Urea Restrains Aldosterone-Induced Development of Peanut Agglutinin–Binding on Embryonic Renal Collecting Duct Epithelia
Karl Schumacher,
Raimund Strehl and
Will W. Minuth
Department of Molecular and Cellular Anatomy, University of Regensburg, Regensburg, Germany
Correspondence to Dr. Karl Schumacher, Department of Molecular and Cellular Anatomy, University of Regensburg, Universitätstrasse 31, D-93053 Regensburg, Germany. Phone: +49-941-943-2875; Fax: +49-941-943-2868;
ABSTRACT. Peanut agglutinin (PNA) represents a commonly usedmarker for -type intercalated (IC) cells and their distributionin the corticomedullary course of the collecting duct (CD) inthe mature rabbit kidney. It has been shown that aldosteroneis able to generate >90% of PNA-binding cells in an embryonicCD epithelium in vitro. In adult kidney, a maximum of only 25%PNA-positive cells is found in the cortical segment of the CD,and PNA-binding completely disappears in the inner-medullaryCD. Molecules that regulate the gradual development of CD-specificcells during organ growth are unknown. In the present experiments,it was found that addition of physiologic concentrations ofurea to the culture medium is able to restrain the action ofaldosterone in embryonic CD epithelia. Urea antagonizes in aconcentration-dependent manner the action of aldosterone finallyleading to only 10% of PNA-binding cells. The data point toa urea-specific effect, because osmolytes such as NaCl and mannitoldid not affect PNA binding. In addition, urea did not influenceexpression of principal-cell typical markers such as AQP2 and3. The findings may explain that a higher number of PNA-positivecells is found in the cortical region of the kidney correlatedwith a low concentration of urea as compared with only few PNA-bindingcells in the medullary CD, where a high concentration of ureaoccurs. Thus, an increasing concentration of urea may triggerthe number of PNA-positive cells in the cortical-medullary courseof the CD during organ development. E-mail: karl.schumacher@vkl.uni-regensburg.de
Embryonic development of the metanephros starts by the morphogenicinteraction between the ureteric bud and the surrounding mesenchyme(1,2). Dichotomous branchings of the ureteric bud epitheliumdetermine first the kidney poles, then the anterior and posteriorhalf of the organ, the number of pyramids, and finally the amountof collecting ducts (CD) that reach the future area cribrosa(1,2). In a subsequent process, the induction of the nephronstakes place. This happens in a highly coordinated process betweenthe CD ampullae as derivatives of the ureteric bud and the capcondensates (3,4). By the exchange of morphogenic signals betweenboth tissues, part of the cap condensate separates and formsat the lateral side of each CD ampulla the pretubular aggregatesas a first visible sign of nephron development (4–8).Organ growth proceeds when the CD ampulla tip divides dichotomouslyand elongates toward the capsule. By this mechanism, each tipof the CD ampulla determines the complete future microarchitectureof the kidney.
The functional development of the CD epithelium occurs in anantiparallel manner as opposed to the growth direction of theCD ampulla (Figure 1). Each CD ampulla in the embryonic partof the developing rabbit kidney consists of three differentzones (9). Zone 1 is located within the tip of the ampulla.It is the niche for epithelial stem cells (9), which are capableof initiating nephrogenesis in the surrounding mesenchymal stemcells of the cap condensate (8,10). Zone 2 comprises the neckof the ampulla. This area appears gray in the light microscopeand contains multiple dividing cells, which cause the elongationof the CD tubule during the dichotomous branching process. Zone3 comprises the shaft of the ampulla and is recognized by theprimary appearance of principal (P) and various types of intercalated(IC) cells (9,11,12).
Figure 1. The antiparallel direction of collecting duct (CD) ampulla growth and CD cell maturation. Zone 1, ampulla tip; zone 2, ampulla neck; zone 3, ampulla shaft.
While the morphogenic signals between the CD ampulla and thecap condensate are being investigated intensively (4–8),little knowledge is available about the mechanism that leadsto the functional development of the CD epithelium. During maturation,the embryonic epithelium within the ampulla shaft develops intoa heterogeneously composed epithelium consisting of functionalP and various types of IC cells (11). This process was investigatedearlier by morphologic (13), immunohistochemical (9), and histochemical(12,14) methods. Especially peanut agglutinin (PNA) served asa histochemical marker to follow the development of -type ICcells (12,14). Although it has been shown in culture experimentsthat -type IC cells may undergo conversion to P cells (15),data obtained from tissue sections of neonatal rabbit kidneyindicated that P and IC cells derive from a common precursorcell located within the CD ampulla by asymmetrical cell division(9,16).
