Macula densa nitric oxide synthase: Expression, regulation, andfunction

CHRISTOPHER S. WILCOX and WILLIAM J. WELCH

Division of Nephrology and Hypertension, Georgetown University Medical Center, Washington, D.C., USA

Macula densa nitric oxide synthase: Expression, regulation, andfunction. The type 1 brain nitric oxide synthase (bNOS) isoformoccurs in macula densa (MD) cells where it functions to vasodilatethe afferent arteriole and blunt expression of tubuloglomerularfeedback (TGF). Dietary salt restriction enhances bNOS expres-sion, yet microperfusion studies with NOS inhibitors imply that itis functionally inactive. We thus assessed the hypothesis thatreduced L-arginine (L-Arg) availability during low salt (LS) intakelimits MD NO generation. Maximal TGF responses were re-corded during Henle’s loop perfusion with artificial tubular fluid(ATF). Microperfusion of L-Arg into the MD of LS, but notnormal or high-salt (HS) rats blunted maximal TGF responses(8.0 6 0.4 to 6.0 6 0.5 mm Hg; N 5 23; P , 0.01). Response toL-Arg was stereospecific, inhibited by coperfusion with mono-methyL-L-arginine (L-NMA), and dependent on system y1 trans-port, because it was blocked by coperfusion with the competitorsL-lysine or L-homoarginine. Absorption of [3H]-L-Arg from theperfused loop, via an L-Arg— or L-homoarginine—inhibitableprocess, was enhanced during HS. Salt restriction thus diminishesTGF attenuation by NO in the MD despite enhanced bNOSexpression because of limited delivery and/or uptake of L-Arg viasystem y1. This defines a novel mechanism of renal microcircu-latory adaptation to salt restriction via L-Arg—dependentchanges in TGF.

In pioneering experiments, Thurau and Schnermannshowed that increasing Na reabsorption at the maculadensa (MD) led to a fall in single-nephron glomerularfiltration rate (SNGFR) [1] that they termed tubuloglo-merular feedback (TGF). This response entails vasocon-striction of the afferent arteriole, but the constrictor medi-ators remain to be defined precisely [2]. Wilcox et al haveproposed recently that nitric oxide (NO) generated in theMD by L-arginine (L-Arg) metabolism via NO synthase(NOS) is a negative mediator blunting expression of TGFresponses [3]. Three or more NOS isoforms are expressedin the normal rat kidney [4]. Brain-type NOS (bNOS) isexpressed densely in MD cells [3, 5, 6]. bNOS is expressed

additionally in some adjacent thick ascending limb (TAL)cells, cells of Bowman’s capsule, some intrarenal nerves,and tubular structures [5, 6]. Both bNOS mRNA [4, 6–9]and nicotinamide adenine dinucleotide phosphate diapho-rase [3] are also expressed in the MD.

MD solute absorption appears to activate NOS andgenerate NO that blunts the expression of TGF-inducedreductions in glomerular capillary hydraulic pressure(PGC). Thus, microperfusion of NG-methyl-L-Arg (L-NMA) into the lumen of the MD or adjacent interstitiumvia the peritubular capillaries enhances the TGF responsein normal rats [3]. This effect of L-NMA is stereospecificand is prevented by coperfusion with excess L-Arg (dem-onstrating dependency on NOS) and by microperfusion offurosemide and thus depends on MD solute reabsorption.Studies with L-NMA or the relatively bNOS-specific inhib-itor 7-nitroindazole have confirmed that blockade of MDNOS enhances rat TGF responses [10]. In a perfused ratjuxtaglomerular apparatus (JGA) preparation, microperfu-sion of the MD by nitro-L-Arg methyl ester (L-NAME)leads to graded constriction of the adjacent afferent arte-riole [11]. This response is dependent on MD soluteabsorption, as it is prevented by MD perfusion with a lowNaCl solution. These studies directly implicate MD NOS inthis response because during MD microperfusion of L-NAME, the vasodilator response to intraarterial infusion ofacetylcholine, which activates endothelial NO generation,is preserved. MD microperfusion with L-Arg in normal ratsblunts TGF [12]. Collectively, these studies demonstrate apowerful MD-afferent arteriolar signaling pathway depen-dent on NO, which vasodilates the afferent arteriole andblunts the expression of TGF. MD-derived NO might alsoblunt the generation of the signal initiating TGF, as NO, orits second messenger cGMP, can inhibit TAL solute reab-sorption [13].

