The salt intake of an organism controls the number of renin-producing cells in the kidney by yet undefined mechanisms. This study aimed to assess a possible mediator role of preglomerular blood pressure in the control of renin expression by oral salt intake. We used wild-type (WT) mice and mice lacking angiotensin II type 1a receptors (AT1a−/−) displaying an enhanced salt sensitivity to renin expression. In WT kidneys, we found renin-expressing cells at the ends of all afferent arterioles. A low-salt diet (0.02%) led to a moderate twofold increase in renin-expressing cells along afferent arterioles. In AT1a−/− mice, lowering of salt content led to a 12-fold increase in renin expression. Here, the renin-expressing cells were distributed along the preglomerular vascular tree in a typical distal-to-proximal distribution gradient which was most prominent at high salt intake and was obliterated at low salt intake by the appearance of renin-expressing cells in proximal parts of the preglomerular vasculature. While lowering of salt intake produced only a small drop in blood pressure in WT mice, the marked reduction of systolic blood pressure in AT1a−/− mice was accompanied by the disappearance of the distribution gradient from afferent arterioles to arcuate arteries. Unilateral renal artery stenosis in AT1a−/− mice on a normal salt intake produced a similar distribution pattern of renin-expressing cells as did low salt intake. Conversely, increasing blood pressure by administration of the NOS inhibitor N-nitro-l-arginine methyl ester or of the adrenergic agonist phenylephrine in AT1a−/− mice kept on low salt intake produced a similar distribution pattern of renin-producing cells as did normal salt intake alone. These findings suggest that changes in preglomerular blood pressure may be an important mediator of the influence of salt intake on the number and distribution of renin-producing cells in the kidney.
- AT1a knockout
the maintenance of salt homeostasis is a central function of the renin-angiotensin-system. A threat to salt balance leads to compensatory changes in renin expression in the kidney in a way that the number of renin-expressing cells increases during salt deficiency and decreases during salt overload (41). It is assumed that the salt-related changes in renin-expressing cells in the kidney are not caused by cell proliferation or apoptosis, but instead result from reversible phenotype changes of preglomerular smooth muscle cells into renin-producing cells and vice versa (34). The intracellular signaling pathways defining the specific phenotype of these cells are yet unknown. Furthermore, it is unclear by which mechanisms salt balance controls the number of renin-producing cells. It has been suggested that the effects of salt balance may affect renin expression in the preglomerular afferent arterioles by autacoids such as nitric oxide (NO) or prostaglandins (42) since salt intake has been shown to control the juxtaglomerular expression of cyclooxygenase-2 (COX-2) (17) and nitric oxide synthase-1 (NOS-1) (37). Inhibition of COX-2 and NOS-1, in fact, attenuates renin expression, but does not abrogate the regulatory effect of salt on renin expression (4, 35). Other candidates that have been considered as possible mediators of salt balance on renin expression are hormones such as atrial natriuretic peptide, which could inhibit renin expression during extracellular volume expansion, or renal nerves, which might stimulate renin expression during salt depletion. However, neither renal innervation, β-adrenoreceptors, nor atrial natriuretic peptides appear to be essential for the regulation of renin expression by salt balance (2, 16, 18, 28). Finally, also a mediator function of renal perfusion pressure is conceivable because the perfusion is known to affect renin expression (29) and the metaplastic phenotype switch in preglomerular arteries (13).
We therefore aimed in this study to assess the role of preglomerular blood pressure for renin expression in response to changes in salt intake. Since the effects of variations of oral salt intake on renin expression in wild-type (WT) mice are rather moderate, we used in addition a mouse model in which changes in oral sodium intake cause marked changes in renin expression (25). Such an enhanced salt sensitivity is observed in mice lacking the AT1a receptors for angiotensin II (8, 25, 30). We have therefore used WT and AT1a-deficient mice to determine the salt-dependent distribution of renin-producing cells in the kidney.
