SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References
  8. Appendix
  9. Supporting Information

Non-Technical Summary  Blood flow in the kidney is tightly regulated. This so-called autoregulation is essential for the function of the kidney as well as for its protection against damage and failure from high blood pressure. Autoregulation is caused by three mechanisms. The signalling molecule nitric oxide (NO) modulates the balance of these mechanisms, blunting the contribution of the fastest mechanism and increasing that of the others. What is unknown is where in the kidney the responsible NO is originating. Our data indicate that the cells of the inner lining of the blood vessels are by far the most important source of NO for this effect compared to other NO-producing cells in the relevant region of the kidney, such as macula densa, smooth muscle or mesangial cells. The findings are important for understanding blood flow autoregulation in the kidney as well as kidney function and failure.

Abstract  Nitric oxide (NO) blunts the myogenic response (MR) in renal blood flow (RBF) autoregulation. We sought to clarify the roles of NO synthase (NOS) isoforms, i.e. neuronal NOS (nNOS) from macula densa, endothelial NOS (eNOS) from the endothelium, and inducible NOS (iNOS) from smooth muscle or mesangium. RBF autoregulation was studied in rats and knockout (ko) mice in response to a rapid rise in renal artery pressure (RAP). The autoregulatory rise in renal vascular resistance within the first 6 s was interpreted as MR, from ∼6 to ∼30 s as tubuloglomerular feedback (TGF), and ∼30 to ∼100 s as the third regulatory mechanism. In rats, the nNOS inhibitor SMTC did not significantly affect MR (67 ± 4 vs. 57 ± 4 units). Inhibition of all NOS isoforms by l-NAME in the same animals markedly augmented MR to 78 ± 4 units. The same was found when SMTC was combined with angiotensin II to reproduce the hypertension and vasoconstriction seen with l-NAME (58 ± 3 vs. 54 ± 7 units, l-NAME 81 ± 2 units), or when SMTC was replaced by the nNOS inhibitor NPA (57 ± 5 vs. 56 ± 7 units, l-NAME 79 ± 4 units) or by the iNOS inhibitor 1400W (50 ± 1 vs. 55 ± 4 units, l-NAME 81 ± 3 units). nNOS-ko mice showed the same autoregulation as wild-types (MR 36 ± 4 vs. 38 ± 3 units) and the same response to l-NAME (111 ± 9 vs. 114 ± 10 units). eNOS-ko had similar autoregulation as wild-types (44 ± 8 vs. 33 ± 4 units), but failed to respond to l-NAME (37 ± 7 vs. 78 ± 16 units). We conclude that the attenuating effect of NO on MR depends on eNOS, but not on nNOS or iNOS. In eNOS-ko mice MR is depressed by NO-independent means.

Abbreviations 
Ang II

angiotensin II

AP

arterial pressure

eNOS

endothelial NOS

iNOS

inducible NOS

i.r.a.

intrarenalarterial

JGA

juxtaglomerular apparatus

MR

myogenic response

nNOS

neuronal NOS

NO

nitric oxide

NOS

nitric oxide synthase

NPA

N-propyl-l-arginine

RAP

renal arterial pressure

RBF

renal blood flow

RPP

renal perfusion pressure

RVR

renal vascular resistance

SMTC

S-methyl-l-thiocitrulline

TGF

tubuloglomerular feedback

1400W

N-aminomethyl-benzyl-acetamidine

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References
  8. Appendix
  9. Supporting Information

Autoregulation in the kidney not only serves to keep renal blood flow (RBF) constant in the presence of changes in arterial pressure, but also controls intravascular pressures in glomerular, peritubular and vasa recta capillaries. Autoregulation therefore impacts on glomerular filtration, proximal reabsorption and medullary perfusion. In addition, it provides a protective shield against hypertensive renal damage (Bidani & Griffin, 2004). The underlying mechanisms are regarded today to depend on the myogenic response (MR) and tubuloglomerular feedback (TGF) (Navar et al. 1996; Just, 2007). In addition, a third regulatory mechanism of unknown origin also contributes (Just et al. 2001; Just & Arendshorst, 2003; Wronski et al. 2003; Just & Arendshorst, 2007; Seeliger et al. 2009; Siu et al. 2009).

The three regulatory mechanisms are in balance with each other. During physiological conditions, MR provides ∼50% of the total regulatory effort, while TGF and the third mechanism contribute around 20–50% each (Schnermann et al. 1984; Wronski et al. 2003; Just, 2007). This balance is crucial because MR, TGF and the third regulatory mechanism have different response times (Holstein-Rathlou & Marsh, 1994; Just, 2007). Accordingly, their relative participation determines the speed of the overall response and hence the spectrum of pressure fluctuations being shielded against. The balance is not fixed, but can be modulated, most importantly by nitric oxide (Wang & Cupples, 2001; Wronski et al. 2003; Just & Arendshorst, 2005; Shi et al. 2006). Inhibition of NO production markedly augments the contribution of MR in RBF autoregulation at the expense of TGF and the third regulatory mechanism in rats (Wang & Cupples, 2001; Wronski et al. 2003; Just & Arendshorst, 2005; Shi et al. 2006) and mice (Just et al. 2009). As MR is the fastest of the regulatory mechanisms, the augmentation of its contribution accelerates the overall response.

What is unclear, however, is the source of NO governing the modulation of MR. NO is produced in the kidney in various vascular, tubular and interstitial cells (Star, 1997; Kone, 1999). Due to the short half-life time of NO, the most likely sources are those in the immediate vicinity of its action. As MR is most prominent in the afferent arteriole (Steinhausen et al. 1989; Imig & Roman, 1992), cells of the juxtaglomerular apparatus (JGA) are the most likely candidates, i.e. macula densa, mesangium, smooth muscle and the endothelium of the afferent arteriole. As we (Just & Arendshorst, 2005) and others (Shi et al. 2006) observed that the NO-dependent modulation of MR is absent or impaired when macula densa function is compromised, we hypothesize that the macula densa is the primary source.

Different cells within the JGA produce NO using certain NOS isoforms (Star, 1997; Kone, 1999). nNOS is expressed almost exclusively in macula densa cells (Bachmann et al. 1995). eNOS is found in the endothelium along the vascular tree including the afferent and efferent arterioles and glomerular capillaries (Ujiie et al. 1994). In contrast to most other organs, iNOS is expressed in the normal kidney without immunological stimulation (Star, 1997; Kone, 1999). Although not unequivocal, iNOS mRNA has been reported in smooth muscle cells of the afferent arteriole (Tojo et al. 1994) and in mesangial cells (Morrissey et al. 1994).

