Evidence for Distinct Vascular and Tubular UreaTransporters in the Pat Kidney1Dominique Promeneur, Germain Rousselet, Lise Bankir, Pascal Bailly, Jean-Pierre Cartron,
Pierre Ripoche, and Marie-Marcelle Trinh-Trang-Tan2
D. Promeneur, L. Bankir, M.-M. Trinh-Trang-Tan, INSERMU 90, HOpital Necker, Paris, France
G. Rousselet, P. Ripoche, Service de Biologie Cellu-laire, CEA/Saclay, Gif-sur-Yvette, France
P. Bailly, J.-P. Carfron, INSERM U 76, Institut National deIa Transfusion Sanguine. Paris, France
(J. Am. Soc. Nephrol. 199#{243};7:852-860)
ABSTRACTFacilitated urea transport has been demonstrated in
several mammalian tissues, including those of thecollecting ducts and red blood cells. Two urea trans-porters have been recently cloned: UT2, expressed inrabbit inner medullary collecting ducts, and HUT1 1,expressed in human erythrocytes. Because of signifi-cant identity (63%) between these Iwo transporters,
and because HUT1 1 is also expressed in the humankidney, they could represent the same transporterwith species-related differences in their sequences. Inthe study presented here, two different cDNA frag-ments, corresponding to the rat equivalents (rUT2 andrUT1 1) of the two previously cloned urea transporters,were isolated by reverse transcription-polymerasechain reaction. These rat probes were used for North-em analysis of RNA extracted from rattissues. From thefollowing findings, the results show that rUT2 and rUT1 1
are two distinct urea transporters: ( 1) The two cDNAfragments isolated in the rat exhibit different sequenc-es; (2) The mRNA for rUT2 is found exclusively in thekidney, with two transcripts (3.2- and 4.4-kilobase(kb)), whereas rUT1 1 (only one transcript, 4.2 kb) ispresent in the brain, spleen, kidney, and testis; (3) Inthe kidney, the inner stripe of the outer medullaexpresses rUT1 1 mRNA and the short transcript of rUT2,whereas the inner medulla expresses rUT1 1 and thetwo rUT2 transcripts; (4) In hydronephrotic kidneys thathave completely lost their tubular epithelium buthave Intact vasculature, rUT2 transcripts are no longerexpressed, whereas expression of rUT1 1 is intensified;
(5) Experimental chronic alterations in urine concen-
1 Received October 23, 1995. Accepted March � 1,1996.
2 Correspondence to Dr. M.-M. Trinh-Trang-Tan, INSERM Unite 90, H#{244}pitalNecker,
161 rue de S#{232}vres,75 743 Paris C#{233}dex15, France.
humans on a normal Western-type diet, as well as in
laboratory rats, urea is by far the dominant soluteexcreted in the urine (approximately 40% to 50% of
total solutes). The unique architecture of the kidney
and a sophisticated vasopressin-dependent mecha-
nism enable mammals to produce hypertonic urine.
Because of the relatively high rate of urea excretionmentioned above and the relatively low concentration
of urea in the blood, the concentration of urea in the
urine accounts for most of the solute-free water reab-sorbed by the kidney. On the other hand, it is wellestablished that urea itself contributes to the overall
urinary concentrating mechanism by being accumu-
lated in the inner medulla and thus permitting more
water to be extracted from the collecting ducts under
the influence of vasopressin ( 1).
It has long been known that urea equilibrates freely
through biological membranes by passive diffusion.
However, this diffusion is relatively slow. The move-ments of urea through collecting ducts, thin limbs of
Henle’s loops, and vasa recta (which contribute to its
accumulation in the renal medulla and to its excretion
in a concentrated urine) (2), as well as those in a few
other cells or tissues, including enythrocytes (3) and
liver (4), proceed at a much faster rate than that
expected from passive diffusion across a lipid bilayer,
and are inhibitable by phboretin (5). This suggests that
urea moves across the plasma membranes of these
cells by “facilitated” diffusion. In addition, a vasopres-
sin-induced facilitated urea transport has been dem-
onstrated in the terminal part of the inner medullary
results from the presence in cell membranes of spe-
cific carrier proteins (2).