The cell biologic mechanism that leads from a precursor cellwithin the ampulla to functional P and various types of IC cellsis unknown (17). It has been shown that alkali feeding of pregnantmothers triggers an earlier appearance of -type IC cells comparedwith controls (18). Furthermore, it was demonstrated that theelectrolyte environment of cultured embryonic CD epithelia isable to influence the development of P and IC cell features(19). Experiments with a tissue-specific model additionallyshowed that application of aldosterone in the culture mediumincreased the amount of PNA-positive cells in maturing CD epithelia(20). More than 90% PNA-positive cells were consistently found,indicating that a feature of -type IC cells is developed (20).However, this high number of PNA-positive cells is an artificialoverexpression under in vitro conditions and is never foundwithin the adult rabbit kidney (12). Here, a maximum of 10 to25% PNA-positive cells was detected in the cortical collectingduct (CCD). Thus, on the one side, aldosterone promotes thedevelopment of PNA-binding cells under in vitro conditions toa high degree, but, on the other side, this amount of cellsis not found in the adult kidney despite the presence of thehormone. Consequently, the action of aldosterone must be restrainedby a growth factor or by the environment during kidney development.
In the present experiments, we show data of embryonic CD epitheliacultured in the presence of aldosterone and urea. We demonstratethat embryonic CD epithelia tolerate a long-term applicationof urea over 14 d. Physiologic concentrations of urea are ableto modulate the development of PNA binding on cells of the developingCD epithelium and restrain the effect of aldosterone.
Tissue Preparation for Histochemistry
One-day-old New Zealand rabbits were anesthetized with etherand killed by cervical dislocation. Both kidneys were removedimmediately. The kidneys were then cut precisely along the corticomedullaryaxis, frozen in liquid nitrogen, and stored at -80°C upto the subsequent treatments.
Tissue Preparation for Generation of Embryonic CD Epithelia
Generation of embryonic CD epithelia was performed by isolatingcortical embryonic explants from the kidneys of newborn NewZealand rabbits (up to 1 d old) according to methods describedearlier (21). The explants consisted of a piece of capsule fibrosawith adherent CD ampullae, S-shaped bodies, and nephrogenicblastema, which were mounted in tissue carriers. For generationof an embryonic CD epithelium, the carriers were placed for24 h in a 24-well plate in a CO2 incubator (5% CO2/95% air).During culture of the explants in Iscove’s modified Dulbecco’smedium (IMDM; Life Technologies BRL Life Technologies, Eggenstein,Germany) including 10% FBS (Boehringer, Mannheim, Germany),an outgrowth of cells from the CD ampullae was observed. Within24 h, the entire surface of the explant (6 mm in diameter) wascovered by a monolayer of polarized CD cells that were positivefor cytokeratin-19.
Perfusion Culture with Generated CD Epithelia
After initiation of culture, the tissue carriers were placedin a perfusion culture container (Minucells and Minutissue,Bad Abbach, Germany, www.minucells.de). Fresh medium was continuouslyperfused for 14 d at a rate of 1 ml/h with an IPC N8 peristalticpump (Ismatec, Wertheim, Germany). For maintaining a constanttemperature of 37°C, the container was placed on a thermoplate (Medax, Kiel, Germany) and covered by a transparent lid.During perfusion culture, IMDM (order no. 21980-032) withoutserum was used as the standard medium. Aldosterone was usedat a concentration of 1 x 10-7 M (Sigma-Aldrich-Chemie, Deisenhofen,Germany). Antibiotic-antimycotic solution (1%; Life TechnologiesBRL-Life Technologies) was added to all culture media. Furthermore,up to 50 mmol/L HEPES (Life Technologies BRL-Life Technologies)was used in the medium to maintain a constant pH of 7.4 in perfusionculture under atmospheric room air (0.3% CO2). Urea was addedat concentrations of 4.5 (serum value), 9, 13.5, and 18 mmol/L.In additional experiments, the same concentrations of urea wereused in combination with 20 mmol/L NaCl. In detail: 20 mmol/LNaCl + 4.5 mmol/L urea, 20 mmol/L NaCl + 9 mmol/L urea, 20 mmol/LNaCl + 13.5 mmol/L urea, and 20 mmol/L NaCl + 18 mmol/L urea.Resulting osmolalities are shown in Table 1.
