Primary Culture and Characterization of HumanRenal Inner Medullary Collecting Duct Epithelial CellsLakshmipathi Khandrika, Fernando J. Kim, Adriano Campagna, Sweaty Koul,Randall B. Meacham and Hari K. Koul*From the Signal Transduction and Molecular Urology Laboratory, Program in Urosciences, Division of Urology, Department of Surgery,School of Medicine (LK, FJK, AC, SK, RBM, HKK) and University of Colorado Comprehensive Cancer Center (HKK), University ofColorado at Denver and Health Sciences Center, Denver, Colorado

Purpose: Our understanding of physiological and pathophysiological events associated with inner medullary collecting ductepithelium is based on studies in cells isolated from mice and rats. We established primary cultures of hIMCD (humanpapillary collecting duct epithelial) cells.Materials and Methods: Normal papillary tissues were dissected from the surgical waste of consenting patients undergoingrenal surgery. Tissues were digested enzymatically. Cells were maintained in Dulbecco’s modified Eagle’s medium supple-mented with glucose and antibiotics. Cultures were treated with ethylenediaminetetraacetic acid and epithelial selectmedium was also used to obtain a pure epithelial culture.Results: The hIMCD cells grew in a monolayer. Cells showed the expression of epithelial specific markers, includingcytokeratin, the tight junction marker zonula occludens 1 and the cytoskeletal protein vimentin. They lacked expression offactor VIII, which is a glycoprotein synthesized by endothelial cells. To our knowledge we also noted for the first timeuroplakin expression in collecting duct epithelial cells. This expression was maintained in primary culture. The hIMCD cellsin culture were highly resistant to hypertonic solutions and they responded to hypertonicity by cyclooxygenase-2 overexpression. Moreover, these cells also survived prolonged periods of hypoxia.Conclusions: To our knowledge this is the first report of successful culture and characterization of primary cultures ofcollecting duct epithelial cells from human renal papillae. These cells will serve as essential tools in helping us fill the gapsin our understanding of the events associated with the physiology and pathophysiology of human renal inner medullarycollecting duct epithelium.

Key Words: kidney; kidney tubules, collecting; cell culture techniques; epithelium, toxicity

The process of urine concentration and maintenance ofwater homeostasis occurs in the renal papillae withthe loop of Henle and collecting duct having a major

role.1 There exists a complex interplay between nephronsand the vasculature to ensure proper kidney functioning.IMCD cells in the papillae are subjected to a harsh livingenvironment with varying concentrations of metabolites inurine2 and low oxygen tension.3 A mechanism by whichIMCD cells are protected from the adverse effects of theurinary environment may be due to the over expression ofprostaglandins. Studies in rats subjected to high sodiumchloride stress have shown over expression of one of theenzymes involved in prostaglandin synthesis, COX-2, inpapillary tissue.4 Notwithstanding anatomical differences inthe kidney among species, hIMCD cells are also subjected to

Submitted for publication August 22, 2007.Study received Colorado Multiple Institutional Review Board ap-

proval.Supported by NIH-DK-RO1-54084 (HK), University of Colorado at

Denver Health Sciences Center School of Medicine and Departmentof Surgery Academic Enrichment Funds.

* Correspondence: Division of Urology, Department of Surgery,School of Medicine, University of Colorado at Denver and HealthSciences Center, 4200 East Ninth Ave., C-317, Denver, Colorado

80262 (telephone: 303-315-2383; FAX: 303-315-1252; e-mail: hari.koul@uchsc.edu).

0022-5347/08/1795-2057/0THE JOURNAL OF UROLOGY®

Copyright © 2008 by AMERICAN UROLOGICAL ASSOCIATION

2057

such a harsh urine environment. However, there are manygaps in our knowledge regarding the mechanisms by whichhuman kidney cells protect themselves from damage posedby the toxic environment.

