Adenosine 1 receptors (A1AR) in the kidney are expressed in the vasculature and the tubular system. Pharmacological inhibition or global genetic deletion of A1AR causes marked reductions or abolishment of tubuloglomerular feedback (TGF) responses. To assess the function of vascular A1AR in TGF, we generated transgenic mouse lines in which A1AR expression in smooth muscle was augmented by placing A1AR under the control of a 5.38-kb fragment of the rat smooth muscle α-actin promoter and first intron (12). Two founder lines with highest expression in the kidney [353 ± 42 and 575 ± 43% compared with the wild type (WT)] were used in the experiments. Enhanced expression of A1AR at the expected site in these lines was confirmed by augmented constrictor responses of isolated afferent arterioles to administration of the A1AR agonist N6-cyclohexyladenosine. Maximum TGF responses (0–30 nl/min flow step) were increased from 8.4 ± 0.9 mmHg in WT (n = 21) to 14.2 ± 0.7 mmHg in A1AR-transgene (tg) 4 (n = 22; P < 0.0001), and to 12.6 ± 1.2 mmHg in A1AR-tg7 (n = 12; P < 0.02). Stepwise changes in perfusion flow caused greater numerical TGF responses in A1AR-tg than WT in all flow ranges with differences reaching levels of significance in the intermediate flow ranges of 7.5–10 and 10–15 nl/min. Proximal-distal single-nephron glomerular filtration rate (SNGFR) differences (free-flow micropuncture) were also increased in A1AR-tg, averaging 6.25 ± 1.5 nl/min compared with 2.6 ± 0.51 nl/min in WT (P = 0.034). Basal plasma renin concentrations as well as the suppression of renin secretion after volume expansion were similar in A1AR-tg and WT mice, suggesting lack of transgene expression in juxtaglomerular cells. These data indicate that A1AR expression in vascular smooth muscle cells is a critical component for TGF signaling and that changes in renal vascular A1AR expression may determine the magnitude of TGF responses.
- single nephron glomerular filtration rate
- smooth muscle
- juxtaglomerular apparatus
tubuloglomerular feedback (TGF) describes a physiological control mechanism in which tubular salt concentration in the region of anatomic contact between the thick ascending limb of Henle and the afferent arteriole is sensed and translated into a signal that modifies the tone of the afferent arterioles. In this feedback system, an increase in salt concentration in the tubulovascular contact region will result in an activation of the smooth muscle cells of the afferent arteriole and thereby in a fall of glomerular capillary pressure, nephron filtration rate, and nephron blood flow. Experimental evidence strongly supports the notion that luminal NaCl concentration changes elicit the local generation of paracrine messengers that are responsible for the vascular effects. A specific role of adenosine as such a local mediator of TGF is suggested by studies in which methylxanthines and specific A1 adenosine receptor (A1AR) antagonists such as 1,3-dipropyl-8-cyclopentylxanthine (DPCPX) or KW-3902 were found to inhibit vascular TGF responses (6, 19, 20). Furthermore, genetic ablation of A1AR resulted in complete loss of TGF function, indicating that adenosine acting on A1AR is an indispensable prerequisite for intact signal transmission during TGF responses (3, 25).
Despite the strong experimental evidence for the involvement of A1AR in mediating TGF, the exact localization of A1AR involved in TGF is not known with certainty. In the renal vasculature, A1AR are present at the glomerular vascular pole and in outer medullary descending vasa recta (5, 8). The localization of A1AR in the most distal portion of the afferent arteriole would be well suited to initiate TGF-dependent vasoconstrictions (24, 31). However, A1AR mRNA was also detected by RT-PCR along most portions of the tubular system, including cortical and medullary thick ascending limbs, and proximal convoluted and straight tubules (30, 36). Functionally, activation of A1AR appears to stimulate electrolyte transport in the proximal tubule and to inhibit it in the thick ascending limb of Henle's loop (2, 34).