The goal of the present study was to elucidate the roles of nNOS, eNOS and iNOS in NO-dependent modulation of MR in RBF autoregulation in vivo. Although the question has been studied before by Shi et al. (2006), who found partial contribution from nNOS, the quantification of MR in that study was done by transfer function analysis. In the present study we used a more direct method for MR quantification. In addition, we employed a two-pronged approach using both pharmacological isoform-selective inhibitors in rats as well as gene targeting in nNOS- and eNOS-deficient mice.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References
  8. Appendix
  9. Supporting Information

Experiments were conducted on 28 male Wistar rats (age 9–15 weeks, body weight 250–500 g, left kidney weight 1.35–2.35 g) and on 29 male and female mice (age 10–47 weeks, body weight 21–39 g, left kidney weight 0.151–0.309 g). Rats were obtained from the breeding colony of Universität Freiburg. Mice were obtained from Universität Magdeburg and transported to Freiburg at least 4 weeks before the experiment. All mice were derived from homozygous breeding of the respective NOS strains (nNOS-ko and eNOS-ko) as developed in the C57BL/6 strain and their wild-type control strain. All rats and mice were fed a standard lab chow with free access to tap water, and were kept on a 12 h–12 h light–dark cycle. Surgery and experimental procedures were similar to earlier reports (Just & Arendshorst, 2005, 2007; Just et al. 2009).

Ethical approval

All experiments and procedures were approved by the governmental oversight authority, Regierungspräsidium Freiburg.

Surgical procedures in rats

After induction of anaesthesia by pentobarbital (50–60 mg (kg body weight)−1) i.p., rats were placed on a temperature-controlled table kept at 37°C. Anaesthetic depth was monitored by cardiovascular and motor responses to ear and toe pinching. Additional doses of pentobarbital (3–8 mg kg−1i.v.) were given as needed. After the experiment the rat was killed by an overdose of pentobarbital (>150 mg kg−1).

The left femoral artery was catheterized (PE-50) to measure blood pressure, and three catheters (PE-10) were placed into the left femoral vein for volume measurements and i.v. administration of pentobarbital and other agents. An isoncotic bovine serum albumin solution (4.75%) was infused initially at 100 μl min−1 until a total dose of 1.25 ml (100 g body weight)−1 had been administered, followed by a maintenance rate of ∼30 μl min−1. The trachea was cannulated (PE-240). Via a midline abdominal incision, the abdominal aorta and left renal artery were exposed. An inflatable vascular occluder (2.0 mm; IVM, Healdsburg, CA, USA) was implanted around the aorta above the left renal artery. A flowprobe (1 RB; Transonic, Ithaca, NY, USA) was placed around the left renal artery and filled with surgical lubricant (Surgilube, Savage Laboratories, Nycomed US Inc., Melville, NY, USA) for ultrasonic coupling. If a renal artery catheter was necessary, the right femoral artery was catheterized (PE-10), the catheter advanced through the aorta and its bent tip directed into the renal artery. A 23-gauge needle with a catheter (PE-50) allowed bladder urine to drain by gravity. Sixty minutes were given for stabilization after surgery.

Surgical procedures in mice

Surgery in mice was similar to rats with the following differences: The initial dose of pentobarbital was 80–90 mg kg−1, maintenance doses were given s.c., and the lethal dose was >200 mg kg−1i.v. Vein and artery catheters were heat-tapered at the tips. Vein catheters were inserted into the jugular vein. The albumin solution was less concentrated (2.4%) and infused at ∼15 μl min−1 throughout the experiment (up to 25 μl min−1 in case of visible peri-aortic lymph leak). Renal artery and aorta were approached retroperitoneally through a left flank incision. A mechanical occluder was used that was custom-made from an 18G spinal tap needle, using a blunted trocar to compress the aorta against the bent tip of the needle via a screw actuator.

Measurements and procedures

Pressure in the left renal artery (RAP) was measured via the femoral artery catheter and a pressure transducer (Statham P23B). Renal blood flow (RBF) was measured via the implanted flowprobe connected to an ultrasound transit-time flowmeter (TS420, Transonic). Zero offset was determined at the end of each experiment after cardiac arrest. In rats, the pressure in the vascular occluder was monitored by a transducer (MSP300, Measurement Specialities, Hampton, VA, USA) to allow for automatic detection of the time points of the rise in RAP. In mice a 1 V signal from a footswitch was used for timing. All data were recorded on a computer at 100 Hz (RAP and RBF) or 10 Hz (occluder or footswitch). To study the autoregulatory response of RBF to a rapid rise in RAP, RAP was reduced by 20 mmHg using the aortic occluder for 60 s and was then rapidly released. Such step reductions and releases of RAP (RAP steps) were repeated every 5 min. At least three RAP steps were made in each experimental period.

Experimental protocols in rats

SMTC (n= 11) To investigate the effect of nNOS-derived NO on MR in RBF autoregulation, the nNOS inhibitor S-methyl-l-thiocitrulline (SMTC) was used. The optimal dose was determined in preliminary experiments (n= 4). RBF and AP were monitored during i.v. infusion of SMTC at increasing doses (2, 4, 8, 16, 32, 64, 128 and 256 μg kg−1 min−1) for ∼5 min each. Reduction of RBF and elevation of AP as indicators for eNOS inhibition were found to commence between 8 and 16 μg kg−1 min−1 (online Supplemental Fig. S1). Accordingly, we chose 10 μg kg−1 min−1 (in 30 μl min−1 saline i.v.) for the experiments. After a control period with at least three RAP steps, SMTC infusion was started followed by three RAP steps. The infusion was stopped 5 min after the 3rd RAP step and 15 min later three further RAP steps were performed as the recovery period. Finally, l-NAME was injected (25 mg kg−1 in 1 ml kg−1 saline i.v.) and 10–15 min later three additional RAP steps were induced to characterize the effect of complete inhibition of all NOS isoforms.

SMTC + Ang II (n= 5) As global NOS inhibition is associated with vasoconstriction and hypertension, which does not occur during selective nNOS inhibition, we combined SMTC with angiotensin II (Ang II) to match AP and RBF to the values ordinarily observed during l-NAME. After the control period, the SMTC infusion was started at 10 μg kg−1 min−1i.v. Fifteen minutes later Ang II was infused at 100 ng kg−1 min−1 in 50 μl min−1 saline i.v. The dose was then adjusted until RBF was reduced 30–50% compared to SMTC alone (average final dose 114 ± 24 ng kg−1 min−1). Five minutes after the final adjustment, three RAP steps were induced. SMTC and Ang II were then stopped and the remainder of the protocol for SMTC alone (see above) was followed, i.e. 15 min equilibration, three RAP steps, l-NAME 25 mg kg−1, 15 min equilibration, three RAP steps.