Table 1. Measured osmolality in the culture medium after addition of 18 mmol/L urea, 20 mmol/L NaCl, and after simultaneous addition of urea and NaCl to the culture medium
Lectin Incubation
Corticomedullary-oriented cryosections (8 µm) of neonatalrabbit kidney and of the generated epithelium were preparedusing a cryomicrotome (Microm, Heidelberg, Germany). The cryosectionswere fixed in ice-cold ethanol and then washed with PBS. Afterincubation in blocking solution (PBS + 1% BSA + 10% horse serum)for 30 min, the sections were exposed to rhodamine-conjugatedPNA or soybean agglutinin (SBA; Vector Laboratories, Burlingame,CA) diluted 1:1000 in PBS for 45 min. After several washes inPBS, the specimens were embedded with Slow Fade Light AntifadeKit (Molecular Probes, Eugene, OR) and analyzed using an Axioskop2 plus microscope (Zeiss, Oberkochen, Germany). Images weremade by a digital camera and thereafter processed with Photoshop5.5 (Adobe Systems, San Jose, CA).
Immunohistochemistry
Cryosections (8 µm thick) of 1-d-old rabbit kidneys werefixed in ice-cold ethanol. After washing with PBS, the sectionswere blocked with PBS containing 1% BSA and 10% horse serumfor 30 min. Cox-1, Cox-2, Aquaporin 2 (AQP2), AQP3 (all obtainedfrom Santa Cruz Biotechnologies, Santa Cruz, CA), mAb anti–Na/K-ATPase,and mAb anti-Troma1 (Development Studies Hydroma Bank; Universityof Iowa, Department of Biological Sciences, Iowa City, IA, undercontract NO1-HD-7-3263 from the NICHD), mAb anti-occludin andanti-Ki67 (Zymed, CA) mAb anti-PCD Amp1 (22), and mAb anti–cytokeratin-19(gift from Prof. Dr. R. Moll, Marburg, Germany) were appliedas primary antibodies for 1 h in blocking solution. The specimenswere incubated for 45 min with donkey anti-mouse or donkey anti-goatIgG FITC-conjugated secondary antibodies diluted 1:200 in PBScontaining 1% BSA (Jackson Immunoresearch Laboratories, WestGrove, PA). The sections were then analyzed using an Axioskop2 plus microscope. Images were made by a digital camera andthereafter processed with Photoshop 5.5.
Quantitative Analysis of Labeled Cells in the Cultured CD Epithelium
To determine the number of immunopositive or lectin labeledcells in the cultured epithelia, we applied a double-labelingprocedure. The epithelia were first labeled with the nuclearmarker propidium iodide (4 µg/ml in PBS; Sigma-Aldrich-Chemie)and then exposed to the respective cellular markers. By thismethod, the number of labeled and unlabeled cells within theepithelium could be determined easily. The mean of immunopositivecells within the epithelium is given in the text.
SDS-PAGE and Western Blotting Experiments
Generated CD epithelia were homogenized in a sample buffer containing2% SDS, 10% glycerin, 125 mM Tris-HCl, and 1 mM EDTA (all obtainedfrom Sigma-Aldrich-Chemie) and centrifuged at 10,000 x g for10 min. The supernatants were used in the following experiments.The amount of proteins was determined by a protein microassay(Bio-Rad Laboratories, Hercules, CA). Thirty-microgram proteinsamples were separated by SDS-PAGE in 10% Laemmli minigels accordingto methods described earlier, which were electrophoreticallytransferred to P-Immobilon membranes (Millipore, Eschborn, Germany).For detecting immunoreactive proteins, the blots were firstblocked (PBS, pH 7.2; 0.05% Tween; Sigma; 10% horse serum, Boehringer,Mannheim, Germany) followed by an incubation for 1 h at roomtemperature with a goat polyclonal antiserum raised againstCox-1, Cox-2, AQP2, and AQP3 diluted 1:200. A horseradish peroxidase–conjugateddonkey anti-goat or anti-mouse Ig antiserum (1:1000; Dianova,Hamburg, Germany) served as detecting antibody applied for 45min as described earlier. Blot development was started by additionof 0.5 mg/ml diaminobenzidine, 0.02% H202, and 0.03% cobaltchloride dissolved in citrate buffer (pH 6.3). Washing the membranein tap water stopped the reaction. Immunoblots were documentedwith a Scan Jet 6200 C (Hewlett Packard). Determination of apparentmolecular weight was performed in conjunction with broad-rangemolecular weight standard proteins (Bio-Rad Laboratories).
Statistical Analyses
All values are presented as mean ± SEM. Levels of significancewere calculated by ANOVA followed by Bonferroni’s testfor multiple comparisons. Differences were considered significantat P < 0.05.