MATERIALS AND METHODS
AT1a−/− mice, originally generated by Ito et al. (15), were obtained from Dr. Thomas Coffman (Dept. of Medicine, Duke University, Durham, NC) and were interbred to generate subsequent generations. WT animals were generated from the F2 generation of crosses between AT1a−/− and 129X1/SvJ mice. Genotyping was performed on tail DNA as described elsewhere (25). All experiments were conducted in male 12- to 20-wk-old homozygous AT1a−/− mice and age-matched WT controls. Animals were kept on a standard rodent chow with free access to tap water. All animal experiments were conducted according to the National Institutes of Health (NIH) guidelines for the care and use of animals in research. The experiments were approved by a local ethics committee.
WT and AT1a−/− mice were assigned to three experimental groups (n = 5/group) receiving different salt diets (low salt 0.02%, normal salt 0.4%, and high salt 4%, wt/wt, NaCl, Ssniff) for 10 days. WT mice received the respective diets for 21 days.
To include the effect of chronically raised arterial blood pressure levels on renin expression, the adrenoceptor agonist phenylephrine was infused at a rate of 29 mg·kg−1·day−1 for 14 days via osmotic minipumps (model 2002, Alzet). The pumps were inserted subcutaneously using a standard surgical approach in AT1a−/− and WT mice under ketamine/xylazine anesthesia. Four days after surgery, mice received a low-salt diet for 10 days. The effect of phenylephrine infusion on blood pressure was determined by tail-cuff manometry. After 14 days, animals were subjected to the sample collection procedure.
N-nitro-l-arginine methyl ester treatment.
To inhibit NO synthesis, five AT1a−/− mice received drinking water containing the NOS blocker N-nitro-l-arginine methyl ester (l-NAME; Sigma-Aldrich) at a concentration of 0.5 mg/ml for 21 days. From day 11 of l-NAME treatment until the time of euthanasia, mice received a low-salt diet (0.02% NaCl).
Unilateral renal artery stenosis.
The left kidneys of three AT1a−/− mice kept on a normal-salt diet were exposed by a small flank incision under sevoflurane anesthetic. The renal artery was dissected from the renal vein and surrounding tissue. A u-shaped silver clip (inner diameter 0.11 mm) was placed around the renal artery. Animals were euthanized 7 days after setting of the clips.
Mice were anesthetized with ketamine/xylazine. Typically, the left renal artery was ligated and the left kidney was removed and snap frozen in liquid nitrogen for subsequent mRNA analysis. The right kidney was perfusion-fixed in situ with 3% paraformaldehyde in PBS for immunohistochemistry and three-dimensional (3D) tissue reconstruction. In mice with left renal artery clips both kidneys of the animal were perfusion fixed.
Blood pressure measurements.
Measurement of the systolic blood pressure in conscious mice was performed noninvasively by tail-cuff manometry (TSE). Before the measurement periods, animals were trained for the experimental procedure by placing them into the holding device on 7 sequential days. For analysis, the values of the last 4 days of measurement were considered.
GAPDH and renin mRNA determination.
Total RNA was isolated from frozen kidneys. GAPDH and renin mRNA determination was performed as described in Machura et al. (21).
As specified in Machura et al. (21), 5-μm-thick serial slides of perfusion-fixed kidneys were stained for α-smooth muscle actin (α-SMA) and renin.
Digitalization of 80–100 serial sections was performed using an AxioCam MRm camera mounted on an Axiovert 200M microscope (Zeiss). Data were imported into Amira 5.3 visualization software (Visage Imgaging). By separation of RGB channels, generated grayscale image subsets for renin and α-SMA were aligned to rebuild the kidney structures. In the segmentation step, the α-SMA and renin data sets were labeled using 8-bit gray values. Volume surfaces and statistics were based upon these segments.
Renin/α-SMA volume ratio.
The immunoreactive renin-to-immunoreactive α-SMA volume ratio was calculated out of 3D kidney reconstructions of five animals of each but the clipped experimental group, where it was calculated out of three animals. A volume segment of each analyzed kidney, containing an interlobular artery with 25 adjacent glomeruli (afferent arterioles) and of similar size, was used for volume measurement of renin- or α-SMA-positive areas by Amira 5.3.