TGF responses are enhanced by dietary salt restriction oractivation of the renin-angiotensin system (RAS) [2, 14].The immunocytochemical expression of bNOS in the MD isenhanced by dietary salt restriction [15]. This effect isdependent partly on enhanced RAS activity, because itcould be blunted by the angiotensin II type 1 (AT1)

Key words: nitric oxide, brain NOS, tubuloglomerular feedback, saltintake, system y1, arginine, L-arginine, L-homoarginine, L-ornithine, L-lysine.

© 1998 by the International Society of Nephrology

Kidney International, Vol. 54, Suppl. 67 (1998), pp. S-53–S-57

S-53

receptor antagonist, losartan, in low-salt rats [15]. Studiesof microdissected vascular poles containing intact MDsegments have shown enhanced mRNA expression forbNOS in rats adapted to low salt intake that parallelsincreases in afferent arteriolar renin gene expression [7, 8].Despite this clear evidence for enhanced expression ofbNOS mRNA and protein in the MD during salt restric-tion, results from functional studies suggest that the MD-afferent arteriolar NO signaling pathway is inactive insalt-restricted rats. Thus, microperfusion of L-NMA intothe MD of salt-restricted rats fails to modify TGF, whereassimilar microperfusions in rats adapted to normal or highsalt intake regularly enhance the response [10, 16]. Thesestudies were designed to examine the hypothesis thatdelivery and/or uptake of L-Arg at the MD limits theblunting of TGF by NO generation during salt restriction.

L-Arg is transported across cell membranes by Na-dependent and -independent pathways, the latter includingsystem y1, for which L-Arg competes with L-ornithine,L-lysine, and L-homoarginine [17–19]. L-Arg is transportedboth into and out of the TAL by a saturable, carrier-mediated process [20, 21].

METHODSMale Sprague-Dawley rats (220 to 300 g) were fed

identical diets (Teklad, Madison, WI, USA) with a low salt(LS; Na content 0.03 g/100 g) or high salt (HS; Na content2.3 g/100 g) content for 8 to 12 days prior to testing. The LSdiet is sufficient for normal growth over this period. Ratswere prepared for micropuncture, microperfusion, andTGF studies, as described in detail previously [22]. In brief,the TGF response was assessed from changes in proximalstop-flow pressure (PSF) recorded via an ultramicropipetteinserted in a proximal tubule upstream from a wax blockand connected to a servo-null micropressure recorder (IPS,La Jolla, CA, USA). TGF was activated maximally bymicroperfusion of artificial tubular fluid (ATF) containingvehicle or drugs from the end proximal tubule into the loopsegment at 40 nl/min. Responses of HS and LS rats werecompared in the following protocols:

Series 1. TGF responses to microperfusion of L-Arg:Effects of salt intake and system y1 transport

TGF responses were contrasted in nephrons of LS rats(N 5 23) or HS rats (N 5 19) during microperfusion ofATF plus vehicle or ATF plus L-Arg (1023 M) perfusedalone. Additional studies were undertaken in nephronsperfused with L-lysine (2 3 1023 M) or L-homoarginine (2 31023 M) perfused alone or with L-Arg as competitiveinhibitors of system y1 transport.