The gradual process of kidney development can be visualizedclearly by histochemical markers. For example, lectin histochemistrydemonstrates the developmental gradient between the outer cortexand the medulla. In the outer cortex, PNA is bound at the basalaspect of each CD ampulla (Figure 2A), whereas in the innercortex, label is detected at the luminal cell side of numerouscells demonstrating the presence of matured -type IC cells (Figure 2B).The inner-medullary CD fails to bind PNA (Figure 2C), becauseonly P cells populate this tubular segment. In contrast, APQ2cannot be detected in the CD ampulla (Figure 2D) but is foundat P cells in the maturing cortical (Figure 2E) and maturedmedullary CD (Figure 2F).
Figure 2. Development gradient within the neonatal rabbit kidney. (A) PNA binding is detectable at the basal aspect of the CD ampulla tip, whereas Aquaporin 2 (APQ2) expression is lacking at this site (D). A faint luminal reaction occurs at the distal end of the CD ampulla shaft region (D, arrow). (B) -type intercalated (IC) cells in the cortical collecting duct (CCD) revealing luminal binding to peanut agglutinin (PNA; arrow). (C) PNA binding is lacking in the medullary collecting duct (MCD), because principal (P) cells expressing AQP2 channels are populating the MCD (F). (E) AQP2 expression in P cells of CCD (arrow). CF, capsula fibrosa; Amp, CD ampulla; Glom, glomerulus; HL, thin limb of loop of Henle.
In perfusion culture experiments of embryonic CD epithelia,it has been shown that aldosterone is able to induce >90%PNA-positive cells within the cultured epithelia (20). However,because in the matured CD of the kidney only 10 to 25% of thecells become PNA positive (12,13), the existence of an antagonizingmechanism must be assumed. In the present experimental series,urea was used to examine its influence on differentiation.
Differentiation Status of the Cultured CD Epithelia
To asses the development of urea-treated embryonic epithelia,we performed immunohistochemistry (Figure 3). Tissue-specificproteins such as Troma1 (Figure 3A) and cytokeratin-19 (Figure 3B)were found in CD epithelia cultured for 14 d in IMDM containing18 mmol/L urea. Detection of occludin (Figure 3C) indicatedthe establishment of a polarized CD epithelium with a luminaland basolateral compartment. The lack of P CD Amp1 (Figure 3D)indicated a downregulation of embryonic features, whereas theabsence of Ki67 label (Figure 3E) demonstrated a postmitoticstatus of the cells. All of these features reveal a cell biologicstatus of the CD epithelium that is identical to characteristicsfound in adult CD.
Figure 3. Developmental status of CD epithelium after culture for 14 d in the presence of 18 mmol/L urea. CD-specific marker such as Troma1 (A) and cytokeratin-19 (B) are found in all cells. Occludin (C) is present, indicating the establishment of a polarized epithelium. The lack of P CD Amp1 (D) shows the downregulation of embryonic features. The absence of Ki67 protein (E) displays the postmitotic status of the epithelial cells. Arrows mark the basal aspect of the CD epithelium.
Culture of Embryonic CD Epithelia in IMDM Containing Aldosterone
Embryonic CD epithelia were cultured for 14 d in serum-freeIMDM containing constantly aldosterone (n = 7). As revealedby histochemistry in this experimental series, >90% PNA-positivecells were detectable in the generated CD epithelia (Figure 4A),whereas <10% AQP2-positive (Figure 5A) and >90% Cox-1–positive(Figure 5B) cells were observed. Labeling with SBA for controlshowed that all cells within the epithelium revealed lectinbinding (Figure 4B).
Figure 4. Influence of urea on PNA and soybean agglutinin (SBA) binding in CD cells after 14 d of culture. (A) In the control (Iscove’s modified Dulbecco’s medium [IMDM] containing aldosterone), PNA binding is visible in >90% of all CD cells (C). Increasing urea concentration decreases PNA binding drastically (E,G). (I) Fewer than 10% of the CD cells are positive after 18 mmol/L urea application, whereas SBA binding is not affected by urea (B, D, F, H, and J). *P < 0.05 compared with control (A).
Figure 5. Influence of urea on AQP2 and Cox-1 expression in CD cells after 14 d of culture. (A) In the control (IMDM containing aldosterone), <10% of CD cells are positive for AQP2, whereas nearly all CD are positive for Cox-1 (B). Increasing urea concentrations do not affect expression of both molecules (C–G).
