nitric oxide enhances paracellular permeability of opossum kidney cells

Kidney International, Vol. 55 (1999), pp. 2215–2223

Nitric oxide enhances paracellular permeability of opossumkidney cells

MINGYU LIANG and FRANKLYN G. KNOX

Departments of Medicine and Physiology and Biophysics, Mayo Clinic and Foundation, Rochester, Minnesota, USA

Nitric oxide enhances paracellular permeability of opossum solute transport in the proximal tubule is not yet clearkidney cells. [1], it is known that the paracellular pathway of the

Background. Nitric oxide (NO) has been shown to be a proximal tubule is unusually leaky compared with otherparacrine/autocrine regulator of proximal tubular transport. Inepithelial tissues [3–5]. This unusual leakiness could po-this study, we investigated the effect of NO on the paracellulartentially facilitate the flux of fluid and solutes across thepermeability of opossum kidney (OK) cells, a proximal tubule

cell line that possesses a leaky paracellular pathway resembling tight junctions. Importantly, studies in our laboratorythat of the in vivo proximal tubule. and those of others suggest that the paracellular trans-

Methods. Paracellular permeability of OK cells cultured on port in the proximal tubule is dynamically regulatedpermeable supports was measured as the apparent paracellular[6–9]. Furthermore, a recent study by Kwon et al showedpermeability coefficient (Papp) for 3[H]-D-mannitol. Changes inthat the paracellular flux in the proximal tubule wascell viability, cellular adenosine triphosphate (ATP) content,

cGMP levels, and lipid peroxidation were assessed. substantially enhanced in human postischemic renal allo-Results. Incubation with 2 mm sodium nitroprusside (SNP), grafts with sustained acute renal failure [10]. These stud-

an NO donor, for 24 hours significantly enhanced the Papp of ies suggest that the paracellular pathway may play anOK cell sheets by 30.6 6 7.0% (N 5 8, P , 0.05). This effectimportant role in proximal tubular transport underwas largely blunted by hemoglobin, a NO scavenger. Cell via-pathophysiological as well as physiological conditions.bility was not compromised. This effect of SNP was concomi-

tant with a moderate reduction of cellular ATP content, an Nitric oxide (NO) is an important biological moleculeincrease in lipid peroxidation, and an increase in cellular cGMP with a wide variety of functions. It is synthesized fromlevels. The antioxidant superoxide dismutase (SOD) signifi- l-arginine by NO synthase (NOS), including endothelialcantly attenuated the effect of SNP on cellular ATP content

NOS (eNOS), neuroneal NOS (nNOS), and inducibleand blunted the increase in Papp caused by SNP. A solubleNOS (iNOS), in many tissues. It has been shown thatguanylate cyclase inhibitor did not affect the effect of SNP on

the Papp. the proximal tubule contains mRNAs encoding iNOSConclusions. NO enhances the paracellular permeability of [11] and perhaps eNOS [12], although no detection of

OK cells possibly via mechanisms involving decreases in cellu- the basal expression of NOS proteins in the proximallar ATP content.tubule has been reported. The existence of iNOS mRNAin the kidney, particularly in the proximal tubule, sug-gests that a large amount of NO could be locally pro-In addition to the much studied transcellular pathway,duced. The existence of NO in the renal cortex was alsothe fluid and solute transport across the proximal tubuleconfirmed by microdialysis studies [13]. According to inepithelium can also take place via the paracellular path-vivo microperfusion studies [14, 15], the direct effect ofway [1, 2]. The paracellular pathway of the proximalsodium nitroprusside (SNP), an NO donor, on the fluidtubule, as that of other epithelial and endothelial tissues,and solute transport in the proximal tubule appears tois gated by tight junctions located at the apical end ofbe dose dependent: A lower dose (1 mm) stimulates thethe paracellular space. Although the quantitative impor-proximal tubule reabsorption, whereas a higher dose (0.1tance of the paracellular pathway in the total fluid andto 1.0 mm) inhibits the proximal tubule reabsorption. Inproximal tubule cell culture studies, NO (1 mm SNP orcytokine-induced NO production) was shown to inhibitKey words: proximal tubule, transport, tight junctions, cytotoxicity,

antioxidants. Na1/H1 exchangers [16] and Na1,K1-ATPase activities[17]; both are important components of the transcellularReceived for publication September 21, 1998transport in the proximal tubule. However, it is notand in revised form December 11, 1999

Accepted for publication January 20, 1999 known whether and how NO affects the paracellulartransport in the proximal tubule. 1999 by the International Society of Nephrology

