Kidney International, Vol. 55 (1999), pp. 1384–1394

Distribution of postsynaptic density proteins in rat kidney:Relationship to neuronal nitric oxide synthase

AKIHIRO TOJO, DAVID S. BREDT, and CHRISTOPHER S. WILCOX

Division of Nephrology and Endocrinology, University of Tokyo, Tokyo, Japan; Department of Physiology,University of California, San Francisco, California, and Division of Nephrology and Hypertension,Georgetown University Medical Center, Washington D.C., USA

Distribution of postsynaptic density proteins in rat kidney: termed PDZ motifs, are multifunctional protein–proteinRelationship to neuronal nitric oxide synthase. interaction domains present in a diverse group of cy-

Background. Neuronal nitric oxide synthase (nNOS) is ex- toskeletal proteins and enzymes [1, 2]. Neuronal or typepressed in skeletal muscle beneath the sarcolemma associatedI nitric oxide synthase (nNOS or NOS I) has a PDZwith dystrophin complex. In brain, nNOS is anchored to synap-motif of approximately 100 amino acids in the aminotic membranes by specific postsynaptic density proteins

(PSD)-95 and PSD-93. We have investigated the cellular and terminus. Recently, Brenman et al have shown that thesubcellular localization of these PSD proteins in the kidney PDZ domain of nNOS binds the PDZ domain of syn-and their relationship to nNOS and the cell membrane. trophin in skeletal muscle [3]. This interaction targetsMethods. Kidneys from male Sprague-Dawley rats were

nNOS to the sarcolemmal dystrophin complex. In brain,processed for light and electron microscopic immunohisto-most of nNOS is found in axon and dendrites in whichchemistry with polyclonal antibodies against PSD and nNOS

proteins. it is concentrated in the thick postsynaptic region of theResults. Western blot analysis of rat kidney revealed a spe- plasma membrane, termed the postsynaptic density [4].

cific band for PSD-93 at the molecular weight of 103 kDa. Two specific molecules of the postsynaptic density,Immunostaining for PSD-93 was located in the thick ascending

PSD-95 and PSD-93 [5, 6], have been identified to associ-limb of the loop of Henle, macula densa cells, distal convolutedate with nNOS in brain. The predicted molecular struc-tubules, cortical collecting ducts, outer and inner medullary

collecting duct, glomerular epithelium, and Bowman’s capsule. ture of PSD-95 includes three PDZ domains, an SrcA pre-embedding electron microscopic immunoperoxidase homology 3 (SH3) domain, and a guanylate kinase-likeprocedure localized PSD-93 to the basolateral membrane of region located in the carboxy-terminus [2, 7]. PSD-93 isthese tubular cells. Using different sized immunogold particles,

60% homologous at the amino acid level with PSD-95a portion of nNOS in the macula densa colocalized withand shares the same domain structure. In the brain, thePSD-93 adjacent to cytoplasmic vesicles and the basolateralPDZ domains of PSD-95 binds to the carboxy-terminalmembrane. In contrast, PSD-95 protein was detected only

weakly in the cortex by Western blot. Immunostaining for tails of certain membrane-associated proteins such as thePSD-95 was located only faintly in the apical membrane of the N-methyl-D-aspartate acid (NMDA) receptor [8], thethick ascending limb, macula densa, distal convoluted tubule, Shaker-type K1 channels [9], and certain inwardly recti-and cortical collecting duct cells.

fying K1 channels [10]. It has been proposed that thisConclusion. PSD-93 is the predominant PSD expressed inprotein–protein interaction between nNOS and mem-the rat kidney. It is located primarily in the basolateral mem-

branes of distal nephron and colocalizes with a pool of nNOS brane-associated ion channels mediated by binding toin cytoplasmic vesicles and basolateral membranes of macula PSD-95 confers spatial orientation to NO generation.densa cells. This may permit selective regulation of nNOS by Ca21

influx through the NMDA receptor.Immunohistochemical studies have shown that nNOS

Specific protein motifs, found in postsynaptic density and PSD-95 are coexpressed in neurons of the hypothala-protein (PSD)-95, discs-large, and zona occludens 1, mus, hippocampus, and cerebellum [5]. In contrast, the

highest levels of PSD-93 occur in Purkinje cells of thecerebellum, which lack nNOS or PSD-95 [5]. On theKey words: nitric oxide, tubular glomerular feedback, dystrophin, distal

convoluted tubule, loop of Henle. other hand, nNOS is expressed in cells outside of thecentral nervous system, such as developing chromaffinReceived for publication August 26, 1998cells of the adrenal glands and secretory cells of theand in revised form November 19, 1998

