|Home | About | Journals | Submit | Contact Us | Français|
Proximal tubule reabsorption is regulated by systemic and intrinsic mechanisms, including locally produced autacoids. Superoxide (O2−), produced by NADPH oxidase (NOX) enhances NaCl transport in the loop of Henle and the collecting duct, but its role in the PT is unclear. We measured PT fluid reabsorption (Jv) in WKY rats and compared that to Jv in SHR, a model of enhanced renal O2− generation. Rats were treated with the NOX inhibitor apocynin (Apo), or with small interfering RNA (siRNA) for p22phox, which is the critical subunit of NOX. Jv was lower in SHR compared to WKY (WKY: 2.4±0.3 vs SHR: 1.1±0.2 nl/min/mm, n=9–11, p<0.001). Apo and siRNA to p22phox normalized Jv in SHR yet had no effect in WKY. Jv was reduced in proximal tubules perfused with S-1611, a highly selective inhibitor of the Na+/H+ exchanger 3 (NHE3), the major Na+ uptake pathway in the proximal tubule, in WKY but not in SHR. Pretreatment with Apo restored an effect of S-1611 to reduce Jv in the SHR (SHR+Apo: 2.9±0.4 vs SHR+Apo+S-1611: 1.0±0.3 nl/min/mm, p<0.001). However, since expression of NHE3 was similar between SHR and WKY, this suggests that O2− affects NHE3 activity. Direct microperfusion of tempol or apocynin into the PT also restored Jv in SHR. In conclusion, O2− generated by NOX, inhibits proximal tubule fluid reabsorption in SHR. This finding implies that PT fluid reabsorption is regulated by redox balance, which may have profound effects on ion and fluid homeostasis in the hypertensive kidney.
In the kidney, the proximal tubule reabsorbs 60–70% of filtered NaCl and fluid. Therefore changes in proximal tubule reabsorption can have profound effects on renal and body fluid balance and may contribute to the development of hypertension. The normal kidney protects against acute increases in blood pressure by excreting NaCl rapidly. The proximal tubule is thought to mediate much of this pressure natriuresis response. In young spontaneously hypertensive rats (SHR), prior to the onset of hypertension, expression of the major Na+ transport systems in the proximal tubules was higher 1 and Na+ excretion was lower compared to normotensive rats (WKY) 2. This was accompanied by an increase in fluid reabsorption in the proximal tubule in young (5 week) pre-hypertensive SHR compared to WKY. These observations suggest that an exaggerated NaCl and fluid reabsorption in the proximal tubule may contribute to the development of hypertension in young SHR, which persists in the adult animal. However, the increased reabsorption seen in young animals is not consistently observed in adult SHR. For example, in 7 and 12 week old SHR, at a time when hypertension was established, baseline Jv in the proximal tubule was lower compared to WKY3. Jv was also lower in adult stroke-prone SHR4, compared to WKY. This may be related to increased reactive oxygen species (ROS), since direct microperfusion of reducing agents corrected this dysfunction4. The SHR is indeed an established model of oxidative stress, with elevated superoxide levels and NADPH oxidase expression in both vascular and renal tissue 5.
The uncertain role of the proximal tubule is partially due to the difficult technologies to measure function. Proximal tubule transport in SHR has been investigated primarily in vitro, with cell preparations, or in vivo by free flow micropuncture. However, free flow measurements, which have characterized proximal tubule function on both WKY and SHR is complicated by interruption of flow to the macula densa, which tends to overestimate proximal tubule flow. Therefore, direct measurements of Jv by in situ microperfusion is the most accurate method to characterize proximal tubule function, since this method separates transport from the confounding effects of flow and tubuloglomerular feedback (TGF). Therefore we measured Jv in the proximal tubule of hypertensive and normotensive rats by direct microperfusion and recollection to test the hypothesis that increased production of O2− impairs proximal tubule function during hypertension in the adult SHR.
Groups of SHR and aged-matched WKY rats were pre-treated with apocynin (16 mg/kg/day), siRNA (25 μg, IV) or vehicle (IV) two days prior to experimentation. The Georgetown Animal Care and Use Committee approved the use of rats in this study. On day 3 animals were anesthetized with thiobarbital (Inactin, 80g/kg IP; Research Biochemicals, Inc, Natick, MA, USA) and prepared for in vivo micropuncture studies. Cannulae were placed in a jugular vein for infusion of fluids and in a femoral artery for the recording of mean blood pressure (MAP) (Powerlab, AD Instruments Inc, Colorado Spring, CO). A tracheotomy tube was inserted. The animals were allowed to breathe room air spontaneously. A catheter was inserted in the bladder and another in the left ureter to collect urine. The left kidney was exposed by a flank incision and stabilized in a Lucite cup (Vestavia, Birmingham Al) mounted on a heated surgical table and bathed in mineral oil maintained at 37°C. After surgical preparation, rats were infused with a solution of 0.154 mol/L NaCl and 1% albumin at 1.5 ml/h to maintain euvolemia 6. Studies were begun after 60 minutes of stabilization.