2215

Liang and Knox: NO enhances paracellular permeability of OK cells2216

The flux of fluid and solutes across the paracellular and HEPES 10, pH 7.4. Cell sheets on Transwell insertspathway is primarily a passive process of which the major were washed with the Na1-free medium. One-half millili-intrinsic determinant is the paracellular permeability. ter and 1.5 ml of the Na1-free medium were then addedIn previous studies, we characterized the paracellular to the apical chamber and basal chamber, respectively.permeability of opossum kidney (OK) cells, a proximal One mCi/ml of 3[H]-D-mannitol was included in the api-tubule cell line, cultured on permeable supports (ab- cal chamber that was thereafter designated as the donorstract; Liang et al, FASEB J 12:A420, 1998). We found chamber, and the basal chamber was designated as thethat OK cell sheets possessed relatively high paracellular receiver chamber. After incubation for 10 minutes, dupli-permeability similar to that of the in vivo proximal tubule cate samples were taken from the donor and the receiver[18]. In this study, OK cell sheets cultured on permeable chambers, and the radioactivity was counted using a scin-supports were used as an in vitro proximal tubule epithe- tillation counter (Beckman LS 6000SC; Beckman Instru-lium model to investigate the effect of NO on the paracel- ments, Fullerton, CA, USA). The time of incubation waslular permeability. We demonstrated that NO enhances confirmed in a preliminary study to be within the initialthe paracellular permeability of the OK proximal tubule linear range for 3[H]-D-mannitol flux. The Papp was thencell line via mechanisms possibly involving decreases in calculated as Papp(cm/second) 5 [CR/CD/t(min)] 3cellular adenosine triphosphate (ATP) content. VR(ml)/[A(cm2) 3 60], where CR was the final 3[H]-D-

mannitol concentration in the receiver chamber, CD wasthe initial 3[H]-D-mannitol concentration in the donorMETHODSchamber, t was the incubation time, VR was the volumeCell cultureof transport medium in the receiver chamber, and A was

Opossum kidney cells were originally obtained from the surface area of cell sheet that was 1 cm2 in our study.the American Type Culture Collection (ATCC, Rockville, Papp calculated this way has the advantage of being inde-MD, USA). The cell line was maintained in a 1:1 mixture pendent of the concentration gradient, time of measure-of Dulbecco’s modified Eagle’s medium (DMEM) and ment, chamber volume, and surface area. All solutionsHam’s F-12 nutrient mixture supplemented with 10% were prewarmed to 378C, and the incubation was per-fetal bovine serum, 100 U/ml penicillin, and 100 mg/ml formed in a 378C incubator.streptomycin. Cells were incubated in 5% CO2/95% air All of the experiments were performed in parallel withat 378C and were subcultured when they reached conflu- controls, and when the effects of various compoundsence by incubation with 0.25% trypsin/0.02% ethylenedi- other than the NO donor were tested, treatment withaminetetraacetic acid (EDTA). For paracellular trans- the NO donor alone was also performed in parallel asport studies, confluent cells were trypsinized as described

“positive controls.” The results were expressed as per-earlier in this article, and were seeded on the inserts

centages of the simultaneously performed controls. The(polyester, 0.4 mm pore size, 1 cm2 growth area) of thecontrol Papp of OK cell sheets for 3[H]-D-mannitol atpermeable Transwell supports. The paracellular perme-378C was approximately 13.70 3 1026 cm/second.ability was assessed after it reached the stable level. OK

cells were used between passages 74 and 91. Adenosine triphosphate assay

The cellular ATP level was measured as previouslyApparent paracellular permeability coefficientdescribed [21] with minor modifications. After treat-(Papp) measurementments, cell sheets were quickly washed with ice-coldOpossum kidney cell sheets were previously deter-PBSA [phosphate-buffered saline (PBS) without Ca21

mined to have a transepithelial electrical resistanceand Mg21]. Cells were collected in distilled water. A(TER) at the level of 10 Vcm2 [19]. Because it has beenportion of the cell suspension was combined with a solu-pointed out that TER at such a low level might no longertion containing trichloroacetic acid (TCA) and EDTAbe reliable as a measurement of paracellular permeabil-(final concentrations 2% and 2 mm, respectively). Theity [20], we chose Papp for 3[H]-D-mannitol as the mea-mixture was briefly sonicated and centrifuged at 4000 3 gsurement of paracellular permeability in our study. D-for five minutes. The supernatant was diluted in 0.1 mmannitol is a hydrophilic and neutral compound thatTris Cl (pH 7.4) and 2 mm EDTA. An aliquot of thedoes not bind to or go across the cell membrane. Becausediluted sample was combined with the HEPES buffera potential influence of the transcellular transport on(containing 25 mm HEPES, 10 mm MgCl2, and 0.02%Papp measurement was noticed in our previous studiesNaN3) and the luciferin-luciferase solution from an ATP(abstract; Liang et al, ibid), a Na1-free medium wasassay kit (Calbiochem, La Jolla, CA, USA). The lightused in the Papp measurement to minimize the potentialemitted was measured with a luminometer (TD-20/20;influence of transcellular transport. The Na1-free me-Turner Designs, Sunnyvale, CA, USA). ATP-free aciddium contained (in mm) choline·Cl 112, KCl 3.5, KH2PO4

1, choline·HCO3 25, CaCl2 1.8, MgSO4 0.5, D-glucose 5, was used as a standard.

Liang and Knox: NO enhances paracellular permeability of OK cells 2217

Lipid peroxidation assay

Lipid peroxidation was assessed by measuring the thi-obarbituric acid reaction substance (TBARS) as pre-viously described [22]. After treatments, the cells werewashed twice with ice-cold PBSA and were collected indistilled water. An aliquot of the sample was added to amixture of 8.1% sodium dodecyl sulfate, 0.8% 2-thio-barbituric acid (TBA), and 20% acetic acid (2:15:15, vol/vol/vol) and was boiled for 60 minutes. The solutionwas cooled to room temperature and extracted with amixture of N-butanol and pyridine (15:1, vol/vol). Theorganic portion was then read at 532 nm with a spectro-photometer. Malonaldehyde bis[dimethyl acetal] wasused as standards.