Cell lines that represent the different regions of thenephron have been established from several species of animalkidneys.5 However, the study of the mechanism of urine for-mation and concentration in the human renal papillae and therole of various urinary metabolites in this process is limiteddue to the paucity of an in vitro model system. Most functionaldata on IMCD cells are derived from studies of mouse IMCDcells in culture.6 Similarly our knowledge of the effect of vari-ous metabolites in distal and collecting duct epithelial cellsthat is associated with renal pathophysiology, eg oxalate neph-rotoxicity, is also extrapolated from mouse (mIMCD) and dog(MDCK) renal epithelial cells.7 The correlation between stud-ies in mice and the physiology of the handling of various sol-utes in the human kidney is possible only by extending thesestudies in a human model system. To our knowledge there is noreport of studies of hIMCD cells in culture to date.

We established primary cultures of hIMCD cells from thesurgical waste of consenting patients undergoing renal sur-gery. The results presented clearly establish conditions forsuccessful primary cultures of epithelial cultures of hIMCD

cells, as demonstrated by marker protein expression. To our

Vol. 179, 2057-2063, May 2008Printed in U.S.A.

DOI:10.1016/j.juro.2008.01.007

HUMAN RENAL INNER MEDULLARY COLLECTING DUCT EPITHELIAL CELLS2058

knowledge we also report for the first time UP expression incollecting duct cells in intact kidney as well as under primaryculture conditions. Moreover, hIMCD cells retain the func-tional characteristics of surviving under hypoxic and hyper-tonic conditions, and responding to hypertonicity by increasedCOX-2 expression. These cells should help us fill gaps in ourunderstanding of the physiological and pathophysiologicalevents associated with collecting duct epithelium.

MATERIALS AND METHODS

Chemicals and ReagentsThe tissue digesting enzymes collagenase, hyaluronidase andDNaseI, other chemicals, and EDTA and Ficoll were obtainedfrom Sigma®. Tissue culture medium, PBS, the antibioticspenicillin and streptomycin, fluorophore tagged secondary an-tibodies and DAPI were obtained from Invitrogen®. MEM D-valine modification medium was obtained from United StatesBiochemical Corp., Cleveland, Ohio. Antibodies against cyto-keratin, vimentin, ZO-1, factor VIII, and UPIa and Ib wereobtained from Santa Cruz Biotechnology, Santa Cruz, Cali-fornia.

Cell Isolation and Purification From TissueTissues were collected from consenting patients in accor-dance with a Colorado Multiple Institutional Review Boardapproved protocol. Portions of normal renal medulla wereisolated from kidney surgical waste tissue. Immediately fol-lowing surgery the tissue was transported from the operat-ing room on ice. It was processed within 15 to 20 minutes ofsurgical removal of the kidney. Papillary tissue was dis-sected from the medulla under sterile conditions in the lab-oratory and chopped into small pieces on ice.

Cells were isolated from tissue essentially using themethod of Zager et al8 with modifications. Briefly, cells wereisolated by digesting tissue in sterile DMEM containingcollagenase type IV (2 mg/ml), hyaluronidase type V (0.2mg/ml) and DNaseI (24 ng/ml) in the presence of penicillin G(150 U) and streptomycin sulfate (150 �g) antibiotics. Diges-tion was continued for 16 to 20 hours at room temperaturewith constant gentle shaking. After digestion the mediumwas filtered through a 100 �m BD Falcon™ nylon cellstrainer to remove undigested tissue and other large debris.The resulting cell suspension was centrifuged at 1,800 rpm for8 minutes at room temperature to pellet the cells. The cellswere washed twice in sterile PBS and recovered by centrifuga-tion, as described. A Ficoll density gradient centrifugation stepwas done to remove cell debris and red blood cells from the cellsuspension.9 Renal cells separated in the gradient werewashed with PBS and recovered by centrifugation.

Cell Culture and PassageCells were seeded in 25 mm flasks in DMEM supplementedwith 10% FBS and antibiotics and maintained at 37C in ahumidified 5% CO2 environment. After 3 days nonadheringcells were removed by washing the cells with sterile PBS.

Removal of Fibroblasts From CultureTo remove fibroblast contamination cultures were treatedwith 0.02% EDTA in PBS for 3 minutes at 37C and then

tapped vigorously to remove the loosely adherent fibro-

blasts.10 When indicated, this procedure was repeated atleast twice.

