG AGA GCT TGG GTC CCT TCA-3' and 5'-CAG ACA CTG AGG TGC TGT TGT CCA-3', corresponding to the nucleotide sequences of +638 to +661 and +901 to +878 of AGT cDNA, respectively. The sense and antisense of the renin primers were 5'-GTC AAA CTT GGC CAG CAT GA-3' and 5'-ATG CCT CTC TGG GCA CTC TT-3', respectively, corresponding to the nucleotide sequences of +661 to +638 and +48 to +67 of AGT cDNA.
Concentrations of hPTH1R (50 pM), renin (50 pM), AGT (50 pM), and GAPDH (4.4 pM) primers were adjusted in preliminary experiments to obtain comparable product amplification despite the different abundance of transcripts. The PCR was run for 30 cycles by repeating denaturation at 94°C for 1 min, annealing at 60°C for 1 min, and polymerization at 72°C for 1 min in the presence of Taq polymerase (Sigma-Aldrich). The last cycle was followed by an additional incubation at 72°C for 8 min. The PCR products were size-fractionated using 1.5% agarose gel electrophoresis. PCR products were identified by their expected size of 817 bp (hPTH1R), 263 bp (AGT), 552 bp (renin), and 415 bp (GAPDH).
For real-time PCR, total renal and liver cDNA were amplified using the LightCycler-FastStart DNA Master SYBR Green kit (Roche Diagnostics, Meylan, France) for quantification of renal renin and liver AGT mRNA. A standard curve was obtained for each cDNA by serial dilutions of mixed cDNA originating from either renal or liver samples. Amplifications were performed in a 20-µl mix, containing 4 mM MgCl2, 0.5 µM of each of the corresponding primer set, 50 ng of cDNA sample, and, according to the manufacturer, TaqDNA polymerase, dNTP (with dUTP instead of dTTP), reaction buffer, and SYBR Green I dye. PCR reactions were run at 95°C for 10 min, followed by 40 cycles of 10 s at 95°C, 5 s at 60°C, and 12 s at 72°C for renin and AGT or 16 s for GAPDH. As a negative control, cDNA was replaced by PCR-grade water. Each sample was quantified with the LightCycler analysis software according to the manufacturer’s protocol.
Heart Rate and BP Determination in Conscious Animals—Analysis of Urinary Parameters
Mean BP and heart rate were measured blindly in control and hPTH1R-treated rats on day 4 before and 6, 12, and 20 d after intravenous injections, with an electrosphygmomanometer (Letica Scientific Instruments LE5002, Barcelona, Spain) using the tail-cuff method. The conscious rats were placed in a thermostatically controlled (37°C) plastic restraint device during measurements. An average of at least five readings were taken for each animal after they became acclimated to the environment.
Control-injected and hPTH1R-DNA–injected Wistar-Han rats were placed in individual metabolic cages with free access to tap water and food 1 d before and 9 and 19 d after plasmid injection. After a 24-h period to acclimatize the rats, urine was collected over another 24-h period for the determination of electrolytes (Na+, K+, Ca2+, PO43–), as well as creatinine. Water and food intake was also assessed. Electrolytes were measured with ion-selective electrodes (EL-Ise; Beckman, Gagny, France), and creatinine was measured (Beckman auto-analyser) by standard colorimetric methods.
Measurement of PRA and Plasma AGT
PRA and plasma AGT were measured on 
1 ml of blood withdrawn from tail vein and collected in an ice-cold EDTA-coated tube. The blood was immediately centrifuged at 4°C for 10 min at 2000 x g. Plasma samples were stored at –20°C until assayed. For PRA, the blood was collected on day 3 before the injection of plasmid, as well as on days 4, 7, 13, and 21 after plasmid injection. The blood for plasma AGT was collected at day 21 after the injection of the plasmid.