The present experiments were performed to determine if enhanced expression of vascular A1AR magnifies the constrictor response to increased macula densa (MD) NaCl concentrations. A four- to sixfold increase in A1AR expression was achieved in transgenic mice in which A1AR cDNA was placed under control of a rat smooth muscle α-actin promoter that has previously been shown to cause lacZ expression in smooth muscle cells exclusively (12). Enhanced constrictor responses of perfused afferent arterioles to administration of an A1AR agonist indicate enhanced expression of A1AR at the expected site. Micropuncture studies showed enhanced TGF responses of stop-flow pressure and increased proximal-distal single-nephron glomerular filtration rate (SNGFR) differences. These data indicate that A1AR expression in vascular smooth muscle cells is a critical component for TGF signaling, and that a change in A1AR expression is a potential determinant of the magnitude of TGF responses.
Generation of transgenic mice.
The mouse A1AR cDNA was linked to a 5.38-kb fragment of the rat smooth muscle α-actin gene containing 2.6 kb of the actin promoter and 2.78 kb of the first intron (pPromIntron; gift of G. K. Owens, University of Virginia). The combination of this promoter region and the first intron, which presumably acts as an enhancer, has been shown to provide robust and specific vascular expression of various transgenes in mice (12). For the generation of the transgenic construct, a pYX-Asc plasmid containing the murine A1AR coding sequence and polyadenylation signal was obtained from LGC Promochem (Teddington, UK). After removing an EcoRI site in the multiple cloning site of pXY-Asc, a NotI/EcoRI/NotI linker was introduced in the NotI site of the multiple cloning site of pXY-Asc. The A1AR transgene as a SalI/EcoRI fragment was cloned into pPromIntron, released from the construct backbone by NotI digestion, and purified for pronuclear injection. The correct plasmid sequence was confirmed by sequencing. One-cell embryos were harvested from superovulated female mice of the FVB/NJ strain. The purified construct DNA was then injected in the pronucleus of a one-cell embryo. Embryos were transferred to foster mothers according to standard protocols. The presence of the transgene was determined by PCR using the primers 5′-TGGCTATCCAGGCTTGTTCCA-3′ (sense) and 5′-GTTAAGCAGATAGTGAGCCT-3′ (antisense). These primers flank a 30-kb intron in the endogenous A1AR gene which is absent in the transgene and therefore permit discrimination between endogenous A1AR and A1AR transgene. Eight independent lines with stable integration of the transgene in the genome were established. Because we did not observe significant differences between transgene-negative FVB/NJ mice from line A1AR-tg4 and line A1AR-tg7, transgene-negative littermates of either line were used as wild-type (WT) controls. Besides A1AR-tg and WT mice, we used for blood pressure telemetry an additional age-matched group of five C57BL/6 mice. In key experiments like measurement of TGF and plasma renin concentration (PRC), A1AR transgenic mice of both lines were used, whereas in complementary studies we only used one A1AR-tg line. In none of the experiments that were performed in both A1AR-tg4 and A1AR-tg7 mice were noticeable differences detected between A1AR-tg lines. Animals used in the experiments were between 4 and 6 mo of age. Mice of both genders were used unless otherwise stated. Animal care and experimentation were approved by the National Institute of Diabetes and Digestive and Kidney Diseases Animal Care and Use Committee and carried out in accordance with the National Institutes of Health principles as outlined in their Guide for the Care and Use of Laboratory Animals.
A1AR mRNA expression.
Total RNA from kidneys, hearts, brains, and renal proximal tubules of transgenic and WT mice was isolated using Trizol reagent (Life Technologies, Carlsbad, CA). For measurement of renal proximal tubular A1AR expression, kidneys were removed, and the cortexes were dissected under a stereomicroscope. Renal cortexes were then minced with a scalpel blade and digested with collagenase. After collagenase treatment, proximal tubules were dissected under a stereomicroscope, the length of dissected segments was determined with a micrometer scale after careful unfolding, and 1.5 mm of tubule per sample was transferred to Trizol reagent, as described in detail elsewhere (29). Total RNA of proximal tubules was extracted using Trizol reagent with the addition of tRNA (1 μg/μl) as a tracer to facilitate RNA precipitation (29). After reverse transcription, real-time RT-PCR was performed using A1AR- and β-actin-specific TaqMan probes in an ABI Prism 7900 Sequence Detection System (Applied Biosystems, Foster City, CA).
Blood pressure telemetry.
