Identification of a contractile function for renal
medullary interstitial cells.
A K Hughes, … , W H Barry, D E Kohan
J Clin Invest.
1995;
96(1)
:411-416.
https://doi.org/10.1172/JCI118050
.
Renomedullary interstitial cells (RMIC) are unique to the renal medulla. By virtue of their
anatomic location and arrangement, RMIC may hinder axial dissipation of the concentration
gradient, thereby aiding urinary concentration. A more active role in urinary concentration
has been postulated on the basis of speculations about RMIC contractile potential,
however, RMIC contraction has not been investigated. To determine if these cells are
contractile, cultured rat RMIC were exposed to endothelin-1 (ET-1), a potent vasoconstrictor
which binds to RMIC, and examined using video microscopy. ET-1 (as low as 10 pM)
caused a slowly developing and dose-dependent reduction in RMIC surface area. ET-1
markedly increased the number and intensity of F-actin microfilament staining.
ET-1-induced RMIC contraction was not altered by nifedipine, was partially reduced by nickel,
and was completely inhibited by H7, indicating that ET-1 action is mediated by protein
kinase C and is partially dependent upon receptor-operated calcium channels. The ET-1
effect does not involve nitric oxide since NG-monomethyl-L-arginine did not alter
ET-1-induced RMIC contraction; in addition, ET-1 had only a minor effect on cGMP levels and no
effect on nitrite production. PGE2 acts in an autocrine manner to dampen ET action since
indomethacin potentiates, while PGE2 inhibits, ET-1-induced RMIC contraction. The
contractile response is not unique to ET-1 since vasopressin also reduces RMIC surface
area and increases F-actin microfiliment […]
Research Article
Identification of
a
Contractile
Function for Renal
Medullary Interstitial Cells
Alisa K. Hughes,*William H. Barry,* and Donald E. Kohan*
Divisions of *Nephrology and
*Cardiology,
Department of Medicine, VeteransAffairsMedical Center and the University ofUtah Schoolof Medicine and the Eccles Programin HumanMolecularBiologyandGenetics, Salt LakeCity, Utah 84132
Abstract
Renomedullary interstitial cells (RMIC)areunique to the renal medulla.Byvirtue of their anatomic locationand
ar-rangement,RMICmayhinder axial dissipation ofthe
con-centration gradient, thereby aiding urinaryconcentration. A moreactive role inurinary concentrationhas been postu-lated on the basis of speculations about RMIC contractile potential, however, RMICcontraction hasnotbeen investi-gated. To determine ifthesecells are contractile, cultured rat RMIC were exposed to endothelin-1 (ET-1), a potent vasoconstrictor which bindstoRMIC,andexamined using videomicroscopy. ET-1 (as low as 10pM) causedaslowly developinganddose-dependentreduction in RMIC surface
area.ET-1markedlyincreasedthenumber andintensityof F-actin microfilament staining. ET-1-induced RMIC
con-tractionwas notalteredbynifedipine,waspartiallyreduced by nickel, and wascompletely inhibited byH7, indicating that ET-1 action is mediated by protein kinase C and is partially dependentupon receptor-operated calcium chan-nels. The ET-1effectdoesnotinvolve nitric oxide since NG-monomethyl-L-argiinine did not alter ET-1-induced RMIC contraction; inaddition, ET-1 hadonly aminor effect on
cGMPlevels andnoeffectonnitriteproduction. PGE2acts
in an autocrine manner to dampen ET action since indo-methacin potentiates, while PGE2 inhibits, ET-1-induced RMIC contraction. The contractile response is not unique
toET-1 sincevasopressinalso reducesRMIC surfacearea
andincreases F-actin microfiliment staining. Thesestudies demonstrate that RMIC in culture are contractile. The
pos-sibility
is raised that contraction ofRMIC plays a role in modifying urinary concentration as well as regulation of other renal medullary functions. (J. Clin. Invest. 1995. 96:411-416.) Keywords:kidney* medulla *urine *concen-tration *prostaglandin E2
Introduction
Renomedullary interstitial cells (RMIC)1 are stellate cells that
areuniquetothe renal medulla andare mostprevalenttowards
Address correspondencetoDonald E. Kohan,Division of Nephrology andHypertension, University ofUtah MedicalCenter, Salt Lake City, UT84132.Phone:801-585-5219; FAX:801-581-4343.
Receivedfor publication5January 1995 andacceptedin revised
form21 March 1995.
1.Abbreviations used in thispaper:ANF,atrialnatriureticfactor; AVP, arginine vasopressin; ET, endothelin; H7, 1-(5-isoquinoline
sulfonyl)-2-methylpiperazine dihydrochloride; NMMA, N
-monomethyl-L-argi-nine monoacetate; NO, nitric oxide; RMIC, renomedullaryinterstitial cells.