PRA and AGT were measured exactly as described previously (37). Briefly, PRA was determined by the ability of circulating renin to generate Ang I from circulating AGT, during a 30-min incubation of plasma at 37°C. AGT was measured in another series of control or hPTH1R-treated rats by converting the entire circulating AGT into Ang I by an excess of renin derived from the addition of rat renal cortex homogenate, over 1 h of incubation at 37°C. In both cases, the conversion of Ang I into Ang II was inhibited by the addition of a cocktail of peptidase inhibitors. Ang I thus generated was then measured by RIA using rabbit anti–Ang I. Antibodies to Ang I were raised in rabbits that were immunized against the peptide coupled to BSA by carbodiimide condensation (37). PRA and AGT concentrations are expressed as nanograms of Ang I generated per milliliter of plasma and per hour of incubation.
Basal, PTHrP(1-36)–, and PTH1R Antagonist–Induced Renin Release from Isolated Perfused Kidney
Three weeks after intravenous injections, control pcDNA3-treated and pcDNA3-hPTH1R–treated rats were anesthetized by intraperitoneal injection of sodium pentobarbital (45 mg/kg). The right kidney was isolated and perfused in an open single-pass circuit exactly as we have described previously (11). The basal release of renin activity was determined after a 60-min perfusion during which the flow was adjusted to achieve a common pressure baseline of 80 mmHg; thereafter, the flow thus adjusted was maintained constant. Renal vascular resistance (RVR) was measured at the end of the first hour of perfusion. Thereafter, perfusion pressure was monitored continuously with a pressure transducer (Staham P23Db; Statham Laboratories, Hato Rey, Puerto Rico). Under these conditions, RVR did not vary more than ±8% over the next 2.5 h of perfusion. Perfusate collections for renin activity determination was started at the end of the first hour of perfusion, every 36 min (at 60, 96, 132, 168, and 204 min), over a period of 6 min. The effects of 100 nM of PTHrP(1-36) or (N10,L11, D-W12)PTHrP(7-34)NH2 (Neosystem, Strasbourg, France) were measured after at least 126 min of perfusion, at a time when basal renin release steady state was reached. PTHrP(1-36) and (N10,L11, D-W12)PTHrP(7-34)NH2 were perfused over a 2- and 6-min period, respectively. Perfusates were collected before and over five to six consecutive 6-min periods after the beginning of peptide infusion. One-milliliter samples from each collected fraction were immediately frozen at –20°C before being assayed for renin activity. The renin activity released by the isolated kidney preparations was determined as the ability of 20 µl of the various perfusate samples to generate Ang I from an excess of partially purified rat AGT, exactly as we have described before (11). The rate of renin activity release was expressed as nanograms of Ang I generated in 1 h of incubation, per minute of perfusion and normalized per gram of kidney weight, taking the weight of the contralateral kidney as the weight of the perfused kidney.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
Within the positive area of the renal cortex (Figure 2L), a number of tubules, mainly the proximal tubules, exhibited granulated fluorescence, despite the above-described very low expression of hPTH1R transcript in total renal cortex. An explanation for this could be that filtered hPTH1R-GFP are reabsorbed into the proximal tubule. However, GFP-derived fluorescence was seen along the entire intrarenal arterial tree, up to the afferent arteriole, an observation consistent with the RT-PCR studies in Figure 1. In the glomeruli, fluorescence, if any, was low and mostly derived from remaining red cells in glomerular capillaries. However, the possibility that some of the glomerular cells expressed the fluorescence transgene as well cannot be excluded.
In the liver (Figure 2, E and F), the portal area, including the portal artery branches, the portal vein branches, and the centrolobular veins, were strongly fluorescent. The biliary ducts were low or negative. Hepatocytes were positive, especially around the centrolobular veins and along the lining of the sinusoids. Gross histologic observation revealed an overall increasing fluorescence gradient from center up to the edge of the liver lobes (Figure 2E).
In brain and cerebellum, fluorescence was lower than in the other organs, in accordance with the RT-PCR studies. In addition to the vessels, which again were constantly highly positive, fluorescence displayed a scattered pattern throughout the brain sections. In the cerebellum (Figure 2O), GFP-positive areas were clearly and consistently seen in the granular layers but never in the molecular layers. Purkinje cells also seemed to express the fluorescence transgene. Sagittal sections showed fluorescence in the parietal area of the cerebral cortex and along the lining of the lateral ventricles, as well as in the choroid plexus (Figure 2N).