The telemetry system of Data Sciences International (St. Paul, MN) was used for these experiments. Transmitters (model TA11PA-C10) were magnetically activated >24 h before implantation. In mice anesthetized with ketamine and xylazine (90 and 10 mg/kg, respectively), the telemeter catheter was inserted in the left carotid artery and advanced in the aortic arch, with the telemeter body positioned in a subcutaneous pocket on the right flank. After 1 wk of recovery, recordings were begun on the morning of the day 7–10, with recording periods of a minimum of 4 days for each animal and data sampling for 10 s every 2 min. Radio signals were detected, processed, and analyzed using a model RPC-1 receiver, a 20-channel data exchange matrix, APR-1 ambient pressure monitor, and a Data Quest ART Silver 2.3 acquisition system. The recording room was maintained at 21–22°C with a 12:12-h light-dark cycle.
Animal preparation for micropuncture experiments.
Mice were anesthetized with 100 mg/kg thiobutabarbital (Inactin) intraperitoneally and 100 mg/kg ketamine subcutaneously. Body temperature was maintained at 38.0°C by placing the animals on an operating table with a servocontrolled heating plate. The trachea was cannulated, and a stream of 100% oxygen was blown toward the tracheal tube throughout the experiment. The left femoral artery was catheterized with hand-drawn polyethylene tubing for continuous measurement of arterial blood pressure. A catheter was also inserted in the right jugular vein for an intravenous maintenance infusion of saline at a rate of 12 μl·g body wt−1·h−1. The bladder was catheterized for urine collections. The left kidney was approached from a flank incision, freed from adherent fat and connective tissue, placed in a Lucite cup, and covered with mineral oil. Because placement of the kidney in the holding cup is easier in the absence of female reproductive organs, all micropuncture studies were done in male mice.
Measurements of stop-flow pressure (PSF) during perfusion of loop of Henle were done as described previously (21, 35). When PSF had stabilized, perfusion rate of the loop of Henle was increased to 30 nl/min, and maximum PSF responses were determined. Perfusion rates then were decreased stepwise to 20, 15, 10, 7.5, 5, and 0 nl/min and maintained until steady states were achieved at each flow rate. The perfusion fluid contained (in mM/l) 136 NaCl, 4 NaHCO3, 4 KCl, 2 CaCl2, 7.5 urea, and 100 mg/100 ml FD&C green (Keystone, Bellefonte, PA).
To determine nephron filtration and absorption rates, mice were infused with 125I-labeled iothalamate (Glofil; Questcor Pharmaceuticals, Hayward, CA) at ∼40 μCi/h. Free-flow micropuncture was subsequently performed according to techniques previously described (4). Briefly, end-proximal and distal segments were identified by injecting a bolus of artificial tubular fluid stained with FD&C green from a 3- to 4-mm tip pipette connected to a pressure manometer. This pipette remained in place during the collections to permit control of intratubular pressure. All proximal collections were done in the last surface segment while distal collection sites were chosen as available. Distal and proximal fluid collections (collection times 3.5–5 min) were done in a paired fashion when possible using oil-filled pipettes. Fluid volume was determined from column length in a constant bore capillary. Samples were transferred to a counting vial, and radioactivity was determined in a gamma counter. Blood samples were collected in heparinized 5-μl microcaps at the beginning and end of micropuncture. 125I-iothalamate radioactivities were measured in duplicate using 500-nl samples of plasma and urine. The period of micropuncture usually did not exceed 60 min.
Plasma chemistry was determined by standard clinical chemistry methods (Department of Laboratory Medicine, Clinical Center, NIH, Bethesda, MD) in plasma collected from the vena cava of freshly anesthetized mice. The plasma volume obtained from a single mouse was between 200 and 300 μl.
Urine osmolarities under ambient conditions and after 48 h of water restriction were determined in spot urine samples that were obtained by transurethral catheterization of female mice. Urine osmolarity was determined by the freezing-point depression method.
Isolated perfused afferent arterioles.