TheJournalof ClinicalInvestigation,Inc. Volume96, July 1995,411-416
thepapillary tip (1, 2). Theyareperhaps best known for their abilitytosynthesize large quantities of vasodepressor lipidsor
lipid precursors, including PGE2 and medullipin I (3, 4). Re-lease of these lipids by RMIC has been postulated toplay a
role inregulating blood pressure and the developmentof hyper-tension (5). RMIC may also be involved in regulating urine concentration, however sucharole for these cellshas not been
well studied. RMIC typically bridge the interstitium between medullary blood vessels and thin limbs of Henle's loops,
form-ing a ladder-like arrangement with the long axis of the cells
perpendiculartothelongaxis of thepapilla (6).This anatomic arrangement suggests a number of functions that RMIC may play inmaintaining urinary concentrating ability. Most appar-ently, they couldprovidestructural support for the medulla and papilla. In addition, because of their orientation, RMIC are
likely to hinder axial diffusion in the medulla (6), thereby limiting dissipation of the solute concentration gradient. In
re-gardstothis latterpoint,it isinterestingto notethat the kanga-roorat, a desert rodent with theabilityto concentrate urineto anunusually high degree, has the greatest abundance of papil-lary RMIC known (7). A third possible function for RMIC has beenpostulated: contraction. These cells havecytoplasmic myoid fibrils (8) and, as mentioned above, are anchored to
adjacent blood vessels and Henle's loops. In 1956, Sternberg and coworkers noted that"should these cells (RMIC) prove to haveacontractilefunction, they might playarole in the regula-tion ofurinary volume,"(7).In1990, Fontouraetal. described atrial natriuretic factor (ANF) receptors on RMIC and noted that the contractilestateof RMIC may influence flow, pressure,
or passive permeability characteristics of the vasa recta or
Henle's loops (9). To date, however, a contractile ability of RMIC hasnotbeenreported. The purpose of the current study was,therefore, todetermine ifRMIC were contractile.
Anumber of vasoactive mediators have been demonstrated
tobind to, and activate signal transduction systems in, RMIC. These include atrial natriuretic factor (9), nitric oxide (10), angiotensinI (4, 11), bradykinin (4), vasopressin (4, 12), and endothelin-1 (ET-1 )(13). Because of ongoing research interest by this laboratory in theproductionand actionsof ET-1 in the renal medulla(14-16), we chose to examine the effect of this peptideonRMIC contraction.ET-1 isa21-amino acidpeptide that is themostpotent vasoconstrictor known (17). It bindsto
high affinity receptorsonRMIC and elicitsanincrease in cyto-solic- free
Ca2+
concentration that is dependent in part upon increases in inositol trisphosphate levels and receptor-operatedCa2+
channels(13).Wereport thatET-1 causes along-lasting and potent contraction of cultured rat RMIC. In addition, the mechanism of ET-1-induced RMIC contraction is explored. Finally, the effect of arginine vasopressin (AVP) on RMICcontraction is examinedtodetermineifthe response is unique
to ET-1.
Methods
Materials. RPMI-1640 and bovine calf serum were obtained from
Labora-tories, Inc., Belmont, CA; human insulin (Humulin) from Eli Lilly,
Indianapolis, IN; N0-monomethyl-L-arginine monoacetate (NMMA)
fromChem-Biochem Research, Salt Lake City, UT;
rhodamine-phalloi-din from Molecular Probes Inc., Grand Junction, OR; and Bradford
reagentfromBio-Rad, Richmond, CA. Cover slips and tissue culture plates wereobtained from Fisher Scientific Co., Santa Clara, CA. All other reagents were from Sigma Chemical Co., St. Louis, MO unless specified otherwise.
Tissueculture. Renomedullary interstitial cells from Sprague-Daw-ley rats were a generous gift of Drs. Edward Nord (State University of NewYork, Stony Brook, NY) and Thomas Maack (Cornell University MedicalCollege, New York, NY). These cells have been extensively
characterized and previously describedin detail (9).Cells were grown inRPMI-1640 containing 20% bovine calf serum and 60 U/ml insulin on25-mm circularglass coverslipsor 24-well plastic plates in a 5%
CO2 environment at 37TC. Experiments were performed at 30-50%
confluenceonpassages 22-27 (18-40h after initialplating). Measurements of changes in cell surface area. On the day of study,
cells were washed twicewith Krebs buffer (145 mM NaCl, 10 mM Hepes, 5 mM glucose, 5 mM KCl, 1 mM Na2HPO4, 2.5 mM CaCl2, 1.8 mMMgSO4, pH7.3) and incubated in KRB at 370Cfor30 min
(preincubation solution). Thecoverslips were then placed in a370C
incubation chamber with a clear glass bottom (18) containing 5 ml KRB(incubationsolution) on the stage of an invertedmicroscope
(Dia-phot, Nikon Inc., Tokyo, Japan). The cellimagewasobtained by 700 nmillumination and collected byaphasecontrast40X objective lens
(NikonPh3 40 DL;Nikon Inc.) and monitored with a video camera
(model 540; Pulnex,Sunnyvale, CA) attached to the video port of the
microscope. Live images wererecorded on a video cassette recorder
(HRD750U; JVC, Salt Lake City, UT) andplayedbackforanalysis. Cell surfaceareachanges weremeasuredbytracingthevideoimaged cell perimeterontotransparentfilm(PP2200; 3M Co., St. Paul, MN).