Accepted for publication November 19, 1998 salivary gland. At these sites, nNOS is coexpressed withPSD-93, but not with PSD-95 [6]. 1999 by the International Society of Nephrology

1384

Tojo et al: PSD and nNOS in kidney 1385

In kidney, nNOS is expressed predominantly in mac- wt/vol bromphenol blue). Samples containing 50 mg pro-tein were separated by SDS-polyacrylamide gel electro-ula densa cells [11–13], but also has a limited expression

at other cells, notably the inner medullary collecting duct, phoresis on an 8.5% gel and electroblotted to nitrocellu-lose membranes. The membranes were incubated withBowman’s capsule, and certain nerve endings within the

kidney [14, 15]. Using electron microscopic (EM) immu- 5% nonfat dry milk in Tris-buffered saline with 0.1%Tween 20 (TBST) for one hour, following overnight in-nohistochemistry, a portion of nNOS in the macula densa

is associated with cytoplasmic vesicles [12, 13]. There cubation with a polyclonal antibody for PSD-93 orPSD-95 at a 1:200 dilution. After rinsing in TBST, mem-is increasing evidence that nNOS in the macula densa

regulates specific functions of the adjacent afferent arte- branes were incubated for one hour with horseradish per-oxidase (HRP)-conjugated anti-guinea pig immunoglobu-rioles, including the tubuloglomerular feedback (TGF)

response [11, 15–17], and the release of renin [18]. The lin antibody for PSD-93 or antirabbit immunoglobulinantibody for PSD-95 at a 1:1000 dilution. After rinsingdistribution and function of PSD proteins in the kidney

have not been studied. Therefore, we investigated the with TBST, blots were detected by 0.8 mm diaminobenzi-dine with 3 mm nickel chloride and 0.3% hydrogen perox-cellular and subcellular localization of two isoforms of

PSD proteins in the rat kidney. We used an EM immuno- ide. Homogenates of whole rat brain were used as posi-tive controls for PSD-93 and PSD-95.cytochemical method to examine the ultrastructural lo-

calization of nNOS and PSD proteins in the macula de-Light microscopic immunohistochemistrynsa cells to test the hypothesis that PSD could function

to anchor nNOS to the basolateral cell membrane or to The tissue was processed for immunohistochemistryas described in detail previously [19, 20]. Wax sectionsintracellular organelles.(2 mm) were dewaxed and were incubated with 3% H2O2

for 15 minutes to eliminate endogenous peroxidase activ-METHODS

ity and thereafter with blocking serum. The sections wereAnimal preparation incubated with a polyclonal antibody directed against

either PSD-93 or PSD-95 [1, 5, 6] at a dilution of 1:200For histochemical studies, five male Sprague-Dawleyrats (180 to 200 g) were fed a standard rat chow (Na for 60 minutes. The guinea pig antiserum to PSD-93 was

raised against a glutathione-S-transferase (GST) fusioncontent 0.3 g · 100 g21) and were given tap water todrink. Rats were anesthetized with pentobarbital (50 mg · protein encoding amino acids 77 to 453 of PSD-93. The

rabbit polyclonal antiserum to PSD-95 was raised againstkg21 i.p.). The kidneys were perfused with phosphate-buffered saline (PBS) followed by perfusion with perio- a GST fusion protein encoding amino acids 1 to 386 of

PSD-95. For PSD-93, the sections were rinsed with TBSTdate-lysine-paraformaldehyde solution through an ab-dominal cannula. Kidney slices were immersed in this and incubated with a HRP-conjugated secondary anti-

body against guinea pig immunoglobulin (Dako, Glos-solution overnight at 48C. Tissue for light microscopicimmunohistochemistry was embedded in wax (Polyeth- trup, Denmark). For PSD-95, the sections were incu-

bated with a biotinylated secondary antibody againstylene glycol 400 disterate; Polysciences Inc., Warrington,PA, USA). For immunogold studies, 1 mm3 blocks of rabbit immunoglobulin (Dako), followed by incubation

with the HRP-conjugated streptavidin solution. HRP la-tissue were embedded in Lowicryl. The remaining tissuewas processed for the pre-embedding immunoperoxi- beling was detected by incubation with peroxidase sub-

strate solution, diaminobenzidine (0.8 mm; Dojindodase method. These methods have been described indetail previously [15, 19, 20]. Laboratories, Kumamoto, Japan). The sections were

counterstained with hematoxylin before being examinedWestern blotting under a light microscope. For the negative control, sec-

tions were processed in the same way but were incubatedThe methods have been described in detail previously[15, 20]. Kidneys were removed immediately after perfu- with TBST buffer instead for the primary antibody.sion with ice-cold saline. Kidney tissue was cut into two