To functionally assess the vascular overexpression of A1AR, the influence of adenosine and of the A1AR agonist N6-cyclohexyladenosine (CHA) on vessel diameter was determined in the isolated perfused afferent arteriole preparation (10). In brief, kidneys were removed and sliced along the corticomedullary axis. Afferent arterioles were dissected at 4°C in albumin-enriched DMEM (0.1%) using sharpened forceps (no. 5; Dumont, Montignez, Switzerland). Tubules were removed except in the region of the thick ascending limb of Henle's loop in the contact region of the glomerulus. Afferent arterioles were identified by preparation of the arterial tree, including the interlobular artery. The afferent arteriole with attached glomerulus was then transferred to a thermoregulated chamber. Arterioles were perfused with use of a perfusion system (Vestavia Scientific, Vestavia Hills, AL) that allowed movement and adjustment of concentric, holding, and perfusion pipettes. The holding pipette (outer diameter 2.13 mm, inner diameter 1.63 mm) had an aperture of roughly 26 μm at the tip and a constriction of ∼20 μm after customizing. The proximal end of the arteriole was aspirated in this pipette. The inner perfusion pipette (OD 1.19 mm, ID 1.02 mm), with a tip diameter of 5 μm, was advanced in the lumen of the arteriole. This pipette was connected to a reservoir containing the perfusion solution. The perfusion pressure in the afferent arteriole was 60 mmHg (15).
Blood was collected from conscious mice by puncture of the submandibular vessels with a 19-gauge needle and collection of ∼20 μl of the emerging blood in an EDTA-containing microhematocrit tube. Red blood cells and plasma were separated by centrifugation; the plasma was ejected in an Eppendorf tube and frozen until used for renin determinations. With the use of a 20-fold dilution of 12 μl of plasma, renin was measured by radioimmunoassay (DiaSorin, Stillwater, MN) as generation of ANG I following addition of excess rat substrate (PRC), with final plasma dilutions varying between 1:500 and 1:1,000. ANG I generation was determined for a 3-h incubation period at 37°C and expressed as hourly average. In each assay, substrate without plasma was incubated for the same time, and any background ANG I formation was subtracted from the plasma-containing samples. In addition, background ANG I levels were determined in a plasma aliquot kept frozen without the addition of substrate until assaying. The MD-dependent inhibition of renin release was assessed by acute salt-loading experiments. In these experiments, aimed to increase salt concentration at the MD, mice received a single intravenous (tail vein) injection of saline (5% of body wt), and blood samples were collected 60 min later.
Data are expressed as means and SE. Statistical comparisons were made by Student's t-test or by ANOVA with a Bonferroni post hoc test when necessary.
A1AR expression in transgenic mice.
To determine the lines with highest transgene expression level, A1AR mRNA expression was measured by real-time RT-PCR in various organs of transgene-positive mice. The two lines that showed the highest A1AR expression levels in the kidney were chosen for the experiments (A1AR-tg4 and A1AR-tg7). Renal A1AR mRNA was 344 ± 42% (n = 8) and 590 ± 43% (n = 8) of transgene-negative mice in lines A1AR-tg4 and A1AR-tg7, respectively. In both lines there was also an increased expression of A1AR mRNA in the heart, but not in the brain (data not shown). Our attempts to demonstrate A1AR protein upregulation by Western blotting were technically unsuccessful since all tested commercial antibodies yielded false-positive bands in A1AR−/− mice.
All experiments were performed in mice derived from A1AR-tg4 and A1AR-tg7 founders using only hemizygotes for transgene transmission. Vascular A1AR-tg transgenic mice were viable and showed no gross anatomic, behavioral, or fertility abnormalities. A panel of blood constituents of A1AR-tg4 and A1AR-tg7 showed no apparent abnormalities when compared with transgene-negative controls (Table 1).
Isolated perfused afferent arterioles.
Addition of CHA, an A1AR-specific agonist, to the perfusion bath reduced the diameters of afferent arterioles from both A1AR-tg and WT mice. However, as shown in Fig. 1, the dose-response relationship was left-shifted in A1AR-tg4 compared with WT. At the highest CHA concentrations (10−5 and 10−4 M), vessel diameters of A1AR-tg4 were reduced by 20.2 ± 4.3 and 19.3 ± 5.4% compared with control conditions, respectively. In contrast, in afferent arterioles from WT mice, a biphasic effect was observed with maximum constrictions at 10−6 M of CHA (−10.3 ± 2.6%) and a decline in constriction at higher concentrations of CHA [10−5 and 10−4 M (−5.5 ± 2 and −4.3 ± 2.2%) compared with control]. Together, these data indicate functional vascular overexpression of A1AR in the afferent arteriole of A1AR-tg mice.
A1AR expression in the proximal tubule.