The tracingswerecut out andweighed. All measurementsof surface
areainvolveddeterminationof the initialcellarea atthetimeofplacing the cellinto the incubation chamber(termed "time 0") followed by
measurementsofthe samecellarea over thenext30-60min. Allresults
werecalculatedbycomparingthesurfaceareaafter various incubation timestotheareaat time 0 in thesamecell.
Experimental protocols formeasuringchangesincellsurfacearea.
Forcontrol measurements, RMIC werepreincubated in KRB for30 min, thenplacedin KRB alone in theincubation chamber for30-60
min. Changes in surface area wereassessed overtheduration ofthe
incubation.Varying concentrations ofET-1or10nM AVPwereadded
immediately after the coverslipswereplacedin the incubation chamber
and baseline cell surface area recorded. Cell surface area was then recordedfor 30-60min in the presence of ET-1 orAVP.Forstudies
ontheeffects ofnifedipine, nickel, NMMA, indomethacin, and 1-(5-isoquinoline sulfonyl)-2-methylpiperazine dihydrochloride (H7) on
ET-1-induced contractions twoprotocols werefollowed. Inthefirst,
theeffects of thesereagents aloneoncell surfacearea wasdetermined
ina manneridenticaltothatforET-1 above. In thesecond, theeffects ofthese reagents on ET-1 actions wasassessed bypreincubating the
cells for 30min with theabovereagents. Thecellswerethenplacedin theincubation chamber,ET-1 added in thecontinued presence of the variouscompounds,andchangesin surfaceareaassessed for 30 minas
described above.Finally, for studies withPGE2-, PGE2andET-1 were
addedatthesametimetocells that had beenpreincubatedin indometh-acin.
StainingofF-actin microfilaments. RMICwerestainedusing
pre-viously describedmethodology (19).Briefly, cells were incubated in KRB aloneorcontaining 10 nM ET-1 or 10nMAVPfor 30 min at
37°C.CoverslipswerethenrinsedinPBS, fixedwith 3.7%formaldehyde
inPBS for 10 minat23°C, rinsed, permeabilizedwith 0.2% Triton
X-100 inPBSfor 1 minat23°C,and rinsedagain. Cellswerethen incu-bated with 5.0U/ml rhodamine-phalloidin in PBS for 30 min in the
darkon arotary shaker, rinsed,and mountedusing 1:1 PBS/glycerol,
and photographed using a microscope (Carl Zeiss, Inc., Thornwood,
NY)equipped withamercurylamp,epifluorescence illumination,and
a camera(Polaroid Corp., Cambridge, MA). All photomicrographs were
obtained under identical conditions.
Determination of cyclic nucleotide accumulation. RvMICin 24-well plates were preincubated with 0.1 mM3-isobutyl-I-methylxanthine for
30min in KRB followed by addition of varyingconcentrations of
ET-1 for 5-ET-10 min at 370C. The mediawas then removed and the cells
treated with 100% ethanol overnight. The ethanol was subsequently evaporated, the samples resuspended in assay buffer andcGMP, and
cAMPdetermined using commercially available radioimmunoassay kits (AmershamCorp., ArlingtonHeights, IL). The remaining cell protein
wassolubilized in 0.1 N NaOH, an aliquot mixed withBradford reagent,
and protein concentration determined by measuring absorbance at 590 nm (20). Allresults were expressed as femtomoles cyclic nucleotide
per microgram total cell protein. As a positive control for cGMP, RMIC were exposed to 10 nM ANF for 10min.
Determination of nitrite levels. RMIC in 24-well plates were
ex-posed to0.1-100nMET-1in KRB at370Cfor 5-30min.After incuba-tion, duplicate 50-Ml aliquots of the supernatants were removed and immediately tested for nitrite (NO-) levels (stable breakdown product of nitricoxide) as previously described (21). To each aliquot, 100 ILI
of Greissreagent (1% sulfanilamide in 30% acetic acid and 0.1%
N-(l-naphthyl)ethylenediamine dihydrochloride in 60% acetic acid in a 1:1 mixture) was added and mixed for one min. Absorbance at 550 nm was immediately measured in a microplate reader (Thermomax;
Molecular Devices Corp., Menlo Park, CA). NO- levels were deter-mined by comparing sample values with a standardcurveestablished with known quantities of sodium nitrite. The cells were solubilized with 0.1 N NaOH and total protein determined as described above.