Values are means ± SE. Differences between experimental groups were analyzed by ANOVA and Bonferroni's adjustment for multiple comparisons. Probability values <0.05 were considered statistically significant.
In WT mice, renin-expressing cells were confined to the juxtaglomerular portion of virtually all afferent arterioles (Fig. 1). Increasing the salt content of the chow from normal (0.4%) to high (4%) for 3 wk led to a ∼50% reduction of both renin mRNA abundance and renin-immunoreactive tissue volumes (Figs. 1 and 2, A and B). Conversely, lowering the salt content of the chow from 0.4 to 0.02% led to a 50% increase in both renin mRNA abundance and renin-immunoreactive tissue volumes (Figs. 1 and 2, A and B). As a consequence, the ratio of renin mRNA over renin-immunoreactive volumes as an indirect measure of the abundance of renin mRNA within renin-expressing cells remained independently of the salt content of the chow (Fig. 2C). The increase in the number of renin-expressing cells in animals on a low-salt diet relative to those on a high-salt diet was due to a moderate retrograde recruitment of renin-expressing cells along the terminal parts of afferent arterioles (Fig. 1).
AT1a-deficient mice kept on standard chow (0.4% wt/wt NaCl) had fivefold increased kidney renin mRNA levels relative to WT mice (Fig. 2A), which paralleled a similar increase in the number of kidney renin-immunoreactive cell volumes (Fig. 2B). Calculating the ratio of renin-immunoreactive volumes over renin mRNA abundance revealed that this ratio was not different between WT and AT1a-deficient mice (Fig. 2C). In AT1a-deficient mice on standard chow (salt 0.4%), renin-expressing cells were found along the whole length of afferent arterioles of superficial but not of juxtamedullary glomeruli (Fig. 3). In addition, renin-expressing cells were also found in interlobular arteries in the superficial cortex region. Feeding a high-salt diet (4%) to AT1a-deficient mice lowered renin mRNA abundance to levels slightly above the values found in WT mice on low-salt chow (Fig. 2A). In parallel, also the renin-immunoreactive cell volume decreased (Fig. 2B). The distribution of renin-producing cells in AT1a-deficient mice on a high-salt diet was again similar to that of WT mice on a low-salt diet (Figs. 1 and Fig. 3). Calculating the ratio of renin mRNA abundance over renin-immunoreactive volume in these kidneys yielded a value that was similar to WT mice (Fig. 2C). Feeding a low-salt diet to AT1a-deficient mice led to a strong increase in renin mRNA (Fig. 2A) that paralleled an increase in the renin-immunoreactive cell volume (Fig. 2B). Renin cells were now spread over the whole length of all afferent arterioles including juxtamedullary afferent arterioles and interlobular arteries down to their origin from arcuate arteries (Fig. 3). The ratio of renin mRNA abundance over renin-immunoreactive volume was again similar to the value found for WT mice (Fig. 2C). Under this condition, we frequently found arterioles with strong renin cell recruitment that were adjacent to arterioles/arteries that showed no signs of renin expression (Fig. 4, A and B).
The changes in renin expression induced by changes in salt intake in both WT and in AT1a-deficient mice paralleled reciprocal changes in blood pressure (Fig. 5). In WT mice, salt-induced changes in systolic blood pressure were very moderate. Systolic blood pressure values during the last 4 days of measurement were lower at an average of 7 mmHg in the low-salt group compared with the high-salt group (P > 0.05). On the contrary, salt intake-related changes in blood pressure were prominent in AT1a-deficient mice, reaching a 30-mmHg systolic blood pressure difference between the high- and low-salt diet (Fig. 5).
In view of the strong renin-stimulatory effect of a low-salt diet in AT1a-deficient mice, we aimed to prevent the fall in blood pressure induced by low salt intake in these animals.