To further assess A1AR expression in the tubular system of the kidney, A1AR mRNA levels were determined in microdissected proximal tubules. A1AR has been shown to be expressed in the proximal tubule, and it accounts for most of the adenosine-dependent effects on tubular salt reabsorption (9, 33). A1AR-to-β-actin mRNA ratios were not different between proximal tubules of A1AR-tg and WT mice, averaging 0.96 ± 0.1 and 0.85 ± 0.1 relative units in A1AR-tg (n = 6) and WT (n = 6), respectively (P = 0.31).
Measurements of PSF responses to a saturating increase of loop of Henle flow rate (0–30 nl/min) in five male WT mice (n = 21 nephrons), six male A1AR-tg4 mice (n = 22 nephrons), and four male A1AR-tg7 mice (n = 12) are shown in Fig. 2. There was a highly significant increase of PSF response magnitudes, from a mean of 8.4 ± 0.9 mmHg in WT to 14.2 ± 0.71 mmHg in A1AR-tg4 and to 12.6 ± 1.2 mmHg in A1AR-tg7 mice. PSF at zero loop flow was not significantly different between mice although it tended to be higher in both A1AR-tg lines (47.7 ± 1.5 mmHg in WT, 50.7 ± 1.4 mmHg in A1AR-tg4, and 51.6 ± 3.1 mmHg in A1AR-tg7). Similarly, mean arterial blood pressure (MAP) during micropuncture was the same in all three groups of mice (100.4 ± 1.7 mmHg in WT, 101.5 ± 1.7 mmHg in A1AR-tg4, and 101.5 ± 2.3 mmHg in A1AR-tg7 mice). Analysis of PSF responses in the different flow ranges (Fig. 3) showed significantly increased responsiveness in the transgenic mice in intermediate flow intervals (7.5–10 and 10–15 nl/min) although responses were numerically greater in transgenic than WT mice in all flow ranges.
Analysis of TGF responses revealed a feature not previously noted or commented upon (Table 2). In 95% of tubules in both WT and transgenic animals, the rapid PSF decline in response to flow stimulation (on response) was observed to stop for 1–2 s (see arrow in Fig. 5) before resuming the fall at a distinctly slower rate (Fig. 4). Fast and slow components contributed approximately one-half to the total TGF response. The return of PSF to baseline upon discontinuing loop perfusion (off response) was slower than the on response, and it also occurred at two different rates. In A1AR-overexpressing mice, total response times were prolonged (both on and off), and this was mainly due to a significant prolongation of the slow response time from 20.4 ± 2.4 to 43.6 ± 3.4 s (P = 0.00001). Furthermore, an increase of the slope of the fast component from 0.92 ± 0.12 to 1.46 ± 0.15 mmHg/s (P = 0.013) contributed to the enhanced responses.
Measurements of SNGFR by free-flow micropuncture in proximal and distal nephron segments were performed to further assess the influence of A1AR overexpression on TGF. In this setting, TGF-dependent regulation of SNGFR is intact during collections from the distal nephron, whereas it is interrupted during collections from the proximal nephron. Although neither proximal nor distal SNGFR values were significantly different between WT (n = 4) and A1AR-tg4 (n = 4; Fig. 5A) mice, the proximal-distal SNGFR difference was markedly augmented in A1AR-tg4 mice (6.25 ± 1.5 nl/min; n = 11) compared with WT (2.6 ± 0.51 nl/min; n = 11; P = 0.034; Fig. 5B), indicating an enhanced TGF-dependent regulation of glomerular filtration rate (GFR) in A1AR-tg.4 compared with WT. Arterial blood pressure in this experimental series was slightly higher in A1AR-tg than in WT mice (109.5 ± 3.1 mmHg vs. 102.4 ± 3.1; P = 0.052) while whole kidney GFR was similar, averaging 1,614.6 ± 212 and 1,776.0 ± 136 μl·min−1·100 g body wt−1 in WT and A1AR-tg mice, respectively. Kidney weight was also similar in WT and A1AR-tg; thus, the GFR-to-kidney weight ratio was indistinguishable between WT and transgenic mice (1,068 ± 99 and 1,150 ± 22 μl·min−1·g kidney−1 in WT and A1AR-tg mice, respectively). WT and A1AR-tg mice did not significantly differ in fluid absorption rates along the proximal tubule (43.4 ± 2 and 40.4 ± 3.8%) and up to the distal puncture site (69.9 ± 2.8 and 68.1 ± 1.8%). The relationship between absolute proximal fluid reabsorption and SNGFR was not measurably different between WT and A1AR-tg mice (Fig. 6).