Series 2. Absorption of [3H]-L-Arg from perfused loop ofHenle (LH): Effects of salt intake and system y1

transportATF containing [3H]-L-Arg (15 cpm/nl; Amersham, Ar-

lington Heights, IL, USA) and [14C]-inulin (15 cpm/nl;

ICN, Costa Mesa, CA, USA) as a recollection marker wasmicroperfused from a precisely calibrated microperfusionpump into the late proximal tubule and collected upstreamfrom an oil block at the early distal tubule. Perfusions weredelivered at 20 nl/min, and collections were made over2-minute periods. To the perfusates were added vehicle,L-Arg (1 mM), or L-homoarginine (1 mM). Aliquots of theperfusate and the collectate were counted in scintillationvials.

Data were analyzed by analysis of variance, and resultswere considered statistically significant at P , 0.05. Dataare presented as mean 6 SEM.

Fig. 1. Percentage changes in maximal tubuloglomerular feedback(TGF) responses of nephrons from rats adapted to low (A) or high (B) saltintakes during orthograde perfusion of the loop of Henle at 40 nl/min withartificial tubular fluid (ATF) containing L-arginine (1023 M), L-lysine(2 3 1023 M), or L-homoarginine (2 3 1023 M), singly or together.Means 6 SEM. ** P , 0.01 versus ATF 1 vehicle.

Wilcox and Welch: Macula densa NOSS-54

RESULTS

There were no significant differences in body wt, kidneyweight, or blood pressure in HS and LS rats. Moreover,PSF during zero loop perfusion was similar (LS, 36.1 6 0.6vs. HS, 37.2 6 0.8 mm Hg; N 5 24; NS). Therefore, TGFresponses are presented as changes in PSF during LHperfusion at 40 nl/min compared with zero perfusions.During LH perfusion with ATF 1 vehicle, maximal TGFresponses, assessed from changes in PSF, were enhanced innephrons of salt-restricted rats (LS, 8.1 6 0.4 vs. HS, 6.6 60.4 mm Hg; P , 0.01).

For series 1, the addition of L-Arg to ATF perfusatesblunted the maximal TGF responses in nephrons of LS rats(Fig. 1A). Microperfusion of L-lysine or L-homoarginine didnot alter TGF responses when given alone but, whencoperfused with L-Arg, blocked the effects of L-Arg inblunting TGF responses. In contrast, L-Arg did not signif-icantly change TGF responses in nephrons of HS rats (Fig.

1B). L-Lysine or L-homoarginine had no effects whethermicroperfused alone or with L-Arg. We conclude that L-Argdelivery and uptake from the tubular lumen via system y1

blunts TGF specifically in nephrons of salt-restricted rats.For series 2, 90% to 110% of the [14C]-inulin microper-

fused from the late proximal tubule was collected at theearly distal tubule in each nephron. Absorption of [3H]-L-Arg from the perfused LH averaged 92% 6 3% and did notdiffer between LS and HS rats. However, the mechanism ofabsorption was strongly salt dependent. Addition of excessL-Arg significantly (P , 0.01) blunted [3H]-L-Arg absorp-tion, but this effect was significantly (P , 0.01) greater inHS than in LS rats. Likewise, addition of L-homoargininesignificantly (P , 0.01) blunted [3H]-L-Arg absorption.This effect also was significantly (P , 0.05) greater in HSthan in LS rats. We conclude that dietary salt restrictiondiminishes the saturable y1 system reabsorptive flux ofL-Arg from the perfused LH.

Fig. 2. Cell diagram showing hypotheses for regulation of macula densa cell brain nitric oxide synthase (bNOS). Solute absorption via an Na1/K1/2Cl,furosemide-inhibitable luminal cotransporter activates bNOS, perhaps via increases in intracellular calcium concentration. L-Arginine uptake into thecell via system y1 is inhibited by L-ornithine, L-lysine, and L-homoarginine and by symmetric and asymmetric dimethylarginine (SDMA and ADMA)and monomethyl L-arginine (L-NMA). ADMA and L-NMA also inhibit bNOS activity. NO generated in the macula densa cell is degraded by interactionwith oxygen radicals (O2

2) to peroxynitrite (ONOO2).