The proximal tubule site was identified, as described previously7, by injections from a “finding” pipette containing dye-stained artificial tubular fluid (ATF). The flow was blocked by injection of T grease (T grade, Apiezon Products, Manchester, UK) via a micropipette (10–12 μm OD) proximal to the perfusion site. The tubule was perfused with a micropipette (8–10 μm OD) connected to a microperfusion pump (model A1400, World Precision Instruments Inc, Sarasota, FL) at 18±3 nl/min. The perfusion solution contained 14C inulin (New England Nuclear, Boston, MA) as volume marker and 0.1% FD&C green dye for identification of the perfused loops. Tubules were perfused for 2–8 minutes prior to fluid collections, which were made at a downstream site with a micropipette (7–10 μm OD) following placement of a column of oil to block downstream flow. Samples were collected for 3–5 minutes and transferred into a constant-bore capillary tube whose length was measured with a micrometer to calculate the tubular fluid volume 8. Thereafter, the samples were injected into scintillation fluid and the 14C activity counted. Collected samples with less than 95% and more than 105% of microperfused inulin were discarded. The amount of microperfused inulin was estimated by the average of 14C-activity in 4 samples microperfused directly into capillary bore tubes at the end of the experiment. To determine the lengths of the perfused segments, tubules were filled with high-viscosity microfil (Flow Tech, Inc. Carver, MA). At the end of the experiments, the kidney was partially digested in 20% NaOH and the casts were measured under a dissecting microscope. The Jv was calculated by the difference in the perfusion rate and the collection rate factored by the length of the nephron: Jv =Vperf (nl/min)−Vcoll (nl/min)/PT length (mm) and expressed nl/min/mm. The composition of the perfusion fluid was as follows in mM: 125 NaCl, 20 NaHCO3, 5 KCl, 1 MgSO4, 2 CaCl, 1 NaHPO4, 5 glucose, and 4 urea. The in vivo nephron microperfusion technique allows control of flow rate, Na+ concentration, osmolality, and extracellular pH. This method is limited by the lack of endogenous agents generated in the proximal tubule, since all perfusions are with artificial tubular fluid containing various antagonists, but it more directly measures the intrinsic transport properties of the epithelium and allows the use of various inhibitors to test the role of key players in regulation of proximal tubule function. In separate nephrons, proximal and distal tubule flows were measured by free-flow micropuncture collections. In these nephrons, a collecting pipette was inserted into a mid-proximal or distal tubule and a small volume of mineral oil was inserted. The oil was allowed to move downstream and tubular fluid was aspirated into the collection pipette for 3–4 minutes. The fluid was transferred to a capillary bore tube and measured by microscopic calibration.
The small interference RNA (siRNA) to rat p22phox (NM-024160) was validated in-vitro as previously reported 9. The target site in p22phox cDNA of the constructs selected is: 299-320 (AAATTACTACGTCCGGGCTGT). The non-silencing control siRNA sequence ATTCTCCGAACGTGTCACGT (catalogue # 1022076; Qiagen, Valencia, CA) has no homology to any sequence in the mammalian genome. TransIT in-Vivo gene delivery system (Mirus, Madison, WI) was used to complex the p22phox siRNA to polymer as per manufacturer’s recommendations. This complexed siRNA was brought to 6 ml volume and delivered via jugular vein in 7 seconds. Kidney cortex was harvested after 48 hrs and saved for gene analysis. In a separate group of SHR, proximal tubules were microdissected for measurement of mRNA.
The kidneys were perfused through the abdominal aorta with 10ml of cold dissection solution to rinse away the blood. The composition of the dissection solution was as follow, in mM: 135 NaCl, 5 KCl, 1 NaH2P04, 1.2 MgS04, 2 CaC12, 6 L-alanine, 10 hydroxyethylpiperazine-N′-2-ethanesulfonic acid, and 5.5 glucose, as well as 0.1% bovine serum albumin (BSA). The kidneys were then perfused with 10 ml of cold dissection solution containing 0.1% collagenase (CLS II, Worthington). Thin sagittal slices were cut from the perfused kidneys and incubated in dissection solution containing 0.1% collagenase at 37°C for 30 min during which the solution was bubbled with 100% oxygen. Microdissection of individual proximal tubule segments was performed in cold dissection solution under a stereomicroscope10.