Culture of Embryonic CD Epithelia in IMDM Containing Aldosterone and Urea
In a next experimental series, generated CD epithelia were culturedin IMDM containing urea concentrations between 4.5 mmol/L (normalserum level) to 18 mmol/L urea (n = 7 in each group). Exposureto increasing urea concentrations resulted in a decreasing numberof PNA-positive CD cells (Figure 4, C, E, G, and I; *P <0.05 compared with control), whereas SBA binding remained unaffected(Figure 4, D, F, H, and J). In detail, as compared with controls(Figure 4A), culture of the epithelia in IMDM with 4.5 mmol/Lurea did not show alterations (Figure 4C), whereas 9 mmol/Lurea leads to 60% (Figure 4E), 13.5 mmol/L urea leads to 30%(Figure 4G), and 18 mmol/L urea leas to only 10% PNA-bindingcells (Figure 4I).
For excluding the possibility that the reduced PNA binding wascaused by the interference of urea with the PNA-binding sites,detection of lectin binding was performed in PBS containing18 mmol/L urea. Compared with control experiments in pure PBS,the presence of urea in the incubation solution did not leadto a reduced PNA binding. This finding clearly demonstratesthat reduced PNA binding obtained in our experiments is causedby a reduced development of glycoproteins or glycolipids.
As shown in Figures 2 and 5, maturing CD epithelia contain AQP2channels and Cox-1 enzyme. To test whether urea is involvedin the development of both proteins, we performed immunohistochemistry.However, the expression of both AQP2 (Figure 5, C, E, G, and I)as well as Cox-1 protein (Figure 5, D, F, H, and J) was unchangedby different urea concentrations.
Culture of Embryonic CD Epithelia in IMDM Containing Aldosterone and NaCl
It was shown in earlier investigations that addition of NaClto IMDM modulates differentiation of embryonic CD epithelia(19). Consequently, 20 mmol/L NaCl was added to IMDM to adaptthe serum level for Na+ in the culture medium (Figure 6; n =7 in each group). The histochemical profile yielded that applicationof additional NaCl in the culture medium did not influence thebinding pattern of PNA (Figure 6A), SBA (Figure 6B), and Cox-1(Figure 7B) in CD epithelia as compared with controls. However,the amount of AQP2-positive cells was moderately increased.In the control group with pure IMDM, <10% AQP2-positive cellswere found (Figure 5A), whereas addition of 20 mmol/L NaCl showedup to 20% of fluorescence cells (Figure 7A).
Figure 6. Influence of NaCl and urea on PNA and SBA binding in CD cells after 14 d of culture. (A) In the control (IMDM containing aldosterone and 20 mmol/L NaCl), PNA binding is visible in >90% of all CD cells (C). Increasing urea concentrations decrease PNA binding drastically despite NaCl addition to the culture medium (E, G). (I) Fewer than 10% of the CD cells are positive after 18 mmol/L urea application, whereas SBA binding is affected neither by urea nor by NaCl (B, D, F, H, and J). *P < 0.05 compared with control (A).
Figure 7. Influence of urea and NaCl on AQP2 and Cox-1 expression in CD cells after 14 d of culture. (A) After addition of 20 mmol/L NaCl to the culture medium, a moderate increase of up to 25% AQP2-positive CD cells are visible (*P < 0.05 compared with IMDM without NaCl; Figure 5A), whereas Cox-1 expression is not altered by NaCl alone or in combination with urea (B, D, F, H, and J). The increase of AQP2 expression is not further affected by increasing urea concentrations (C, E, G, I).
Culture of Embryonic CD Epithelia in IMDM Containing Aldosterone, NaCl, and Urea
Addition of both 4.5 mmol/L urea and 20 mmol NaCl to IMDM reflectsthe concentration status, which is found for these moleculesin serum (n = 7 in each group). Under these culture conditions,we found the same data as obtained without urea (Figures 6, C and D, and 7, C and D). However, addition of higher concentrationsof urea (9, 13.5, or 18 mmol/L) in combination with 20 mmol/LNaCl leads to the same effect (Figure 6, E, G, and I) as seenwithout 20 mmol/L NaCl (Figure 4, E, G, and I), where the amountof PNA-binding CD cells was drastically reduced. In comparisonwith PNA-binding occurrence of SBA binding (Figure 6), AQP2(Figure 7) and Cox-1 enzyme (Figure 7) was not affected underthis experimental set-up.