Effect on Urinary Parameters
Food and water intake, as well as diuresis and electrolyte excretions, are shown in Table 1. None of these parameters was affected by in vivo hPTH1R gene delivery as compared with control rats that were treated with empty pcDNA3.
|
|
We again took advantage of the in vitro kidney preparations to analyze the effect of hPTH1R gene delivery on RVR during the fourth week after plasmid injection. As shown in Figure 4A, vascular resistance in kidneys derived from hPTH1R gene–treated rats decreased by 34% as compared with control animals (P < 0.001). Whereas having no effect in kidneys that were prepared from control rats, blockade of the PTH1R using 10–7 M of a PTH1R antagonist (40) produced a reproducible slow increase in basal perfusion pressure of the kidneys that were derived from hPTH1R gene–treated rats (Figure 4, B and C). Perfusion pressure returned to baseline as soon as the antagonist infusion was stopped.
|
|
60% (P < 0.001) in hPTH1R gene–treated rats, as compared with control rats. Consistently, the AGT transcripts in liver were also clearly reduced in hPTH1R gene–treated rats as compared with the control rats (Figure 5C). Quantification by real-time PCR analysis confirmed this reduction: AGT mRNA in the liver of pcDNA3-hPTH1R–treated rats indeed represented only 21.1 ± 8.8% (n = 4) of the AGT mRNA present in the liver of control animals. However, GAPDH did not change significantly. We next asked whether the delivery of hPTH1R would affect the release of renin activity from kidney. Renal renin release, which accounts for all of the renin in the plasma (41), is influenced by a number of circulating and neural factors, intrarenal baroreceptors, and changes in the delivery of salt to the distal tubule through macula densa–mediated mechanisms. The intrinsic basal ability of the kidney to secrete renin activity therefore was measured, using the isolated kidney preparations, perfused extracorporally at constant flow and under nonfiltering and stable pressure conditions. This was done during the fourth week after plasmid injection. Immediately after the first hour of perfusion, the kidneys that were prepared from hPTH1R gene–treated animals exhibited an unexpected fourfold decrease in the rate of renin release, as compared with control rats (Figure 6A). However, after 2 h of perfusion, the rate of renin activity release reached comparable steady-state baselines in control and hPTH1R gene–treated rats. Moreover, both semiquantitative and quantitative RT-PCR analyses displayed comparable level of renin mRNA expression in the renal cortex in the fourth week after plasmid injection (Figure 6B).
|
We therefore asked the question of the effect of endogenous PTHrP on the release of renin activity from isolated kidneys that overexpress the PTH1R in their vasculature. For this purpose, we infused the 10–7 M of the PTH1R antagonist (N10, L11, D-W12)PTHrP(7-34)NH2 during 6 min (Figure 6C). In marked contrast to the stimulatory effect of exogenously added PTHrP on renin release, the renin release in response to the antagonist was also increased by 93% (P < 0.01) in control rats, indicating that endogenous PTHrP actually inhibits the release of renin. However, whereas hPTH1R cDNA delivery decreased the response to added PTHrP, the antagonist consistently tended to increase this effect by 39% (P = 0.05 to 0.1).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Because of these differences, our initial focus was on the tissue distribution of the PTH1R tagged with GFP. In accordance with previous studies (33), a single injection of naked hPTH1R cDNA resulted 21 to 25 d later in generalized expression of the foreign gene at both the mRNA and the protein levels, especially in the liver, as well as all vascular beds that have been analyzed. We assume, given the physiologic responses described herein, that hPTH1R expression was present as early as days 3 to 4. The kidney, heart, cerebellum, and brain also expressed the hPTH1R. In all of these organs, the PTH1R has been shown to be expressed physiologically (7,8). This widespread expression of the PTH1R is in some ways similar to the Qian transgenic mouse (widespread expression in the arterial media in all organs) but very different in most others (endothelial expression; the parenchymal expression in liver, brain, renal tubules, and epicardial as well as endocardial expression in the heart). This complex pattern of expression makes a clear overall understanding of the vascular effects of the PTH1R difficult. However, these studies serendipitously reveal, as discussed below, a number of interesting and potentially important regulatory roles for the PTH/PTHrP/PTH1R system in the cardiovascular system, including the RAS. In addition and as anticipated, these studies resulted in an animal model in which renal hemodynamic control and homeostasis can be studied easily in relation to the PTH/PTHrP/PTH1R system in otherwise healthy rats. Moreover, this approach clearly is one that can be used to study a number of other cardiovascular-regulating genes in the systemic and particularly the intrarenal arterial system of healthy rats. In the paragraphs that follow, we focus first on the expected observations and next on the unexpected observations resulting from PTH1R overexpression in the rat.