Epithelial Selection and Cell PassageTo select epithelial cells specifically the cells were main-tained in MEM eagle D-valine modification medium at 37Cin a humidified 5% CO2 environment.11 A confluent mono-layer of papillary cells was passaged with 0.25% trypsin-EDTA and incubation at 37C for 6 minutes. The cells wereremoved by gently tapping the sides of the flask and thenseeded at appropriate cell density in a new flask.

Expression of Specific MarkersMarker protein expression was visualized by antibodies di-rected against the markers. In these studies mouse mono-clonal antibody against human cytokeratin proteins (cytoker-atins 4, 5, 6, 8, 10, 13 and 18), rabbit polyclonal antibodiesagainst human ZO-1, vimentin and type 1 factor VIII relatedantigen light chain, and goat polyclonal antibodies againstUPIa and UPIb were used.

Briefly, cells were grown to about 50% confluence on8-well chamber slides (Nalge® Nunc™), washed with sterilePBS, fixed for 20 minutes in 3.7% formaldehyde and perme-abilized for 5 minutes at �20C in 70% acetone, and 30%methanol mixture. After blocking with 10% serum the cellswere incubated with an appropriate dilution of primary an-tibody for 1 hour at room temperature. Excess antibody wasremoved by washing the cells twice in PBS for 5 minuteseach. Bound antibody was detected by incubation with FITCor Cy3 conjugated secondary antibodies. Nuclei were coun-terstained with DAPI (10 �g/ml) for 5 minutes and thenwashed with PBS again. Images were captured using anRXDA immunofluorescence microscope (Leica, Solms, Ger-many). Appropriate negative controls were included to as-certain positive antibody reactivity.

Hematoxylin and eosin staining of paraffin embedded renalpapillary tissue was done using standard procedures.12 Anti-gen retrieval for immunostaining was performed by heatingthe slides in citrate buffer (10 mM sodium citrate and 1 mMEDTA, pH 6.2) for 15 minutes in a microwave oven and thenhybridizing with antibodies, as previously described.13 Pri-mary antibodies were detected by fluorescent labeled second-ary antibodies and images were captured as described.

Hyperosmolarity ExperimentsIn these studies the confluent monolayer of cells was ex-posed to increasing concentrations of sodium chloride inHanks balanced salt solution for 2 hours. Cell viability wasdetermined by crystal violet staining, as previously de-scribed.12 Changes in cell morphology upon exposure to in-creasing osmolal solutions was detected by staining with 10ng/ml FITC conjugated wheat germ agglutinin (Invitrogen).Wheat germ agglutinin selectively binds to N-acetylglu-cosamine and N-acetyl neuraminic acid (sialic acid) resi-dues, which are commonly present in cell membranes.

Hypoxia/Reoxygenation EnvironmentThe cultures were placed in a modular incubator and flushedwith 5% CO2 balanced N2 gas to simulate hypoxic condition.The cells were incubated at 37C for 24 hours and then

allowed to reoxygenate for an additional 24 hours.

HUMAN RENAL INNER MEDULLARY COLLECTING DUCT EPITHELIAL CELLS 2059

Cell Survival MeasurementsCell survival was measured by MTT assay14 or crystal violetstaining, as described previously.12 For MTT assay at theend of predetermined times the medium was removed and100 �l 1 mg/ml MTT solution were added and incubated at37C for 3 hours. Sodium dodecyl sulfate (100 �l 10%) made in0.01 N HCl was added and incubated overnight to extract theformazan crystals that had formed in the cells. Color intensitywas measured at 540 nm using an MRX multiwell reader(Dynex Technologies, Chantilly, Virginia). For crystal violetstaining cells were stained with 0.4% crystal violet in 0.2 Mcitric acid for 1 hour. Excess crystal violet was removed bymultiple washes in deionized water. Crystal violet incorpo-rated in cells was extracted overnight with 2-ethoxy ethanoland quantitated by measuring absorbance at 560 nm.

Gene Expression AssayGene expression was monitored by relative quantitativepolymerase chain reaction. Briefly, RNA was extracted fromhIMCD cells using an RNEasy® mini kit and cDNA was pre-pared using an iScript™ second strand cDNA synthesis kit.