NPA (n= 7) As an alternative nNOS inhibitor we used Nω-propyl-l-arginine (NPA). NPA has a higher selectivity for nNOS (150-fold vs. eNOS, 3000-fold vs. iNOS) than SMTC (10- and 400-fold, respectively) (Huang et al. 1999). The disadvantage is the higher difficulty compared to SMTC to confirm efficient nNOS inhibition. In a previous in vivo study 17 μg kg−1 min−1i.v. of NPA induced effective nNOS inhibition in the renal medulla (Kakoki et al. 2001). To be on the safe side we infused NPA into the renal artery and chose 30 μg kg−1 min−1i.r.a. (dissolved in saline, infused at 30 μl min−1). The protocol included a control period, a NPA period 10 min after starting the NPA, and a l-NAME period 15 min after stopping NPA and injecting l-NAME (25 mg kg−1). Each period consisted of three RAP steps. To demonstrate efficiency we infused 3000 μg kg−1 min−1 NPA in ∼100 μl min−1i.r.a. for 3–10 min in three additional rats. This induced a reduction of RBF presumably due to eNOS inhibition. Vasoconstriction reached >60% of that induced later by l-NAME in the same animal, even though NPA had not reached its maximum effect during the infusion (online Supplemental Fig. S2). AP increased by 6, 14 and 11 mmHg in the three rats despite i.r.a. administration.

1400W (n= 5) To study the contribution of iNOS, the iNOS inhibitor N-[(3-aminomethyl)benzyl]acetamidine (1400W) was used. The dose was taken from Shi et al. (2006). After a control period, 1400W was infused for 30 min at 30 μg kg−1 min−1i.r.a. (in 30 μl min−1 saline). A hundred seconds after stopping the infusion three RAP steps were performed. Finally, l-NAME was administered followed 15 min later by three RAP steps.

Experimental protocols in mice

nNOS-ko As a genetic approach to the role of nNOS in RBF autoregulation nNOS-deficient mice (nNOS-ko; Huang et al. 1993; n= 9) were studied and compared to age- and gender-matched wild-type mice (n= 6) from the same colony. The mice were shipped from the colony directed by G.K. in Magdeburg to Freiburg as homozygous breeding pairs; from these, one generation was bred in Freiburg. Experiments were made on breeders and descendants. nNOS-ko and paired wild-type animals were studied in alternating sequence. RAP steps were induced during control conditions and in most mice also 10 min following injection of l-NAME (25 mg kg−1 in 1 ml kg−1 saline i.v.).

eNOS-ko Similar experiments were made in eNOS-deficient mice (eNOS-ko, n= 9) generated by Gödecke et al. (1998) and age- and gender-matched wild-type mice of the C57Bl6 strain (n= 5). The mice were bred in Magdeburg and shipped to Freiburg several weeks before the experiment. eNOS-ko and paired wild-type mice were studied in alternating sequence using the same protocol as for nNOS-ko mice. Although there was some variation between baseline levels of the wild-types for the nNOS group and those for the eNOS group (Tables 4 and 5), the identical treatment and alternate study sequence within each group should have focused the pairwise comparison fairly well upon the respective gene knockout.

Table 4.  Baseline haemodynamic parameters in mice
Experimental periodnMAP (mmHg)RBF (ml min−1 gKW−1)HR
  1. *,**:P < 0.05, P < 0.01 versus wild-type; †,‡: P < 0.05, P < 0.01 versus n/eNOS-ko control. RBF normalized to grams kidney weight (gKW).

Wild-type683 ± 16.1 ± 0.5574 ± 15
Wild-type +l-NAME6115 ± 3*,‡2.6 ± 0.3*,†552 ± 9
nNOS-ko control984 ± 26.8 ± 2.1597 ± 14
nNOS-ko +l-NAME9113 ± 3**,‡2.8 ± 0.4*,‡565 ± 16
Wild-type599 ± 37.7 ± 0.7529 ± 11
Wild-type +l-NAME4116 ± 4*4.7 ± 0.8526 ± 16
eNOS-ko control9111 ± 3*7.1 ± 0.6543 ± 26
eNOS-ko +l-NAME6106 ± 27.9 ± 0.8517 ± 38
Table 5.  Autoregulatory parameters in mice
Experimental periodnMyogenic response (AR units)1Tubuloglomerular feedback (AR units)1Third mechanism (AR units)1Total effective autoregulation (% of perfect)
  1. 1Units of autoregulation, i.e. improvement of autoregulation within the respective time window in % of perfect autoregulation (see Methods). *,**: P < 0.05, P < 0.01 versus wild-type; †, ##: P < 0.05, P < 0.01 versus n/eNOS-ko control; &,&&: P < 0.05, P < 0.01 versus n/eNOS-ko +l-NAME.

Wild-type638 ± 343 ± 637 ± 380 ± 4
Wild-type +l-NAME6114 ± 10*,‡9 ± 3**,‡3 ± 4**,†98 ± 8
nNOS-ko control936 ± 456 ± 633 ± 493 ± 9
nNOS-ko +l-NAME9111 ± 9**,‡21 ± 5*,‡−1 ± 6**,‡106 ± 11
Wild-type533 ± 459 ± 519 ± 596 ± 2
Wild-type +l-NAME478 ± 16*,†,&19 ± 3*,†,&4 ± 785 ± 3
eNOS-ko control944 ± 859 ± 422 ± 9100 ± 7
eNOS-ko +l-NAME637 ± 752 ± 930 ± 589 ± 11

Albumin, l-NAME and Ang II were obtained from Sigma-Aldrich (Taufkirchen, Germany); SMTC, NPA and 1400W were from Enzo Life Sciences (Lörrach, Germany).