Arterial blood pressure in conscious mice.
MAP and heart rate measured over 24 h in conscious mice by radiotelemetry showed no significant differences between WT and A1AR-tg mice. As seen in a low-resolution time course, MAP and heart rate tended to be higher in A1AR-tg than in WT in most 2-h periods with significance being reached in time periods before light on and lights off (Fig. 7). No differences in blood pressure were detected between animals from either the A1AR-tg4 or the A1AR-tg7 line. Locomotor activity tended to be higher in WT than transgenic mice and can therefore not explain the differences in blood pressure and heart rate. A noteworthy difference between mice of the FVB strain used in the present study (both WT and A1AR-tg) and other strains (like C57BL/6) is the 6-h shift in blood pressure periodicity that leads to the increase in MAP and heart rate in the afternoon that normally is cued to turning the lights off (Fig. 8). In addition, the MAP and heart rate rise shortly before lights on are much more pronounced in FVB than C57BL/6 mice (n = 5, Fig. 8).
Urine osmolarity and concentrating ability.
Ambient urine osmolarity was not different between A1AR-tg and WT mice. As summarized in Fig. 9, it averaged 1,294 ± 163 mosmol/kgH2O in WT (n = 13) and tended to be higher in both A1AR-tg4 (1,678 ± 214 mosmol/kgH2O; n = 10) and A1AR-tg7 (1,668 ± 276 mosmol/kgH2O; n = 9) without reaching significance (P = 0.16 and 0.23 vs. WT). Water restriction (48 h) markedly increased urine osmolarity in all genotypes with concentrating ability being similar in A1AR-tg4 (4,499 ± 122 mosmol/kgH2O, P = 0.53 vs. WT), A1AR-tg7 (4,469 ± 172 mosmol/kgH2O, P = 0.66 vs. WT), and WT (4,356 ± 172 mosmol/kgH2O).
Basal PRC (ng ANG I·ml−1·h−1) and, by inference, renin secretion were slightly increased in the A1AR-tg lines compared with WT mice, averaging 1,918 ± 193 in A1AR-tg4 (n = 10), 1,983 ± 666 in A1AR-tg7 (n = 9), and 1,251 ± 142 in WT (n = 13; P = 0.01 and 0.24 vs. WT; Fig. 10) mice. Previous data had shown that absence of A1AR abrogates the inhibitory effect on renin secretion elicited by acute intravenous injection of 5% of body weight saline (7). After saline infusion (60 min), PRC declined to 1,018 ± 295, 912 ± 295, and 714 ± 120 ng ANG I·ml−1·h−1 in A1AR-tg4, A1AR-tg7, and WT mice, respectively (Fig. 10). Thus overexpression of A1AR did not augment the inhibitory response to acute volume expansion compared with WT [decreases of 55% (P = 0.28 vs. WT), 47% (P = 0.46 vs. WT), and 35% in A1AR-tg4, A1AR-tg7, and WT, respectively].
A critical role of A1AR in TGF signaling is supported by the consistent TGF inhibitory effects of both receptor antagonists and genetic A1AR deletion (3, 20, 25). However, both approaches do not unambiguously localize the site of A1AR involved in TGF signaling. The focus of the present study was to investigate the role of vascular A1AR in the mediation of TGF. The major finding is that, in mice with vascular overexpression of A1AR, TGF responsiveness is significantly enhanced, establishing that vascular A1AR are responsible for the vasoconstrictor responses produced by increases in tubular salt concentration.
To address the role of vascular A1AR in the TGF pathway, experiments were performed in transgenic mouse lines in which the placement of A1AR cDNA downstream of the smooth muscle α-actin promoter/enhancer caused enhanced A1AR expression in the vasculature. The promoter employed in this study has been successfully used before to obtain high and specific expression levels of transgenes in smooth muscle cells (12). LacZ expression under the control of the α-actin promoter/enhancer in adult mice was found to be restricted to smooth muscle cells in various organs like esophagus, stomach, intestine, lung, and nearly all blood vessels (12). Of note, high expression levels of lacZ were observed in the renal vasculature, including arteries and arterioles (12). In contrast to other smooth muscle-specific promoters like smooth muscle MHC and smooth muscle-22, expression was found to be homogeneous between individual cells, with basically all smooth muscle cells being transgene-positive. The α-actin promoter/enhancer used to express lacZ conferred a high level of reproducibility of transgene expression across multiple independent founder lines, indicating that regulatory sequences within the promoter operate rather independently of the genomic insertion locus (12).