Electron microscopic immunoperoxidase procedureportions, cortex and inner medulla. The tissue was ho-mogenized on ice with a Teflon glass tissue homogenizer Sections (50 mm) of each kidney were cut on a vibra-

tome and were processed for pre-embedding EM immu-(Iwaki, Chiba, Japan) in 4 ml buffer containing 50 mmTris pH 7.4, 0.2 mm ethylenediaminetetraacetic acid noperoxidase as described previously [15, 20]. The sec-

tions were washed with 50 mm NH4Cl in PBS and were(EDTA), 2 mm leupeptin, 50 mm dithiothreitol, and 0.2mm phenylmethylsulfonyl fluoride. Homogenates were incubated with 0.1% bovine serum albumin in TBST for

30 minutes, followed by incubation with the primarycentrifuged at 12,000 g for 20 minutes. The supernatantwas diluted in sodium dodecyl sulfate (SDS) sample antibody against PSD-93 at 1:100 dilution overnight at

48C. For a negative control, other sections were incu-buffer (0.5 m Tris-HCl, pH 6.8, 10% vol/vol glycerol,10% wt/vol SDS, 5% vol/vol b-mercaptoethanol, 0.05% bated without the antibody. After rinsing with TBST,

Tojo et al: PSD and nNOS in kidney1386

Fig. 1. Western blots of PSD-93 in the kidney. Aprominent band for PSD-93 is seen in the kidneycortex and inner medulla at 103 kDa. The antibodyto PSD-95 recognizes a 95 kDa band in brain andshows a faint band in renal cortex, but not in innermedulla.

the sections were incubated with HRP-conjugated sec- body (Amersham Life Science, Buckinghamshire, UK).The sections were washed with buffer, incubated withondary antibody against guinea pig immunoglobulin

(Dako) for two hours. The sections were rinsed and fixed 2% glutaraldehyde in PBS for 30 minutes, rinsed withdistilled water, counterstained with uranyl acetate andwith 1% glutaraldehyde in PBS for one hour and rinsed

again, and the HRP labeling was detected by incubation lead citrate, and examined with the EM.with peroxidase substrate solution, diaminobenzidine(0.8 mm; Dojindo Laboratories). The sections were post-

RESULTSfixed in 2% osmium tetroxide in 0.1 m sodium cacodylate

Protein expression for PSD-93 and PSD-95buffer for one hour at 48C and were embedded in epoxyresin. Sections (1 mm) were cut on an ultramicrotome Western blotting was used to detect PSD-93 and

PSD-95 protein expressed in homogenates of wholeand were stained with toluidine blue for light microscopy.Ultrathin sections for electron microscopy were cut simi- brain, kidney cortex, and inner medulla. The antibody

for PSD-93 applied to protein extracted from the brain,larly, stained with lead citrate, and photographed on atransmission EM (Hitachi H-7000, Tokyo, Japan). renal cortex, or medulla produced a prominent band cor-

responding to that anticipated for PSD-93 protein with aPostembedding immunogold procedure molecular weight of 103 kDa (Fig. 1). In the protein

extracted from the brain, the antibody for PSD-93 recog-Ultrathin sections of the Lowicryl blocks were cut ona ultramicrotome, mounted on colloidin-coated nickel nized additional bands at a molecular weight of around

75, 93, and 150 kDa after a long exposure, but thesegrids, and processed for immunogold labeling as de-scribed previously [15, 20]. Briefly, the sections were extra bands were weakly expressed in the protein ex-

tracted from the kidney. Western blotting of proteinincubated with 0.1 m NH4Cl for one hour and were rinsedwith buffer solution (0.02 m Tris HCl, 0.15 m NaCl, 0.05% extracts prepared similarly and probed with an antibody

of PSD-95 showed less prominent expression, but a 95Tween 20, adjusted to pH 7.2) for 15 minutes. The sec-tions were incubated with both primary antibodies at kDa band corresponding to PSD-95 was seen for protein

extracts from brain and renal cortex, but not from me-48C overnight: guinea pig polyclonal antiserum againstPSD-93 at a dilution of 1:200 and mouse monoclonal dulla. The antibody for PSD-95 recognized an additional

band at a molecular weight of around 75 kDa protein,antibody against nNOS (Transduction Laboratories,Lexington, KY, USA) at a dilution of 1:100. After wash- which has been also noted previously [1, 6], in the protein

extracted from both the brain and renal cortex.ing the sections with buffer, they were incubated for twohours with both secondary antibodies at a dilution of