Among the findings that were predictable was the progressive decrease in mean BP that became clearly different from control animals 3 wk after the administration of the plasmids in conscious animals. This decrease was confirmed independently in anesthetized animals. The decrease in BP in conscious rats was greater here than in the transgenic mice of Qian et al. (19), suggesting that an abrupt overexpression of the PTH1R in vessels is able to decrease BP, presumably in the absence of the compensatory mechanisms that normally take place during development in transgenic mice. This hypotensive effect of local PTH1R delivery to the arterial wall is consistent not only with the previously acknowledged role of PTHrP as a local vasorelaxant and vasodilatory agent (42) but also with the decrease in PRA reported herein and discussed below. In addition to the direct action of the PTHrP/PTH1R system on blood vessels, other studies have demonstrated a central pressor action of direct intraventricular injection of PTHrP, an effect mediated by sympathetic activation (43). In support of this possibility, in the present studies, the expression of the in vivo–delivered hPTH1R cDNA was detected in brain, including the periventricular area. Whether overexpression of PTH1R in brain could result in PTHrP-mediated pressor effect, which in turn would buffer the direct dilatory effects of PTHrP in vessels, requires further studies.
A second finding was predictable from both the SMP-8–transgenic mice (19) and our own previous studies (33). In these latter studies, plasmids of the hPTH1R cDNA were delivered to spontaneously hypertensive rats, a genetic model of hypertension. In this model, hPTH1R gene delivery decreased the intrinsic RVR in a manner similar to that seen in the present normotensive Wistar rats. It is very likely that the reduction in renal vascular tone was due to the vasodilatory action of endogenous PTHrP produced in intrarenal arterial wall. This is supported by the vasoconstrictive response to the PTH1R antagonist not observed in control-treated rats. The pressure increase thus induced was slow, reaching 3 to 4 mmHg after 6 min of antagonist infusion and rapidly returned to baseline upon antagonist withdrawal. It is likely that a longer infusion period would produce higher vasoconstrictive responses, but this kind of experiment would require prohibitively expensive amounts of antagonist. Although the expression of hPTH1R fused to GFP was clearly detected in the juxtaglomerular afferent arteriole, local alteration of the RAS, if any (see below), would not influence the vascular tone of the isolated perfused kidney, as locally produced AGT is very low. However, because hPTH1R-GFP was detected in the intima, endothelium-derived vasoactive compounds, including nitric oxide (NO), could contribute to the decrease of renal vascular tone upon hPTH1R gene delivery. In support of this possibility, PTHrP and PTH1R have been shown to be expressed in endothelial cells (44,45). Moreover, PTHrP has been proved to increase the release of NO from endothelial cells that may be involved in the PTH1R-mediated vasorelaxant effect of PTHrP (4,19,46). Whether NO-synthase inhibitors would limit the decrease of renal vascular tone in transfected animals is an interesting question that requires further investigation.
The present studies yielded a number of surprises as well. Whereas both PTH(1-34) and PTHrP(1-36) previously and repeatedly have been demonstrated to induce tachycardia both in vivo and in isolated perfused hearts (1,47), in the present studies, in vivo transfer of PTH1R unexpectedly led to a slowing of the heart rate. The mechanisms underlying this negative chronotropic effect are not clear. Because Ang II has been demonstrated to cause chronotropic effects in the heart (48), it is possible that the decrease in PRA reported herein and discussed below is responsible for the slight decrease in heart rate 3 wk after hPTH1R cDNA delivery. It is also possible that the bradycardia resulted from direct central nervous system effects of PTH1R delivery or from interference with the peripheral functions of the autonomic nervous system. Alternatively, direct PTH1R-mediated inotropic effects on the myocardium or indirect effects via increased coronary perfusion (49) could have induced a reflex bradycardia. Further studies are obviously required to assess these hypotheses.