Expression of Cox-2 mRNA and GAPDH was monitoredby relative quantitative polymerase chain reaction usingcertain primer sets, including Cox-2 forward primer5=AGGGCCAGCTTTCACCAAC-3= and reverse primer5=GGAGGATACATCTCTCCATCAA-3=, and GAPDH for-ward primer 5=ACCACAGTCCATGCCATCAC-3= and re-verse primer 5=TCCACCACCCTGTTGCTGTA-3=). Cox-2and GAPDH transcripts were separated by agarose gel elec-trophoresis and expression was quantitated by densitome-try. The ratio between Cox-2 and GAPDH expression levelswas determined to normalize the expression levels of Cox-2transcripts.

RESULTS

Culture of hIMCD Cells in Epithelial SelectionMedium was Necessary for Epithelial SelectionDespite careful dissection of the epithelial region fibroblastgrowth could not be prevented (fig. 1, A). Fibroblasts, whichwere identified by their spindle shape and unidirectionalgrowth, were detected in the initial stages of new cultures.They are an inherent problem in all new cultures from tissueisolates. As a first step, the cultures were deprived of fibro-blastic growth by EDTA treatment, as described. In addi-tion, selective epithelial cell growth was promoted usingESM nutrient medium, which selectively supports epithelial

FIG. 1. Phase contrast images of renal papillary epithelial cells in

culture. A, initial collecting duct cell cultures show contamination withfibroblasts. B, pure hIMCD cells cultures with epithelial morphology.

cell growth.11 After repeated EDTA treatments and succes-sive passages in ESM the cultures were determined to beprimarily epithelial, as determined by their morphology andgrowth characteristics (fig. 1, B). Cells in primary culturemaintained their epithelial integrity, as assessed by trypanblue exclusion, cell morphology and contact inhibition, as istypical of epithelial cells in culture. Moreover, epithelialcells continued to grow up to 5 passages, after which thecells lost their growth potential.

Primary hIMCD Cell Cultures ShowedExpression of Specific Epithelial MarkersThe cells in primary culture were further probed for expres-sion epithelial specific markers using immunofluorescencemicroscopy. The hIMCD cells expressed cytokeratins, a class

FIG. 2. Renal papillary cells in culture expressed epithelial cellmarkers. hIMCD cells were grown on 8 chamber slides and fixed informaldehyde. Epithelial cell marker protein cytokeratin, cell junc-tion protein ZO-1 and vimentin were hybridized using primary anti-bodies for respective proteins and detected with Cy3 conjugated sec-ondary antibodies. Cells in primary cultures were negative for factorVIII, suggesting that cultures were devoid of any endothelial cells.

of intermediate filaments necessary to maintain the overall

increased cell number compared to cells grown in normoxic environ-ment. Asterisk indicates 1-way ANOVA p �0.05.

medium. Relative survival at higher osmolarity was compared to that at iwith FITC conjugated wheat germ agglutinin.

HUMAN RENAL INNER MEDULLARY COLLECTING DUCT EPITHELIAL CELLS2060

cellular structural integrity (fig. 2, A). Cells also expressedthe tight junction protein ZO-1 and the intermediate fila-ment protein vimentin (fig. 2, B and C). Moreover, cells werefound to be negative for factor VIII related antigen, theendothelial cell marker (fig. 2, D). These results are consis-tent with pure epithelial cultures.

Primary hIMCD Cell CulturesWere Adapted for Survival in HypoxiaCells were grown in a hypoxic environment to assess theirability to survive in low oxygen conditions. The hIMCD cellscontinued to survive in a low oxygen environment (fig. 3).Moreover, we observed a statistically significant increase incell numbers following hypoxia as well as reoxygenationcompared to that in cells in normoxia for the entire studyduration (about 40% and 80% more, respectively). Theseresults indicate that hIMCD cells demonstrate adaptabilityto diverse environmental conditions with a preference for ahypoxic environment.

ty changes. A, hIMCD cells were incubated in Hanks balanced saltstained with crystal violet following 24 hours in normal growth