Light microscopic distribution of PSD-93 and PSD-951:50. There were 10 nm gold-labeled goat anti-guineapig IgG antibody (Aurion, Wageningen, The Nether- Light microscopic observation of 2 mm wax sections

demonstrated immunostaining for PSD-93 in cells of thelands) and 30 nm gold-labeled goat antimouse IgG anti-

Tojo et al: PSD and nNOS in kidney 1387

Fig. 2. Light micrographs illustrating immunostaining for PSD-93 in the rat kidney. Immunostaining for PSD-93 is observed in the thick ascendinglimb of loop of Henle (✩ in panel a), macula densa (→ in a), distal convoluted tubules (m in panels a and c), cortical collecting ducts (panel d),and the inner medullary collecting ducts (panel e). In the negative control, no immunostaining was observed in these structures (panels b and f).Magnifications are: 3250 (a), 3440 (b), 3400 (c), 3300 (d), 3400 (e, f).

macula densa (Fig. 2a), the thick ascending limb of (Fig. 2 b, f). Light microscopic examination of 1 mmEpon sections showed predominant immunostaining ofHenle, the distal convoluted tubule (Fig. 2c), the cortical

collecting duct (Fig. 2d), the outer medullary collecting the basolateral regions of cells of the macula densa (Fig.3a) and outer and inner medullary collecting ducts (Fig.duct, and the inner medullary collecting duct (Fig. 2e).

The proximal tubules, thin descending limb of loop of 3b). PSD-93 was also expressed in the Bowman’s capsuleand glomerular epithelial cells (Fig. 3a).Henle, and renal vasculature did not show significant

staining for PSD-93. PSD-93 immunoreactivity was posi- Immunostaining for PSD-95 was less intense but waslocated faintly in cell of the glomerular capillary walltive in the adventitia of renal arteries. The negative con-

trols showed no staining of any structures in the kidney (Fig. 4a), the apical membrane of the distal convoluted

Tojo et al: PSD and nNOS in kidney1388

tubule including the macula densa (Fig. 4a), and some of PSD-93 and nNOS were found to be very closelycolocalized within the cells. These results confirm previ-cells in the cortical collecting duct (Fig. 4 b, c). There

was no significant staining for PSD-95 in the proximal ous findings in macula densa cells that nNOS is expressedpredominantly in the cytoplasm [12, 13], but that a com-tubules and renal vasculature. In the medulla, PSD-95

was not detected in cells of the outer and inner medullary ponent is expressed in association with small cytoplasmicvesicles and basolateral membranes [12]. The double-collecting ducts, but was detected in the thin descending

limb of the loop of Henle (Fig. 4d) and the thick as- labeling method applied to macula densa cells clearlydetected a population of nNOS that is very intimatelycending limb of the loop of Henle.colocalized with PSD-93 along the cytoplasmic vesicles

Electron microscopic distribution of PSD-93 and and over the basal infoldings of basolateral membranes.neuronal nitric oxide synthase The mRNA for PSD-95 is expressed in neurons of the

brain, but has not been identified in the kidney [2, 5]. WeWhen examined by EM using the pre-embeddingmethod, PSD-93 was detected predominantly in the ba- detected a small amount of PSD-95 protein expression in

the cortex of rat kidney by Western blot analysis. Theresolateral membrane and cytoplasmic vesicles of cells ofthe macula densa (Fig. 5a), in the basal infolding and was faint immunostaining for PSD-95 in the apical mem-

branes of cells of the thick ascending limb of the loopvesicles of the distal tubule (Fig. 5b), and in the basolat-eral membrane of the inner medullary collecting duct of Henle, macula densa, distal convoluted tubule, cortical

collecting duct, and along the glomerular capillary wall.(Fig. 5c). There was little staining in the apical mem-brane, mitochondria, or nuclei of these cells. The function of PSD in the kidney is unknown. In the

central nervous system, PSD-95 interacts with nNOS andThe immunogold double EM-labeling method wasused to colocalize PSD-93 and nNOS in macula densa the NMDA receptor [8], thereby regulating this signaling