Statistics. All results are compared by ANOVA. P < 0.05 was deemed significant as determined by the Scheffe F-test. All data are expressed as mean±SEM.
Results
RMIC contraction. Incubation of RMIC in KRB alone for up
to 60min hadnosignificant effect oncell surface area. ET-1
(10 nM) caused a slowly developing reduction in surface area that was maximal (75.3±1.9% of original cell surfacearea) 30
min after addition of the peptide (Fig. 1 B). The response
to ET-1 was dose-dependent with a significant, albeit small, contraction occurring at 10 pM ET-1 (Fig. 1 A). Cells were
stained for F-actin filaments (with rhodamine-phalloidin) after a30-min incubationin KRB alone or containing 10nM ET-1.
Asis apparent from Fig. 3, ET-1 caused a marked increase in the number and the intensity of F-actin filaments in RMIC, providing further support for ET-1 -stimulated contractile activ-ity byRMIC.
AVP(10 nM) also contracted RMIC (Fig. 2). The response to AVPwas maximal by 15 min and persisted for at least 1 h. This was faster than the contraction observed with ET-1; the reasons for the different response times is uncertain, but could reflect differences in ligand stimulation of factors modifying RMIC contraction such as PGE2 (see below). Similar to ET-1, AVP caused a marked increase in the number and intensity of F-actin microfilaments (Fig. 3).
100
so.
40
-A
o* a
a
co
C
o
a
8
CO
0.01 0.1 10
ET-1 (nM)
30
Minutes
Figure1. Effect of ET-1onsurfaceareaofratmedullary interstitial cells.(A)Cellareawasdeterminedimmediatelybeforeand 30mm
after addition of0.01-10 nMET-1.(B)Cellareawasdetermined
immediatelybefore and then 10-60 min after addition of10 nM ET-1.n=3-4eachdatapoint. *P<0.01; * *P<0.005; f P<0.001;
allvscontrol and f P<0.005vs0.01 nMET-1.
Cinhibitor,hadnoeffectonunstimulatedRMIC surface area,
butcompletelyblockedthe contractileresponsetoET-1.Hence,
ET-1 stimulated RMICcontractionisdependentinparton volt-age-insensitive extracellular calcium entry-and is mediated
through activation ofproteinkinase C.
ET-1 stimulates RMIC production of PGE2 (13, 22), an eicosanoidwhichcaninhibit thevasoconstrictiveeffectof ET-1 (23). To determine ifcyclooxygenase products modify the
contractile responsetoET-1 inRMIC, cells weretreated with indomethacin followedby addition of ET-1 (Fig. 5). Indometh-acin alone cause a slight, but significant, increase in RMIC
surface area, suggesting that basal cyclooxygenase products may tonically contract these cells. In contrast, indomethacin potentiated thecontractile effect ofET-1, suggesting that
ET-1 stimulation ofacyclooxygenase productdecreases the
con-tractileresponsetothepeptide.Theeffect ofPGE2onthe
ET-1-induced contraction in the presence of indomethacin was
examinedsince this isthemajoreicosanoidproduced by RMIC.
As shown in Fig. 5, PGE2 markedly reduced the contractile
responsetoET-1, supportinganinhibitory role forthe
prosta-noid.
120
1004
so
40-20.
0
* *
0f
0 10 20 0o 40 50 So
Minutes
Figure2. Effectofvasopressinonsurfaceareaofratmedullary intersti-tial cells.Cell area wasdeterminedimmediately before and then
10-60 min-after addition of10nM AVP. n = 7each datapoint. **P
<0.005; *P <0.001; bothvscontrol.
Nitric oxide (NO) is a well known antagonist of ET-1, inhibitingboth ET-1 production (24) and smooth muscle
con-traction (25). To examine whether NO could be modifying theET-1 response,RMICwerepreincubatedwithNMMA,an
antagonist of NO action. The dose ofNMMA used has been previouslyshown to completelyblockendogenous NO produc-tionbyanumberofcell types(21). NMMA hadnoeffecton
basal RMIC surfaceareanordid italterET-1 -induced
contrac-tion (Fig. 4), suggestingthatNO isnotinvolved inmediating the'ET-1 response. Nitrite levelswereexamined in supernatants of cells incubated with media aloneorcontaining 10nMET-1 for up to30 min, however no nitrite was detected under any conditions. This assay isrelativelyinsensitive since theGreiss reagent detectsonly microgramquantitiesof nitrite andcould, therefore,havemissed less dramatic alterationsinendogenous NOproduction. To further examineanalteration inendogenous NO levels, the effect ofET-1 oncGMP (themediatorofNO actions) accumulationinRMICwasalsoquantitated.ET-1 did
cause a very modest reduction in cGMP levels, but only at
relatively highconcentrations(10-100nM) of thepeptide (Fig. 6 B). ThecGMP system wasclearlyresponsivein these cells
asANFmarkedlyincreasedcGMP levels. These data suggest, thatasmall reduction in NOmight playarelativelyminor role inmediatingET-l-induced RMICcontraction,but isunlikely
tobe ofphysiologic significance.