Fig. 1. Sodium nitroprusside (SNP) time- and concentration-depen-Lactate dehydrogenase assaydently enhanced the paracellular permeability coefficient (Papp) of opos-

After treatments, the culture media in the apical and sum kidney (OK) cell sheets. OK cell sheets cultured on permeablesupports were incubated with culture medium alone (time 0) or withthe basal chambers were collected. The lactate dehydro-culture medium containing SNP at the indicated concentrations for 1,genase (LDH) activity in the medium was measured by2, 6, or 24 hours. The Papp for 3[H]-D-mannitol was measured as described

using an LDH assay kit from Sigma Chemical Co. (St. in the Methods section. Symbols are: (d) SNP, 0.5 mm; (s) SNP, 2mm; (.) SNP, 5 mm; (dashed line), reference line. N $ 6 for each timeLouis, MO, USA). The cell sheets were collected, brieflypoint; *P , 0.05 m vs. culture medium alone (time 0).sonicated, and used for the total cellular LDH activity

measurement. The LDH released into the medium wasexpressed as a percentage of the total cellular LDH ac-tivity. Statistics

The data are shown as means 6 sem. The N valuesNO22 assay

that are shown represent the numbers of cell sheets di-NO22 accumulation in the culture medium was mea-

vided into several experiments with each one done insured with the Griess reaction. An aliquot of the sampleduplicate or triplicate. The unpaired t-test was used whenwas mixed with the Griess reagent containing 1% (wt/comparing two groups. When comparing multiple groups,vol) sulfanilamide, 0.1% (wt/vol) naphthylethylenedia-one-way analysis of variance was performed, followedmine di-HCl, and 5% (vol/vol) phosphoric acid (85%).by Dunnett’s or Student-Newman-Keuls test. A P valueThe mixture was incubated at room temperature for 10of less than 0.05 was considered significant.minutes and read at 570 nm with a microplate reader.

NaNO2 was used as standards.

RESULTScGMP assayEffect of sodium nitroprusside on the Papp of opossumAfter the medium was removed, the cells were quicklykidney cellswashed with ice-cold PBSA and collected in 10% TCA.

Sodium nitroprusside, which spontaneously releasesAfter brief sonication, the suspension was centrifugedNO, was used as a NO donor in this study. Figure 1at 4000 3 g for five minutes. The supernatant was trans-depicts the time- and concentration-dependent effect offerred to another tube. TCA in the supernatant wasSNP on the paracellular permeability of OK cell sheetsextracted with water-saturated ether. After acetylation,cultured on permeable Transwell supports. At the con-the cGMP was measured using an enzyme immunoassaycentration of 0.5 mm, SNP did not significantly alter thekit (Cayman Chemical, Ann Arbor, MI, USA). The pro-Papp of the OK cell sheets. The Papp was significantlytein concentration was measured with a Bio-Rad proteinenhanced by 20.6 6 5.8% after the cells were incubatedassay kit (Bradford method) using bovine serum albuminwith 2 mm SNP for 24 hours (N 5 6, P , 0.05 vs. control).as standards.This was concomitant with the accumulation of NO2

2,Materials one of the NO metabolites, in the culture medium, which

at the end of the incubation reached 38.50 6 1.33 mm in3[H]-D-mannitol was purchased from Sigma Chemicalthe apical chamber and 34.77 6 0.80 mm in the basalCo. Permeable Transwell supports and other plastic cellchamber (N 5 6). At 5 mm, SNP enhanced the Papp to aculture supplies were from the Corning Costar (Cam-greater extent (increased by 38.3 6 6.8% after 24 hr ofbridge, MA, USA). The cell culture medium and other

reagents were from Sigma. incubation, N 5 6, P , 0.05 vs. control). Furthermore,


Fig. 2. The enhancing effect of sodium nitroprusside (SNP) on the Papp Fig. 3. Incubation with 2 mM sodium nitroprusside (SNP) for 24 hourswas blunted by hemoglobin. Opossum kidney (OK) cell sheets cultureddid not increase the lactate dehydrogenase (LDH) release. Opossumon permeable supports were incubated with culture medium alonekidney (OK) cell sheets were incubated with culture medium (control),(control, only the error bar is visible), culture medium containing 2 mmculture medium containing 0.5 mm, 2 mm, or 5 mm SNP for 24 hours.SNP, or culture medium containing 2 mm SNP plus 3.2 mg/ml hemoglo-Activities of LDH released into the apical and basal chambers werebin (Hb) for 24 hours. The Papp for 3[H]-D-mannitol was measured asmeasured, combined, and expressed as percentages of the total cellulardescribed in the Methods section. N 5 8; *P , 0.05 vs. control; #P ,LDH activity (N 5 6; *P , 0.05 vs. control).0.05 vs. SNP alone.