Wilcox and Welch: Macula densa NOS S-55

DISCUSSION

The finding that bNOS is heavily expressed in the MDduring dietary salt restriction [15], but apparently does notfunction to produce NO, because TGF is unaltered duringNOS blockade [10, 16] implies that the NOS expressedmight not be functional. However, these studies demon-strate a robust blunting of TGF by microperfusion of L-Arginto the MD of LS rats. This response to MD perfusionwith L-Arg apparently entails NO generation, as it isstereospecific and blocked by coperfusion of L-NMA [23].Therefore, the NOS expressed during salt restriction isfunctional but apparently is limited by L-Arg availability.These single-nephron studies have parallels with whole-kidney results in which dietary salt restriction reducesplasma levels and rates of excretion of the NO metabolites,NO2 1 NO3 and cGMP, enhances the renal vasoconstrictorresponse to NOS blockade with L-NAME [24], and in-creases the NOS-dependent renal vasodilator response toL-Arg [25]. These whole-kidney data complement the morespecific single-nephron studies and indicate that dietary saltrestriction diminishes NO generation via an L-Arg—de-pendent mechanism. Rat plasma [L-Arg] averages 80 mM

during salt restriction but increases significantly to 160 mM

during salt loading [26], demonstrating a systemic effect ofsalt intake on arginine metabolism. However, the contrastbetween this evidence of L-Arg—stimulated NO genera-tion and the extracellular concentrations that are one ortwo orders of magnitude above the Km for NOS measuredin purified systems has been considered paradoxical [27].Nevertheless, the kidney generates methylarginine deriva-tives that both inhibit arginine uptake from the LH viasystem y1 and inhibit NOS activity in the MD [28].Immunocytochemical studies have located the MD as a siteof expression of the enzyme NG-NG-dimethylarginine dim-ethylaminohydrolase that recycles dimethylarginines to cit-rulline [28]. Therefore, cellular levels of L-Arg in the MDmay be limited because of competition with dimethylargi-nine for uptake, and the intracellular NOS activity in vivomay be inhibited by the presence of these methylarginines.This could account for a higher apparent Km for L-Arg invivo.

This finding of a saturable, carrier-mediated L-Arg re-absorptive pathway in the LH, for which L-homoargininecompetes, confirms previous studies [21]. Our data areconsistent with system y1 transport, in which L-Arg com-petes with other cationic amino acids [17–19]. Tracerquantities of L-Arg were almost completely absorbed fromthe LH, regardless of salt intake. However, the pathways ofabsorption differed significantly. During dietary salt load-ing, absorption through the system y1 transport pathwaywas increased substantially. Because luminal L-Arg uptakeinto MD cells for NO generation also depends on system y1

transport, the results suggest that, during HS, the robustL-Arg uptake into the MD via system y1 accounts for the

lack of response to provision of further L-Arg via thetubular lumen. In contrast, during salt restriction, systemy1 uptake of L-Arg is greatly diminished. This is com-pounded by reduced plasma L-Arg concentration [26] and,probably, enhanced reabsorption of L-Arg proximally viaNa1-linked absorption [29], which would limit argininedelivery to the MD segment and uptake into cells. Thiscould account for the requirement of extra L-Arg during LSto generate NO via MD bNOS.

In conclusion, these studies demonstrate that MD NOgeneration is limited during salt restriction. This is impor-tant physiologically because a limited NO generation couldcontribute to TGF enhancement and thereby to renalvasoconstriction. The failure of NO generation cannot beascribed to decreased NOS expression but rather to de-creased L-Arg availability. Reduced availability can beascribed to a reduced plasma level, enhanced proximalreabsorption, and diminished uptake in the LH and MDcells via sodium-independent system y1. These findingsthat L-Arg levels and transport are regulated and thatarginine can act systemically and within the kidney todetermine NOS activity, elevate it to a level of a hormoneinvolved in body fluid volume homeostasis. These hypoth-eses are summarized in Figure 2.

ACKNOWLEDGMENTS

These studies were supported by grants from the NIDDK to CSW (DK49870 and DK 36079) and by funds from the George E. Schreiner Chairof Nephrology. We thank Marly Davidson for preparing the manuscript.