system and facilitating efficient Ca21 activation of nNOScells. It revealed that some of the immunogold label fornNOS (larger particles in Fig. 6) was colocalized with following NMDA receptor activation [2]. In skeletal

muscle cells, the nNOS protein is anchored to fast twitchPSD-93 (smaller particles in Fig. 6) along the basolateralmuscle fibers by dystrophin [21]. A lack of dystrophinmembrane or vesicles (arrows in Fig. 6), but some werein the muscles of patients with Duchenne muscular dys-distinct from PSD-93 and appeared in the cytoplasm oftrophy causes nNOS to translocate from the sarcolemmathe macula densa cells (Fig. 6).to the cytosol [3]. This translocation is associated withThe distribution of PSD-93, PSD-95, and nNOS isa loss of nNOS function [3]. Thus, PSD is implicated insummarized in Table 1.the efficient activation of nNOS and in the vectorialrelease of NO. Recently, Wu et al reported that NO

DISCUSSION generated by constitutive NOS enhances Ca21/calmodu-The primary new findings are that immunoreactive lin-dependent phosphorylation of proteins in the PSD

PSD-93 occurs in cells of the thick ascending limb of the and that this phosphorylation cascade can inactivateloop of Henle, the macula densa, the distal convoluted nNOS [22]. This intriguing finding suggests further thattubule, the cortical, outer, and inner medullary collecting the PSD complex and nNOS may also interact via feed-ducts, the glomerular epithelium, and Bowman’s capsule. back inhibition.Immunohistochemical expression of nNOS is prominent By analogy with previous findings in the central ner-in cells of the macula densa, Bowman’s capsule, and vous system and skeletal muscle, PSD family proteinsweakly in inner medullary collecting duct [11, 12, 15]. may regulate nNOS activity in macula densa cells of theTherefore, PSD-93 colocalizes with nNOS at all of the kidney. Studies from several laboratories have shownsites of nNOS expression in the kidney. However, PSD- that local blockade of nNOS in the juxtaglomerular ap-93 was also expressed in cells of the distal convoluted paratus by microperfusion of the NOS inhibitors,tubule, the thick ascending limb of the loop of Henle, NG-methyl-l-arginine (L-NMA), nitro-l-arginine, or theand cortical and outer medullary collecting ducts, which relatively selective nNOS specific inhibitor, 7-nitro-are not prominent sites for nNOS expression [11, 12, 15]. indazole into the macula densa segment or surrounding

Electron microscopic immunocytochemistry of macula interstitium enhances the TGF response [reviewed indensa and inner medullary collecting duct cells revealed 23]. These responses to NOS inhibitors are stereospecificthat PSD-93 is expressed prominently along the basolat- and are prevented by excess l-arginine. Therefore, theyeral membrane, where it is closely associated with a pool presumably indicate inhibition of NOS [11]. The TGFof cytoplasmic vesicles. nNOS is expressed predomi- response entails a predominant vasoconstriction of thenantly in the cytoplasm but is also associated with cyto- afferent arteriole that reduces the glomerular capillaryplasmic vesicles and the basolateral cell membrane. In- pressure and the nephron glomerular filtration rate.deed, by applying the double immunogold particle- Thus, in the blood vessel wall, the release of nitric oxide

(NO) in the juxtaglomerular apparatus functions to mod-labeling technique to macula densa cells, a population

Tojo et al: PSD and nNOS in kidney 1389

Fig. 3. Light micrographs from 1 mm sections of pre-embedding electron microscopic (EM) samples. These illustrate immunostaining forPSD-93 in macula densa (a) and the inner medullary collecting ducts (b). Magnification 3590 (a), 3470 (b).

Fig. 4. Light micrographs illustrating immunostaining for PSD-95 in the rat kidney. Immunostaining is observed in cells of the glomerularendothelium, macula densa, apical membrane of some cells in the distal convoluted tubules (b) and cortical collecting ducts (d). In the medulla,the inner medullary collecting ducts do not stained for PSD-95, but there is some staining in cells of the thin descending limb of loop of Henle(d). Magnifications are 3380.

ulate the vasoconstrictor responses of the adjacent by endothelial cell NOS, the NO that blunts the TGFresponse in the juxtaglomerular apparatus derivessmooth muscle cells during engagement of TGF. How-

ever, unlike the blood vessel wall where NO is generated mainly or exclusively from nNOS.