As shown in Table 1, addition of urea or NaCl alone or in combinationwith the culture medium resulted in changes of osmolality. Becauseaddition of NaCl did not reduce the amount of PNA-binding cells(Figure 6A), the decrease of PNA binding in urea-containingmedium must be a specific effect that is not be caused by osmolality.Control experiments with addition of 20 mmol/L mannitol to IMDMalso showed no effect on PNA binding.
It can be concluded that aldosterone induces and urea restrainsthe development of PNA binding on cells of the embryonic CDepithelium. As shown in Western blotting experiments, the expressionof Na/K-ATPase (Figure 8A), Cox-1 (Figure 8B), Cox-2 (Figure 8C),and AQP3 (Figure 8E) was not altered by addition of increasingurea as well as by 20 mmol/L NaCl or by simultaneous application.The moderate 20% increase of AQP2-positive cells after NaCladdition to IMDM shown by immunohistochemistry (Figure 6A) couldnot be demonstrated in related immunoblotting experiments (Figure 8D).
Figure 8. Western blotting results obtained from CD epithelia cultured in the presence of urea or NaCl or in simultaneous application after 14 d of perfusion culture. As positive control served a total kidney extract. Na/K-ATPase (A), Cox-1 (B), Cox-2 (C), and AQP3 (E) are detectable in the cultured epithelia under different conditions, and their expression is not affected by urea or NaCl. Reaction for AQP2 in the CD epithelia is not observed in the Western blotting experiments; however, it is found in the kidney extract (D).
Urea is the end product of nitrogen metabolism and has to beeliminated through the urine to avoid damage to the tissueswithin the organism. The interstitial accumulation of urea inthe kidney is an important prerequisite for the formation andelimination of a concentrated urine (23). This capacity developsuntil the end of organ growth (24). In the adult cortex, peritubularurea concentration should be fairly equal to plasma urea, exceptin the medullary rays, where it might be a little higher. Inthe outer stripe of the medulla, however, the concentrationis much higher than in the cortex because of the urea returningfrom the inner medulla in the ascending vasa recta, which arevery abundant in this region. In the inner stripe, the interstitiumsurrounding the tubules has a higher urea concentration thanin the cortex. Finally, in the inner medulla, the urea concentrationis much higher than in the outer part. Typically, urea concentrationin the papilla may vary from 500 to 1500 mmol/L according tothe intensity of the urine concentrating activity (25,26).
During the development of the organism, tissues are exposedto different concentrations of urea. After birth, renal CD epitheliahave to withstand the highest urea concentrations within theorganism. Pilot experiments were conducted recently to evaluatethe osmotic tolerance of adult renal inner medullary epithelialcells. In these experiments, cells were isolated from adultanimals and kept in culture under numerous passages. Experimentalexposure was performed by adding NaCl and urea to the culturemedium until 1640 mosmol/kg H2O was reached (27). During thistreatment, only 30% of cells survived for 24 h. However, whenthe exposure to urea was made in a linear increase over 20 h,nearly 90% of the cells stayed viable 24 h later. It was concludedthat gradual changes in osmolality allow cells to tolerate highamounts of urea up to 1640 mosmol/kg H20.
In the present experiments, we used a culture protocol mimickingtissue-specific characteristics. Renal CD epithelia were generatedfrom embryonic CD ampulla cells on a kidney-specific collagenoussupport. During perfusion culture, a polarized CD epitheliumwas established, which in analogy to the adult kidney did notshow any more mitotic cells (Figure 3E) (19). Applying physiologicconcentrations of urea over 14 d, we analyzed in the presentexperiments the histochemical profile of generated CD epithelia.The cultured embryonic CD epithelia were able to tolerate anenvironment that contained 4.5 to 18 mmol/L urea for at least14 d in culture in excellent condition. According to morphologiccriteria, qualitative differences were not observed betweenurea-treated specimens and controls.
We initially exposed the epithelia to 4.5 mmol/L urea, whichcorresponds to physiologic serum values. As shown in Figure 4C,this urea concentration does not influence the aldosterone-induceddevelopment of PNA binding in the CD epithelium. However, increasingconcentrations of 9, 13.5, or 18 mmol/L urea lead to a downregulationof PNA binding (Figures 4 and 6). Whereas culture media withouturea showed 90% PNA-positive cells, application of urea (Figure 4, E, G, and I)or urea in combination with additional NaCl(Figure 6, E, G, and I) to the medium resulted in a clear decreaseof 10% PNA-positive cells. Thus, long-term application of ureaover 14 d to the culture medium is a suitable method to affectthe aldosterone-induced development of PNA-binding cells ingenerating renal CD epithelia. Because we do not have functionaldata yet, we cannot ascribe the generated PNA binding exclusivelyto adult -type IC cells (12). The cell biological mechanismunderlying this development remains unclear. It has been shownthat the typical IC-cell distribution in the CD of adult kidneysis the result of apoptosis (28). However, it is unlikely thaturea affects the apoptotic pathway in the presented experimentalset-up. PNA binding develops continuously, and the maximum isreached at day 10 of the culture period (16). At this time,proliferation has stopped in the epithelium and the cell numberremains constant during the proceeding culture period.