FIG. 3. Renal papillary cells in culture were resistant to hypoxia.hIMCD cells were incubated in hypoxic medium under 5% CO2balanced N2 gas environment for 24 hours and then reoxygenated.Cell survival was determined by MTT assay. hIMCD cells hadenhanced growth potential in hypoxic environment, as seen by

FIG. 4. Renal papillary cells in culture were highly resistant to osmolarisolution with increasing sodium chloride concentration for 2 hours and

nitial osmolarity of 300 mOsm. B, staining for membrane glycolipids

HUMAN RENAL INNER MEDULLARY COLLECTING DUCT EPITHELIAL CELLS 2061

Primary hIMCD Cell CulturesWere Resistant to HypertonicityA characteristic of epithelial cells in the papillae is theirability to survive in hyperosmolar solutions. Results showedthat hIMCD cells had excellent survival under hyperosmo-lar conditions with 100% survival at up to 700 mOsm and80% survival at 800 mOsm, compared with that at the initialosmolarity of 300 mOsm (fig. 4, A). However, there was adrastic decrease in cell survival in medium at above 900mOsm. Even at those extreme conditions at least 30% of thecells survived.

Moreover, these cells maintained membrane integrity un-der high osmolar solutions, although at higher concentra-tions they appeared large and bulged (fig. 4, B). Takentogether these results demonstrate the resilience of primarycultures of hIMCD cells over a wide range of osmolal condi-tions.

Primary hIMCD Cell Cultures Responded toHypertonicity by COX-2 Gene Over ExpressionFigure 5 shows the up-regulation of COX-2 mRNA levels inhIMCD cells upon exposure to hyperosmolal solutions for 2hours. hIMCD cells showed an increase in COX-2 mRNAlevels with increasing osmolarity of up to 600 mOsm. More-over, Cox-2 expression directly correlated with the amountof osmolarity.

UP was Selectively Expressed in hIMCD CellsUPs are specific markers for urothelial differentiation. Im-munostaining of cultured hIMCD cells revealed the presence

FIG. 5. Renal papillary cells responded to hyperosmolarity by in-creasing COX-2 expression. hIMCD cells were exposed to increasingsodium chloride concentration in medium for 2 hours. Total RNAwas extracted. Cox-2 and GAPDH gene expression was monitored.A, Cox-2 and GAPDH expression after exposure to increasing osmo-

larity. B, Cox-2-to-GAPDH expression relative ratio. Asterisk indi-cates expression vs that at initial 300 mOsm p �0.05.

of UPs, as represented by staining for UPIa and Ib (fig. 6, A).To confirm the expression of UP proteins in IMCD cellstissue sections from normal renal papillary tissue were im-munostained (fig. 6, B). Cytokeratin, and UPIa and Ib werefound to be co-localized in the papillary tubules (fig. 6, C).These results show that hIMCD cells are distinct from prox-imal tubular epithelial cells since UPs are not expressed inthe proximal tubular epithelium.

DISCUSSION

Isolated hIMCD cells were initially cultured in T25 cm2

flasks in DMEM medium with 4.5% glucose and 10% serum.Cells showed a good growth potential but cultures wererapidly taken over by fibroblast cells that could have carriedover from the originally isolated tissue. Fibroblasts have agrowth advantage compared to epithelial cells in DMEMand consequently the percent of fibroblastic cells increasedwithin a few days of growing the epithelial cells in culture(fig. 1, A).

To overcome this limitation it has been suggested thattreatment with EDTA selectively impairs the ability of fi-broblasts to adhere and, thus, they can be removed by suchtreatments.10 In our hands repeated treatment of the cellswith EDTA did in fact remove many fibroblasts but suchtreatment was not sufficient to completely remove all fibro-blasts. Moreover, in DMEM the fibroblasts that remainedcontinued to divide more rapidly than the neighboring epi-thelial cells and their number kept increasing with time. Toovercome this limitation we took advantage of the propertyof epithelial cells to selectively use unnatural D-valine.11

Epithelial cells have been shown to have the enzyme D-amino acid oxidase and use D-valine medium, whereas fi-broblasts lacking the enzyme are unable to grow in suchmedium. Although D-valine does not act as a selectivepoison for fibroblasts, it limits their growth potential andeventually all fibroblasts die. Thus, by using EDTA treat-ment to remove fibroblasts and growing the cells in ESMwe were successful in obtaining a pure epithelial cellculture (fig. 1, B).