Cell viabilitythe effect of 5 mm SNP was significant after a shorter The integrity of cell sheet is a prerequisite for mean-

ingful paracellular permeability measurements. Exces-incubation time (increased by 15.6 6 4.8% after six hrsive NO has been shown in many cases to be cytotoxic.of incubation, N 5 6, P , 0.05 vs. control), which wasTherefore, it is critical to rule out the impairment of cellconcomitant with the accumulation of 26.16 6 0.89 mmviability by SNP, which would compromise the cell sheetNO2

2 in the apical chamber.integrity and result in false increases in paracellular per-Sodium nitroprusside, once dissolved in solution, re-meability. This was achieved in this study by using two

leases other substances such as CN2 in addition to NO. approaches. First, the release of LDH from the cells intoTherefore, it is important to confirm that the observed the medium was measured as an indicator of cytotoxicity.effect of SNP was mediated by NO. This was achieved As shown in Figure 3, incubation with either 0.5 mm orin this study by testing the combined effect of SNP and 2 mm SNP for 24 hours did not change LDH release,hemoglobin, a NO scavenger. As shown in Figure 2, whereas incubation with 5 mm SNP caused a significant

increase in LDH release from 4.17 6 0.37% to 8.29 6incubation with 2 mm SNP alone for 24 hours enhanced0.90% (N 5 6, P , 0.05). This indicates that incubationthe Papp by 30.6 6 7.0% (N 5 8, P , 0.05 vs. control).with either 0.5 mm or 2 mm SNP for 24 hours did notParallel coincubation with 2 mm SNP and 3.2 mg/mlhave significant cytotoxicity, whereas 5 mm SNP did.hemoglobin for 24 hours resulted in a Papp increase ofSecond, the increase of the Papp caused by incubation of11.1 6 5.0% (N 5 8, P , 0.05 vs. either SNP alone or 2 mm SNP for 24 hours was completely reversed within

control), which was 36% of that caused by SNP alone. 24 hours after returning the cell sheets to the cultureHemoglobin (3.2 mg/ml) alone did not significantly affect medium without SNP (Fig. 4). In contrast, the Papp wasthe Papp. The fact that hemoglobin did not completely unable to recover when the concentration of SNP wasabolish the effect of SNP might suggest possible roles of increased to 5 mm (Papp increased by 38.3 6 6.8% after 24

hr of treatment and 40.8 6 3.0% after 24 hr of recovery,other substances released from SNP. Although we areN 5 6, both P , 0.05 vs. control). This is in agreementnot able to rule out this possibility completely based onwith the LDH release data and further supports that thethese data, it is worth mentioning that the NO scavengingOK cell viability was not significantly compromised afterability of hemoglobin may not be sustained for the entire24 hours of incubation with 2 mm SNP.

course of the relatively long incubation time involved inthis study. In any case, the enhancing effect of SNP Mechanisms for the enhancing effect of sodium

nitroprusside on the Papp of opossum kidney cellson the Papp was largely blunted by hemoglobin, whichindicated that at least the major component of this effect cGMP is one of the most common mediators of NO

effects. It is natural to speculate that cGMP might alsoof SNP was mediated by NO.


Table 2. Effects of sodium nitroprusside (SNP) on paracellularpermeability, cellular adenosine triphosphate (ATP) content,

and lipid peroxidationa

CellularPapp (N $ 6) ATP (N 5 6) TBARS (N 5 6)

Treatment % of control

Control 100.061.4 100.0 67.0 100.0 611.7SNP 0.5 mm 99.667.5 94.6 68.4 215.6636.7b

SNP 2 mm 120.064.8b,c 41.765.3b,c 400.7654.5b,c

SNP 5 mm 143.165.9b,c,d 16.562.3b,c,d 395.0623.3b,c

a Opossum kidney cell sheets were incubated with SNP at the indicated concen-trations for 24 hours. The average absolute values of cellular ATP content andthiobarbituric acid reaction substance (TBARS) for control group were 3.60nmol/mg protein and 3.64 nmol/mg protein, respectively. Similar differencesbetween groups were observed when expressing these values as nmol/cell sheet.

b P , 0.05 vs. controlc P , 0.05 vs. SNP 0.5 mmFig. 4. The change of Papp caused by 2 mM sodium nitroprusside (SNP)d P , 0.05 vs. SNP 2 mmwas fully recovered within 24 hours. Opossum kidney (OK) cell sheets

were incubated with culture medium alone or culture medium con-taining 2 mm SNP for 24 hours. The Papp was measured and shown as“treatment.” The cell sheets were then returned to culture medium foranother 24 hours, and the Papp was measured again and shown as a “24-

SNP (Table 1) makes it difficult to rule out the role ofhour recovery.” The first bar (only the error bar is visible) of eachgroup represents the control cells. The second bar of each group repre- cGMP completely.sents cells treated with SNP (N 5 6; *P , 0.05 vs. control). It has been shown in several types of epithelial cells

that depletion of cellular ATP causes dramatic increasesin paracellular permeability. Therefore, we examinedwhether SNP caused changes in the cellular ATP contentTable 1. Effect of sodium nitroprusside (SNP) on paracellular

permeability may not be mediated by cGMPa in OK cell sheets. As shown in Table 2, incubation with0.5 mm SNP for 24 hours did not significantly alter eitherPapp Cellular cGMP

Treatment % of control, N 5 8 % of control, N 5 3 the Papp or the cellular ATP content. When the concentra-tion of SNP was increased to 2 mm, the Papp significantlyControl 100.061.1 100.066.3

SNP 130.166.3b 209.069.9b increased by 20.0 6 4.8%. In the meantime, the cellularSNP 1 ODQ 126.468.2b 150.764.6b,c

ATP content was significantly decreased to 41.7 6 5.3%a Opossum kidney cell sheets were incubated with 2mm SNP or 2 mm SNP 1 of control (N 5 6, P , 0.05 vs. either control or 0.530 mm ODQ for 24 hours. The average absolute value of cellular cGMP content

was 1.01 pmol/mg protein. Abbreviations are: cGMP, cyclic GMP; Papp, apparent mm). Increasing the SNP concentration to 5 mm furtherparacellular permeability coefficient for 3[H]-D-mannitol; ODQ, 1H-[1,2,4] oxa- decreased the cellular ATP content to 16.5 6 2.3% ofdiazolo-[4,3-a]quinoxalin-1-one.

b P , 0.05 vs. control control (N 5 6, P , 0.05 vs. either control, 0.5 mm, orc P , 0.05 vs. SNP alone 2 mm) and caused a significantly larger increase in the Papp

of 43.1 6 5.9%. Therefore, it appears that the increases inPapp correlate with the decreases in cellular ATP content.