In the present study, a three- to fivefold overexpression of the mRNA for A1AR was found by RT-PCR in kidney and heart, but not in brain. Absence of A1AR upregulation in brain probably reflects the fact that A1AR in brain are predominantly located on neuronal and glial cells. Overexpression of A1AR in renal afferent arterioles was established by the demonstration that A1AR activation with CHA caused greater reductions in the diameter of isolated perfused afferent arterioles from A1AR-tg than from WT mice. These results are qualitatively in agreement with earlier observations showing that A1AR-specific agonists such as CHA cause a dose-dependent vasoconstriction of perfused afferent arterioles from rabbits and mice (11, 32). The present study revealed a marked dissociation of the dose-response relationship of CHA-induced vasoconstriction between A1AR-tg and WT. Whereas the vasoconstrictor effect of A1AR agonists was partially reversed in WT at concentrations higher than 10−6 M, there was no such reversal in A1AR-tg up to a concentration of 10−4 M. We assume that the reversal at higher CHA concentrations reflects loss of selectivity and interaction of CHA with vasodilator A2AR that are also present in afferent arterioles (5, 13, 27). An increased number of A1AR would be expected to relatively suppress the dilator effect of CHA spillover on A2AR and therefore shift the dose-response relationship toward extended vasoconstriction.
The present results establish that modulating the expression level of A1AR in vascular smooth muscle cells is sufficient to alter TGF responsiveness. In contrast, proximal fluid absorption was found to be similar between WT and transgenic mice. These observations are important since expression of A1AR has been found virtually along the entire renal tubule, including proximal tubule, thick ascending limb, and cortical collecting duct (8, 30, 36). It was therefore possible that A1AR activation may indirectly affect TGF responses by modulating tubular NaCl transport and salt delivery to the MD (28). Previous support for a role of vascular A1AR in TGF mediation has come from the observation that nonspecific adenosine receptor and A1AR-specific antagonists reduced or even abolished TGF responses when administered intravenously (6, 14, 34). Furthermore, tubular application of the A1AR-specific antagonist DPCPX inhibited TGF responses of neighboring nephrons, suggesting that DPCPX can leave the tubular lumen and react with A1AR outside of the tubular lumen (20).
The enhanced TGF response in A1AR-overexpressing mice was not associated with major changes of GFR under resting conditions. However, when salt delivery to the MD was acutely interrupted by withdrawal of fluid proximal to the MD, SNGFR increased significantly more than in WT mice, thus revealing a tonic GFR-suppressing effect that depends on the A1AR expression level. These results suggest that the A1AR density on afferent arterioles may be a determinant of TGF responsiveness. It is unclear whether the enhanced TGF response in A1AR-tg mice is only due to an augmented constrictor response of afferent arterioles to adenosine. It has been proposed that efferent arteriolar dilatation contributes to the fall in glomerular pressure with increased NaCl delivery to the MD (17), an effect possibly mediated by A2bAR receptors (1). Our data show that the enhancement of TGF in A1AR-tg compared with WT was most pronounced in the intermediate flow ranges, and not significant in the high flow range, suggesting that at high interstitial adenosine concentrations in the juxtaglomerular (JG) apparatus the balance of A1AR and A2b in the efferent arteriole might be shifted toward A1AR-dependent vasoconstriction. This would more effectively counteract A2bAR-mediated vasodilatation of the efferent arteriole in A1AR-tg than WT mice.
