Kidney International, Vol. 61 (2002), pp. 1003–1010

Cl� channels in basolateral TAL membranes. XVI.MTAL and CTAL cells each contain the mRNAsencoding mmClC-Ka and mcClC-Ka

MARINA V. MIKHAILOVA, CHRISTOPHER J. WINTERS, and THOMAS E. ANDREOLI

Division of Nephrology, Department of Internal Medicine, University of Arkansas College of Medicine, andMedical Research Service, The John L. McClellan Veterans Hospital, Little Rock, Arkansas, USA

Cl� channels in basolateral TAL membranes. XVI. MTAL and 7–9], apical K� conductance [10, 11] and basolateral Cl�

CTAL cells each contain the mRNAs encoding mmClC-Ka conductance in MTAL segments [12]. In microperfusedand mcClC-Ka. CTAL segments, ADH has no effect on net salt absorp-Background. Our prior data indicate that two separate but

tion, net K� secretion or basolateral Cl� efflux [4–6].homologous basolateral chloride (Cl�) channels, mmClC-KaCultured mouse MTAL and CTAL cells utilized rou-and mcClC-Ka, are the principal mediators of net Cl� absorp-

tion in mouse medullary thick ascending limb (MTAL) and tinely in our laboratory exhibit the same functional heter-cortical thick ascending limb (CTAL) cells, respectively. In the ogeneity as do microperfused mouse MTAL and CTALpresent studies, we evaluated the possibility that there might

segments, particularly with respect to the effects of thebe translational or post-translational suppression of mmClC-Kaadenylate cyclase cascade on net Cl� influx and effluxand mcClC-Ka activity in CTAL and MTAL cells, respectively.

Methods. Polymerase chain reaction (PCR) fragments were (Methods section) [1–6]. Individual basolateral Cl� chan-prepared that were highly specific for either mmClC-Ka or nels from cultured MTAL cells and CTAL cells, studiedmcClC-Ka, the cDNAs encoding mmClC-Ka and mcClC-Ka, in bilayers or by patch-clamping, also exhibit this func-respectively.

tional heterogeneity [1–3]. Cl� channels from MTALResults. Using reverse transcription (RT)-PCR with thesecells are gated either by raising cytosolic Cl� (K1/2 � 10highly specific products, mRNAs specific for non-homologous

channel sequences in either mmClC-Ka or mcClC-Ka were pres- mmol/L) or, at 2 mmol/L cytosolic Cl�, by [protein kinaseent in both MTAL and CTAL cells. A � adenosine 5�-triphosphate (PKA � ATP)] [3, 13].

Conclusions. Both mouse MTAL and CTAL cells contain These two results are unique among the ClC family ofthe mRNAs encoding mmClC-Ka and mcClC-Ka. There mayCl� channels, and neither of these effects is obtained inbe translational or post-translational suppression of mmClC-Ka

activity in CTAL cells, and of mcClC-Ka activity in MTAL basolateral Cl� channels from cultured mouse CTALcells. cells [1, 2].

We recently cloned two highly homologous cDNAs,mmClC-Ka from cultured mouse MTAL cells, and

This article evaluates whether pre- or post-transcrip- mcClC-Ka from cultured mouse CTAL cells [2]. Thesetional events are responsible for the fact that basolateral cDNAs encode two 75 kD Cl� channels, mmClC-KaCl� channels in mouse medullary thick ascending limb and mcClC-Ka, respectively [2]. In cultured MTAL and(MTAL) and cortical thick ascending limb (CTAL) cells CTAL cells transfected with antisense oligonucleotidesdiffer functionally [1–3]. The rationale for our studies specific either for the mmClC-Ka cDNA (mmAntisense)depended on the following considerations. or for the mcClC-Ka cDNA (mcAntisense), mmAnti-

Microperfused mouse MTAL and CTAL segments are sense inhibited by 50% synthesis of mmClC-Ka infunctionally heterogeneous [4–6]. Antidiuretic hormone MTAL cells but had no effect on mcClC-Ka synthesis in(ADH) enhances apical Na�/K�/2Cl� admittance [5, CTAL cells; mmAntisense also suppressed bumetanide-

insensitive conductive 36Cl� efflux in MTAL but notCTAL cells. Likewise, mcAntisense inhibited synthesisKey words: chloride channels, thick ascending limb, cloning DNA,

transport. of mcClC-Ka by 40% in CTAL cells but not mmClC-Kain MTAL cells; mcAntisense also blocked bumetanide-Received for publication June 14, 2001insensitive 36Cl� efflux in CTAL but not MTAL cells [2].and in revised form August 28, 2001

Accepted for publication October 24, 2001 These data [2] are consistent with the view that mmClC-Ka and mcClC-Ka are the cardinal physiologic mediators 2002 by the International Society of Nephrology

1003

Mikhailova et al: Cl� channels in basolateral TAL membranes1004

of net basolateral membrane Cl� efflux in MTAL and in the MTAL with a weak signal in the CTAL [28]. Noneof these studies evaluated the roles of either ClC-K1 orCTAL cells, respectively [1, 2].