Table 2 also showed the changes in lipid peroxidationmediate the enhancing effect of NO on the Papp of OKcaused by SNP. SNP caused substantial increases in lipidcell sheets. This does not seem to be the case. As shownperoxidation in OK cell sheets, an effect of excessivein Table 1, 24-hour incubation with 2 mm SNP did sig-NO that has also been observed in many other cells andnificantly increase the cellular cGMP level to 209 6 9.9%tissues. Different from the changes in Papp and cellularof control (N 5 3, P , 0.05 vs. control). When theATP content, the increase in lipid peroxidation was sig-increase of cGMP was significantly blunted by the coin-nificant at a 0.5 mm concentration, whereas 5 mm didcubation with a selective soluble guanylate cyclase inhibi-not cause a further increase in lipid peroxidation com-tor, 1H-[1,2,4]oxadiazolo-[4,3-a]quinoxalin-1-one (ODQ),pared with 2 mm.at 30 mm, however, the increase in Papp was unaltered.

It is known that excessive NO can react with superox-An acute stimulation of OK cells by SNP for 10 minuteside anion (O2

2), lead to the formation of several toxiccaused a slightly greater increase in the cellular cGMPfree radicals, and cause oxidative injury in biologicallevel to approximately 2.5-fold of the control. At thismembranes and other targets. Indeed, these data didtime point, 30 mm ODQ completely abolished the in-suggest the existence of oxidative challenges resultingcrease of cGMP caused by SNP. These data suggest thatfrom the SNP treatment. Therefore, in the next step, wecGMP is probably not the mediator of the enhancingtested whether antioxidants could alter the effect of SNPeffect of SNP on the Papp of OK cell sheets. However, theon the Papp of OK cell sheets. As shown in Figure 5A, bothfact that ODQ did not completely abolish the increase of

cellular cGMP level caused by 24-hour incubation with O22 scavenging enzyme superoxide dismutase (SOD; 300


Fig. 5. Antioxidants blunted the enhancing effect of sodium nitroprus-side (SNP) on the Papp of opossum kidney (OK) cell sheets. OK cellsheets were incubated with culture medium alone (control, only theerror bar is visible), culture medium containing 2 mm SNP, 2 mm SNPplus 300 U/ml superoxide dismutase (SOD), or 2 mm SNP plus 500 mg/ml vitamin E for 24 hours. (A) Effect of SOD and vitamin E on theincrease of Papp caused by SNP (N 5 4 to 12). (B) Effect of SOD andvitamin E on the reduction of cellular adenosine triphosphate (ATP)content caused by SNP (N 5 6). (C) Effect of SOD and vitamin E onthe lipid peroxidation caused by SNP. TBARS is thiobarbituric acidreaction substance. N 5 6; *P , 0.05 vs. control; #P , 0.05 vs. SNPalone.

U/ml) and lipid soluble antioxidant vitamin E (500 mg/ the impact of the relatively long incubation time shouldbe kept in mind. There was a slight tendency for theml) significantly blunted the enhancing effect of SNP on

the Papp. The increase in Papp caused by 2 mm SNP (24- lipid peroxidation to decrease when SOD was present,hr incubation) was attenuated by approximately 50% by but this did not reach statistical significance.the coincubation with either SOD or vitamin E. The Papp

in the presence of SNP plus SOD or SNP plus vitaminDISCUSSIONE was still significantly higher than control. In the mean-

In this study, we demonstrated that exogenous NOtime, SOD also significantly attenuated the reducing ef-reduced the cellular ATP content and enhanced the par-fect of SNP on the cellular ATP content (Fig. 5B), raisingacellular permeability of OK proximal tubule cells. Anti-the cellular ATP content from 41.7 6 5.3% of controloxidants ameliorated the decrease in cellular ATP and(N 5 6) to 62.0 6 6.4% of control (N 5 6, P , 0.05 vs.blunted the increase in paracellular permeability.either control or SNP alone). Vitamin E also tended to

Nitric oxide has been shown to be involved in theattenuate the reducing effect of SNP on the cellular ATPregulation of tight junctions, the structure controllingcontent. Although this effect of vitamin E did not reachparacellular permeability, and actin cytoskeleton, thestatistical significance, the cellular ATP was in fact raisedstructure closely related to tight junction regulation. Into an extent (58.9 6 7.4%) rather close to that causedisolated rat hepatocyte couplets, the NO donor SNP wasby SOD (Fig. 5B). As expected, vitamin E, a lipid solubleshown to potentiate the apical cytoskeleton contractionantioxidant, significantly attenuated the lipid peroxida-and increase the tight junction permeability [23]. Intion caused by SNP (Fig. 5C), although it did not com-

pletely abolish the increase in lipid peroxidation. Again, Caco-2BBe intestinal epithelial cells, NO donors were