For this purpose, AT1a−/− mice on a low-salt diet (0.02%) were treated either with the NOS inhibitor l-NAME or with the α-adrenergic agonist phenylephrine. Both drugs significantly increased blood pressure values (Fig. 6C). Renin mRNA abundance and the renin-immunoreactive cell volume in AT1a−/− mice treated with a low-salt diet plus l-NAME or with phenylephrine were significantly lower than in mice receiving no additional drug treatment (Fig. 6, A and B). The ratio of renin mRNA abundance over renin-immunoreactive volume in AT1a−/− mice treated with a low-salt diet plus l-NAME or with phenylephrine was again similar to WT control values (not shown). The distribution of renin-expressing cells in these mice was similar to that found in AT1a-deficient mice on a normal-sodium diet (Fig. 7, A–D). In a complementary approach, we reduced renal blood pressure in AT1a−/− mice on a normal-salt diet by placement of an unilateral renal artery clip. This maneuver produced a strong increase in the number of renin-expressing cells (Fig. 7, E and F) in a distribution pattern that was similar to that induced by a low-salt diet. Contralateral renin expression was similar to that for nonclipped AT1a-deficient mice and was not suppressed. Systolic blood pressure values of clipped AT1a-deficient mice (106 ± 4 mmHg, means ± SE, n = 3) measured at the day of euthanasia were also not different from nonclipped mice (Fig. 5).
This study was done to obtain information about the mechanisms by which salt intake into an organism determines the number of renin-producing cells in the kidney. In accordance with other studies (33), we found that the effect of changes in oral salt intake on renin-expressing cells is relatively small in normal, untreated animals (Fig. 1). Therefore, it is more difficult to derive clear principles of salt-regulated renin expression in WT animals. We therefore additionally used the AT1a−/− mouse model, in which renal renin expression is strongly regulated by the rate of oral salt intake (8, 25). This animal model is not confounded by a possible direct effect of AT1a receptors on renin-producing cells, since experiments with AT1a chimeric mice (26), cross-transplantation studies with AT1a-deficient mice (9), and studies with mice deficient for aldosterone synthase (23) indicate that the enhancement of renin expression in AT1a-deficient mice is not causally related to the lack of AT1a receptors on renin-producing cells themselves. Now, with this model we were able to clearly quantify the effects of graded salt intake on the number and distribution of renin-producing cells. A first result of our study was that salt intake changed the renin-immunoreactive 3D volumes, likely reflecting the number of renin-expressing cells, in parallel with renin mRNA abundance. This finding suggests that salt balance regulates steady-state renin production mainly by modulating the number of renin-producing cells rather than by graded up- and downregulation of renin gene transcription within individual cells. Although an altered expression per cell cannot be excluded, the assumption of such an “all or none” state of renin expression is in good accordance with previous studies (11, 31). Thus salt balance appears to regulate renin expression primarily by inhibiting or stimulating the phenotype switch of renin precursor cells (myofibroblasts, smooth muscle cells) into renin-producing cells (34). There are not many factors known to cause such a reversible phenotype switch. The best documented among these is the perfusion pressure of the preglomerular arteriolar tree (12, 20, 36), which falls from the renal arterial pressure down to the glomerular pressure mainly along the interlobular arteries and afferent arterioles. It has been hypothesized that the increasing blood pressure in the developing kidney is a major factor causing the characteristic developmental shift of renin expression from larger vessels to the end of afferent arterioles (6, 10). The unique position of renin-producing cells in the adult kidney may be due to the fact that the intraluminal pressure at the juxtaglomerular pole of afferent arterioles is lowest within the preglomerular vasculature. Any fall in renal perfusion pressure in the adult kidney appears to cause retrograde recruitment of renin-expressing cells by a reversible phenotype switch of precursor cells (12, 20, 36).
Our data now suggest that there is no major effect of salt intake on renin expression that is additional or superior to the effect of pressure changes associated with changes in salt intake. Two lines of evidence support the conclusion that the recruitment pattern of renin-expressing cells in response to changes in oral salt intake, particularly in AT1a-deficient mice, are related to pressure-regulated renin cell recruitment. First, salt-regulated renin cell recruitment follows a distal-proximal recruitment pattern that is inversely related to the supposed preglomerular blood pressure profile in which blood pressure falls from the arcuate arteries along the interlobular arteries and afferent arterioles down to the glomerular pressure (1, 3, 38). Recruitment occurs first in afferent arterioles of the superficial cortex and finally in juxtamedullary afferent arterioles, which directly branch off from arcuate arteries. This conclusion is supported by the results of 3D analysis, showing that the recruitment pattern of renin is in accordance with this expected but technically not measureable pressure gradient along the preglomerular tree. Second, experimental modulations of renal perfusion pressure modulate salt-regulated renin expression in a fashion that is not distinguishable from changes in salt intake itself.