The kidney cortex and proximal tubules were harvested after 48 hrs and RNA was extracted using RNAqueous-4PCR kit (Ambion, Austin, TX). The cDNA was synthesized using superscript III with random hexamer (Invitrogen, Gaithersburg, MD). The gene expression for p22phox was assessed with real time PCR (ABI 7700, Foster City, CA), using FAM labeled p22phox Taqman probe assay (Rn00577357_m1, ABI, Foster City, CA) multiplexed with VIC labeled 18s control probe. Relative amounts of mRNA, normalizedby 18S rRNA, were calculated from threshold cycle numbers (CT, i.e., 2−ΔΔCT).
The kidney cortex was placed in RIPA lysis buffer containing protease inhibitors: PMSF 100 μg/ml, leupeptin 5 μg/ml, aprotenin 5 μg/ml, and sodium fluoride 1mM/ml. The dissected sections with RIPA lysis buffer were homogenized in Fastprep Bio101 (Thermo, Milford, MA) and the tubes were spun at 12,500 rpm for 15 minutes in a cold centrifuge. The supernatant was aliquoted and frozen at −80°C for future analysis. Protein concentrations were determined by Bio-rad Protein Assay Reagent (Bio-rad, Hercules, CA). Protein lysate (100 μg) for each sample was denaturedin boiling water for 5 min. After denaturation the lysate was placed on ice for 5 minutes and loaded on to 12.5% SDS-PAGE gel (Bio-Rad, Hercules, CA). The gel was transferred to nitrocellulose membrane (Bio-rad, Hercules, CA). The membrane was blocked with 5% Blotto milk. The primary antibody for NHE3 2 μg/ml (ab22765) was from Abcam Inc., Cambridge, MA and the antibody for NHERF2 was the kind gift of Dr James B. Wade and Paul A. Welling (University of Maryland, Baltimore, MD). The secondary antibodies (1:10,000) were peroxidase-labeled goat anti-rabbit (KPL, Gaithersburg, MD). The blots were probed with Beta Actin (Sigma, St. Louis, MO) for equal loading. Densitometry for western analysis was performed with Image J program (NIH, Bethesda, MD) for NHE3 and NHERF2.
The effects of apocynin, siRNA to p22phox, tempol and S-1611 in WKY and SHR were analyzed by 2 × 2 ANOVA, with post-hoc testing. Values are means ± SE. P < 0.05 was considered statistically significant.
The effect of the inhibition of NHE3 was tested by direct microperfusion of S-1661 (10−5 M), a highly selective inhibitor to NHE3 in Groups 1–4 above.
Jv of the proximal tubule was measured in WKY and SHR treated with siRNA directed to p22phox, a critical subunit of NADPH oxidase or scrambled siRNA (scr). Groups 5–8: WKY + scr; WKY + siRNA; SHR + scr; SHR + siRNA. Solutions (6 ml) containing siRNA were injected into cannulated jugular veins of anesthetized (1% isofluorane) rats. Animals were allowed to recover and prepared for microperfusion analysis after 48 hours.
Mean arterial blood pressure (MAP), measured under anesthesia was higher in SHR compared to WKY (Table 1). Treatment for 2 days with Apo had no effect on MAP in either strain. Proximal flow (VPT), measured by free flow collections in separate nephrons, was similar between WKY and SHR, but was decreased by Apo in SHR only (p<0.001). Distal flow (VD) and urine flow were similar in all groups.
To determine the mechanism of this dysfunction, we tested the activity of NHE3, a major Na+ uptake path in the proximal tubule. Jv was reduced by direct microperfusion of the NHE3 selective inhibitor, S-1611 (10−5 M, maximally effective dose) in WKY as expected, but not in SHR (Figure 4). This suggests that the dysfunction is due to reduction in NHE3 activity. When the reduced Jv of SHR was restored by Apo, direct microperfusion of S-1611 lowered Jv, similar to its effect in WKY (Figure 5). This suggests that the dysfunction of NHE-3 is linked to NAPDH oxidase generated O2−.