The plasmid delivery approach used herein allowed us to uncover close endogenous interactions that exist between the PTH/PTHrP/PTH1R system and the RAS. These studies again generated another set of unanticipated results. First, 3 wk after plasmid injection, overexpression of the PTH1R led to a decrease of the PRA. This is in marked contrast with the previously described stimulatory effect of exogenous PTH and PTHrP on the release of renin (11). Surprising, in the spontaneously hypertensive rat strain, the peripheral delivery of hPTH1R gene under identical conditions led to an increase of PRA (33). In this previous study, the increase of PRA may have been responsible for the absence of a beneficial effect of PTH1R cDNA delivery on BP. Unfortunately, however, neither renal renin release nor AGT was examined in these earlier studies in hypertensive rats. The profound alteration of the RAS in the spontaneously hypertensive rat model of genetic hypertension may explain this discrepancy.
In principle, the mechanisms by which overexpression of the PTH1R decreases PRA might involve either a decrease in renin release from the kidney or a decrease in AGT from liver, or both. In the isolated perfused kidney after 1 h of perfusion, the release of renin from control rats was high and rapidly decreased to steady-state baseline after 2 h of perfusion. This pattern of renin release from isolated perfused kidney preparation is typical (11) and reflects anesthesia and surgical stimulation of renin (50). It is interesting that this initial elevation in renin release was not seen in hPTH1R-treated rats, suggesting that overexpression of the PTH1R protected the rat from the effects of anesthesia and surgery. This point is discussed below together with the abolition of both stress-induced increase of heart rate and PRA. The steady-state levels at which the release of renin declined were comparable in control and hPTH1R-treated rats. This level of release represented an authentic active secretion and not merely residual leakage from injured juxtaglomerular cells, as it could still be stimulated by the 
-agonist isoproterenol, an archetypal stimulator of renin (results not shown). Taken together, these results strongly indicate that under baseline conditions, hPTH1R gene delivery does not affect the release of renin activity from the kidney. Importantly, this was further supported by the comparable level of renin mRNA expression in the renal cortex in control and hPTH1R gene–treated animals. Therefore, an alteration of the rate of renin release from kidney seems not to be responsible for the decrease of PRA after hPTH1R cDNA delivery.
A second unanticipated observation was that pharmacologic antagonism of the PTH1R did not inhibit but instead stimulated the release of renin in both control and PTH1R-treated rats. Among the questions arising from this unexpected finding, two are most obvious. First, does the PTH1R antagonist [N10, L11, D-W12]PTHrP(7-34)NH2 possess biologic activity on juxtaglomerular secreting cells, and, hence, do the renin-secreting juxtaglomerular cells express a different PTHR? To our knowledge, there are no published data addressing the possibility of a PTHR activated in vitro by N-terminal truncated PTH or PTHrP peptides in any cell. Second, is there a pool of intrarenal PTH1R that mediates the release of a renin inhibitor and that somehow overcompensates for the direct stimulatory effect of PTHrP on juxtaglomerular cells? As mentioned above, in addition to its interaction with juxtaglomerular PTH1R, locally produced PTHrP binds to PTH1R in endothelium, which in turn would produce the release of NO. It has been well documented previously that NO can exert both an inhibitory and a stimulatory effect on the release of renin. In fact, experiments with isolated kidneys have provided evidence that NO stimulates renin release but only under conditions of low extracellular calcium (51). Moreover, in primary cultured juxtaglomerular cells, addition of NO donors again inhibited renin release (52). In any case, we observed in recent preliminary experiments that L-nitroarginine methylester, a common NO synthase inhibitor, indeed markedly increased steady-state release of renin activity from the isolated kidney of normal rats perfused under conditions of normal calcium concentration (2 mM Ca2+), under the same stable conditions of flow and pressure used herein (results not shown). It therefore seems conceivable that the inhibitory effect of PTH1R-mediated NO release from the endothelium might overcompensate for the direct stimulatory effect of PTHrP on renin secretion from juxtaglomerular cells. Such a hypothesis might apply to the lower stimulatory potency of exogenous PTHrP on renin release in hPTH1R-transfected rats as well. High concentration (10–7 M) of exogenously added PTHrP—which most likely is never reached endogenously—presumably would increase the proportion of activated endothelial PTH1R, compared with juxtaglomerular PTH1R. Clearly, however, neither of these questions can be answered at present, and future experiments that test the effect of NO blockade on renin release, in vitro and in vivo, induced by both PTH1R agonist and antagonist in pcDNA3-hPTH1R–treated rats will confirm or refute such a possibility.