shown to dilate tight junctions and deplete cellular ATP causes increases in tight junction permeability. In fact,ATP depletion has been widely used as a maneuver to[21]. NO donors were also shown to potentiate the en-

hancing effect of H2O2 on the paracellular permeability investigate the regulation of tight junctions in numerousstudies [27–29]. In the ATP depletion model, the cellularof endothelial cells [24] and alter the cytoskeleton ar-

rangement in glomerular epithelial cells [25]. In this ATP content is usually rapidly decreased to less than 5to 10% of the basal level. In our study, the cellular ATPstudy, we showed that the NO donor SNP significantly

enhanced the paracellular permeability of OK cell was reduced to approximately 40 (by 2 mm SNP) or 60%(by SNP plus SOD or vitamin E) of control, and thissheets. As characterized in our previous study, this cell

line, commonly used as a proximal tubule cell line in moderate reduction of cellular ATP was accompaniedby significant increases in paracellular permeability. Thisgeneral, also possesses a leaky paracellular pathway re-

sembling that of the in vivo proximal tubule (abstract; is consistent with the study by Salzman et al, in which asimilar decrease of cellular ATP content in Caco-2BBeLiang et al, ibid) [18]. This effect of SNP on the paracellu-

lar permeability of OK cell sheets was mediated by NO cells was accompanied by significant increase in tightjunction permeability [21]. Several studies have shownreleased because it was largely blunted by the NO scav-

enger hemoglobin. The LDH measurements and the re- that NO inhibits enzymes in the tricarboxylic acid cycleand the mitochondrial respiration chain, which are re-versibility study confirmed that the increase of paracellu-

lar permeability caused by SNP was not a false increase in quired for effective ATP production [30–32]. This couldexplain the reduction of cellular ATP content by NOpermeability resulting from impairment of cell viability.

The increase in paracellular permeability of OK cell observed in this study.In this study, the enhancing effect of SNP on the Pappsheets caused by NO was accompanied by several bio-

chemical alterations, including a reduction in cellular was observed only when SNP was given at mm concentra-tions and when the incubation time was longer thanATP content, an increase in lipid peroxidation, and an

accumulation of cGMP. Of these changes, the reduction several hours. This suggests that the enhancing effect ofNO on the proximal tubule paracellular permeabilityin cellular ATP content appears to be the mechanism

for the enhancing effect of NO on the paracellular per- may exist only when there is prolonged overproductionof NO. As mentioned earlier in this study, the kidney,meability. This is supported by several facts. First, the

dose-dependent increases in the Papp consistently corre- including the proximal tubule itself, contains iNOSmRNA [11], which is capable of producing a large quan-lated with the dose-dependent decreases in cellular ATP

content (Table 2). Second, SOD significantly blunted tity of NO. In fact, iNOS activity and NO production inthe kidney have been shown to increase substantiallythe increase of Papp by NO, an effect accompanied by a

significant attenuation of the decrease of cellular ATP under a variety of conditions. These include hypoxia/ischemia [26, 33], hypertension [34], anemia [35], in-content, but not accompanied by significant attenuation

of the increase of lipid peroxidation (Fig. 5). Third, vita- flammation [36, 37], stimulation with cytokines [38], stim-ulation with atrial natriuretic factor or angiotensin IImin E blunted the increase of Papp to a similar degree as

SOD did, although it attenuated the lipid peroxidation [39], etc. Under these conditions, the NO productionwas increased by several to more than 20-fold. Becauseto a much larger degree. Fourth, attenuating the cGMP

accumulation by ODQ did not blunt the increase of Papp. the basal renal cortical interstitial NO22 concentration

is at the mm level [13], the dose of NO used in this studyHowever, a closer inspection of the data revealed thatODQ did not completely abolish the accumulation of (resulting in the accumulation of approximately 30 mm

NO22) is reasonably obtainable in in vivo kidneys whencGMP. Moreover, the blunting effects of vitamin E on

the increases of Papp and lipid peroxidation were statisti- iNOS is up-regulated.The overproduction of NO in the kidney is thoughtcally significant, although its attenuating effect on cellu-

lar ATP reduction was not. Therefore, although the re- to be involved in diseases such as acute renal failure [40].A study by Noiri et al showed that in vivo targetingduction of cellular ATP content is more likely to be the

mechanism for the enhancing effect of NO on the Papp, of iNOS with oligodeoxynucleotides protected the ratkidney against ischemia [41], suggesting that the overpro-the roles of lipid peroxidation and cGMP cannot be

completely ruled out at this stage. Enhanced protein duction of NO by iNOS may play a pathogenesis role inkidney ischemic injury. Chiao et al also reported thatnitrosylation, concomitant with increases in iNOS ex-

pression, has been observed in the outer medulla of a-melanocyte–stimulating hormone inhibited iNOS in-duction in kidneys and protected against renal injurypostischemic rodent kidneys [26]. Whether the genera-

tion of peroxynitrite from NO and O22 and the subse- after ischemia in mice and rats [26]. Moreover, a recent

study by Kwon et al showed that the paracellular perme-quent protein nitrosylation are involved in the enhancingeffect of NO on the Papp of OK cells remains to be an- ability of the proximal tubule was substantially enhanced

in human postischemic renal allografts with sustainedswered.It is well established that the depletion of cellular ATP acute renal failure [10]. Putting these results and those


5. Seely JF: Variation in electrical resistance along length of ratof this study together, it seems reasonable to speculateproximal convoluted tubule. Am J Physiol 225:48–57, 1973

that the overproduction of NO could causally contribute 6. Ramsey CR, Berndt TJ, Knox FG: Effect of volume expansionto the increase in the paracellular permeability of the on the paracellular flux of lanthanum in the proximal tubule.