It is important to point out that the initial presence offibroblasts varied among nephrectomy samples. In sampleswith a low number of fibroblasts pure cultures were easilyobtained in ESM alone without any EDTA pretreatmentsnecessary to remove fibroblast growth. Thus, careful moni-toring of cultures and the timely removal of any unwantedfibroblasts are essential to successfully establish primarycultures of epithelial cells from renal papillary tissue. Ifexperimental conditions permit, it would be advisable tomaintain these cultures in ESM, as in the current study.

Early cultures showed dispersed growth but they hadrapid growth potential, attaining confluence after a week.Confluent cultures contained the typical cobblestone mor-phology characteristic of many epithelial cell cultures andthey were contact inhibited. A major limitation of papillarycells is their loss of growth after 5 or 6 passages. Cells growwell even at a low glucose concentration and, although fi-broblasts were not detected at subsequent passages, ESMwas used throughout to maintain the epithelial phenotype.

Primary hIMCD cultures showed growth and morpholog-ical characteristics of typical epithelial cells in culture. Tofurther characterize the cells immunostaining for specific

epithelial cell markers was done. Positive reactivity against

in fotin, a

HUMAN RENAL INNER MEDULLARY COLLECTING DUCT EPITHELIAL CELLS2062

vimentin and cytokeratin proteins further confirmed theepithelial nature of the primary cultures of hIMCD cells.Lack of staining for factor VIII, which is a specific endothe-lial cell marker, also suggested that the cultures were of asingle type. This reactivity pattern is similar to that ob-served in the HK-2 human proximal tubule epithelial cellline.15

Epithelial cell cultures are known to form tight junctionsamong individual cells, as shown in cultures of rat papillarycollecting duct cells.16 Epithelial cells in culture in the cur-rent study were derived from the papillae of human kidneysand stained positive for ZO-1, which is one of the proteinsfound in cell junctions. Cells in culture showed dispersedgrowth in the early days and immunostaining detected aneven distribution of the cell junction protein ZO-1 in thecytoplasm. Steimer et al reported a similar distribution ofZO-1 in early porcine alveolar epithelial cell cultures withlittle distribution at the cell border, although confluent cul-tures showed continuous staining along the cell circumfer-ence.17

Epithelial cells lining the IMCD are constantly bathed inurine and as such they are subjected to fluctuations in os-molarity compared to the rest of the cells of the kidney. Theability of these cells to survive in a harsh environmentsuggests a high level of adaptability of these cells.2 Weobserved that IMCD cells could survive in hyperosmolarsolutions (fig. 4, B). Perturbations in this cellular environ-ment may lead to the accumulation of undesirable toxinsthat are otherwise excreted in the urine and may affect theability of cells to continue to grow. To this end we observedthat there was no loss of cell viability even in 700 mOsmsolution for 2 hours and about 50% of the cells survived up to4 hours even at 800 mOsm (data not shown). However, these

FIG. 6. UP expression in renal papillary epithelial cells and humanconfirmed by immunostaining normal human renal medulla fixedsection. H & E, reduced from �10. C, immunostaining for cytokera

cells were not subjected to these hyperosmolar solutions for

such long periods as in the human kidney, although theability to survive for extended periods demonstrated theadaptability of these cells to the harsh urine environment.When cells exposed to hyperosmolar solutions for 2 hourswere allowed to return to normal medium, they continued togrow and divide, indicating that periods of exposure to hy-perosmolar solutions do not irreversibly impair the growthproperty of these cells. These responses to hyperosmolarityare in agreement with those in several previous studies inintact tubules or cells isolated from mice and rats.18

While many mechanisms have been proposed that allowcells to survive hypertonic conditions, to our knowledge theirrelevance to primary hIMCD cell cultures is not known. Onestudy showed that COX-2, an enzyme involved in the synthesisof membrane prostaglandins, is up-regulated in mIMCD cellsupon exposure to hyperosmolarity.6 Our results of the induc-tion of COX-2 expression under hypertonic conditions are inagreement with this study (fig. 5). Prostaglandins are known tofunction in water diuresis in rat renal medullary interstitialcells.4 However, we have not directly evaluated the role ofCOX-2 expression in protecting hIMCD cells against alter-ations in medium osmolarity.