Data analysis

The 100 Hz data of RAP and RBF were smoothed by a sliding average over 50 values each. Renal vascular resistance (RVR) was calculated as renal perfusion pressure (RPP)/RBF, where RPP = RAP – 4 mmHg. These data sets of RAP, RBF and RVR were downsampled to 10 Hz. Short segments were extracted into single files containing the last 10 s before each RAP reduction, and the segment from 10 s before through to 120 s after release. The exact time points for release of RAP were derived from the occluder or footswitch signal. The autoregulatory response of RVR was normalized to regulatory efficiency as a per cent of perfect autoregulation, with 100% denoting a RVR adjustment matched to keep RBF constant in face of a change in RAP; 0% indicates unchanged RVR, i.e. absence of autoregulation; <0% means that RVR paradoxically increased with increasing pressure. Total RBF autoregulation was calculated as (RVRend– RVRpre)/[(RPPend/RBFpre) – RVRpre]*100, where subscript ‘pre’ denotes averages for the last 10 s before release of the aortic occluder, and ’end’ the average during 90–120 s after the step increase in RAP. The time-course of RVR was normalized by the same formula, substituting RVRend by RVR of each time point. As described earlier (Just & Arendshorst, 2007; Just et al. 2009), the regulatory strength of each mechanism was derived from the improvement of autoregulatory efficiency within specified time intervals after the pressure step: MR was estimated from the change in RVR between t0= 0–1 s and t1= 4–7 s after the pressure step, TGF from t1= 4–7 s to t2= 25–35 s in rats or to t2= 20–30 s in mice, and the third mechanism from t2= 25–35 s in rats or from t2= 20–30 s in mice to t3= 90–120 s. Identical time windows were used during all experimental conditions regardless of potential changes in the dynamics of the autoregulatory mechanisms. Note that the strength of the MR was calculated starting from the initial reduction of RVR at t0= 0–1 s, whereas total autoregulatory efficiency was calculated from RVR before the pressure rise (t=−10–0 s). Therefore, the sum of all three mechanisms exceeds 100%. To avoid confusion, autoregulatory strengths are denoted as units rather than %, even though they do reflect autoregulatory improvement as % of perfect autoregulation. Analysis of RVRmin showed that RVRmin was either not significantly altered by the interventions, or if so could not explain the observed changes in MR strength (for more details see online Supplemental Table S1). The speeds of the autoregulatory mechanisms were determined from the first derivative of RVR, as calculated by the Savitzky–Golay algorithm (window size 11 points, 3rd order). The average speed at 0.4–2 s after the RAP step was taken to reflect MR, that at 8–16 s was taken for TGF, and at 30–60 s for the third mechanism. Results were normalized to baseline RVR and expressed as a per cent of RVR change per second.

Statistical analysis

Statistical significance was tested by ANOVA for repeated measures in conjunction with Holm–Sidak or Tukey post hoc test (SigmaStat 3.5, SPSS Inc., Chicago, IL, USA). In case of non-normal distribution Tukey test for multiple comparisons was used as a post hoc test. A P value <0.05 was considered statistically significant. Data are presented as mean ± SEM.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References
  8. Appendix
  9. Supporting Information

We evaluated RBF autoregulation and the strength of the autoregulatory mechanisms based on the dynamic changes of RVR in response to a rapid 20 mmHg step increase of RAP. The autoregulatory response pattern of RVR as a function of time is shown in the control traces of Figs 1 and 2. As described previously, the RVR response during control conditions is characterized by the following features: with the step increase of RAP, RVR initially fell, and subsequently rose to provide a final level of autoregulation of virtually perfect autoregulation within the next 120 s (99 ± 3% of perfect autoregulation in the pooled control responses of all 28 rats). Within the first 5–10 s RVR rose quickly to reach a level of 37 ± 3% of perfect autoregulation. After a short delay, a secondary rise of RVR began, which brought RVR to 79 ± 3% by 25 s after the pressure step. Finally, a slow third phase of RVR adaptation between 25 and 120 s after the pressure step improved autoregulation to its final level. The transition between the initial and the secondary response, determined from the turning point of RVR changes, was found at 6.6 ± 1.1 s in all control rats. The initial response is considered to reflect MR, the secondary rise is ascribed to TGF, and the slow response to an unidentified third regulatory mechanism. Note that the action of MR is calculated beginning from the initial minimum of RVR. This reflects the fact that the autoregulatory mechanisms need to overcome the initial reduction of RVR (presumably due to passive distention of the resistance vessels), before they can contribute to effective autoregulation. Accordingly, the sum of the strengths of all three regulatory mechanisms exceeds 100 units, i.e. 100% of perfect. Even though these units do reflect autoregulatory improvement as a per cent of perfect autoregulation, data are expressed in units to avoid confusion with total effective autoregulation (see also Methods section).

image

Figure 1. Renal autoregulatory response to a step increase in renal arterial pressure in rats. Influence of nNOS inhibition by SMTC as compared to global NOS inhibition by l-NAME A, time-course of renal arterial pressure (RAP) during a step increase of RAP in rats induced by rapid release following a reduction for 60 s. RAP is released at time t= 0 s. B, corresponding time-course of renal blood flow. C, corresponding time-course of calculated renal vascular resistance. Data are shown during initial control conditions (open circles), during infusion of nNOS-inhibitor SMTC (filled circles), and during global inhibition of all NOS isoforms by l-NAME (small circles with dashed line). Mean (continuous or dashed lines with symbols) ± SEM (dotted lines), n= 11. For clarity only every 5th value of those calculated and analysed is plotted.

Download figure to PowerPoint

image

Figure 2. Renal autoregulatory response in rats. Effects of nNOS inhibition by SMTC combined with Ang II, nNOS inhibition by NPA and iNOS inhibition by 1400W Time-courses of renal vascular resistance in response to a step increase in renal arterial pressure in rats initiated at time t= 0 s. Data are shown during control conditions (open circles), during nNOS or iNOS inhibition as described below (filled circles) and during global NOS inhibition by l-NAME (small circles with dashed line). A, filled circles depict nNOS-selective inhibition by SMTC combined with infusion of Ang II to reproduce the vasoconstriction and hypertension observed during inhibition of all NOS isoforms, n= 5. B, nNOS-selective inhibition by NPA, which is more selective than SMTC, n= 7. C, iNOS-selective inhibition by 1400W, n= 5. Mean (continuous or dashed lines with symbols) ± SEM (dotted lines). As in Fig. 1 only every 5th value is plotted.

Download figure to PowerPoint

To dissect the contributions of nNOS, eNOS and iNOS we employed both a pharmacological approach in rats using the NOS inhibitors SMTC, NPA and 1400W, and a genetic approach using knockout mice deficient for nNOS and eNOS. In each case the effect of a certain NOS isoform upon MR in RBF autoregulation was compared to the influence of global blockade of all NOS isoforms by l-NAME.