APPENDIX

Abbreviations used in this article are: ATF, artificial tubular fluid;bNOS, brain nitric oxide synthase; HS, high salt; JGA, juxtaglomerularapparatus; LH, loop of Henle; L-Arg, L-arginine; L-NMA, monomethyL-L-arginine; L-NAME, nitro-L-Arg methyl ester; LS, low salt; MD, maculadensa; NO, nitric oxide; NOS, nitric oxide synthase; PGC, glomerularcapillary hydraulic pressure; PSF, proximal stop-flow pressure; RAS,renin-angiotensin system; SNGFR, single-nephron glomerular filtrationrate; TAL, thick ascending limb; TGF, tubuloglomerular feedback.

Reprint requests to Christopher S. Wilcox, M.D., Ph.D. Chief, Division ofNephrology and Hypertension, Georgetown University Medical Center, 3800Reservoir Road, N.W., PHC F6003, Washington, D.C. 20007, USA.E-mail: wilcoxch@gunet.georgetown.edu

REFERENCES

1. THURAU K, SCHNERMANN J: Die Natriumkonzentration an den Mac-ula densa Zellen als regulierender Faktor fur das Glomerulumfiltrat(Mikropunktionsversuche). Klin Wochenschr 43:410–413, 1965

2. BRIGGS JP, SCHNERMANN J: Control of renin release and glomerularvascular tone by the juxtaglomerular apparatus, in Hypertension:Pathophysiology, diagnosis, and management, edited by LARAGH JH,BRENNER BM, New York, Raven, 1995, p 1359

3. WILCOX CS, WELCH WJ, MURAD F, GROSS SS, TAYLOR G, LEVI R,SCHMIDT HH: Nitric oxide synthase in macula densa regulates glo-merular capillary pressure. Proc Natl Acad Sci USA 89:11993–11997,1992

4. MOHAUPT MG, MADSEN KM, WILCOX CS, KONE BC: Molecularcharacterization and localization of three nitric oxide synthase iso-forms in the rat kidney. (abstract) Endothelium-J Endothel Cell Res:Third International Meeting Biology of Nitric Oxide S25, 1993

Wilcox and Welch: Macula densa NOSS-56

5. TOJO A, ZHANG L, WILCOX CS, WELCH WJ, TISHER CC, MADSEN KM,GROSS SS, SCHMIDT HHHW: Differential localization and regulationof nitric oxide synthase isoforms in the juxtaglomerular apparatus.(abstract) J Am Soc Nephrol 3:554, 1992

6. BACHMANN S, BOSSE HM, MUNDEL P: Topography of nitric oxidesynthesis by localizing constitutive NO synthases in mammaliankidney. Am J Physiol 268:F885–F898, 1995

7. BOSSE HM, BOHM R, RESCH S, BACHMANN S: Parallel regulation ofconstitutive NO synthase and renin at JGA of rat kidney under variousstimuli. Am J Physiol 269:F793–F805, 1995

8. SINGH I, GRAMS M, WANG W, YANG T, KILLEN P, SMART A,SCHNERMANN J, BRIGGS JP: Coordinate regulation of renal expressionof nitric oxide synthase, renin, and angiotensinogen mRNA by dietarysalt. Am J Physiol 270:F1027–F1037, 1996

9. TERADA Y, TOMITA K, NONOGUCHI H, MARUMO F: Polymerase chainreaction localization of constitutive nitric oxide synthase and solubleguanylate cyclase messenger RNAs in microdissected rat nephronsegments. J Clin Invest 90:659–665, 1992

10. WELCH WJ, WILCOX CS: Role of nitric oxide in tubuloglomerularfeedback: Effects of salt intake. Clin Exp Physiol Pharmacol 24:582–586, 1997

11. ITO S, REN Y: Evidence for the role of nitric oxide in macula densacontrol of glomerular hemodynamics. J Clin Invest 92:1093–1098, 1993