However, the antisense oligonucleotides specific for ClC-K2 in net Cl� transport by MTAL or CTAL cells,or the effects of the adenylate cyclase cascade on netmmClC-Ka and mcClC-Ka did not completely suppress

the appearance of 75 kD bands in Western blots of Cl� transport [24–28].In our prior experiments, we assigned the functionalMTAL and CTAL, respectively. Since mmClC-Ka and

mcClC-Ka are both 75 kD proteins, the data cited above roles for Cl� efflux in MTAL and CTAL cells to mmClC-Ka and mcClC-Ka [1–3, 13, 14]. In the present studies,[2] do not exclude the possibility that both MTAL and

CTAL cells might contain both mmClC-Ka and mcClC-Ka, we evaluated whether the mRNAs for mmClC-Ka andmcClC-Ka might be present both in adenylate cyclase-with the mmClC-Ka isoform inactive in CTAL cells and

the mcClC-Ka isoform inactive in MTAL cells. This cir- insensitive CTAL cells and adenylate cyclase-sensitiveMTAL cells [1, 2].cumstance would be comparable to that which is ob-

tained in apical membranes of proximal convoluted tu- To test this possibility, polymerase chain reaction(PCR) fragments of mmClC-Ka and mcClC-Ka were pre-bules (PCT), where a number of Na�/H� exchangers

(NHE) exist [14, 15], but NHE3 appears to be the sole pared using primers highly specific for each of thesetwo cDNAs. Using RT-PCR with these highly specificisoform mediating Na�/H� exchange [16–18].

It is pertinent to compare these results with those primers, DNA fragments specific for non-homologouschannel sequences in either mmClC-Ka or mcClC-Kapublished by others. As noted previously [1, 19], mouse

mmClC-Ka is �94% homologous with rat rClC-K2 [1, 2, were present both in MTAL and in CTAL cells. Thus,both MTAL and CTAL cells contain the mRNAs encod-20, 21] and 80% homologous both with human hClC-Ka

and hClC-Kb [1, 2, 22]. rbClC-Ka, cloned by us from rabbit ing mmClC-Ka and mcClC-Ka, while mmClC-Ka andmcClC-Ka are the physiological mediators of net Cl�outer medulla [19], is 81% homologous with mmClC-Ka

[1] and with rClC-K2 [20, 21], and approximately 85% efflux in MTAL and CTAL cells, respectively [1–3]. Thathomologous to human hClC-Ka and hClC-Kb [1, 2, 22]. is, there may be translational or post-translational sup-Finally, Uchida has indicated that mcClC-Ka (GenBank pression of mmClC-Ka functional activity in CTAL cells,accession number AF124848) is the mouse homolog of and of mcClC-Ka functional activity in MTAL cells.rClC-K1 [21].

Tissue localization studies of these channels have beenMETHODSproblematic. In non-quantitative reverse transcription–

Mouse kidney cortex and outer medulla were dis-polymerase chain reaction (RT-PCR) studies using ratsected freehand as described previously [19]. The MTALkidney, Kieferle et al identified rClC-K1 in microdis-and CTAL cells have been isolated from SV40 transgenicsected CTAL segments but provided no data on MTALmice, and used at passages 15 through 35 as describedsegments [22]. Using non-quantitative RT-PCR, Vander-in our previous publications [1–3]. We recognize that,walle localized mRNAs for both rClC-K1 and rClC-K2when using cultured cell lines, there is always the possi-in outer medullary late proximal tubules, MTAL andbility of cellular dedifferentiation or cross-contamina-CTAL, and along the entire collecting duct [23]. In thetion. It is therefore pertinent to stress again, as noted inMTAL and CTAL, these workers identified eitherthe introductory comments, that these cultured cell linesrClC-K1 and/or rClC-K2, since the polyclonal antibodyshare, nearly identically with respect to net Cl� transport,used in these studies could not distinguish betweenthe properties of microperfused mouse MTAL andrClC-K1 and rClC-K2. Neither study provided functionalCTAL segments.data regarding the roles of rClC-K1 or rClC-K2 in net