To confirm that the reduction in Jv of the proximal tubule was linked to NADPH oxidase, we repeated these experiments in rats treated with siRNA direct to p22phox, the critical subunit of NADPH oxidase. mRNA expression of p22phox, relative to expression of 18S, was reduced by siRNA treatment by 62±4% (n=6) in the renal cortical tissue and by 66±7% in microdissected proximal tubules (n=4) (Figure 6). Jv in SHR was restored by siRNA to p22phox, but unaffected in WKY (Figure 7). MAP was not affected by siRNA to p22phox: (MAP scr siRNA: 118±3 vs MAP siRNA p22phox: 115±3 mmHg, ns)
The major new finding of this study is that fluid reabsorption in the proximal tubule of adult SHR is impaired compared to WKY. Further, this reduced function can be completely restored by inhibition of O2−. Treatments that targeted O2− in general (tempol) and the more specific source of O2−, NADPH oxidase (Apo, siRNA) restored the lower Jv of the PT in SHR. Two results suggest that these effects are independent of the high blood pressure in SHR. First, Apo and siRNA to p22phox did not lower BP in SHR, yet Jv was corrected. Second, clamping of the renal artery in SHR to reduce renal perfusion pressure to normotensive levels had no effect on the reduced Jv.
To evaluate the mechanism of the reduced Jv, PTs were perfused with S-1611, a highly selective inhibitor of Na+/H+ exchanger-3 (NHE3). S-1611 reduced Jv in WKY PT by 50%, but surprisingly had no effect on the reduced Jv in SHR. However, when Jv was restored in SHR nephrons with Apo, subsequent microperfusion of S-1611 reduced Jv by 50%, similar to its effect in WKY. Since there was no difference in the expression of NHE3 in the renal cortex between strains, these data suggest that the difference in Jv between SHR and WKY is linked to NHE3 activity, rather than content. NHE3 activity may depend on the NHE regulatory factors (NHERF). We showed that NHERF-2 expression was higher in SHR than WKY, which may account for reduced NHE3 activity in SHR. Further, Apo reduced NHERF2 expression, consistent with restoration of PT function. However, both Apo and tempol, microperfused acutely into the proximal tubule restored Jv in the SHR. This suggests that O2− also regulates Jv directly and perhaps independently of NHE-3 and NHERF-2. Combined these results suggests that Jv is regulated by a complex interaction of regulatory factors (NHERF-2), which translocates NHE-3 and more acutely by O2− inactivation of NHE-3. The acute effect of O2− may not be due to translocation of NHE3, but further studies will be required to determine this mechanism.
Young SHR have an enhanced proximal tubular reabsorption and sodium uptake1, based on reports of greater NHE3 activity and lower Na+ excretion. Also, Jv measured by split-drop micropuncture, which measures the time of injected fluid uptake in the PT was also higher in 5 week SHR3. These studies have popularized the concept that Na+ retention in young SHR contributes to the development of hypertension. Yet this overactive PT does not persist into adulthood. The reduced Jv in PT of SHR in this study contrasts with the findings in other studies that did not detect differences in PT flow or reabsorption between adult SHR and WKY12. In fact, many studies do not report differences in Na+ excretion or transporter activities between adult SHR and normotensive rats. However, direct measurements have shown that Jv is lower in adult hypertensive rats compared to normotensive rats 4. Therefore, our observations with microperfusion and recollection in the SHR confirm these earlier studies that used the split-drop methodology. Further, Wu and Johns showed the reduced Jv could be restored by local delivery of exogenous superoxide dismutase (SOD)4. We confirmed this observation using the membrane-permeant antioxidant tempol and apocynin. Further, we have extended the studies on the role of O2− in regulation of proximal tubule function by NHE3 and NHERF2 expression.
As noted above, PT flow was similar in SHR compared to normotensive rats12. Indeed we also failed to see differences in PT flow between WKY and SHR when measured by free flow collections in separate nephrons (Table 1). However, when O2− was suppressed by Apo, PT flow was substantially reduced in the SHR PT, consistent with increased Jv in this group. The failure to detect basal differences in PT flow in hypertension remains unclear. This could be due to the method used in this study, which blocks flow to the macula densa and tends to increase GFR and therefore, PT flows in both strains. Additionally, similar flow in this segment could be real and the differences in transport function between these strains occurs in segments not evaluated in this study.