Two such urea transporters have recently been
cloned and characterized. Using expression-cloning inXenopus oocytes, You and coworkers isolated a cDNA
from rabbit renal medulla (UT2) that encodes for a 397
amino acid protein described as the vasopressin-sen-
sitive urea transporter of the IMCD (7). This trans-
porter is highly expressed in rabbit IMCD and in
another structure in the renal outer medulla, as well
as in the rabbit colon (7). By screening a cDNA library
of human bone marrow cells, Olives et al. subse-
quently cloned and characterized a cDNA (HUT 11)
coding for a 39 1 amino acid protein, that exhibits 63%
rbtJT2
rUT2
rUTh
HUT11
rbUT2
rUT2
rUTh
HUThh
rbUT2
rUT2
rUTh 1
HUT11
rbUT2rUT2
rtJThh
HUT11
�ALGTIFSK WDLPVFTLPF NI�VTLXLAA
. . . .s.vOhigo 2.S.
. . . .NSML M.LSM. .5.
P11I�G1!LI� LFI�P.LICL_HAAIGSTM�M L�ALLTIATP1
V. . .V I S
. . . .A. . .C. ILL. . . .M LL.V I.G.SL.A..
G. ILL. . . .M LL.I A.G.SLSA..
LI�LRAI2- -�I�QVYGCDN
E. .KSL. -- . .V. .1
E. KS. .VG. .V. .1
Promeneur et al
Journal of the American Society of Nephrology 853
sequence identity with UT2 and a similar predicted
membrane topology (8). Both transporters include
similar glycosylation sites and several cysteine resi-
dues, seven ofwhich lie in an equivalent position. The
functional characterization of HUT1 1 suggests that it
represents the human erythrocyte urea transporter(8). It has been demonstrated that HUT1 1 protein is
carried by the Kidd antigen (9). HUT 1 1 mRNA was alsodetected in the human (tumoral) kidney (8) and inendothelial cells of vasa recta (10).
Because UT2 and HUT1 1 exhibit significant identityand are both expressed in kidney, they could repre-
sent the same transporter with species-related differ-
ences in their sequences. Our results reveal that thesetwo transporters are indeed different for the following
reasons: (1 ) Using RT-PCR with primers correspond-
ing to sequences specific for either UT2 or HUT1 1 , we
were able to clone two cDNA fragments correspondingto two mRNA in the same species, i.e. , the rat. (2) The
expression of the two mRNA in several tissues and in
the four regions of the kidney is different. (3) ThesemRNA are affected differently by experimental maneu-
vers designed to differentiate vascular and tubularstructures in the kidney. (4) Their expression is influ-
enced differently by chronic changes in the hydration/
vasopressin axis and in resulting urine concentrating
activity.
MATERIAL AND METHODS
Cloning of the Rat Equivalents of Rabbit UT2and Human UT1 1
Degenerated oligonucleotides were designed from aminoacid sequences specific for either rabbit UT2 (oligo 1 and 4) or
human UT1 1 (oligo 2 and 3) (Figure 1). Reverse transcrip-
tion-polymerase chain reaction (RT-PCR) was performed. aspreviously described ( 1 1), on total RNA from rat kidney. PCR
consisted of 30 cycles with an hybridization temperature of58#{176}C(for UT2) or 42#{176}C(for UT1 1). The faint bands obtainedafter migration on an agarose gel were picked out andreamplifled in the same conditions. The product of theseamplifications were then blunt-ended with T4 DNA polymer-
ase (Beckman, Gagny, France), and subcloned into pBlue-script II SK-plasmid (Stratagene, La Jolla, CA), digested withEcoRV. In each case, several clones were sequenced usingthe dideoxy chain termination method ( 12) and T7 polymer-ase (Pharmacia, Saint Quentin en Yvelines, France) to selectclones without Taq polymerase-induced mutations. OnerUT2 and one rUT1 1 clone with the wild-type sequence wereused for the Northern blotting experiments. The size of thetranscripts detected by Northern analysis with these clones
was evaluated by comparison with that of 1 8 and 28 5 RNA
and with two commercial size markers (RNA Ladder GibcoBRL and RNA Molecular Weight Marker II; Boehringer,Mannheim, Germany). DNA precipitations. digestions, and
clonings were performed according to standard protocols( 1 2). Restriction and modification enzymes were obtainedfrom New England Biolabs (Beverly. MA).