Thus a fall in renal perfusion pressure by unilateral renal artery stenosis in AT1a-deficient mice strongly enhanced renin cell recruitment, whereas contralateral renin expression remained unaltered relative to nonclipped mice. Normally, contralateral renin expression is suppressed after unilateral renal artery stenosis as a consequence of the rising blood pressure. In AT1a-deficient mice, blood pressure does not increase after renal artery stenosis (7), as also confirmed in this study. To increase blood pressure in AT1a-deficient mice, we used chronic infusion of phenylephrine or feeding with l-NAME. Phenylephrine, which is thought to have no direct effect on renin-expressing cells, clearly attenuated renin cell recruitment in AT1a-deficient mice kept on a low-salt diet. Also, l-NAME treatment attenuated renin cell recruitment in AT1a-deficient mice kept on a low-salt diet. We are aware that the findings obtained with l-NAME treatment do not allow an unequivocal conclusion, since NO itself may exert an influence on renin gene expression. If so, then only renin expression in the proximal but not in the distal parts of the preglomerular vascular tree would appear to be sensitive toward NO.
In WT mice, we only found rather moderate blood pressure effects of commercial low (0.02%)- and high-salt (4%) diets, which is in accordance with previous studies (22, 24). In parallel, our data also agree with other studies reporting that kidney renin mRNA levels differ by a factor of only two between mice on low- and high-salt diets (16, 19, 39). Apparently, moderate salt-induced changes in blood pressure paralleled only moderate changes of renin cell recruitment.
In view of the salt-sensing properties of macula densa cells, it is assumed that the macula densa signals adapt the activity of the renin system to the rate of dietary salt intake. Experimental data supporting a role of the macula densa for regulation of renin expression by salt intake, however, are scarce. On the contrary, salt regulation of renin expression is maintained also in nonfiltering kidneys, in which the salt-sensing function of the macula densa should be blunted (32, 43, 44). Moreover, COX-2-derived prostanoids, which are considered as important signaling molecules of the macula densa, are not essential for salt-regulated renin expression (14, 27). Similar findings were reported for nitric oxide produced by the neuronal isoform of NOS localized in the macula densa (5, 40). Our observation made in AT1a-deficient mice that arterioles with strong renin cell recruitment were directly adjacent to arterioles/arteries that showed no signs of renin expression (Fig. 4) suggests that signals essential for induction of renin expression by salt depletion likely act in a short-distance range. Therefore, it appears not very likely that macula densa signals can account for the metaplastic transformation of renin cells in the proximal interlobular arteries.
In summary, our data are compatible with the concept that blood pressure changes in the preglomerular arteriolar tree importantly contribute to the effect of salt balance on the recruitment of renin-producing cells in the kidney.
This work was financially supported by Deutsche Forschungsgemeinschaft, Sonderforschungsbereich 699.
No conflicts of interest, financial or otherwise, are declared by the authors.
Author contributions: K.M., B.N., and A.K. provided conception and design of research; K.M., B.N., D.S., R.K., and A.G. performed experiments; K.M., B.N., and D.S. analyzed data; K.M., B.N., and A.K. interpreted results of experiments; K.M. and B.N. prepared figures; K.M., B.N., and A.K. drafted manuscript; K.M., B.N., and A.K. edited and revised manuscript; K.M. and A.K. approved final version of manuscript.
We thank Jurgen Schnermann for a critical reading of the manuscript and for helpful discussions. The technical assistance provided by Anna M'Bangui and Marcela Loza-Hilares is gratefully acknowledged.
- Copyright © 2012 the American Physiological Society