Inhibition of nNOS by SMTC (Fig. 1, Table 2), did not show statistically significant augmentation of the strength of MR: during control conditions MR improved autoregulation by 57 ± 4 units, reflecting its baseline strength; during SMTC infusion the strength of MR was 67 ± 4 units (Fig. 1, Table 2). This was not significantly different from the control. In the subsequent recovery period MR strength was 57 ± 3 units. Finally, inhibition of all NOS isoforms by l-NAME enhanced MR to provide an improvement of autoregulation by 78 ± 4 units (Fig. 1, Table 2). The strength of TGF was 37 ± 2 and 36 ± 3 units during the two control periods before and after SMTC, did not significantly change during SMTC (35 ± 2 units), but was reduced to 15 ± 2 units by l-NAME (Fig. 1).

Table 2.  Autoregulatory parameters in rats
Experimental periodnMyogenic response (AR units)1Tubuloglomerular feedback (AR units)1Third mechanism (AR units)1Total effective autoregulation (% of perfect)
  1. 1Units of autoregulation, i.e. improvement of autoregulation within the respective time window in % of perfect autoregulation (see Methods). *,**: P < 0.05, P < 0.01 versus Control; †,‡: P < 0.05, P < 0.01 versus isoform-specific inhibitor group.

Control11 57 ± 437 ± 221 ± 499 ± 5
SMTC11 67 ± 435 ± 216 ± 3103 ± 4
Control II11 57 ± 336 ± 318 ± 395 ± 4
l-NAME11 78 ± 4**15 ± 2**,‡9 ± 2*94 ± 4
Control554 ± 743 ± 522 ± 396 ± 5
SMTC + Ang II558 ± 335 ± 326 ± 784 ± 6
Control II543 ± 644 ± 424 ± 491 ± 4
l-NAME581 ± 2**,‡19 ± 2**,†5 ± 3*,†88 ± 6
Control756 ± 740 ± 418 ± 494 ± 7
NPA757 ± 541 ± 417 ± 4100 ± 10
l-NAME779 ± 4*,†18 ± 3**,‡5 ± 495 ± 9
Control555 ± 448 ± 324 ± 3110 ± 3
1400W550 ± 146 ± 529 ± 4108 ± 5
l-NAME581 ± 3**,‡26 ± 68 ± 692 ± 8

These effects of NOS inhibition on the strength of the autoregulatory mechansisms were also reflected in their speeds. During control conditions the maximum velocity of RVR change in the MR time window reached 4.9 ± 0.4% s−1 (Table 3). This speed of MR was not affected by SMTC (5.6 ± 0.5% s−1), but was markedly accelerated by l-NAME (9.2 ± 0.9% s−1, P < 0.01, Table 3). Speeds of TGF and the third mechanism were 0.8 ± 0.1% s−1 and 0.03 ± 0.01% s−1 (Table 3). The speed of TGF was significantly blunted by l-NAME (0.2 ± 0.1% s−1, P < 0.01), but not by SMTC (0.6 ± 0.1% s−1, Table 3). Note also that MR displayed oscillations during l-NAME (Fig. 1).

Table 3.  Speed of autoregulatory mechanisms in rats
Experimental periodnMyogenic responseTubuloglomerular feedbackThird mechanism
  (% change of renal vascular resistance s−1)
  1. *,**: P < 0.05, P < 0.01 versus Control; †,‡: P < 0.05, P < 0.01 versus isoform-specific inhibitor group.

Control11 4.88 ± 0.440.799 ± 0.090.033 ± 0.01
SMTC11 5.57 ± 0.490.654 ± 0.110.033 ± 0.01
Control II11 4.64 ± 0.400.666 ± 0.090.021 ± 0.01
l-NAME11 9.24 ± 0.89**0.195 ± 0.08**,‡0.044 ± 0.02
Control53.89 ± 0.790.986 ± 0.150.036 ± 0.01
SMTC + Ang II54.78 ± 0.71*0.732 ± 0.130.032 ± 0.02
Control II53.43 ± 0.710.712 ± 0.050.038 ± 0.01
l-NAME58.1 ± 1.99**,‡0.097 ± 0.09**,†0.004 ± 0.02
Control74.34 ± 0.600.619 ± 0.15−0.015 ± 0.03
NPA74.65 ± 0.740.685 ± 0.160.014 ± 0.02
l-NAME79.0 ± 0.98**,‡0.117 ± 0.07**,‡0.018 ± 0.03
Control54.49 ± 0.420.822 ± 0.100.006 ± 0.01
1400W54.49 ± 0.300.874 ± 0.110.006 ± 0.02
l-NAME512.5 ± 0.6**,‡0.249 ± 0.07**,‡−0.014 ± 0.02

Since global NOS inhibition produces renal and systemic vasoconstriction, which does not occur with selective inhibition of nNOS, we also investigated the effect of SMTC under similar degrees of hypertension and renal vasoconstriction as those customarily seen with l-NAME. This was achieved by combining SMTC with Ang II (Fig. 2A). RBF decreased from 9.8 ± 0.5 ml min−1 g−1 during control to 5.6 ± 0.3 ml min−1 g−1 during SMTC + Ang II, i.e. to a level virtually identical to that during l-NAME in the same experiments (6.6 ± 0.4 ml min−1 g−1, Table 1). Likewise, AP was increased from 112 to 132 mmHg during Ang II + SMTC and to 134 mmHg during l-NAME. Nevertheless, MR (58 ± 3 vs. control 54 ± 7 units) and TGF (35 ± 3 vs. 43 ± 5 units) were not significantly altered by SMTC + Ang II as compared to control, although there were slight trends for an augmentation of MR and reduction of TGF. Subsequent administration of l-NAME induced the expected effects (MR 81 ± 2 units, TGF 19 ± 2 units, Fig. 2A, Table 2). The speed of MR was slightly increased by SMTC + Ang II (4.8 ± 0.7 vs. 3.9 ± 0.8% s−1, P < 0.05, Table 3). l-NAME, however, markedly accelerated MR compared to the control (8.1 ± 2.0% s−1, P < 0.01), which was also significantly faster than during SMTC+AngII (4.8 ± 0.7% s−1, P < 0.01).

Table 1.  Baseline haemodynamic parameters in rats
Experimental periodnMAP (mmHg)RBF (ml min−1 gKW−1)HR (beats min−1)
  1. *,**: P < 0.05, P < 0.01 versus Control; †,‡: P < 0.05, P < 0.01 versus isoform-specific inhibitor group. RBF normalized to grams kidney weight (gKW).