More specifically, net Cl� uptake in both MTAL andCl� absorption in MTAL or CTAL cells, or of the actionCTAL cells is bumetanide-sensitive. It is augmented byof the adenylate cyclase cascade on these channelsdb-cAMP in MTAL but not CTAL cells [1, 2, 14]. Like-[22, 23].wise, apical Cl� entry in microperfused MTAL andUchida and colleagues identified rClC-K1 in rat thinCTAL segments is bumetanide- or furosemide-sensitiveascending limbs [24, 25] and provided evidence for the[4–7], and ADH or db-cAMP enhance apical Cl� admit-role of ClC-K1 in transepithelial net Cl� absorption bytance in microperfused MTAL segments but not in mi-thin limbs by producing nephrogenic diabetes insipiduscroperfused CTAL segments [4–7, 12]. Second, basolat-in knockout mice lacking the ClC-K1 [26]. Using theeral Cl� efflux is bumetanide- or furosemide-insensitivesame knockout mice, Kobayashi et al used immunofluo-in cultured CTAL and MTAL cells, as it is in the corre-rescence to localize ClC-K2 to MTAL and CTAL cellssponding microperfused segments [4, 6]. Third, net baso-as well as to distal nephron cells [27]. Because the micelateral Cl� conductance in MTAL but not CTAL cellswere ClC-K1 knockouts, the presence of ClC-K1 inis enhanced by raising cytosolic Cl� concentrations [1, 2,MTAL cells could not be evaluated. Likewise, Yoshi-

kawa et al found rClC-K2 and the mRNA for rClC-K2 13] or, at 2 mmol/L cytosolic Cl�, by (ATP � PKA)

Mikhailova et al: Cl� channels in basolateral TAL membranes 1005

[3, 13]. These findings are unique among the ClC familyof Cl� channels. Likewise, in microperfused MTAL butnot CTAL segments, the adenylate cyclase cascade in-creases apical Cl� entry directly and, by raising cytosolicCl� concentrations, increases basolateral Cl� conduc-tance [4, 5, 12]. Fourth, both cultured MTAL cells [3]and cultured CTAL cells [1] express Tamm-Horsfall pro-tein, an explicit marker for thick ascending limb cells.

Finally, the Cl� channels fused into bilayers from cul-tured MTAL and CTAL cells are basolateral in origin,as in microperfused MTAL and CTAL segments [4–6],and they share the properties of basolateral vesicles fusedinto bilayers from fresh medullary vesicles [3, 13, 14].These facts have been confirmed by duplicating the Cl�

transport results in vesicles from outer medulla or cul-tured MTAL and CTAL cells by excised inside-out baso-lateral patch clamping studies in cultured MTAL [3] andCTAL [1] cells.

In short, the cells can be identified as thick limb cellsby Tamm-Horsfall staining [1, 3]. There is no detectablecross-contamination between the cultured MTAL andCTAL cells with respect to Cl� transport pathways.These cells share the physiologic properties of basolat- Fig. 1. A representative beta-actin reverse-transcription-polymerase

chain reaction (RT-PCR) control experiment yielding a predicted DNAeral Cl� transport present in the corresponding micro-fragment 400 base pairs in length.perfused tubule segments [1–6, 14].

Isolation of poly(A)� mRNA

Total RNA was prepared as described previously [1]. PCR DNA fragment of mmClC-Ka was a 409 base pair se-quence corresponding to amino acids 313–450 of mmClC-Poly (A)� mRNA was purified from total RNA on oligo

(dT)cellulose (Gibco BRL, Life Technologies Inc., Gai- Ka [1]. For mcClC-Ka, oligonucleotides 5�-TGTTCTGTCAGCGAAATTTTCTCCGC-3� (base numbers 1047–thersburg, MD, USA). Final RNA concentrations were

determined spectrophotometrically. 1072) and 5�-GATGGCAGCTCCGATGATGAATATA-3�(base numbers 1432–1457) were used as sense and anti-

Design of specific primers sense primers, respectively. The anticipated PCR DNAfragment of mcClC-Ka was a 410 base pair sequenceFour PCR primers with similar parameters of reassoci-

ation partially determined by high homology between corresponding to amino acids 349–486 of mcClC-Ka.As controls, we used primers for mouse beta-actinthe two ClC sequences were designed using an Oligo 4-S

analysis program (National Bioscience Inc., Plymouth, (5�-CGCTGCGCTGGTCGTCGACAACGGCT-3� and5�-CATACAGGGACAGCACAGCCTGGAT-3�). AsMN, USA), based on the previously described [1] mouse

cDNA clones mmClC-Ka (GenBank accession number indicated in Figure 1, the control RT-PCR product wasalso 400 bp long. This same control was used in all ofAF-124847) and mcClC-Ka (GenBank accession number