Tojo et al: PSD and nNOS in kidney1390

Fig. 5. Electron micrograph illustrating PSD-93 immunostaining using the pre-embeddingimmunoperoxidase procedure. Cells of themacula densa (a) and the distal convolutedtubule (b) are immunoreactive for PSD-93on the basolateral membrane and basal in-foldings and on the membranes of cytoplasmicvesicles. Cells of the inner medullary collect-ing duct (c) show a basolateral localizationof PSD-93. Magnifications are: 317,000 (a),320,000 (b), 32000 (c).

There are several unusual properties of nNOS func- Nephrol 8:312A, 1997). However, whereas the TGF re-sponse of the normotensive rat nephron is enhanced bytion in the MD and its regulation of TGF that await a

clearer explanation. For example, a comparison of the microperfusion of 7-nitroindazole into the macula densasegment, the SHR nephron is unresponsive. Similarly,spontaneously hypertensive rat (SHR) and the normo-

tensive Wistar-Kyoto (WKY) control shows enhanced expression of nNOS in the kidney or macula densa ofSprague-Dawley rats is enhanced during salt restrictionnNOS mRNA and protein in the kidney of hypertensive

rat, matched by robust nNOS immunoreactivity in the [19, 24, 25], yet the normal enhancement of TGF re-sponse during microperfusion of L-NMA into the maculamacula densa cells (abstract; Wilcox et al, J Am Soc

Tojo et al: PSD and nNOS in kidney 1391

Fig. 5. (continued)

densa is lost [16, 17, 23]. In both of these circumstances, ula densa cells is altered with salt intake or hypertensionand might thereby contribute to the regulation of nNOSnNOS expression is apparently fully dissociated from

nNOS function. Further studies will be required to deter- activity.The distribution of PSD-93 over in the basolateralmine whether expression or function of PSD-93 in mac-

Tojo et al: PSD and nNOS in kidney1392

Fig. 6. Electron micrograph illustrating double labeling for PSD-93 and nNOS in a macula densa cell. PSD-93 is labeled with 10 nm and nNOSwith 30 nm immunogold. Some of the nNOS labels are colocalized with PSD-93 labels at the cytoplasmic vesicles or basolateral membrane (arrows).Magnification is 333,000.

Table 1. Sites of immunoreactive PSD expression in the rat kidney and their relationship to immunoreactive nNOS expression

PSD-93 PSD-95 nNOS

Subcellular Subcellular SubcellularNephron site Expression location Expression location Expression location

Bowman’s capsule 1 2 1glomerulus 1 1 2PT 2 2 2DL 2 1 AP 2TAL 1 1 AP 2a

MD 1 BL, CV 1 AP 1 C, CV, BLDCT 1 BL, CV 1 AP 2CCD 1 1 AP 2OMCD 1 BL 2 2IMCD 1 BL 2 1Vasculature 2b 2 2

Abbreviations are: PT, proximal tubule; DL, thin descending limb of loop of Henle; TAL, thick ascending limb of loop of Henle; MD, macula densa; DCT, distalconvoluted tubule; CCD, cortical collecting duct; OMCD, outer medullary collecting duct, IMCD, inner medullary collecting duct; BL, basolateral membrane; C,cytoplasm; CV, cytoplasmic vesicle; AP, apical membrane. Site of expression and subcellular location of nNOS are from previous publications [11–13].

a A few cells in TAL show nNOS immunoreactivityb PSD-93 immunoreactivity is positive in the adventitia

Tojo et al: PSD and nNOS in kidney 1393

membrane of cells of the distal convoluted tubule, corti- REFERENCEScal collecting duct, and outer medullary collecting duct 1. Cho KO, Hunt CA, Kennedy MB: The rat brain postsynaptic

density fraction contains a homolog of the Drosophila discs-largesuggests that it may have additional functions, as thesetumor suppressor protein. Neuron 9:929–942, 1992are not prominent sites for nNOS expression. PSDs can

2. Christopherson KS, Bredt DS: Nitric oxide in excitable tissue:cluster ion channels at the cell membrane [2, 9, 10]. In Physiological roles and disease. J Clin Invest 100:2424–2429, 1997

3. Brenman JE, Chao DS, Xia H, Aldape K, Bredt DS: Nitricthe brain, PSD proteins have been implicated in theoxide synthase complexed with dystrophin and absent from skeletalanchoring and activation of K1 channels [9, 10]. The muscle sarcolemma in Duchenne muscular dystrophy. Cell 82:743–

distal nephron segments express disparate K1 channels 752, 19954. Aoki C, Fenstemaker S, Lubin M, Go CG: Nitric oxide synthaseon the apical and basolateral membranes [26]. An in-

in the visual cortex of monocular monkeys as revealed by light andwardly rectifying K1 channel (Kir 2.3) is located on the electron microscopic immunocytochemistry. Brain Res 620:97–113,1993basolateral membrane and regulates membrane poten-