The number of PNA-positive cells varies in the different regionsof the rabbit kidney. In the outer cortex, a maximum of PNA-positivecells is found, whereas their number decreases toward the innercortex (12,13). In contrast, in the medulla, PNA-positive cellscannot be detected. It is unknown by which mechanisms the cellulargradient is developed, showing numerous PNA-binding cells inthe outer CCD, few in the inner CCD, and none in the medullaof the kidney (Figure 2). The present culture experiments pointout that an area-dependent urea accumulation may be one of thefactors influencing the amount of PNA-binding cells in the respectivesegment. Urea concentration is low in the cortex but increasestoward the medulla. Thus, low urea concentration in the outercortex parallels with numerous PNA-positive cells, whereas increasingurea concentration in the inner cortex shows a decline. It remainsunclear whether urea is the only substance restraining the amountof PNA-positive cells during development. Although we did notyet perform physiologic experiments to test typical functionalparameters of the generated epithelia, it is obvious that PNAbinding on CD cells is modulated by increasing concentrationsof urea under in vitro conditions. However, whether the samemechanism is acting in the developing organ remains to be elucidated.
Acknowledgments
The technical assistance of Lucia Denk and the comments of Dr.L. Bankir are gratefully acknowledged.
Saxen L: Organogenesis of the Kidney, Cambridge, Cambridge University Press, 1987
Ekblom P: Renal development. In: The Kidney: Physiology and Pathophysiology, edited by Seldin DW, Giebisch G, New York, Raven, pp 475–501, 1992
Al-Awqati Q, Goldberg MR: Architectural patterns in branching morphogenesis in the kidney. Kidney Int 54: 1832–1842, 1998[CrossRef][Medline]
Sariola H: Nephron induction revisited: From caps to condensate. Curr Opin Nephrol Hypertens 11: 17–21, 2002[CrossRef][Medline]
Plisov SY, Yoshino K, Dove LF, Higinbotham KG, Rubin JS, Perantoni AO: TGF2, LIF and FGF2 cooperate to induce nephrogenesis. Development 128: 1045–1057, 2001[Abstract]
Kispert A, Vainio S, McMahon AP: Wnt-4 is a mesenchymal signal for epithelial transformation of metanephric mesenchyme in the developing kidney. Development 125: 4225–4234, 1998[Abstract]
Barasch J, Pressler L, Connor J, Malik A: A ureteric bud cell line induces nephrogenesis in two steps by two distinct signal. Am J Physiol 271: F50–F61, 1996
Kloth S, Gmeiner T, Aigner J, Jennings ML, Röckl W, Minuth WW: Transitional stages in the development of the rabbit collecting duct. Differentiation 63: 21–32, 1998[Medline]
Oliver JA, Barasch J, Yang J, Herzlinger D, Al-Awqati Q: Metanephric mesenchyme contains embryonic renal stem cells. Am J Physiol Renal Physiol 283: F799–F809, 2002[Abstract/Free Full Text]
Kaissling B, Kriz W: Structural analysis of the rabbit kidney. Adv Anat Embryol Cell Biol 6: 1–123, 1979
Satlin LM, Matsumoto T, Schwartz GJ: Postnatal maturation of rabbit collecting duct III. Peanut lectin-binding intercalated cells. Am J Physiol 262: F199–F208, 1992
Evan PE, Satlin LM, Gattone VH II, Connors B, Schwartz GJ: Postnatal maturation of rabbit renal collecting duct II. Morphological observations. Am J Physiol 261: F91–F107, 1991
Minuth WW, Rudolph U: Successive lectin-binding changes within the collecting duct during post-natal development of the rabbit kidney. Pediatr Nephrol 4: 505–509, 1990[CrossRef][Medline]
Fejes-Toth G, Naray-Fejes-Toth A: Evidence for conversion of intercalated cells (ICC) to principal cells (PC) [Abstract]. J Am Soc Nephrol 1: 717, 1990
Aigner J, Kloth S, Jennings ML, Minuth WW: Transitional differentiation patterns of principal and intercalated cells during renal collecting duct development. Epithelial Cell Biol 4: 121–130, 1995[Medline]
Benchimol C, Zavilowitz B, Satlin LM: Developmental expression of ROMK mRNA in rabbit cortical collecting duct. Pediatr Res 47: 46–52, 2000[Medline]