A side observation in the present experiments is the discovery of a biphasic nature of the maximum TGF response. The critical observation is that, in 95% of nephrons, the initial rapid fall in PSF comes to a brief, but clearly discernible, halt that is followed by a continued fall in pressure at about one-third of the initial speed. Increased expression of A1AR did not change the biphasic nature of the response, but it markedly prolonged the slow component and the total duration of the response. Detection of the biphasic response was probably facilitated by a higher time resolution of the recording. Nevertheless, inspection of numerous earlier recordings from both rats and mice showed a biphasic PSF response in most nephrons as well. Understanding the causes for the biphasic response will require further work, but it suggests that TGF is the summation of the effect of two mechanisms with different temporal characteristics. Because A1AR−/− mice do not show TGF responses, one would conclude either that both mechanisms are A1AR-dependent or that an A1AR-dependent fast component triggers a second slower response not occurring without A1AR initiation. It is conceivable that the fast component reflects direct constriction of the distal afferent arteriole, which then serves as the trigger for secondary adjustments that enhance constriction with a relatively slow time constant.
Like TGF, MD control of renin secretion has been found to require intact A1AR signaling. In A1AR-deficient mice, the suppression of renin secretion in response to increases in tubular salt concentration was shown to be absent (7). Similarly, in the isolated perfused kidney, pharmacological blockade of A1AR by DPCPX or genetic A1AR deficiency prevented the suppression of renin secretion after increases in renal perfusion pressure (22). The present studies show that vascular overexpression of A1AR did not alter MD control of renin secretion, since the suppression produced by acute salt loading was essentially unchanged. These data suggest that the α-actin promoter probably did not drive A1AR transgene expression in renin-producing JG cells, in accordance with previous reports indicating that JG cells in the adult mouse kidney do not show immunostaining for smooth muscle α-actin (18, 23). Whereas suppression of renin secretion by acute volume expansion was not affected by A1AR overexpression, PRC under baseline conditions tended to be higher in A1AR-tg mice compared with WT. This observation is unexpected, since A1AR activation elicits inhibition of renin release, and it indicates the operation of an indirect effect. One possibility is that increased vascular resistance upstream of the JG cells may mimic reduced renal perfusion pressure and result in a local baroreceptor-dependent stimulation of renin secretion.
Telemetric measurements of cardiovascular function and spontaneous locomotor activity showed discrete differences between WT and A1AR-tg mice consisting of a tendency for higher MAP in the second half of the light period and at the end of the dark period. Heart rates were found to be significantly elevated in the same time periods, suggesting a major resetting of the arterial baroreceptor. Locomotor activity cannot account for these differences, since it was consistently lower in transgenic than WT mice. In contrast to these small differences between WT and A1AR-tg strains, we observed a major shift in circadian phases in both groups of animals that is presumably related to the FVB/NJ strain used in these studies. Previous studies by us and other investigators have shown that rhythmic changes of blood pressure and heart rate rhythms are strictly governed by the daily light cycle. In contrast, both groups of FVB mice showed an increase in blood pressure and heart rate in the middle of the light period and a decline in the middle of the dark period. Thus FVB mice have an ∼6-h phase shift in blood pressure and heart rate rhythms that appears independent of light cues. Furthermore, there was a marked anticipatory increase in blood pressure and heart rate shortly before light onset, especially in the A1AR-tg mice. An abnormal rhythmic behavior of wheel running activity has been reported previously in FVB mice (16). Running behavior was more irregular and fragmented with considerably more activity in the daytime period. It is possible that the rhythm abnormalities are related to the visual impairment in FVB mice that is caused by the expression of a mutant phosphodiesterase 6b (Pde6brd1) and the resulting early onset retinal degeneration (26).
In summary, we generated transgenic mouse lines with vascular overexpression of A1AR using a smooth muscle α-actin promoter/enhancer. Vascular overexpression of A1AR resulted in a markedly enhanced TGF responsiveness of PSF and SNGFR, indicating that A1AR expression in vascular smooth muscle cells is critical for TGF signaling.
This work was supported by the intramural research program of the NIDDK and by a grant from the Deutsche Forschungsgemeinschaft (SFB 699/A4). C. S. Wilcox was supported by NIH Grants DK-36079, DK-49870, and HL-68686-01.
We are grateful to Dr. G. K. Owens (University of Virginia, Charlottesville, VA) for kindly donating the SMA promoter construct and to Huiyan Lu [National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institutes of Health (NIH), transgenic core facility] for performing pronuclear microinjections. We gratefully acknowledge Veronika Kattler (Institute of Physiology, University of Regensburg) for expert technical assistance.
Present address of M. Oppermann: Children's Hospital, University Medical Center, Regensburg, Germany.
- Copyright © 2009 American Physiological Society