AF-124848). Sequence selective primers were designed our RT-PCR experiments.Figure 2 includes the RT-negative controls for the RT-by using sequences that had identical homology for the

corresponding cDNA, either mmClC-Ka or mcClC-Ka, PCR reactions and the PCR reaction of mouse genomicDNA. Without RT, no product was obtained using mRNAbut low homology with respect to the non-correspond-

ing cDNA. Each primer had no homology with any as the template. Figure 2 also shows that the PCR of ge-nomic DNA resulted in products greater than 2 kb bothGenBank sequence except those for mmClC-Ka and

mcClC-Ka. for mmClC-Ka and mcClC-Ka. Thus, the latter productsdiffered appreciably in size from the 400 base pair RT-The primer sequences, with the mismatches for the

non-corresponding cDNAs indicated by underlining, were PCR products (Fig. 2). Table 1 summarizes these results.The sizes of the individual fragments are identical to theas follows. For mmClC-Ka, oligonucleotides 5�-ACTA

CAGCCAGCGGACTTTCTTATTT-3� (base numbers corresponding DNA sequences of mouse genomic DNA(GenBank accession numbers AC079572 and AC079559,940–965) and 5�-AATGACAGCTCCATAGACAAAG

ATG-3� (base numbers 1325–1349) were used as sense respectively). Alignment of the DNA fragments obtainedonly with these primers with the genomic sequence showedand antisense primers, respectively. Thus, the anticipated

Mikhailova et al: Cl� channels in basolateral TAL membranes1006

Table 1. Size comparison of the DNA fragments obtained usingRT-PCR with mRNA or PCR with genomic DNA

DNA fragment size

Reaction mmClC-Ka mcClC-KaTemplate type primer primer

mRNA RT-PCR 400 base pairs 400 base pairsGenomic DNA PCR 2.4 kb 2.2 kb

RT-PCR and PCR were carried out as described in the Methods section usingprimers specific either for mmClC-Ka or mcClC-Ka.

used with success by multiple groups of investigators[29–32].

Fig. 2. Representative reactions using an mRNA template with (RT-RT-PCR reactionsPCR) and without (PCR) reverse transcriptase on the primers specific

for mmClC-Ka (mm) or mcClC-Ka (mc). Lanes 5 and 6 show a represen- The RT-PCR reactions were performed using a Genetative PCR reaction using genomic DNA.

AMP Gold RNA PCR Core Kit (PE Biosystems, Fos-ter City, CA, USA). cDNA was synthesized on 0.2 �gof poly(A)� mRNA isolated either from homogenatesof mouse inner kidney cortex or outer medulla, or fromthat each of the genes contained three introns betweencultured mouse CTAL or MTAL cells. In all cases, thethe primers used, thus excluding the possibility of inad-reaction volume was 20 �L containing 15 units of Multi-vertent amplification and that the length of the corre-Scribe reverse transcriptase incubated at 42�C for 15sponding genomic sequences was 2.4 and 2.2 kb forminutes. PCR was performed as described above at ammClC-Ka and mcClC-Ka, respectively.final concentration of 200 �mol/L of each dNTP, 0.15�mol/L sense primer, 0.15 �mol/L antisense primer, andQuantitative RT-PCR analysis of mmClC-Ka and2.5 units of AmpliTag Gold DNA Polymerase (PerkinmcClC-Ka mRNA levels expressed in mouse kidneyElmer) in a total volume of 50 �L. For these experimen-The amplification conditions needed for a quantifiabletal conditions, the magnitude of DNA fragments ob-PCR product were determined as follows. Linearizedserved in our densitometry data (Figs. 4, 5, and 6) wasplasmids containing cloned sequences for mmClC-Kaproportional to the mRNA abundance complementaryand mcClC-Ka (0.05 to 50 ng/sample) were used as tem-to the cDNA synthesized by reverse transcription [33, 34].plates [1]. PCR reactions were performed using Gene-

Amp PCR System 9600 (Perkin-Elmer). The annealingtemperatures varied between 50� and 65�C, and the num- RESULTSber of cycles examined was in the range of 15 to 45. Specificity of primersEach PCR cycle included 20 seconds of denaturation at The goals of our experiments were to determine if the94�C, 20 seconds of annealing at 60�C, and 180 seconds mRNAs encoding mmClC-Ka and mcClC-Ka [1] wereof extension at 72�C. The products were separated on present in inner cortical and outer medullary tissue ho-1% agarose gels and visualized by ethidium bromide mogenates and in cultured MTAL and CTAL cells. Thestaining. The relative band intensities corresponding to approach depended on preparing one pair of primersPCR products of expected molecular mass were quanti- completely homologous only to mcClC-Ka (which there-fied by a densitometer (Eagle Eye II; Stratagene, La Jolla, fore could amplify only mcClC-Ka templates) and a pairCA, USA) using a 100 bp Mass Standard (New England of primers completely homologous to mmClC-Ka (whichBiolabs, Beverly, MA, USA). could amplify only mmClC-Ka templates).