Distribution in Different Tissues and in theDifferent Zones of the Normal Kidney
This study was conducted in adult male Sprague-Dawleyrats. Total RNA was extracted from kidney, liver, salivaryglands, duodenum, colon, spleen. gastrocnemius muscle,brain, testis, and lung (about 500 mg of each tissue). us-
ing the acid-guanidium-isothlocyanate-phenol-chloroforme
(AGIPC) method (13).For intrarenal localization, the kidneys were taken out
while the rats were under general anesthesia. The left kidneywas removed first after its artery and vein were clamped to
(87) LT.W M. .L L. . . . .Y. . .AT. . . V.M F CA. . .T. . .F
AAXV�AALTN VLSVF�L21�C TWPFC]� (343)
. . .L. . . A. M 1�O1igo 4
T. .L.VGMA. FMAEV. . .A L (346)
TGHYNL!12T TLLQ�VSSVP_NITWSEIQVP
K. . . .AVTT. .�. .DV...
T. . .S K.FN �. . .LSAL
p. . .A K.VI.ITTA. �. .DLSAL
DSZXF�LCGF NSTLACIAVG GMFYVITWQT_HLLAVA.CAIS�
:: i IKID. .S. .W. . � I. . . .Oligo 3j
ED W. . � M. . . .MAL. . . .1 . . . . LG. ...
Figure 1 . Molecular characterization of rat equivalent probes for UT2 and UT1 1 . A fragment of the rabbit UT2 (rbUT2) is shown(residues 86 to 343). The corresponding fragments of human UT1 1 (HUT1 1) (residues 87 to 346) and of the two rat equivalentscloned in the study presented here (rUT2 and rUT1 1 , see Results) are shown for comparison. The residues conserved In the foursequences are indicated in bold type and underlined. Conserved residues in HUT1 1 , rUT2, and rUT1 1 are indicated by dots. Theoligonucleotides used for RT-PCR for either UT2 (oligo 1 and 4) or UT1 1 (oligo 2 and 3) are indicated In boxes. Putativeglycosylafion sites are double-underlined.
Renal (Vascular and Tubular) Urea Transporters
854 Volume 7 ‘ Number 6 ‘ 1996
prevent the consequences of severe hemorrhage on the right
kidney (sudden fall in blood pressure followed by prompt and
intense vasopressin release). Both kidneys were immediately
placed in ice-cold saline and sliced with a razor blade. Thewhole inner medulla (including papifia) (IM) and representa-tive fragments of cortex and of the two subregions of theouter medulla, outer (OS) and inner stripe (IS), were carefully
sampled. Figure 2 illustrates the criteria adopted to recognizethe limits between these zones (also see Reference 14). RNA
was extracted from approximately 70 mg of tissue of each
zone in 0.75 mL of RNA NOW5 (Biogentex, Seabrook, TX).
Study in Hydronephrotic Kidneys
Hydronephrosis of the left kidney was induced in five maleSprague-Dawley rats (body weight, approximately 270 g)
according to the procedure of Steinhausen et a!. ( 15). While
the rats were under pentobarbital anesthesia, the left kidneywas exposed through laparotomy. The left ureter was ligatedwith a silk thread to interrupt urine drainage permanently.The left renal artery was occluded for 60 min with a smallclamp (MC 43; Mona, Paris, France). After the clamp wasremoved, the abdominal incision was closed and the ratswere placed in normal cages with standard laboratory rat
chow and tap water ad libitum. After 9 wk, the rats were
anesthetized and both kidneys were removed and weighed.Total RNA was extracted separately from the right and left
kidneys of each rat by the AGIPC method.
G
MR
VB
AV
Figure 2. Diagram of the different zones of the rat kidneyshowing the different nephron segments in each zone. Onfresh kidney sections, the limit between cortex (C) and outerstripe (OS), which have the same tubular composition andthus the same natural color, is not easy to recognize. Understereomicroscopic observation, It is possible to distinguishthe reddish glomeruli (G) present only in the cortex. Arcuateveins (AV), running at the corticomedullary border, formwide lacunas, which were used for defining the line alongwhich C and OS were separated In this study. Medullary rays(not discernible on fresh tissue) are indicated (MR). The innerstripe (IS) Is clearly discernible by its reddish appearance(because of the very dense capillary plexus surrounding thetubules in this region), contrasting with the white inner me-dulla (IM). Vascular bundles (VB) can be observed underthe stereomicroscope as reddish stripes running through theIS.
Study in Rats with Different Urine ConcentratingActivity
Urinary concentrating activity (UCA) was altered chroni-
cally by modifying the normal balance between vasopressin
secretion and thirst. Twelve male Sprague-Dawley rats (mi-
tial body weight. approximately 230 g) were housed in mdi-vidual metabolic cages and divided into four groups of three
rats each. In two groups. UCA was reduced by increasing
daily water intake moderately (MWI) or highly (HWI). This
was achieved by mncorporatmng the daily food ration in awater-rich agar gel. bringing either 1 .6 mL water (MW!) or3.2 mL water (HWI) and 40 mg agar per g food ( 1 6). In another
group (dDAVP), UCA was increased by chronic infusion of200 ng/day ofdDAVP ( 1 -desamino 8-D-arginine vasopressin;Ferrmng, Malm#{246}, Sweden), a synthetic V2 agonist of vasopres-
sin (AVP) devoid ofvasopressor effects. The drug was infusedvia osmotic mmnipumps implanted mntraperitoneally under
ether anesthesia ( 1 7). No intervention was performed in ratsin the last group, which thus had a normal water intake(NWI) and UCA, depending on their own balance between
thirst and vasopressmn secretion. All rats had free access totap water during the entire study.