The papillary region is most hypoxic areas of the humankidney and as such cells from this region should be able towithstand prolonged periods of hypoxia. Our studies dem-onstrate that primary cultures of hIMCD cells showed ex-cellent survival in hypoxic environment and they continuedto grow upon reoxygenation (fig. 3). These results are inagreement with those in previous studies by Campos et alshowing that IMCD cells from rats were resistant to hypoxiafor 24 hours.19 Due to the constant presence of urine in thepapillary tip the cells depend on the capillary blood flow toobtain the required oxygen. Thus, the ability of primary

l papillae. A, hIMCD cells in culture. Papillary UP expression wasrmaldehyde and paraffin embedded. Reduced from �63. B, tissuend UPIa and Ib expression. Reduced from �63.

rena

cultures of hIMCD cells to withstand hypoxia may suggest

HUMAN RENAL INNER MEDULLARY COLLECTING DUCT EPITHELIAL CELLS 2063

that IMCD cells are well adapted to ensure proper function-ing even at the low oxygen pressure typical of the renalpapillae.

UPs are the markers of urothelial differentiation andthey are expressed only in urothelial cells during advancedstages of differentiation. They are major integral membraneproteins and they have a role in strengthening the urothe-lium.20 The UP family comprises 4 proteins and in theurothelium UPIa and Ib form tightly packed structures withUPII and UPIII. The hIMCD cells in culture showed stain-ing for UPIa and Ib proteins (fig. 6, A). This staining wasspecific and not found in cortical epithelial cells (data notshown). To our knowledge this is the first report of thepresence of UPs in hIMCD cells. We extended these obser-vations to evaluate UP expression directly in tissue sectionsfrom the renal medullary region. Results confirmed thatindeed UPIa and Ib are also expressed in intact kidney andexpression is limited to epithelial cells in the collecting ducts(fig. 6, C). While several roles have been attributed to UPs,including acting as a barrier to the external environment,the role of UPs in the ability of hIMCD cells to survive in thehyperosmolar environment was not directly evaluated in thecurrent study. Additional studies are needed to addressthe role of UPs in hIMCD cells.

CONCLUSIONS

To our knowledge this is the first report of successful estab-lishment of a primary culture from hIMCD cells. IsolatedhIMCD cells showed a characteristic growth pattern typicalof other epithelial cell cultures. The hIMCD cells retainedmorphological as well as functional features of collectingduct epithelium. Establishment of the hIMCD cell primaryculture should help us fill the gaps in our understanding ofthe physiological and pathophysiological conditions associ-ated with epithelial cells lining the hIMCD.

Abbreviations and Acronyms

COX � cyclooxygenaseDAPI � 4,6-diamidino-2-phenylindole

DMEM � Dulbecco’s MEMEDTA � ethylenediaminetetraacetic acid

ESM � epithelial selection mediumFITC � fluorescein isothiocyanate

GAPDH � glyceraldehyde-3-phosphatedehydrogenase

hIMCD � human IMCDIMCD � inner medullary collecting ductMEM � modified Eagle’s medium

PBS � phosphate buffered salineUP � uroplakinZO � zonula occludens

REFERENCES

1. Taniguchi J and Imai M: Computer analysis of the significanceof the effective osmolality for urea across the inner medul-lary collecting duct in the operation of a single effect for thecounter-current multiplication system. Clin Exp Nephrol2006; 10: 236.

2. Neuhofer W and Beck FX: Survival in hostile environments:strategies of renal medullary cells. Physiology (Bethesda)

2006; 21: 171.

3. Miller RL and Kohan DE: Hypoxia regulates endothelin-1 pro-duction by the inner medullary collecting duct. J Lab ClinMed 1998; 131: 45.