Sincealterations in cAMPaccumulation have been impli-cated incellcontraction, the effect of ET-1 on RMIC cAMP levelswasexamined. ET-1,atdoseswhichclearly elicited
con-traction, had noeffecton RMIC cAMP accumulation (Fig. 6 A), indicating that this is not themechanismby which ET-1
-induced RMIC contractionoccurs.
Discussion
Thecurrentstudy demonstratesthatRMICarecontractile cells. ET-1 andAVP cause along-lasting reductionin RMICsurface
area that isassociated with marked increasesinthe numberand
intensity of F-actin microfilament staining. The finding that cultured RMICarecontractile suggests that this isanimportant physiologic property of RMIC in vivo, particularly since cells 0
co
a
.5
o
U
2
it
co
a
ID
.C
0o
Li
0
t
Figure 3.Photomicrographs of rat medullary interstitial cells exposed to media alone (A) and(C),containing 10 nM ET-1 (B), or 10 nM AVP
(D)for 30min.Cellswerestained withrhodamine-phalloidinandphotographed under fluorescent microscopy using identicalexposure conditions. Bothphotomicrographsare x250.Arepresentative of fiveexperiments isshownforET-l andthreeexperimentsfor AVP.
in culture tend to lose their contractility. Although it is not possible to draw conclusions about the physiologic relevance of RMIC contraction from the current study, one must consider how thecontractile state of these cells could potentially impact
on medullary function. RMIC are intimately associated with loops of Henle and blood vessels and, to a lesser extent, with medullary collecting ducts (2). Consequently, RMIC form a bridgepredominantly between Henle's loops and blood vessels thatare orientedwith the long axis of the RMIC perpendicular
tothelong axis of the papilla (6). This results in a ladder-like
structure in which RMIC could limit axial diffusion (6). How
then couldRMIC contraction modify renal medullary function? First, RMIC contraction may reduce medullary interstitial
vol-ume by pullingHenle's loops, blood vessels, and to a lesser
extentcollecting ducts, closerto oneanother. This wouldresult
in less distance for water and solutes to diffuse, facilitating exchange between the tubules and the blood. Whether this mechanism could be physiologically important is, however, un-certain. Wexler and colleagues have proposed a model to ex-plain medullary urine concentration that assumes asignificant physical separation between the collecting duct and Henle's
limbs (26). In contrast, anatomical studies have demonstrated arelatively smalldistancebetweencollecting ductsand Henle's
limbs (27), raising doubts as to the role of distance between these tubule segments in modulating urine concentration. This
issue remains unresolved, however, since RMIC contraction during the processof tissue fixation could alter medullary geom-etry.Further,thearchitecture of the renal medulla is notoriously difficulttopreserveduring fixation due to the high tonicity (2).
0
o CS 'a
0
8
Co .9 (uA
Control NIF Ni- N&M H7
Figure 4.Effect of10
MM
nifedipine(NIF),1mMNi, 0.5mMNMMA, and 10 uMH7 on 10 nMET-l -inducedcontractionbyratmedullary interstitial cells. InstudieswithoutET-1, cellarea wasdeterminedim-mediatelybeforeand 30 minafter addition ofthe above reagents. In
studies with ET-1, cellswerepreincubated withthe above reagentsfor 30 minfollowed by addition ofET-1 for30 min in the presenceof the
various reagents. Cellarea wasdetermined immediately beforeand 30 minafter addition ofET-1. n= 3-5 eachdatapoint. *P< 0.001vs
controlandvs sameconditionin theabsence of ET-1; #P< 0.005
vscontrol and vsNIF alone; f P< 0.005vscontrolorET-1 alone
and P< 0.025vs Nialone; **P< 0.001 vsET-l alone.
a 2
6-I
0
a
:LC
0
0C
A
i
0.