Control11 111 ± 38.8 ± 0.5408 ± 8
SMTC11 113 ± 38.5 ± 0.5396 ± 8
Control II11 109 ± 39.0 ± 0.6405 ± 7
l-NAME11 135 ± 2**,‡5.5 ± 0.4**,‡374 ± 8**,‡
Control5112 ± 39.8 ± 0.5414 ± 6
SMTC + Ang II5132 ± 2**5.6 ± 0.3**380 ± 49
Control II5109 ± 311.1 ± 0.6*425 ± 16
l-NAME5134 ± 3**6.6 ± 0.4*,†370 ± 11
Control7112 ± 39.6 ± 0.7433 ± 14
NPA7114 ± 210.0 ± 0.7442 ± 15
l-NAME7142 ± 2**,‡5.8 ± 0.5**,‡405 ± 14**,‡
Control5110 ± 310.0 ± 0.5435 ± 10
1400W5104 ± 39.6 ± 0.6430 ± 12
l-NAME5136 ± 3**,‡5.0 ± 0.3**,‡386 ± 10**,‡

Inhibition of nNOS by NPA revealed no significant changes in MR (57 ± 5 vs. 56 ± 7 units) and TGF (41 ± 4 vs. 40 ± 4 units, Fig. 2B). Complete NOS inhibition by l-NAME increased MR to 79 ± 4 units and reduced TGF to 18 ± 3 units. AP or RBF were not affected by NPA (Table 1). The speed of MR was not affected by NPA (4.7 ± 0.7 vs. 4.3 ± 0.6% s−1), but was markedly accelerated by l-NAME (9.0 ± 1.0% s−1, P < 0.01, Table 3).

Inhibition of iNOS by 1400W did not show statistically significant alterations in MR (50 ± 1 vs. 55 ± 4 units) or TGF (46 ± 5 vs. 48 ± 3 units) compared to control conditions. In the same experiments, inhibition of eNOS by l-NAME led to an augmentation of MR to 81 ± 3 units and a reduction of TGF to 26 ± 6 units (Fig. 2C). AP or RBF were not affected by 1400W (Table 1). Likewise, the speed of MR was not affected by 1400W (4.5 ± 0.3 vs. 4.5 ± 0.4% s−1) but strongly accelerated by l-NAME (12.5 ± 0.6% s−1, P < 0.01, Table 3).

The genetic elimination of nNOS in nNOS-ko mice did not result in any alteration of MR compared to wild-type mice (36 ± 4 vs. 38 ± 3 units, Figs 2A and B, Table 5). There was a non-significant trend for a larger TGF (56 ± 6 vs. 43 ± 6 units, P > 0.06, Table 5, online Supplemental Fig. S6A), which may reflect the well-established NO-dependent modulation of TGF (Wilcox et al. 1992; Ito & Ren, 1993; Thorup & Persson, 1994). Baseline values of AP, RBF, kidney and body weight also did not differ between nNOS-ko and wild-type mice (Table 4). Inhibition of the remaining NOS isoforms, i.e. eNOS and iNOS, by l-NAME exerted the same augmentation of MR in nNOS-ko and wild-type mice (111 ± 9 vs. 114 ± 10 units, Fig. 3A and B, Table 5), similar to the results in rats. The speed of MR did not differ between nNOS-ko and wild-type mice (5.3 ± 0.5 vs. 5.6 ± 0.6% s−1) and was augmented by l-NAME in both genotypes (to 14.7 ± 1.2% s−1 and 13.3 ± 1.3% s−1 in nNOS-ko and wild-type mice, P < 0.01, Table 6).

image

Figure 3. Renal autoregulatory response in wild-type, nNOS-ko and eNOS-ko mice Time-courses of renal vascular resistance in response to a step increase in renal arterial pressure. A, pooled data from all wild-type mice before (open circles, n= 11) and during l-NAME (small circles with dashed line, n= 10). B, nNOS-deficient mice before (filled circles, n= 9) and during l-NAME (small circles with dashed line, n= 9). C, eNOS-deficient mice before (filled circles, n= 9) and during l-NAME (small circles with dashed line, n= 6). Mean (continuous or dashed lines with symbols) ± SEM (dotted lines). Only every 5th value is plotted.

Download figure to PowerPoint

Table 6.  Speed of autoregulatory mechanisms in mice
Experimental periodnMyogenic responseTubuloglomerular feedbackThird mechanism
  (% change of renal vascular resistance s−1)
  1. *,**: P < 0.05, P < 0.01 versus wild-type; †, ‡: P < 0.05, P < 0.01 versus n/eNOS-ko control; &,&&: P < 0.05, P < 0.01 versus n/eNOS-ko +l-NAME.

Wild-type65.57 ± 0.620.73 ± 0.130.001 ± 0.05
Wild-type +l-NAME614.67 ± 1.15**,‡0.27 ± 0.060.069 ± 0.06
nNOS-ko control95.34 ± 0.491.04 ± 0.210.033 ± 0.03
nNOS-ko +l-NAME913.26 ± 1.29*,†0.33 ± 0.100.052 ± 0.05
Wild-type53.67 ± 0.361.16 ± 0.04−0.042 ± 0.03
Wild-type+l-NAME410.8 ± 2.37**,‡,&&0.37 ± 0.13**,†−0.026 ± 0.05
eNOS-ko control94.11 ± 0.740.93 ± 0.100.037 ± 0.03
eNOS-ko +l-NAME64.08 ± 0.860.74 ± 0.14−0.027 ± 0.03

eNOS-ko mice were hypertensive as expected (Table 4). RBF displayed the same autoregulatory response (online Supplemental Fig. S3B) and hence the same MR strength as in wild-type mice (44 ± 8 vs. 33 ± 4%, Fig. 3C, Table 5). However, global inhibition of the remaining NOS activity by l-NAME failed to alter MR (37 ± 7 units; n= 6) in eNOS-ko mice, whereas it induced the expected augmention of MR in the wild-type animals (78 ± 16 units, P < 0.05 vs. control; Fig. 3C, Table 5). Similarly, the speed of MR did not differ between eNOS-ko and wild-type mice (4.1 ± 0.7 vs. 3.7 ± 0.4% s−1, Table 6). l-NAME induced the expected acceleration of MR in the paired wild-types (10.8 ± 2.4% s−1, P < 0.01), but failed to alter MR speed in eNOS-ko mice (from 4.1 ± 0.9% s−1, P > 0.9, Table 6).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References
  8. Appendix
  9. Supporting Information

The present study shows that the modulation of renovascular MR and RBF autoregulation by NO does not depend on nNOS or iNOS, but seems to be exclusively due to eNOS. This conclusion was supported both pharmacologically in rats and genetically in gene-targeted mice, and results did not depend on baseline renal vasoconstriction or arterial pressure. Given the expression patterns of the NOS isoforms, the findings indicate that NO-dependent modulation of MR is governed by the endothelium, but not the macula densa, smooth muscle or mesangial cells. An additional finding was that a compensatory factor appears to exist in eNOS-deficient mice capable of depressing MR independent of NO.