Narbaitz RA, Tran NT, Levine DZ: Differentiation of renal intercalated cells in fetal and postnatal rats. Anat Embryol 183: 353–361, 1991[Medline]
Steiner P, Strehl R, Kloth S, Tauc M, Minuth WW: In vitro development and preservation of specific features of collecting duct epithelial cells from embryonic rabbit kidney are regulated by the electrolyte environment. Differentiation 62: 193–202, 1997[CrossRef][Medline]
Minuth WW, Fietzek W, Kloth S, Stöckl G, Aigner J, Röckl W, Kubitza M, Dermietzel R: Aldosterone modulates the development of PNA binding cell isoforms within renal collecting duct epithelium. Kidney Int 44: 337–344, 1993[Medline]
Minuth WW: Neonatal rabbit kidney cortex in culture as tool for the study of collecting duct formation and nephron differentiation. Differentiation 36: 12–22, 1987[Medline]
Strehl R, Kloth S, Aigner J, Steiner P, Minuth WW: PCD: Amp1, a new antigen at the interface of the embryonic collecting duct epithelium and the nephrogenic mesenchyme. Kidney Int 52: 1469–1477, 1997[Medline]
Sands JM: Molecular approaches to urea transporters. J Am Soc Nephrol 13: 2795–2806, 2002[Abstract/Free Full Text]
Kim YH, Kim DU, Han KH, Jung JY, Sands JM, Knepper MA, Madsen KM, Kim J: Expression of urea transporters in the developing rat kidney. Am J Physiol Renal Physiol 282: F530–F540, 2002[Abstract/Free Full Text]
Yancey PH, Burg MB: Distribution of major organic osmolytes in rabbit kidneys in diuresis and antidiuresis. Am J Physiol 257: F602–F607, 1989
Bankir L, Tringh-Trangh-Tan MM: Renal urea transporters. Direct and indirect regulation by vasopressin. Exp Physiol 85S: 243S–252S, 2000
Cai Q, Michea L, Andrews P, Zhang Z, Rocha G, Dmitrieva N, Burg MB: Rate of increase of osmolarity determines osmotic tolerance of mouse inner medullary epithelial cells. Am J Physiol Renal Physiol 283: F792–F798, 2002[Abstract/Free Full Text]
Kim J, Cha JH, Tisher CC, Madsen KM: Role of apoptotic and nonapoptotic cell death in removal of intercalated cells from developing rat kidney. Am J Physiol 270: F575–F592, 1996[Medline]
Received for publication May 6, 2003.
Accepted for publication July 24, 2003.
Top
Abstract
Introduction
-type intercalated (IC) cells and their distribution in the corticomedullary course of the collecting duct (CD) in the mature rabbit kidney. It has been shown that aldosterone is able to generate >90% of PNA-binding cells in an embryonic CD epithelium in vitro. In adult kidney, a maximum of only 25% PNA-positive cells is found in the cortical segment of the CD, and PNA-binding completely disappears in the inner-medullary CD. Molecules that regulate the gradual development of CD-specific cells during organ growth are unknown. In the present experiments, it was found that addition of physiologic concentrations of urea to the culture medium is able to restrain the action of aldosterone in embryonic CD epithelia. Urea antagonizes in a concentration-dependent manner the action of aldosterone finally leading to only 10% of PNA-binding cells. The data point to a urea-specific effect, because osmolytes such as NaCl and mannitol did not affect PNA binding. In addition, urea did not influence expression of principal-cell typical markers such as AQP2 and 3. The findings may explain that a higher number of PNA-positive cells is found in the cortical region of the kidney correlated with a low concentration of urea as compared with only few PNA-binding cells in the medullary CD, where a high concentration of urea occurs. Thus, an increasing concentration of urea may trigger the number of PNA-positive cells in the cortical-medullary course of the CD during organ development. E-mail: karl.schumacher@vkl.uni-regensburg.de 
, maturing CD epithelia contain AQP2 channels and Cox-1 enzyme. To test whether urea is involved in the development of both proteins, we performed immunohistochemistry. However, the expression of both AQP2 (Figure 5, C, E, G, and I) as well as Cox-1 protein (Figure 5, D, F, H, and J) was unchanged by different urea concentrations. 