J Am Soc Nephrol 9:1147–1152, 1998proximal tubule in postischemic kidneys with sustained7. Jacobson HR: Altered permeability in the proximal tubule re-

acute renal failure. However, although the paracellular sponse to cyclic AMP. Am J Physiol 236:F71–F79, 1979permeability of OK cells resembles that of the in vivo 8. Grandchamp A, Boulpaep EL: Pressure control of sodium reab-

sorption and intercellular backflux across proximal kidney tubule.proximal tubule, precautions should be taken when ex-J Clin Invest 54:69–82, 1974trapolating these in vitro results to the in vivo proximal 9. Humbert F, Grandchamp A, Pricam C, Perrelet A, Orci L:

tubule. Morphological changes in tight junctions of Necturus maculosusproximal tubules undergoing saline diuresis. J Cell Biol 69:90–96,The relative increase in the Papp of OK cells observed1976in this study is modest compared with the relative 10. Kwon O, Nelson WJ, Sibley R, Huie P, Scandling JD, Dafoe

changes in paracellular permeability generally observed D, Alfrey E, Myers BD: Backleak, tight junctions, and cell-celladhesion in postischemic injury to the renal allograft. J Clin Investin other cell types [21, 27, 28]. However, it is noteworthy101:2054–2064, 1998that the basal Papp of OK cells is much higher than those 11. Ahn KY, Mohaupt MG, Madsen KM, Kone BC: In situ hybridiza-

other cell types. Furthermore, the change of Papp ob- tion localization of mRNA encoding inducible nitric oxide synthasein rat kidney. Am J Physiol 267:F748–F757, 1994served in this study is, in fact, similar to the changes

12. Ujiie K, Yuen J, Hogarth L, Danziger R, Star RA: Localizationobserved in in vivo proximal tubules under various con- and regulation of endothelial NO synthase mRNA expression inditions [6, 7, 10], which tends to support the relevance rat kidney. Am J Physiol 267:F296–F302, 1994

13. Zou A, Cowley AW: Nitric oxide in renal cortex and medulla.of the data from the present study.Hypertension 29:194–198, 1997In summary, we demonstrated that NO enhanced the 14. Wang T: Nitric oxide regulates HCO3

2 and Na1 transport by aparacellular permeability of the OK proximal tubule cell cGMP-mediated mechanism in the kidney proximal tubule. Am J

Physiol 272:F242–F248, 1997line via mechanisms possibly involving decreases in cellu-15. Eitle E, Hiranyachattada S, Wang H, Harris PJ: Inhibitionlar ATP content. This effect of NO might have significant of proximal tubular fluid absorption by nitric oxide and atrial

implications in pathophysiological conditions such as re- natriuretic peptide in rat kidney. Am J Physiol 274:C1075–C1080,1998nal ischemic injury.

16. Roczniak A, Burns K: Nitric oxide stimulates guanylate cyclaseand regulates sodium transport in rabbit proximal tubule. Am J

ACKNOWLEDGMENTS Physiol 270:F106–F115, 199617. Guzman NJ, Fang M, Tang S, Ingelfinger JR, Garg LC: Auto-

This work was supported by a National Heart, Lung, and Blood crine inhibition of Na1/K1-ATPase by nitric oxide in mouse proxi-Institute Grant HL-55594. The authors gratefully acknowledge Ms. mal tubule epithelial cells. J Clin Invest 95:2083–2088, 1995Jennifer M. Gross for excellent technical support and Ms. Joanne 18. Preisig PA, Berry CA: Evidence for transcellular osmotic waterZimmerman for secretarial assistance. The authors also thank the labo- flow in rat proximal tubules. Am J Physiol 249:F124–F131, 1985ratory group for helpful discussion and critical reading of the manu- 19. Schwegler JS, Heuner A, Silbernagl S: Electrical properties ofscript. cultured renal tubular cells (OK) grown in confluent sheets. Pflug-

ers Arch 415:183–190, 1989Reprint requests to Franklyn G. Knox, M.D., Ph.D., Departments of 20. Balda MS, Gonzalez-Mariscal L, Matter K, Cereijido M, An-

Medicine and Physiology and Biophysics, Mayo Clinic and Foundation, derson JM: Assembly of the tight junction: The role of diacylglyc-Rochester, Minnesota 55905, USA. erol. J Cell Biol 123:293–302, 1993

21. Salzman AL, Menconi MJ, Unno N, Ezzell RM, Casey DM,Gonzalez PK, Fink MP: Nitric oxide dilates tight junctions and

APPENDIX depletes ATP in cultured Caco-2BBe intestinal epithelial mono-layers. Am J Physiol 268:G361–G373, 1995