Io
A
12-6
3
Control 1Mo ETl1 10 no ET-1
20- e .*
lam
126
12 ^
4-*
0 0.01 0.1 0 100 AN
ET-1 (no)
Figure6. Effect ofET-1
oncAMP (A)orcGMP
(B) accumulation in rat
medullaryinterstitial
cells. Cellswere
preincu-bated with0.1mM
3-iso-butyl-l-methylxanthine
for30min,followedby addition of ET-1 or10 nMatrialnatriuretic
pep-tide(ANP)for10min. n = 8eachdatapoint
ex-ceptn = 4 forANP. * P <0.01 vscontrol; ** P < 0.001vscontrol.
caused by RMIC contraction could reduce the energy required
tomaintain thehighsoluteconcentrations, particularlytowards thepapillary tip. Put another way, ifpapillary volume isless, there is less soluterequired tomaintain ahypertonic environ-mentand less energy neededtodrivethe inner medullary
con-centratingmechanism. Viewed in the above context,onewould predict that RMIC contraction would, therefore, favor urine concentration and medullary waterreabsorption. Other effects of RMIC contraction must also be considered. Since a single RMICcanencircle upto30%ormore of the basement
mem-brane ofagiven cross-section ofaloop of Henleor vasa recta
(1,2),it ispossible that,aspreviously speculated byFontoura andcoworkers (9), RMICcontraction could alter tubulefluid
orbloodflow, permeability characteristics of thevasa recta or
thin limbs, andeven interstitial, tubule, orvascular pressures. The challenge will be to design in vivo studies in which the
120U
so
40~~~~~~~
a
40
Figure5.Effect of 10
jLM
indomethacin (INDO)and1ILM
PGE2on10nMET-l-induced contractionbyrat
medul-lary interstitialcells. For
INDOalonestudies, cell
area wasdetermined im-mediatelybefore and30 min after addition of INDO. For ET and ET
+PGE2 studies,cells werepreincubatedwith INDO for 30 min, followed by addition of ET-l and/or PGE2 for 30 min.Cellarea wasdetermined in these latter studies immediately before
and 30minafteradditionofET-1 and/or PGE2.n= 3-6 each data
point. *P<0.001vscontrol; **P<0.025vscontrol; # P<0.001 vsINDO aloneor vscontroland P<0.01 vsET-l alone; f P<
0.005 vscontrol, ET-l alone, and INDO +ET-1.
role of RMIC contraction can be tested, however, thecurrent
findingthatRMIC do indeedcontractshouldprovide impetus
tothedevelopment of relevant techniques.
ET-1 has been demonstrated to bind to RMIC and elicit increases inthe concentration of cytosolic-free Ca2 , inositol
trisphosphate, and PGE2 (13, 22, 28). In thecurrentstudy ET-1 was shown tocontract RMICat aconcentration as lowas 10
pMandhalf-maximal contractionat - 100pM. This concentra-tion rangefor ET- 1action is somewhat lower than thatdescribed
byWilkes and coworkers for the initial riseincytosolic Ca2+ concentration and inositol trisphosphate levels in RMIC (0.5
and 1 nMET-1,respectively) (13). This group did note,
how-ever,that a secondsustainedrise incytosolic-free Ca2+ concen-tration in RMIC was associated with activation of receptor-operatedcalciumchannelsandoccurredatET-1concentrations as low as 1 fm. Our finding that nickel, but not nifedipine, inhibitsET-1-inducedcontraction lends further support to the notion that ET-1 activatesdihydropyridine-insensitive, receptor-operatedcalciumchannelsin RMIC.
The current study explored several potential mechanisms
for ET-1-induced RMIC contraction. ET-1-induced RMIC
contraction is dependent upon activation of protein kinase C
since H7completely blocksthe ET-1 effect. This finding isin
agreement withprevious data showing that protein kinase C is
animportant mediator ofatleast someaspects of ET-1 -induced signal transduction in RMIC (activation ofphospholipase D) (28). ET-1 didnot altercAMP levels andonly modestly
re-duced cGMP levels at concentrations ofET-1 that were well above thoserequired to elicit RMIC contraction. Nitricoxide
does not appear to mediate the ET-1 response in these cells since NMMA hadnoeffectonRMICcontractionafter simula-tionwith ET-1.Perhaps of greatest interestwasthefinding that indomethacin potentiatesET-1 action, while addition of
PGE2
reduces ET-1-elicited RMIC contraction. Takentogetherwith previousstudiesshowingthat ET-1 stimulatesPGE2 production
I
a
Is
8
0
Codol Er NM - i
by RMIC (13, 22),thesedataindicatethat PGE2 is a negative modulatorofET-l-stimulated RMIC contraction.
ThatET-1regulates RMICcontraction suggests an interest-ing and complex system of ET-1 action in the inner medulla. ET-1 is not produced by RMIC (13); rather, it derives from
adjacent inner medullary collectingduct (IMCD) ( 14) and
pos-sibly endothelial cells (29). In the IMCD, ET-1 inhibits vaso-pressin-stimulated water (30, 31), and possibly sodium (32), reabsorption. In contrast, we speculate that ET-1, by contracting
RMIC, would enhance water reabsorption. RMIC express only
ETA receptors (13), while ET-1 exerts its effects on IMCD
cells through activation of
ETB
receptors (33). Hence, the neteffect of ET-1 action in the innermedulla may depend upon the relative degreeof activation of the two endothelin receptor subtypes. It should also be noted thatthese considerations do
not take into account ET-1 actions onvasa recta and medullary blood flow.