Taken together, because renin gene expression and both steady-state basal renal renin release and PTH1R antagonist-stimulated renin release were not statistically different in control and PTH1R gene–treated rats, overexpression of the PTH1R along the intrarenal arterial tree did not affect the overall renin expression and secretion from kidney. In addition, in marked contrast to what was expected from the previously described stimulatory effect of exogenous PTHrP on renin release (11), the present findings strongly suggest an overall inhibitory effect of endogenous PTHrP on renin release. Taken together, these results also strongly indicate that the decrease in PRA was not due to a decrease in renal renin release.
Although having no effect on renal renin release, PTH1R overexpression led to a marked decrease of both AGT mRNA in the liver and AGT protein in the plasma. Previous studies have demonstrated that extrahepatic sources of AGT do not contribute significantly to the circulating pool of AGT (53). In these previous studies, the deletion of hepatic AGT mRNA resulted in a 90 to 100% decrease in circulating AGT in mice. This decrease occurred despite the presence of abundant AGT in the kidney and other organs, suggesting that AGT locally synthesized in other tissues is not released into the circulation and that most, if not all, of the AGT in the systemic circulation is derived from the liver (53). Moreover, it is well accepted that AGT circulates at concentrations slightly below the Km value of renin for the conversion of AGT into Ang I (54). Therefore, in the absence of any change in renal renin release (see above) and given that the kidney is the main source of circulating renin, in the present studies, liver-derived AGT seems to be the rate-limiting determinant responsible for the decrease of PRA measured 3 wk after the plasmid injection.
What, then, are the mechanisms by which overexpression of the PTH1R decreases AGT mRNA in the liver? The understanding of such mechanism(s) is complicated by the number of physiologic systems that are susceptible to overexpress the PTH1R. Nevertheless, because the liver is thought to be the primary transfected site after tail-vein delivery of cDNA (including arterial cells, hepatocytes, and Kupffer cells) and because the PTH1R is normally highly expressed in the liver (55), it seems reasonable to hypothesize that the overexpression of the PTH1R in the liver was responsible for the downregulation of the AGT mRNA. PTHrP is not expressed in the liver during normal adult life (56), suggesting that PTH might be the PTH1R ligand responsible for the downregulation of AGT. Furthermore, hepatic PTHrP is induced in a number of physiopathologic situations, including inflammation (57), which might induce PTHrP to become a locally produced ligand of the hepatic PTH1R to modulate AGT synthesis. Clearly, whether and how PTH or PTHrP regulates the synthesis of AGT in the liver is a new question that, until now, has never been explored.
In the liver, AGT is encoded by a highly inducible gene, regulated by a variety of physiologic hormone systems (54). One of the regulators of AGT gene transcription in the liver is Ang II itself (54). Ang II not only stimulates the synthesis of AGT but also stabilizes AGT mRNA (58). However, in the present studies, the decrease in PRA and hence Ang II rather seemed to be a consequence of the downregulation of AGT gene or destabilization of the AGT mRNA in the liver. Alternatively, in rodents, adrenal gland–derived glucocorticoids and mineralocorticoids, including aldosterone, are among the most potent stimulators of AGT expression in the liver (54). Whether hepatic AGT is downregulated by the decrease of circulating gluco- and mineralocorticoids after transfection and expression of the hPTH1R in the adrenal gland is an interesting question that also requires further study. Finally, the AGT gene or AGT mRNA-regulatory proteins may be downstream targets of the PTH1R that is normally present in the hepatocyte. Overall, our data provide initial demonstration that the PTH/PTHrP/PTH1R system is able to interact with the production of AGT in the liver.