Abbreviations used in the article are: ATP, adenosine 59-triphos- 22. Ohkawa H, Ohishi N, Yagi K: Assay for lipid peroxidase in animalphate; eNOS, endothelial nitric oxide synthase; Hb, hemoglobin; iNOS, tissues by thiobarbituric acid reaction. Anal Biochem 95:351–358,inducible nitric oxide synthase; LDH, lactate dehydrogenase; NO, nitric 1979oxide, NOS, nitric oxide synthase, nNOS, neuronal nitric oxide syn- 23. Burgstahler AD, Nathanson MH: NO modulates the apicolat-thase; OK, opossum kidney; Papp, paracellular permeability coefficient; eral cytoskeleton of isolated hepatocytes by a PKC-dependent,SNP, sodium nitroprusside; SOD, sodium dismutase; TBAR, thiobarbi- cGMP-independent mechanism. Am J Physiol 269:G789–G799,turic acid reaction; TBARS, thiobarbituric acid reaction substance; 1995TCA, trichloracetic acid; TER, transepithelial electrical resistance. 24. Okayama N, Kevil CG, Correia L, Jourd’heuil D, Itoh M, Gris-

ham MB, Alexander JS: Nitric oxide enhances hydrogen perox-ide-mediated endothelial permeability in vitro. Am J PhysiolREFERENCES273:C1581–C1587, 1997

25. Sharma R, Lovell HB, Wiegmann TB, Savin VJ: Vasoactive1. Garcia NH, Ramsey CR, Knox FG: Understanding the role ofparacellular transport in the proximal tubule. News Physiol Sci substances induce cytoskeleton changes in cultured rat glomerular

epithelial cells. J Am Soc Nephrol 3:1131–1138, 199213:38–43, 19982. Spring KR: Routes and mechanism of fluid transport by epithelia. 26. Chiao H, Kohda Y, McLeroy P, Craig L, Housini I, Star RA:

a-Melanocyte-stimulating hormone protects against renal injuryAnnu Rev Physiol 60:105–119, 19983. Boulpaep EL, Seely JF: Electrophysiology of proximal and distal after ischemia in mice and rats. J Clin Invest 99:1165–1172, 1997

27. Canfield PE, Geerdes AM, Molitoris BA: Effect of reversibletubules in the autoperfused dog kidney. Am J Physiol 221:1084–1096, 1971 ATP depletion on tight junction integrity in LLC-PK1 cells. Am J

Physiol 261:F1038–F1045, 19914. Lutz MD, Cardinal J, Burg MB: Electrical resistance of renalproximal tubule perfused in vitro. Am J Physiol 225:729–734, 1973 28. Mandel LJ, Bacallao R, Zampighi G: Uncoupling of the molecu-


lar “fence” and paracellular “gate” functions in epithelial tight nitric oxide synthase in iron-deficiency anemia. Kidney Int 52:195–201, 1997junctions. Nature 361:552–555, 1993

36. Kashem A, Endoh M, Yano N, Yamauchi F, Nomoto Y, Sakai29. Tsukamoto T, Nigam SK: Tight junction proteins form large com-H: Expression of inducible-NOS in human glomerulonephritis: Theplex and associate with the cytoskeleton in an ATP depletionpossible source is infiltrating monocytes/macrophages. Kidney Intmodel for reversible junction assembly. J Biol Chem 272:16133–50:392–399, 199616139, 1997

37. Salvemini D, Seibert K, Masferrer JL, Misko TP, Currie MG,30. Stueher DJ, Nathan CF: Nitric oxide: A macrophage productNeedleman P: Endogenous nitric oxide enhances prostaglandinresponsible for cytostasis and respiratory inhibition in tumor targetproduction in a model of renal inflammation. J Clin Invest 93:1940–cells. J Exp Med 169:1543–1555, 19891947, 199431. Stadler J, Billiar TR, Curran RD, Stuehr DJ, Ochoa JB, Sim-

38. McLay JS, Chatterjee P, Nicolson G, Jardine AG, McLay NG,mons RL: Effect of exogenous and endogenous nitric oxide onRalston SH, Grabowski P, Haites NE, MacLeod AM, Hawks-mitochondrial respiration of rat hepatocytes. Am J Physiol worth GM: Nitric oxide production by human proximal tubule260:C910–C916, 1991 cells: A novel immunomodulatory mechanism? Kidney Int

32. Dimmeler S, Lottspeich F, Brune B: Nitric oxide causes ADP- 46:1043–1049, 1994ribosylation and inhibition of glyceraldehyde-3-phosphate dehy- 39. McLay JS, Chatterjee PK, Mistry SK, Weerakody RP, Jardinedrogenase. J Biol Chem 267:16771–16774, 1992 AG, McLay NG, Hawksworth GM: Atrial natriuretic factor and

33. Yu L, Gengaro PE, Niederberger M, Burke TJ, Schrier RW: angiotensin II stimulate nitric oxide release from human proximalNitric oxide: A mediator in rat tubular hypoxia/reoxygenation in- tubular cells. Clin Sci 89:527–531, 1995jury. Proc Natl Acad Sci USA 91:1691–1695, 1994 40. Edelstein CL, Ling H, Schrier RW: The nature of renal cell

34. Vaziri ND, Ni Z, Oveisi F: Upregulation of renal and vascular injury. Kidney Int 51:1341–1351, 1997nitric oxide synthase in young spontaneously hypertensive rats. 41. Noiri E, Peresleni T, Miller F, Goligorsky MS: In vivo targetingHypertension 31:1248–1254, 1998 of inducible NO synthase with oligodeoxynucleotides protects rat

kidney against ischemia. J Clin Invest 97:2377–2383, 199635. Ni Z, Morcos S, Vaziri ND: Up-regulation of renal and vascular

nitric oxide enhances paracellular permeability of opossum kidney cells

Documents