Figure 3 illustrates the linear regions for PCR product The results in Figure 4 illustrate the high degree ofaccumulation, either mmClC-Ka or mcClC-Ka, with var- specificity of these PCR reactions. The primers comple-ying amounts of DNA (ng/sample) or with respect to mentary to mcClC-Ka generated the expected �400 basethe number of cycles. In all cases, the experiments were pair cDNA fragment of mcClC-Ka and no detectablecarried out in the linear regions of PCR product accumu- fragment of mmClC-Ka, while the primers complemen-lation, 5 to 50 ng DNA/sample and 15 to 25 cycles, both tary to mmClC-Ka generated the expected �400 base pairfor mmClC-Ka and mcClC-Ka. Since the experiments fragment of mmClC-Ka but no detectable mcClC-Kawere carried out within the linear range of DNA synthe- fragment (Fig. 4A). The densitometric tracings in Figuresis, our experimental approach may be termed semiquan- 4B show quantitatively that the crossover of PCR prod-

ucts was at background levels. Thus, the primers selectedtitative. This approach is well established and has been

Mikhailova et al: Cl� channels in basolateral TAL membranes 1007

Fig. 3. Conditions for linear PCR product ac-cumulation: The amount of cDNA/sample andthe number of cycles.

(Methods section) were virtually uniquely complemen- the cultured MTAL and CTAL cells had virtually thesame Cl� transport characteristics as their microperfusedtary either for mcClC-Ka or for mmClC-Ka (Fig. 4).counterparts. Accordingly, we repeated the experiments

RT-PCR: Mouse inner cortex and outer medulla shown in Figure 5 using poly(A)�mRNA isolated eitherfrom cultured mouse CTAL cells or from cultured mouseFigure 5A shows that, using poly(A)�mRNA eitherMTAL cells.from cortex or medulla, the RT-PCR reactions yielded

The results of these experiments are shown in Figure 6.�400 bp DNA fragments of both mcClC-Ka andFigure 6A shows that, when the PCR primers specificmmClC-Ka, the former considerably more in abundancefor mcClC-Ka were used, relatively dense �400 base pairthan the latter. The densitometric data presented in Fig-DNA fragments were obtained using poly(A)�mRNAure 5B confirm this view quantitatively: the DNA frag-from either CTAL cells or MTAL cells. Using the PCRment for mcClC-Ka was at least fourfold greater thanprimers specific for mmClC-Ka, relatively faint �400 basethat for mmClC-Ka in homogenates either from innerpair DNA fragments were detected using poly(A)�-cortex or from outer medulla.mRNA from either MTAL or CTAL cells. The densito-metric tracings in Figure 6B illustrate these data quanti-RT-PCR: Cultured mouse CTAL and MTAL cellstatively. Using either poly(A)�mRNA from culturedIt could be argued that the results presented in Figure 5CTAL or MTAL cells, there was a three- to fourfoldwere non-specific because homogenates of outer medullagreater yield of 400 base pair fragments using PCR primer

and inner cortex might contain contaminating amounts pairs specific for mcClC-Ka instead of mmClC-Ka.of MTAL and CTAL cells, respectively. However, cul-tured mouse CTAL and MTAL cells were obtained from

DISCUSSIONmicrodissected mouse CTAL and MTAL segments [1],and their Cl� transport properties differ qualitatively The experiments reported in this paper were designed

to evaluate further the factor or factors responsible for[1–3, 13]. Moreover, as noted in the Methods section,

Mikhailova et al: Cl� channels in basolateral TAL membranes1008

Fig. 4. cDNA fragments obtained using mcClC-Ka and mmClC-Katemplates and primers specific either for mcClC-Ka (labeled c) ormmClC-Ka (labeled m). (A) A 1% agarose gel stained with ethidiumbromide. (B) Quantitative densitometry tracings in all experiments Fig. 5. RT-PCR DNA fragments obtained using poly(A)�mRNA(N � 5). from either inner cortex or outer medulla and primers specific either

for mcClC-Ka (labeled c) or mmClC-Ka (labeled m). (A) A 1% agarosegel stained with ethidium bromide. (B) Quantitative densitometry trac-ings of RT-PCR products in all experiments (N� 5).

the functional heterogeneity between microperfusedMTAL and CTAL segments [4, 6], with particular em-

three sets of studies have correlated ClC channel locationphasis on Cl� transport pathways. ClC channel localiza-in thin or thick ascending limb cells with net Cl� efflux.tion in CTAL and MTAL segments using polyclonal