5. Brenman JE, Chao DS, Gee SH, McGee AW, Craven SE, Santil-tial [27]. The expression of PSD-93 on the basolaterallano DR, Wu Z, Huang F, Xia H, Peters MF, Froehner SC,

membrane of the macula densa, outer and inner medul- Bredt DS: Interaction of nitric oxide synthase with the postsynap-tic density protein PSD-95 and a1-syntrophin mediated by PDZlary collecting ducts, and cells of thick ascending limbdomains. Cell 84:757–767, 1996and cortical collecting duct could be related to this K1

6. Brenman JE, Christopherson KS, Craven SE, McGee AW,channel. A ROMK (Kir 1.1) inwardly rectifying ATP- Bredt DS: Cloning and characterization of postsynaptic density

93, a nitric oxide synthase interacting protein. J Neurosci 16:7407–sensitive K1 channel is expressed on the apical mem-7415, 1996brane of cells of thick ascending limb, macula densa, 7. Naisbitt S, Kim E, Weinberg RJ, Roa A, Yang FC, Craig AM,

distal convoluted tubule, cortical collecting duct, and Sheng M: Characterization of guanylate kinase-associated protein,a postsynaptic density protein at excitatory synapses that interactsouter medullary collecting duct [28, 29]. The weak ex-directly with postsynaptic density-95/synapse-associated protein

pression of PSD-95 that was apparent on the apical mem- 90. J Neurosci 17:5687–5696, 19978. Kornau HC, Schenker LT, Kennedy MB, Seeburg PH: Domainbranes of these cells could be related to the this channel.

interaction between NMDA receptor subunits and the postsynapticIn the inner medullary collecting duct, nNOS is ex-density protein PSD-95. Science 269:1737–1740, 1995

pressed diffusely throughout the cytoplasm, whereas 9. Kim E, Niethammer M, Pothschild A, Jan YN, Sheng M: Cluster-ing of Shaker-type K1 channels by direct interaction with a familyPSD-93 is expressed predominantly over the basolateralof membrane-associated guanylate kinases. Nature 378:85–88, 1995membrane cell membrane. Further studies are necessary 10. Cohen NA, Brenman JE, Snyder SH, Bredt DS: Binding of the

to explain the possible interactions of PSD protein with inward rectifier K1 channel Kir 2.3 to PSD-95 is regulated byprotein kinase A phosphorylation. Neuron 17:759–767, 1996nNOS or K1 channels in the kidney.

11. Wilcox CS, Welch WJ, Murad F, Gross SS, Taylor G, Levi R,In conclusion, an isoform of the PSD, PSD-93, is ex- Schmidt HHHW: Nitric oxide synthase in macula densa regulatespressed in cells of the thick ascending limb of the loop glomerular capillary pressure. Proc Natl Acad Sci USA 89:11993–

11997, 1992of Henle, macula densa, distal convoluted tubule, cortical12. Tojo A, Gross SS, Zhang L, Tisher CC, Schmidt HHHW, Wilcox

and medullary collecting ducts, Bowman’s capsule and CS, Madsen KM: Immunocytochemical localization of distinct iso-forms of nitric oxide synthase in the juxtaglomerular apparatus ofglomerular epithelium. It is located prominently overnormal rat kidney. J Am Soc Nephrol 4:1438–1447, 1994the basolateral membrane of the macula densa cells.

13. Bachmann S, Bosse HM, Mundel P: Topography of nitric oxidePSD-93 is colocalized with nNOS in cells of the macula synthesis by localizing constitutive NO synthases in mammalian

kidney. Am J Physiol 268:F885–F898, 1995densa, where it might function to anchor nNOS to the14. Terada Y, Tomita K, Nonoguchi H, Marumo F: Polymerase chainbasolateral membrane and internal organelles. The other

reaction localization of constitutive nitric oxide synthase and solu-isoform, PSD-95, is less strongly expressed. There is faint ble guanylate cyclase messenger RNAs in microdissected rat neph-

ron segments. J Clin Invest 90:659–665, 1992immunostaining of the apical membrane of cells of the15. Tojo A, Welch WJ, Bremer V, Kimoto M, Kimura K, Omata M,thick ascending limb of the loop of Henle, distal convo- Ogawa T, Vallance P, Wilcox CS: Colocalization of demethylat-

luted tubule, macula densa, and cortical collecting duct. ing enzymes and NOS and functional effects of methylargininesin rat kidney. Kidney Int 52:1593–1601, 1997These data suggest an interaction in the kidney between