As described before by others (Wang & Cupples, 2001; Wronski et al. 2003; Shi et al. 2006) and ourselves (Just & Arendshorst, 2005; Just et al. 2009), global inhibition of all NOS isoforms by l-NAME strongly augmented and accelerated MR at the expense of TGF and the third autoregulatory mechanism, thereby speeding up the overall response. This was observed in all experiments of the present study in rats and mice except for eNOS-ko mice (Figs 1–3). l-NAME also induced oscillations of MR in all experiments in rats as well as in mice. The same was observed previously (Just & Arendshorst, 2005; Just et al. 2009), and is also reflected by an elevation of the admittance gain at the MR frequency range (∼0.1 Hz) in transfer function analyses (Just et al. 1999; Shi et al. 2006), suggesting a buffering effect of NO.

To determine the influence of nNOS in this modulatory influence of NO, we initially used the nNOS-selective inhibitor SMTC, since it is established, water soluble, and can conveniently be shown to reach effective dosage due to moderate selectivity over eNOS. The optimal dose at the borderline to non-specific eNOS inhibition was determined in preliminary experiments (online Supplemental Fig. S1). This dose of SMTC did not significantly augment or accelerate MR, except for small trends. Thus, although a minor contribution cannot be excluded, these results clearly speak against a major role of nNOS.

Global NOS inhibition by l-NAME is accompanied by renal and systemic vasoconstriction as well as by elevation of arterial pressure, none of which occurs with nNOS inhibition alone. If either of these factors should be required for an influence of nNOS-derived NO, it would have been missed by pure nNOS inhibition. Therefore, we studied the effect of SMTC also in combination with a vasoconstrictor dose of Ang II. Under this regime, RBF and RAP reached similar levels as those observed during l-NAME. Nevertheless, MR, TGF and RBF autoregulation were not altered compared to control conditions. Although the speed of MR was slightly accelerated by SMTC + Ang II, this effect was considerably smaller than that induced by l-NAME.

During SMTC alone, we observed small, non-significant trends for an augmentation and acceleration of MR. Since the selectivity over eNOS is only 10-fold (Huang et al. 1999) and baseline RBF tended to be reduced as well, the most likely explanation seems to be unspecific inhibition of eNOS. To more reliably assess the role of nNOS we used the nNOS inhibitor NPA, which has a much higher selectivity for nNOS. To demonstrate efficient nNOS inhibition with NPA we tested a dose that clearly reduced RBF as indication for non-specific eNOS inhibition; 3000 μg kg−1 min−1 caused a vasoconstriction reaching at least 60% of that induced by l-NAME in the same animals (online Supplemental Fig. S2). Considering the 150-fold selectivity of NPA for nNOS over eNOS (Huang et al. 1999), the dose of 30 μg kg−1 min−1 employed in our autoregulation experiments should have provided substantial if not complete nNOS inhibition. However, neither strength nor speed of MR, nor other aspects of RBF autoregulation were affected by NPA.

iNOS has been reported to be expressed in the normal kidney in afferent arteriolar smooth muscle and mesangial cells. However, iNOS inhibition by 1400W did not affect MR or RBF autoregulation, thus ruling out iNOS contribution, in congruence with a previous report (Shi et al. 2006).

These results refute a significant role of nNOS or iNOS in the modulatory influence of NO. By virtue of exclusion, this points towards eNOS as the pivotal isoform. Since a pharmacological tool for selective inhibition of eNOS is not available, we studied the role of nNOS and eNOS in gene-targeted mice. As described before (Just, 2007; Just & Arendshorst, 2007; Just et al. 2009), RBF autoregulation in wild-type mice resembled that in rats, including the augmentation of MR in response to global NOS inhibition (Just et al. 2009).

In nNOS-ko mice, RBF autoregulation did not differ significantly from wild-type animals, and global NOS inhibition enhanced MR similar to wild-type mice and rats. This corroborates the pharmacological findings in rats, that nNOS is not required for the modulatory influence of NO. However, several splice variants for nNOS exist (Brenman et al. 1997). Some of these variants, such as nNOS-β, are lacking exon 2, which is targeted in the nNOS-ko mice used in our study (Huang et al. 1993). Hence this variant would be spared from the gene knockout (Brenman et al. 1997; Wang et al. 1999). Previous studies found significant expression of nNOS-β mRNA in the kidney, although it was questioned whether this derived from the macula densa (Bachmann & Oberbaumer, 1998). However, more recent work did detect nNOS-β mRNA in isolated macula densa cells (Lu et al. 2010). Thus, NO production from the macula densa may have been only partially reduced in our nNOS-ko mice. This mitigates the conclusive value of our findings of unaltered autoregulation in these animals. Nevertheless, the results are at least compatible with a negligible role of nNOS in the modulation of MR and in this respect are consistent with our pharmacological findings in rats.

Given the evidence for eNOS in the modulatory effect of NO, one would expect a strong MR in eNOS-ko mice. However, autoregulation was virtually identical in eNOS-ko and wild-type mice. Nevertheless, l-NAME completely failed to augment or accelerate MR in these animals. Together, these results demonstrate that eNOS is absolutely required for the modulatory effect of NO. Furthermore, the small baseline strength of MR in eNOS-ko mice cannot be due to compensatory upregulation of nNOS or iNOS. In contrast, a compensatory factor seems to exist that attenuates MR by NO-independent means.

Taken together, in our in vivo setting, the modulatory effect of NO on MR and RBF autoregulation appears to be exclusively due to eNOS-derived NO, presumably from the afferent arteriolar endothelium (Ujiie et al. 1994; Star, 1997; Kone, 1999). In contrast, nNOS from the macula densa (Bachmann et al. 1995), or iNOS from smooth muscle (Tojo et al. 1994) or mesangium (Morrissey et al. 1994) do not contribute. These results differ from those of Shi et al. (2006) who found a partial contribution of nNOS. The reason for the difference is not clear. Shi et al. used a more indirect, frequency-domain method for quantification of MR. However, their criteria are sound and it seems surprising that the more indirect method would pick up rather than miss slight additional effects. Other differences could also play a role, such as the type of anaesthaesia (isoflurane vs. pentobarbital), the nNOS inhibitor used (LVNIO vs. SMTC and NPA), or hypothetical variations in sodium balance or housing conditions. Differences between rat strains appear less likely as an explanation, considering our consistent findings even between two species. Notwithstanding the quantitative differences regarding nNOS, both studies clearly support the importance of eNOS for the modulatory effect of NO and refute a role for iNOS. In our hands – supported both pharmacologically and genetically, and consistent within rats and mice – the contribution of nNOS seems to be negligible.