Uchida and colleagues have provided evidence for theantibodies, whether in our laboratory [1, 2, 19, 28] or inlocalization of the ClC-K1 channel in thin ascending limbthe laboratories of others [20–25], has been problematiccells, and knockout mice lacking the ClC-K1 channelbecause of the high degrees of homology among thesedevelop overt nephrogenic diabetes insipidus [21, 24–26].channels.These data are thus consistent with the possibility thatMoreover, tissue localization of various ClC channelsClC-K1 mediates net Cl� efflux in the thin ascending

to ascending limb cells is not sufficient to identify such limb. Second, Simon et al have provided evidence indi-channels as mediating net Cl� influx in ascending limbs. cating that mutations in CLCNKB produce Bartter’s syn-For example, apical membranes of proximal tubules con- drome Type III in humans, presumably in the MTALtain multiple NHE isoforms [15–17], but only NHE3 ap- [36]. CLCNKB appears to be the human homolog ofpears to mediate Na�/H� exchange in these membranes rClC-K2 [27], mmClC-Ka [2] and rbClC-Ka [19]. Third,[18, 35]. Thus, the assignment of ClC channels as media- as noted earlier in this article, our previous studies withtors of net Cl� efflux in ascending limb cells, in our view, basolateral vesicle fusion into bilayers [1, 3, 13, 14, 19],

patch clamp data [1, 3] and antisense oligonucleotiderequires functional correlation. To our knowledge, only

Mikhailova et al: Cl� channels in basolateral TAL membranes 1009

analysis), Virlon et al found strikingly high levels ofmessage for mcClC-Ka in MTAL cells from mouse kid-ney [37]. However, earlier studies provided no informa-tion concerning the mRNA for mmClC-Ka [36].

Vandewalle et al, using RT-PCR, identified themRNAs for rClC-K1 and rClC-K2 in MTAL and CTALcells [23]. However, because these workers used a poly-clonal antibody reacting both to rClC-Ka and rClC-Kb,it was not possible to distinguish between these two chan-nels in MTAL and CTAL [23]. Uchida et al, who denotedClC-K1 as the rat homolog to mcClC-Ka, localizedrClC-K1 exclusively in the rat inner medulla [21, 24, 25].Evidently there is accord between the present results,those of Virlon et al [36], and those of Kieferle et al [22],and a partial lack of accord between the present studies,the Virlon studies [36] and the Kieferle studies [22] withrespect to the reports of Uchida et al [21, 24, 25].

There also is a partial lack of accord between theresults of Yoshikawa et al [28] and some of the datacited above. For example, Yoshikawa et al found themRNA for rClC-K2 in MTAL cells but not the mRNAfor rClC-K1 [28], termed by Uchida as the rat homologto mcClC-Ka [21]. Obviously, these results differ fromthe results presented in Figure 6 and from the results ofVirlon et al [37]. Further, Yoshikawa et al found a veryweak mRNA signal for rClC-K2 in the CTAL [28], whileour experiments showed a diminished but clearly identi-fiable signal for mmClC-Ka, the mouse homolog torClC-K2 [2, 27], in CTAL cells (Fig. 6). These issuesthus require further clarification.

It is also reasonable to infer from the present datathat the lack of functional expression of mmClC-Ka inCTAL cells, or of mcClC-Ka in MTAL cells, is referableto post-transcriptional events, operating singly or in uni-Fig. 6. RT-PCR DNA fragments obtained using poly(A)�mRNA

from either CTAL or MTAL cells and primers specific either for son, which may differ in MTAL and CTAL cells. SomemcClC-Ka (labeled c) or mmClC-Ka (labeled m). (A) A 1% agarose of the potential classes of explanations in either or bothgel stained with ethidium bromide. (B) quantitative densitometry trac-

cell types might include suppression of mRNA transla-ings of RT-PRC products in all experiments (N � 5).tion for mmClC-Ka and mcClC-Ka synthesis in CTALand MTAL cells, respectively; suppression of traffickingof channel proteins to membranes; or suppression, eitherby inhibitors or lack of co-factors, of functional expres-data [2] are consistent with the view that mmClC-Ka andsion of mmClC-Ka and mcClC-Ka in CTAL and MTALmcClC-Ka are the principal mediators of net Cl� effluxcells, respectively. Obviously, added data are requiredin cultured mouse MTAL and CTAL cells, respectively.to evaluate these and other possibilities.The present experiments evaluated whether cultured

MTAL and CTAL cells contained the requisite mRNAACKNOWLEDGMENTSsequences for encoding both mmClC-Ka and mcClC-Ka.