16. Wilcox CS, Welch WJ: TGF and nitric oxide: Effects of salt intakePSD-93 and nNOS, at least in the macula densa, and and salt-sensitive hypertension. Kidney Int 49(Suppl 55):S9–S13,raise the possibility that this PSD protein family may 1996

17. Welch WJ, Wilcox CS: Macula densa arginine delivery and uptakehave a wider function in relationship to ion channels.in the rat regulates glomerular capillary pressure: Effects of saltintake. J Clin Invest 100:2235–2242, 1997

18. He XR, Greenberg SG, Briggs JP, Schnermann JB: Effect ofACKNOWLEDGMENTSnitric oxide on renin secretion. II. Studies in the perfused juxta-

This work was supported by grants from the NIDDK (DK49870 glomerular apparatus. Am J Physiol 268:F953–F959, 1995and DK36079) and by funds from the George E. Schreiner Chair of 19. Tojo A, Madsen KM, Wilcox CS: Expression of immunoreactiveNephrology. nitric oxide synthase isoforms in rat kidney: Effect of dietary salt

and losartan. Jpn Heart J 36:389–398, 1995Reprint requests to Dr. Christopher S. Wilcox, Division of Nephrol- 20. Bremer V, Tojo A, Kimura K, Hirata Y, Goto A, Nagamatsu

ogy and Hypertension, Georgetown University Medical Center, 3800 T, Suzuki Y, Omata M: Role of nitric oxide in rat nephrotoxicReservoir Road N.W., PHC F6003, Washington D.C. 20007, USA. nephritis: Comparison between inducible and constitutive nitric

oxide synthase. J Am Soc Nephrol 8:1712–1721, 1997E-mail: wilcoxch@gunet.georgetown.edu

Tojo et al: PSD and nNOS in kidney1394

21. Kobzik L, Reid MB, Bredt DS, Stamler JS: Nitric oxide in skeletal of renin and brain-type nitric-oxide-synthase (bNOS) mRNA lev-els in rat kidneys. Pflugers Arch Eur J Physiol 432:394–400, 1996muscle. Nature 372:546–548, 1994

26. Hirsch J, Schlatter E: K1 channels in the basolateral membrane22. Wu K, Xu JL, Suen PC, Huang YY, Mount HT: Nitric oxideof rat cortical collecting duct. Kidney Int 48:1036–1046, 1995increases calcium/calmodulin-dependent phosphorylation of pro-

27. Le Maout S, Brejon M, Olsen O, Merot J, Welling PA: Basolat-teins in the postsynaptic density of adult rat cerebral cortex. Braineral membrane targeting of a renal-epithelial inwardly rectifyingRes Mol Brain Res 40:22–26, 1996potassium channel from the cortical collecting duct, CCD-IRK3,23. Welch WJ, Wilcox CS: Tubuloglomerular feedback and maculain MDCK cells. Proc Natl Acad Sci USA 94:13329–13334, 1997densa-derived NO, in Nitric Oxide and the Kidney: Physiology 28. Boim MA, Ho K, Shuck ME, Bienkowski MJ, Block JH, Slightom

and Pathology, edited by Goligorsky MS, Gross SS, New York, JL, Yang Y, Brenner BM, Hebert SC: ROMK inwardly rectifyingChapman & Hall, 1997, pp 216–232 ATP-sensitive K1 channel II: Cloning and distribution of alterna-

24. Singh I, Grams M, Wang WH, Yang T, Killen P, Smart A, tive forms. Am J Physiol 268:F1132–F1140, 1995Schnermann J, Briggs JP: Coordinate regulation of renal expres- 29. Xu JZ, Hall AE, Peterson LN, Bienkowski MJ, Eessalu TE,sion of nitric oxide synthase, renin, and angiotensinogen mRNA Hebert SC: Localization of the ROMK protein on the apical mem-by dietary salt. Am J Physiol 270:F1027–F1037, 1996 brane of rat kidney nephron segments. Am J Physiol 273:F739–

F748, 199725. Schricker K, Potzl B, Hamann M, Kurtz A: Coordinate changes