4. Pucci ML, Endo S, Nomura T, Lu R, Khine C, Chan BS et al:Coordinate control of prostaglandin E2 synthesis and up-take by hyperosmolarity in renal medullary interstitialcells. Am J Physiol Renal Physiol 2006; 290: F641.

5. Toutain H and Morin JP: Renal proximal tubule cell culturesfor studying drug-induced nephrotoxicity and modulation ofphenotype expression by medium components. Ren Fail1992; 14: 371.

6. Rauchman MI, Nigam SK, Delpire E and Gullans SR: Anosmotically tolerant inner medullary collecting duct cellline from an SV40 transgenic mouse. Am J Physiol 1993;265: F416.

7. Moryama MT, Domiki C, Miyazawa K, Tanaka T and SuzukiK: Effects of oxalate exposure on Madin-Darby canine kid-ney cells in culture: renal prothrombin fragment-1 mRNAexpression. Urol Res 2005; 33: 470.

8. Zager RA, Schimpf BA and Gmur DJ: Physiological pH. Effectson posthypoxic proximal tubular injury. Circ Res 1993; 72:837.

9. Chatterjee S, Gupta P and Kwiterovich PO Jr: Separation ofhuman urinary proximal tubular cells from familial hyper-cholesterolemic homozygotes by Ficoll gradient centrifuga-tion. Morphological and biochemical characteristics. Vir-chows Arch B Cell Pathol Incl Mol Pathol 1984; 45: 365.

10. Sordillo LM, Oliver SP and Akers RM: Culture of bovine mam-mary epithelial cells in D-valine modified medium: selectiveremoval of contaminating fibroblasts. Cell Biol Int Rep1988; 12: 355.

11. Hawksworth GM: Isolation and culture of human renal corti-cal cells with characteristics of proximal tubules. MethodsMol Med 2005; 107: 283.

12. Maroni PD, Koul S, Chandhoke PS, Meacham RB and KoulHK: Oxalate toxicity in cultured mouse inner medullarycollecting duct cells. J Urol 2005; 174: 757.

13. Sheriffs IN, Rampling D and Smith VV: Paraffin wax embed-ded muscle is suitable for the diagnosis of muscular dystro-phy. J Clin Pathol 2001; 54: 517.

14. van de Loosdrecht AA, Beelen RH, Ossenkoppele GJ, Broek-hoven MG and Langenhuijsen MM: A tetrazolium-basedcolorimetric MTT assay to quantitate human monocyte me-diated cytotoxicity against leukemic cells from cell linesand patients with acute myeloid leukemia. J ImmunolMethods 1994; 174: 311.

15. Ryan MJ, Johnson G, Kirk J, Fuerstenberg SM, Zager RA andTorok-Storb B: HK-2: an immortalized proximal tubuleepithelial cell line from normal adult human kidney. Kid-ney Int 1994; 45: 48.

16. Husted RF, Hayashi M and Stokes JB: Characteristics of pap-illary collecting duct cells in primary culture. Am J Physiol1988; 255: F1160.

17. Steimer A, Laue M, Franke H, Haltner-Ukomado E and LehrCM: Porcine alveolar epithelial cells in primary culture:morphological, bioelectrical and immunocytochemical char-acterization. Pharm Res 2006; 23: 2078.

18. Pihakaski-Maunsbach K, Tokonabe S, Vorum H, Rivard CJ,Capasso JM, Berl T et al: The gamma-subunit of Na-K-ATPase is incorporated into plasma membranes of mouseIMCD3 cells in response to hypertonicity. Am J PhysiolRenal Physiol 2005; 288: F650.

19. Campos SB, Ori M, Dorea EL and Seguro AC: Protective effectof L-arginine on hypercholesterolemia-enhanced renal ische-mic injury. Atherosclerosis 1999; 143: 327.

20. Garcia-Espana A, Chung PJ, Zhao X, Lee A, Pellicer A, YuJ et al: Origin of the tetraspanin UPs and their co-evolutionwith associated proteins: implications for UP structure and

function. Mol Phylogenet Evol 2006; 41: 355.