This work was supported with resources and the use of facilities atThe results in Figures 4 to 6 show that both renal innerthe Central Arkansas Veterans Healthcare System and was supportedcortex, and, more specifically, cultured mouse CTAL and by NIH Grant 5 R01 DK25540 to T.E. Andreoli. A preliminary reportof these studies has appeared elsewhere (abstract; Mikhailova et al;MTAL cells, contained the mRNAs encoding mmClC-World Congress of Nephrology, J Am Soc Nephrol 12:36A–37A, 2001).Ka and mcClC-Ka. That is, the transcriptional processesWe are grateful for the technical assistance provided by Ms. Anna

for producing mmClC-Ka and mcClC-Ka are intact Grace Stewart and for assistance in preparing this manuscript providedby Ms. Clementine M. Whitman.in both MTAL and CTAL cells. In each instance, the

relative abundance of the mRNA encoding mcClC-Ka Reprint requests to Thomas E. Andreoli, M.D. Department of Inter-nal Medicine, University of Arkansas College of Medicine, 4301 Westwas perceptibly greater than that encoding mmClC-KaMarkham, Slot 640, 308-C Shorey Building, Little Rock, Arkansas(Methods section) [29–34]. In partial accord with this 72205, USA.E-mail: AndreoliThomasE@uams.eduview, using serial analysis of gene expression (SAGE

Mikhailova et al: Cl� channels in basolateral TAL membranes1010

dependent proton secretion in the mouse proximal convolutedREFERENCEStubule. J Clin Invest 105:1141–1146, 2000

18. Wang T, Yang CL, Abbiati T, et al: Mechanism of proximal tubule1. Winters CJ, Reeves WB, Andreoli TE: Cl� channels in basolateralbicarbonate absorption in NHE3 null mice. Am J Physiol 277:F298–TAL membranes. XIII. Heterogeneity between basolateral MTALF302, 1999and CTAL Cl� channels. Kidney Int 55:593–601, 1999

19. Zimniak L, Winters CJ, Reeves WB, Andreoli TE: Cl� channels2. Winters CJ, Zimniak L, Mikhailova MV, et al: Cl� channels inin basolateral renal medullary vesicles. X. Cloning of a Cl� channelbasolateral TAL membranes. XV. Molecular heterogeneity betweenfrom rabbit outer medulla. Kidney Int 48:1828–1836, 1995cortical and medullary channels. J Membr Biol 177:221–230, 2000

20. Adachi S, Uchida S, Ito H, et al: Two isoforms of a chloride3. Reeves WB, Winters CJ, Filipovic DM, Andreoli TE: Cl� chan-channel predominantly expressed in thick ascending limb of Henle’snels in basolateral renal medullary vesicles. IX. Channels fromloop and collecting ducts of rat kidney. J Biol Chem 269:17677–mouse mTAL cell patches and medullary vesicles. Am J Physiol17683, 1994(Renal, Fluid Electrol Physiol) 269:F621–F627, 1995 21. Uchida S: In vivo role of ClC chloride channels in the kidney.4. Friedman PA, Andreoli TE: CO2-stimulated NaCl absorption in Am J Physiol (Renal, Fluid Electrol Physiol) 279:F802–F808, 2000

the mouse renal cortical thick ascending limb of Henle. Evidence 22. Kieferle S, Fong P, Bens M, et al: Two highly homologous mem-for synchronous Na�/H� and Cl�/HCO exchange in apical plasma bers of the ClC chloride channel family in both rat and humanmembranes. J Gen Physiol 80:683–711, 1982 kidney. Proc Natl Acad Sci USA 91:6943–6947, 1994

5. Hebert SC, Andreoli TE: Effects of antidiuretic hormone on 23. Vandewalle A, Cluzeaud F, Bens M, et al: Localization andcellular conductive pathways in mouse medullary thick ascending induction by dehydration of ClC-K chloride channels in the ratlimbs of Henle. II. Determinants of the ADH-mediated increases kidney. Am J Physiol (Renal, Fluid Electrol Physiol) 272:F678–in transepithelial voltage and in net Cl� absorption. J Membr Biol F688, 199780:221–233, 1984 24. Uchida S, Sasaki S, Furukawa T, et al: Molecular cloning of a

6. Hebert SC, Culpepper RM, Andreoli TE: NaCl transport in chloride channel that is regulated by dehydration and expressedmouse medullary thick ascending limbs. I. Functional nephron predominantly in kidney medulla. J Biol Chem 268:3821–3824, 1993

25. Uchida S, Sasaki S, Nitta K, et al: Localization and functionalheterogeneity and ADH-stimulated NaCl cotransport. Am J Phys-characterization of rat kidney-specific chloride channel, ClC-K1.iol (Renal, Fluid Electrol Physiol) 10:F412–F431, 1981J Clin Invest 95:104–113, 19957. Sun A, Grossman EB, Lombardi MJ, Hebert SC: Vasopressin

26. Matsumura Y, Uchida S, Kondo Y, et al: Overt nephrogenicalters the mechanism of apical Cl� entry from Na�:Cl� to Na�:diabetes insipidus in mice lacking the ClC-K1 chloride channel.K�:2Cl� cotransport in mouse medullary thick ascending limb.Nat Genet 21:95–98, 1999J Membr Biol 120:83–94, 1991