Another striking new concept gained from present studies is that the PTH/PTHrP/PTH1R system might be involved in various stress-induced responses. As a rule, the animals react to stressful stimuli with a number of physiologic responses, including increases in BP, heart rate (59), and PRA (60). In the present studies, the rats underwent stressful manipulations, too, including tail-vein blood sampling, repeated stays in metabolic cages, and BP measurements. Consistently, control rats exhibited increases in heart rate and PRA over the first 4 to 6 d after plasmid injection. Both parameters returned more or less rapidly to baseline thereafter, presumably through learning and habituation processes. It is interesting that these stress responses were blunted or even abolished in rats that overexpressed the PTH1R. In addition, the anesthesia-induced renin release mentioned above disappeared in PTH1R gene–treated rats. In recent years, Ang II has been recognized as one of the stress hormones that participate in various stress-induced responses, including sympathetic and neuroendocrine responses. For instance, central blockade of Ang II receptor results in the inhibition of the sympathetic and neuroendocrine responses to stress (59). It therefore is tempting to anticipate that the decrease in Ang II formation in response to hPTH1R gene delivery was responsible for the inhibition of the stress response on the heart rate and renal renin release. Alternatively, because PTH1R was overexpressed in several brain regions (see Figure 2, N and O), a direct central action of PTHrP cannot be ruled out. In any event, the present studies revealed the possible involvement of the PTHrP/PTH1R system in modulating cardiovascular stress–induced responses.
In conclusion, we demonstrate that a single tail-vein injection of hPTH1R or GFP-tagged hPTH1R cDNA in rats results 3 wk later in generalized expression of the foreign gene at both the mRNA and the protein levels, especially in vessels, liver, heart, kidney, and central nervous system. Overexpression of the PTH1R in these organs, where it is expressed physiologically, consistently decreased BP and renal tone. Unexpected, however, hPTH1R overexpression decreased heart rate. Moreover, overexpression of the hPTH1R decreased PRA via the decrease of liver production and circulating AGT. These studies also revealed that endogenous PTHrP actually inhibits renin release and that hPTH1R overexpression tends to increase that effect. Thus, sudden overexpression of the PTH1R in adult healthy rats profoundly alters the CVS, RVS, and RAS, which strongly supports the endogenous PTH/PTHrP/PTH1R system as a crucial regulator of heart and vascular tone. Importantly, these results also revealed that PTH1R-mediated mechanisms might have protective effects against stress-induced cardiovascular responses, including the increases in heart rate, renal renin release, and PRA.
|
Acknowledgments |
|---|
We warmly thank Dr. C. Silve (INSERM U 426, Paris, France) for kindly providing the EGFP-N2-hPTH1R plasmid construct. We also gratefully acknowledge Mr. J. Steger and Mrs. C. Coquart (Medical School Pharmacology Institute, ULP, Strasbourg, France) for help in the determinations of renin activities and AGT, and the Laboratoire d’Analyses de Biologie Médicale (CHU, Strasbourg, Dr. M. Offner) for measuring blood and urinary parameters. Mrs. D. Kuhlwein is thanked for secretarial help. We also warmly thank Prof. Andrew F. Stewart (University of Pittsburgh School of Medicine, Pittsburgh, PA) for helpful and stimulating discussions during the preparation of this manuscript.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, n = 13) or pcDNA3 that contained the full length of the hPTH1R gene (
; n = 13). The perfusion flow was adjusted during an initial 60-min equilibration period to achieve a stable pressure baseline at 
; n = 15) or pcDNA3-hPTH1R (•; n = 15). P < 0.05 in ANOVA for repeated measures test, versus naïve rats, 4 d before plasmid injection (*) or the corresponding time control that received empty plasmid (#). (B) Angiotensinogen (AGT) concentration in plasma determined on blood samples that were taken from tail vein of rats that were treated with empty pcDNA3 (n = 4) or pcDNA3-hPTH1R (n = 4) 21 d earlier. *P < 0.05 in t test, versus control kidney preparations. (C) RT-PCR analysis comparing product intensities of the AGT (263 bp) and GAPDH (415 bp) genes in renal cortex of rats, to which plasmids were delivered intravenously 3 wk before. Each lane represents a separate rat.