12. THORUP C, PERSSON AEG: Macula densa-derived nitric oxide inregulation of glomerular capillary pressure. Kidney Int 49:430–436,1996

13. BAILLY C, NEANT F, AMIEL C: Cyclic GMP mediates the inhibition byplatelet-activating factor of Cl reabsorption in the cTAL. (abstract)J Am Soc Nephrol 4:450, 1992

14. THURAU K: Renal hemodynamics. Am J Med 36:698–719, 196415. TOJO A, MADSEN KM, WILCOX CS: Expression of immunoreactive

nitric oxide synthase isoforms in rat kidney: Effects of dietary salt andlosartan. Jpn Heart J 36:389–398, 1995

16. WILCOX CS, WELCH WJ: TGF and nitric oxide: Effects of salt intakeand salt-sensitive hypertension. Kidney Int 49(Suppl 55):S9–S13, 1996

17. SCHMIDT HHHW, BAEBLICH SE, ZERNIKOW BC, KLEIN MM, BOHMEE: L-arginine and arginine analogues: Effects on isolated bloodvessels and cultured endothelial cells. Br J Pharmacol 101:145–151,1990

18. SCHMIDT K, KLATT P, MAYER B: Characterization of endothelial cellamino acid transport systems involved in the actions of nitric oxidesynthase inhibitors. Mol Pharmacol 44:615–621, 1993

19. SCHMIDT K, LIST BM, KLATT P, MAYER B: Characterization ofneuronal amino acid transporters: Uptake of nitric oxide synthaseinhibitors and implication for their biological effects. J Neurochem64:1469–1475, 1995

20. DANTZLER WH, SILBERNAGL S: Amino acid transport by juxtamedul-lary nephrons: Distal reabsorption and recycling. Am J Physiol 255:F397–F407, 1988

21. SILBERNAGL S, VOLKER K, DANTZLER WH: Cationic amino acid fluxesbeyond the proximal convoluted tubule of rat kidney. Pflugers Arch429:210–215, 1994

22. WELCH WJ, WILCOX CS: Potentiation of tubuloglomerular feedbackin the rat by thromboxane mimetic: Role of macula densa. J Clin Invest89:1857–1865, 1992

23. WELCH WJ, WILCOX CS: Salt-sensitive regulation of PGC by maculadensa L-arginine delivery. (abstract) J Am Soc Nephrol 5:615, 1994

24. DENG X, WELCH WJ, WILCOX CS: Renal vasoconstriction duringinhibition of NO synthase: Effects of dietary salt. Kidney Int 46:639–646, 1994

25. DENG X, WELCH WJ, WILCOX CS: Renal vasodilation with L-arginine:Effects of dietary salt. Hypertension 26:256–262, 1995

26. DENG X, WELCH WJ, WILCOX CS: Plasma arginine concentration andrenal nitric oxide generation during changes in dietary salt andangiotensin II. (abstract) J Am Soc Nephrol 6:658, 1995

27. ARNAL JF, MUNZEL T, VENEMA RC, JAMES NL, BAI C, MITCH WE:Interactions between L-arginine and L-glutamine change endothelialNO production: An effect independent of NO synthase substrateavailability. J Clin Invest 95:2565–2572, 1995

28. TOJO A, WILCOX CS, BREMER V, KIMOTO M, WELCH WJ, KIMURA K,OMATA M, OGAWA T, VALLANCE P: Methyl arginines in rat kidney:Comparison of immunohistochemical localization of demethylatingenzyme NG, NG-dimethylarginine dimethylaminohydrolase (DDAH)with nitric oxide synthase, and effects on arginine transport. (abstract)J Am Soc Nephrol 7:1574, 1996

29. SILBERNAGL S, DEETJEN P: Molecular specificity of the L-argininereabsorption mechanism: Microperfusion studies in the proximaltubule of rat kidney. Pflugers Arch 340:325–334, 1973

Wilcox and Welch: Macula densa NOS S-57