How is the PT regulated by O2−? One possibility is the interaction of O2− and nitric oxide (NO). NO inhibits Na+ uptake in the TALH13 via its effects on cGMP and on the density of membrane channels. However, NO promotes Na+ and/or fluid reabsorption in the proximal tubule14. Jv, measured by the same method in this study was lower in iNOS and nNOS knockout mice15, 16. Jv in normal rat proximal tubule is also reduced by NOS inhibition17. Taken together, these results suggest that endogenous NO enhances fluid transport in the proximal tubule 18. Therefore O2−, which reduces bioavailable NO should inhibit fluid reabsorption in the proximal tubule, which is consistent with our finding. However, NOS expression in the proximal tubule is not consistently shown, therefore the source of NO is not known. Yet there is considerable evidence that proximal tubule is constantly exposed to NO, since NOS is expressed in adjacent tissue and cells. In the kidney eNOS is expressed predominantly in the renal vascular endothelial cells 19, 20. Also nNOS is expressed in the epithelium, Bowman’s capsule and especially in the macula densa 21–24, which is in close contact with the proximal tubule in the renal cortex. iNOS is widely expressed in the tubule epithelia, including the proximal tubule, the thick ascending limb of Henle’s loop, and distal convolute tubule 19, 24. Data supporting the effects of NO to enhance Na+ uptake is difficult to obtain, since systemic administration of NOS inhibitors increases BP and renal perfusion pressure 25–27. However, an acute dose of an iNOS inhibitor, but not an nNOS inhibitor, both of which did not increase MAP, increased Na+ excretion 2-fold in rats28. Also, we showed that microperfusion of the proximal tubule with a non-selective NOS inhibitor reduced Jv by 30–40%, suggesting that NO promotes Na+ uptake in the proximal tubule17.
Another possibility suggested by our data is that O2− inhibits NHE3 activity. NHE3 is a major Na+ uptake mechanism in the proximal tubule and increased activity of NHE3 in the renal brush border membrane (BBM) may be involved in the pathogenesis of hypertension 29–32. However, in elegant experiments, McDonough and colleagues show that NHE3 is inactivated by translocation 33, 34. Therefore reduced reabsorption in proximal tubule of SHR may be due to a relocation of the NHE3, which would alter function 34, 35. We showed that in the SHR, NHE3 activity is suppressed and can be restored by several strategies. Therefore it is possible that high renal levels of O2− lead to translocation of NHE3 and antioxidants reactivate this pathway. Further, the regulatory factor for NHE3, NHERF2 may participate in the redox regulation of NHE3. Increased NHERF2 leads to translocation and inactivation of NHE336, 37. We confirmed that NHERF2 is higher in SHR 38. This increase is due to higher levels of O2−, since Apo reduced NHERF2 levels and increased NHE3 expression in the renal cortex and improved Jv in the PT in the SHR.
Renal levels of O2− and NADPH oxidase (NOX) are higher in SHR compared to WKY 39, 40. In addition, in this study we show that p22phox mRNA is expressed in microdissected proximal tubules and is reduced by systemic infusion of siRNA to p22phox. Also, direct microperfusion of tempol and apocynin into the proximal tubule suggests that the increased O2− in the SHR is derived from NOX in the proximal tubule. Consistent with this view, we showed that the SHR kidneys also had greater expression of NO synthase, yet bioavailable NO was lower in SHR41. Treatment with tempol increased NO function and presumably NO levels, indicating the powerful scavenging effects of O2−.
Further explanation of these results is required, since Na+ and fluid excretion of age-matched SHR and WKY are similar12. We suggest that downstream sites of Na+ reabsorption are increased by O2−, which compensates for the dysfunction in the PT. Garvin and colleagues have shown that O2− increased Na+ reabsorption in the TALH and the collecting duct, which could normalize excretion. The shift in nephron relative function can potentially alter oxygen consumption, since the PT is more O2 efficient for Na+ transport than downstream segments. Therefore, the increased O2− in SHR may contribute to the reduced oxygen efficiency and renal hypoxia seen in this strain40.
In conclusion, fluid uptake in the proximal tubule of adult SHR is impaired due to increased NADPH oxidase-dependent O2−. Our data suggests that O2− reduces NHE3 activity, possibly due to increased expression of NHE3 regulatory factor, but also inhibits fluid uptake independently of these pathways. These data suggest that increased ROS alters proximal tubular reabsorption during hypertension and impacts renal regulation of fluid and solute balance.
These data suggest that during hypertension responsibility for Na+ and fluid reabsorption along the nephron changes. Reabsorption in the proximal tubule is reduced and is increased downstream. This has an impact on oxygen usage for Na+ uptake since the PT is relatively more efficient than more distal segments. Therefore to preserve normal Na+ excretion, the kidney burns more oxygen as it transfers work during chronic increase in blood pressure. This may be a normal adaptation as the kidney acts to maintain stable Na+ balance. Increased superoxide generated during hypertension mediates this shift and therefore may be an adaptive process that allows the kidney to preserve normal function at the expense of more oxygen usage.
Source of funding
The studies reported in this manuscript were support by the National Institutes of Health (NIH) grants HL089583 and DK072183.
Conflict of interest disclosure: NONE.