Given the high diffusibility of NO in living tissue, such preferential susceptibility to NO from different sources may seem surprising. However, several factors could promote an endothelial predominance: The overall quantity of NO from the endothelium might exceed that from the macula densa, possibly accentuated by convection of NO from upstream vascular segments through the blood stream (Smith et al. 2003).The diffusion path from the endothelium to the smooth muscle is probably shorter and not interspersed with mesangial cells, as compared to that from most macula densa cells (Barajas, 1979). Hypothetical polarized expression of aquaporins (Herrera & Garvin, 2007) could further facilitate diffusion from the endothelial side. On the other hand, the efficiency of NO signalling might be differentially modified by a factor with differential distribution but less diffusibility than NO, such as superoxide. Scavenging of bioavailable NO by superoxide could reduce the influence of macula densa-derived NO more than that from the endothelium. Interestingly, biologically relevant NO scavenging in the JGA (Wilcox & Welch, 2000; Ren et al. 2002) was indeed found to occur on the tubular but not the endothelial side of the afferent arteriole (Liu et al. 2004). Alternatively, superoxide could be part of the signalling pathway of NO. If the local level of NO would exceed that of superoxide, the scavenging reaction between the two would have more relevance for the bioavailability of superoxide than for that of NO. Since superoxide has been shown to exert direct, NO-independent effects on TGF (Liu et al. 2004) and other renal vasoconstrictor responses (Majid et al. 2005; Just et al. 2007, 2008), NO might act by muffling of such influences through ‘reverse’ scavenging of superoxide by NO. Previous work indicated ‘reverse’ scavenging in the renal vasoconstrictor response to endothelin (Just et al. 2008) and preliminary data suggest an even larger role in the NO-dependent modulation of RBF autoregulation (Just & Arendshorst, 2008).

An unexpected finding is the normal strength of MR in eNOS-ko mice. This indicates a compensatory factor depressing MR in eNOS-ko mice. Alternatively, a factor could be missing in eNOS-ko mice that is essential for normal NO-dependent modulation of MR. The nature of such factors remains unclear at present and warrants further investigation. Given the previous observations that the modulatory influence of NO is partly (Shi et al. 2006) if not entirely (Just & Arendshorst, 2005) prevented when the macula densa is rendered non-functional by furosemide, it is tempting to assume that the factors depressing MR in the absence of eNOS and in the absence of macula densa function are the same. The present results showing the modulating influence being independent from nNOS rule out our initial hypothesis that the macula densa dependency might be due to a macula densa-derived source of relevant NO. Another factor that could depress MR is prostaglandin E2, as it is produced by macula densa cells (Peti-Peterdi et al. 2003), upregulated by low tubular NaCl concentration or furosemide (Peti-Peterdi et al. 2003), and capable of dilating the afferent arteriole (Chatziantoniou & Arendshorst, 1992). Another possibility is a factor required for NO-dependent modulation but absent during furosemide or eNOS deficiency. A very attractive candidate is superoxide, since NAD(P)H-oxidase, the superoxide-producing enzyme, is expressed in macula densa cells (Zhang et al. 2009) and stimulated by elevation of tubular NaCl concentration (Liu et al. 2007). Since furosemide inhibits NaCl entry into the macula densa (Castrop & Schnermann, 2008), it would be expected to reduce local superoxide production and might thereby prevent NO-dependent modulation. Preliminary data show that NAD(P)H oxidase inhibition indeed reversed the augmentation of MR induced by NOS inhibition (Just & Arendshorst, 2008). Whether the factor(s) compensating the modulation of MR in eNOS-ko mice are the same as those causing the present and previous observation that baseline RBF is the same in eNOS-ko and wild-type animals (Ortiz & Garvin, 2003), remains open at this time.

Whatever the compensatory mechanism in eNOS-ko mice, it seems desirable or at least permissible for eNOS-ko mice to maintain MR at the same level as wild-type mice. Inasmuch as a stronger MR provides faster and thus more complete shield against hypertensive renal damage, eNOS-ko mice might be prone to chronic renal failure, having similar autoregulation as wild-type mice, but higher arterial pressure (Bidani & Griffin, 2004). Although in general, eNOS-ko mice do not seem to particularly suffer from renal failure (Huang, 2000), their susceptibility does appear to be enhanced when exposed to additional renal challenges such as reduced renal mass (Nakayama et al. 2009) or diabetes mellitus (Zhao et al. 2006). Thus, in the absence of other risk factors, eNOS-ko mice might be able to afford maintaining a normally sized MR in terms of renal failure, which might be of advantage in terms of other reasons, perhaps improved pressure natriuresis.

Perspectives

The virtually exclusive influence of eNOS points to an important role of the endothelium in modulating RBF autoregulation. This influence may thus be compromised in pathological situations associated with endothelial dysfunction, such as diabetes mellitus. The preferential influence of NO coming from the endothelial side raises the question of the factors causing such polarized susceptibility, including the possibility of differential interaction between NO and superoxide. Such factors might play a role in extrarenal tissues as well. The negligible contribution of nNOS excludes macula densa-derived NO production as the reason for the macula densa dependence of NO-dependent modulation, prompting continued search for other explanations. Future identification of the factors causing compensatory depression of MR in eNOS-ko mice will probably have importance beyond renal physiology, as it may help to understand why NO-dependent modulation of MR typically occurs in certain vascular beds and segments, but not in others (de Wit et al. 1998).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References
  8. Appendix
  9. Supporting Information

Appendix

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References
  8. Appendix
  9. Supporting Information

Author contributions

Conception of the project was spearheaded by A.J. Project refinement and design of experiments was done by M.D. and A.J. with important genetic advice from G.K. All mice were bred and provided by G.K. All experiments, data collection, analysis, and interpretation were performed at Universität Freiburg. Some mice were studied by A.J.; all other experiments and all analyses were done by M.D. The manuscript was drafted by M.D. and A.J. with essential discussion and revising from G.K. All authors approved the final version.

Acknowledgements

This study was made possible by funding from the Freiburg Institute of Advanced Studies (FRIAS) of the Universität Freiburg.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References
  8. Appendix
  9. Supporting Information

Fig. S1

Fig. S2

Fig. S3

Fig. S4

Fig. S5

Fig. S6

Table S1

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors

FilenameFormatSizeDescription
TJP_4696_sm_SuppMat.pdf866KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.