AVPalso contractsRMIC in culture.This demonstrates that
RMICcontraction is notunique to ET-1. In addition, it raises the interesting possibilitythat RMICcontraction mightcontribute to
AVP-mediated increases in urineconcentration. Detailed
stud-ies on the mechanism ofAVP-stimulated RMIC contraction of
ET-1 are clearly needed.
Insummary, the current study demonstrates that RMIC are contractilecells. ET-1 elicits RMIC contraction through activa-tion ofprotein kinaseC andreceptor-operated calciumchannels. AVP alsocontracts RMIC; the mechanism of this effect requires furtherinvestigation. In addition, it will be of great interest to determine the effect of other agents that bind to RMIC, includ-ing angiotensin II and ANF, on RMIC contraction. Further, a role forPGE2 in modifying the effects of these agents on RMIC contraction merits examination. Finally, studies are needed which aredesigned to investigate the role of RMIC contraction in regulating renal medullary function.
Acknowledgments
The authorsthank Dr. Rex Jamison for his helpful comments. This workwas supported in part by Merit Review and Career Devel-opmentAwards from the Department ofVeteransAffairs and by
Na-tional Institutes of Health grant R29 DK44440 (allto D. E. Kohan).
References
1. Osvaldo,L., and H. Latta. 1966. Interstitial cells of the renal medulla. J. Ultrastruc. Res. 15:589-613.
2. Bohman,S.-O. 1974. Theultrastructureof the rat renal medulla as observed after improvedfixation methods. J. Ultrastruc. Res. 47:329-360.
3. Muirhead, E. E. 1991. Themedullipinsystem of blood pressure control. Am. J.Hypertens. 4:556S-568S.
4. Zusman, R. M., and H. R. Keiser. 1977. Prostaglandin biosynthesis by
rabbitrenomedullary interstitialcells in tissue culture. J. Clin. Invest.
60:215-223.
5. Muirhead, E. E. 1990. Discovery of the renomedullary system of blood pressure control and its hormones. Hypertension (Dallas). 15:114-116.
6. Lemley, K. V., and W. Kriz. 1991. Anatomy of the renal interstitium. KidneyInt. 39:370-381.
7.Sternberg, W. H., E. Farber, and C. E. Dunlap. 1956.Histochemical localiza-tion of specificoxidative enzymes: R. Localization of diphosphopyridine nucleo-tide andtriphosphopyridine nucleotide diaphorases and thesuccindehydrogenase
system in thekidney. J. Histochem. Cytochem. 4:266-283.
8. Harper, J. T., H. Puchttler, S. N. Meloon, and M. S. Terrry. 1970. Light-microscopic demonstration of myoid fibrils in renal epithelial, mesangial and interstitial cells. J. Microsc. (04).91:71-85.
9. Fontoura,B. M. A., D. R. Nussenzvieg, K. M. Pelton, and T. Maack. 1990.
Atrial natriureticfactor receptors in cultured renomedullary interstitial cells. Am. J.Physiol. 258:C692-C699.
10.Ujiie,K., L.Hogarth, R. Danziger, J. G. Drewett, P. S. T. Yuen, I.-H. Pang, and R. A. Star. 1994. Homologous and heterologous desensitization of a
guanylylcyclase-linked nitric oxidc receptor in cultured rat medullary interstitial cells.J. Pharmacol. Exp. Ther. 270:761-767.
11.Zhuo, J., D. Alcorn, J. McCausland, and F. A.0. Mendelsohn. 1994. Localizationand regulationofangiotensinII receptors in renomedullary interstitial cells. Kidney Int. 46:1483-1485.
12.Beck, T. R., A.Hassid,andM. J. Dunn. 1980. The effect of arginine vasopressin and its analogs on the synthesis of prostaglandin E2 by rat renal medullary interstitial cells inculture, J. Pharmacol. Exp. Ther. 215:15-19.
13.Wilkes,B.M.,A.S. Ruston, P. Mento, E.Giraldi,D.Hart, M. V.Molen, R.Barnett,andE. P. Nord. 1991.Characterizationof endothelin 1 receptor and
signal transductionmechanisms in rat medullaryinterstitialcells. Am. J.Physiol.
260:F579-F589.
14. Kohan, D. E.1991. Endothelinsynthesis byrabbit renal tubule cells. Am. J.Physiol.261:F221-F226.
15.Kohan,D. E., and E.Padilla. 1993. Osmolarregulationof endothelin-I productionbyratinnermedullary collecting duct.J.Clin. Invest.91:1235-1240. 16. Kohan, D. E., and A. K. Hughes. 1993. Autocrine role of endothelin in ratIMCD: inhibition ofAVP-inducedcAMPaccumulation. Am. J.Physiol. 265:F129-F129.