27. Kobayashi K, Uchida S, Mizutani S, et al: Intrarenal and cellular8. Plata C, Meade P, Hall A, et al: Alternatively spliced isoformlocalization of CLC-K2 protein in the mouse kidney. J Am Socof apical Na�-K�-Cl� cotransporter gene encodes a furosemide-Nephrol 12:1327–1334, 2001sensitive Na�-Cl� cotransporter. Am J Physiol (Renal, Fluid Elec-

28. Yoshikawa M, Uchida S, Yamauchi A, et al: Localization of rattrol Physiol) 280:F574–F582, 2001CLC-K2 chloride channel mRNA in the kidney. Am J Physiol9. Kim G-H, Ecelbarger CA, Mitchell C, et al: Vasopressin in- (Renal, Fluid Electrol Physiol) 276:F552–F558, 1999creases Na-K-2Cl cotransporter expression in thick ascending limb 29. Tena-Sempere M, Manna PR, Zhang F-P, et al: Molecular mecha-

of Henle’s loop. Am J Physiol (Renal Fluid Electrol Physiol) 45: nisms of leptin action in adult rat testis: Potential targets for leptin-F96–F103, 1999 induced inhibition of steroidogenesis and pattern of leptin recep-

10. Reeves WB, McDonald GA, Mehta P, Andreoli TE: Activation tor messenger ribonucleic acid expression. J Endocrinol 170:413–of K� channels in renal medullary vesicles by cAMP-dependent 423, 2001protein kinase. J Membr Biol 109:65–72, 1989 30. Das SJ, Olsen I: Up-regulation of keratinocyte growth factor and

11. Ecelbarger CA, Kim G-H, Knepper MA, et al: Regulation of receptor: A possible mechanism of action of phenytoin in woundpotassium channel kir 1.1 (ROMK) abundance in the thick as- healing. Biochem Biophys Res Comm 282:875–881, 2001cending limb of Henle’s loop. J Am Soc Nephrol 12:10–18, 2001 31. Ogretmen B, Kraveka JM, Schady D, et al: Molecular mechanisms

12. Molony DA, Reeves WB, Hebert SC, Andreoli TE: ADH in- of ceramide-mediated telomerase inhibition in the A549 humancreases apical Na�,K�,2Cl� entry in mouse medullary thick as- lung adenocarcinoma cell line. J Biol Chem 276:24901–24910, 2001

32. Ohno J, Horio Y, Sekido Y, et al: Telomerase activation and p53cending limbs of Henle. Am J Physiol (Renal, Fluid Electrol Phys-mutations in urethane-induced A/J mouse lung tumor develop-iol) 21:F177–F187, 1987ment. Carcinogenesis 22:751–756, 200113. Winters CJ, Reeves WB, Andreoli TE: Cl� channels in basolat-

33. Chelly J, Montarras D, Pinset C, et al: Quantitative estimationeral renal medullary membrane vesicles. IV. Interactions of Cl�

of minor mRNAs by cDNA-polymerase chain reaction. Applica-and cAMP-dependent protein kinase with channel activity.tion to dystrophin mRNA in cultured myogenic and brain cells.J Membr Biol 122:89–95, 1991Eur J Biochem 187:691–698, 199014. Winters CJ, Zimniak L, Reeves WB, Andreoli TE: Cl� channels

34. Sucher NJ, Deitcher DL, Baro DJ, et al: Genes and channels:in basolateral renal medullary membranes. XII. Anti-rbClC-KaPatch/voltage-clamp analysis and single-cell RT-PCR. Cell Tissueantibody blocks mTAL Cl� channels. Am J Physiol (Renal, FluidRes 302:295–307, 2000Electrol Physiol) 273:F1030–F1038, 1997 35. Wu M-S, Biemesderfer D, Giebisch GH, Aronson PS: Role of

15. Sardet C, Franchi A, Pouyssegur J: Molecular cloning, primary NHE3 in mediating renal brush border Na�-H� exchange. Adapta-structure, and expression of the human growth factor-activatable tion to metabolic acidosis. J Biol Chem 271:32749–32752, 1996Na�/H� antiporter. Cell 56:271–280, 1989 36. Simon DB, Bindra RS, Mansfield TA, et al: Mutations in the

16. Tse C-M, Brant SR, Walker MS, et al: Cloning and sequencing chloride channel gene, CLCNKB, cause Bartter’s syndrome typeof a rabbit cDNA encoding an intestinal and kidney-specific Na�/ III. Nat Genet 17:171–178, 1997H� exchanger isoform (NHE3). J Biol Chem 267:9340–9346, 1996 37. Virlon B, Cheval L, Buhler J-M, et al: Serial microanalysis of

17. Choi JY, Shah M, Lee MG, et al: Novel amiloride-sensitive sodium- renal transcriptions. Proc Nat Acad Sci USA 96:15286–15291, 1999