17.Yanagisawa, M.,H.Kurihara,S.Kimura,Y.Tomobe,M.Kobayashi, Y.
Mitsui,Y.Yazaki,K.Goto,and T. Masaki. 1988. A novel potent vasoconstrictor
peptide produced byvascular endothelial cells. Nature(Lond.). 332:411-415.
18.Kohmoto, O.,H.Ikenouchi,Y.Hirata,S.-I.Momomura,T.Serizawa,and W. H. Barry. 1993. Variable effects ofendothelin-1 on [Ca2+],transients,pHi,
and contraction in ventricular myocytes. Am. J.Physiol. 265:H793-H800.
19.Simonson, M.,and M. J.Dunn. 1990.Endothelin-1stimulates contraction of ratglomerularmesangialcells andpotentiates 3-adrenergic-mediated cyclic
adenosinemonophosphate accumulation.J.Clin.Invest. 85:790-797. 20.Bradford,M. M. 1976. Arapidand sensitive method forquantitationof
microgram quantitiesofprotein utilizingtheprincipleofprotein-dye binding. Anal.Biochem. 72:248-254.
21.Markewitz,B.A.,J. R.Michael,and D. E. Kohan. 1993.Cytokine-induced expression ofanitric oxidesynthaseinratrenal tubule cells. J. Clin. Invest. 91:2138-2143.
22.Barnett,R.L.,L.Ruffini,D.Hart,P.Mancuso,and E. P.Nord. 1994. Mechanism of endothelinactivation of phospholipase A2in rat renalmedullary
interstitial cells. Am. J.Physiol. 266:F46-F56.
23.Miura, K., T.Yukimura,Y.Yamashita,T. Shimmen,M.Okumura, S.
Yamanaka,M.Imanishi,and K.Yamamoto. 1991. Renal and femoral vascular responses toendothelin-1 indogs:role ofprostaglandins. J. Pharmacol. Exp. Ther. 256:11-17.
24.Boulanger,C.M.,and T. F.Luscher.1991.Hirudin and nitrates inhibit the thrombin-induced release of endothelin from theintact porcineaorta. Circ. Res.68:1768-1772.
25.Madeddu,P.,V.Anania,P. P.Parpaglia,C.Troffa,A.Pazzola,A.Soro,
P. Manunta, G.Tonolo,M. P.Demontis,M. V.Varoni,etal. 1991.
Endothelin-1-inducedrenal vasoconstriction is bluntedby enalaprilatand enhancedbyEDRF
antagonist inawakenormotensiverats.Clin. Invest. Med. 14:600-606.
26.Wexler, A.,R.Kalaba,and D. Marsh.1991.Three-dimensional anatomy and renal concentrating mechanism. I. Modeling results. Am. J. Physiol. 260:F368-F383.
27.Han, J.,K.Thompson,C.-L.Chou,and M.Knepper. 1992.Experimental
testsof three-dimensional model ofurinary concentratingmechanism. J. Am. Soc.
Nephrol. 2:1677-1688.
28.Friedlaender,M.M.,D.Jain,Z.Ahmed,D.Hart,R. L.Barnett,and E. P. Nord.1993. Endothelin activation ofphospholipaseD:dual modulationby protein
kinaseC and Ca2.Am.J.PhysioL 264:F845-F853.
29.Pupilli, C.,M.Brunori,N.Misciglia,C.Sell,L.Ianni,M.Yanagisawa,
M.Mannelli,and M.Serio. 1994.Presenceand distribution ofendothelin-1gene
expressionin humankidney.Am. J.Physiol.267:F679-F687.
30.Nadler,S. P.,J. A. Zimplemann, and R. L. Hebert. 1992. Endothelin inhibitsvasopressin-stimulatedwaterpermeabilityinratterminal innermedullary collectingduct. J. Clin. Invest. 90:1458-1466.
31.Oishi, R.,H.Nonoguchi,K.Tomita,and F. Marumo. 1991. Endothelin-1 inhibits AVP-stimulatedosmoticwaterpermeabilityinratmedullarycollecting
duct. Am. J.Physiol.261:F951-F956.
32.Zeidel,M.L.,H. R.Brady,B. C.Kone,S. R.Gullans,and B. M. Brenner. 1989.Endothelin,apeptideinhibitorofNa+-K+-ATPasein intact tubular
epithe-lialcells. Am. J.Physiol.257:C1l10-C1107.
33.Kohan,D.E.,E.Padilla,and A. K.Hughes.1993. Endothelin B receptor mediates ET-1 effectsoncAMPand PGE2 accumulation inratIMCD. Am. J.