We noticed in a previous study that rats offered water-enriched food tend to eat more than rats offered dry food. Toensure the same food intake in all rats, they were all given
only 16 g dry food per day (or, for HWI and MW!, the
equivalent in agar gel), I.e. , an amount slightly lower than thespontaneous intake of rats fed dry powder ad Itbitum (ap-proximately 18 g/day). Dry agar was added in an amount
appropriate to the dry food in NW! and dDAVP rats so that all
rats ate the same amount of agar per day (0.64 g/day). The
food used in this study was a synthetic powdered foodcontaining 23% protein in the form of casein (INRA; La
Mini#{232}re, Guyancourt, France).
Rats were maintained under these conditions for 5 wk.
C They were weighed every week and grew normally. Twenty-four h urine samples were collected for the last 3 days. Urine
Os flow rate and osmolality (freezing point method; Roebling
IS osmometer, Berlin, Germany) were measured, and averages
IM for the 3 days were calculated for each rat. A blood samplewas collected on the last day. Urea concentration in urine
was measured on an automatic analyzer (Hitachi 717;
Boehringer. Mannheim, Germany). Kidneys were removedand weighed. The total RNA was extracted from whole kid-neys by the AGIPC method.
Northern Analysis
Localization and quantification of the message for UT2 and
UT! 1 was performed using the two rat clones obtained asdescribed above. The full-length eDNA encoding rat y-glu-
tamyltranspeptidase, used as a marker of tubular epithe-
hum, was kindly provided by Dr. G. Guellaen ( 18). Thehuman probe for glyceraldehyde 3-phosphate dehydroge-
nase (G3PDH), used as reference, was purchased from Clon-tech (Palo Alto, CA).
Probes were radiolabeled by the random-priming tech-nique (Megaprime labeling kit; Amersham, Les Ulis, France)using [a-32P] dCTP to a specific activity of approximately 2 XiO#{176}cpm/�tg cDNA.
RNA samples ( 1 5 �g) were denatured and separated byelectrophoresis on agarose gel ( 1 .2%) containing 0.6 M form-
aldehyde. Equal RNA loading was verified by visual mnspec-
tion after coloration with ethidium bromide. The RNA weretransferred overnight from gel to nitrocellulose membranes
(Hybond C-Extra, Amersham), which were then baked in a
rUT2
rUTh
28 S
18 S
28 S
Rat 1 2 3
�CJ) *(/) >,C/)oOS�oOcQ� oocQ�
rUT2
rUThh
28 S
18S
Promeneur et aI
Journal of the American Society of Nephrology 855
vacuum oven (2 h at 80#{176}C). Blots were placed in a glass
hybridization tube containing 10 mL medium composed of
5 x SSPE (see composition below), 50% formamide, 0.5%
sodium dodecyl sulfate (SDS), 1 mg/mL BSA, Ficoll 400, andpolyvinylpyrrolidone, and 40 j.�g/mL denatured salmonsperm DNA. Prehybridization was performed at 42#{176}Cfor 4 hin a hybridization oven. The radiolabeled probe was thenadded to prehybridization medium (approximately 4 X 106cpm/mL) and membranes were incubated overnight at 42#{176}C.
The blots were washed twice at 25#{176}Cin 2 X SSPE and 0.1%SDS for 15 mm. and once at 55#{176}Cin 1 x SSPE and 0. 1% SDSfor 30 mm. The fmal wash was carried out at 55#{176}Cin 0. 1 xSSPE and 0. 1% SDS for 30 min. The composition of 1 X SSPEwas: NaCl, 150 mM; NaH2PO4, 10 mM; EDTA, 1 mM (pH 7.4).
The blots were exposed for 1 to 6 days to autoradiographic
film (Hyperfilm-MP, Amersham), with intensifying screens, at-80#{176}C. Expression of mRNA was quantified by densitometry
(Deskan II and ScanAnalysis; Biosoft, Cambridge, UnitedKingdom) after the autoradiograms were scanned (Scan Jet;Hewlett Packard, Orsay, France). The results are expressedas arbitrary units of density.
RESULTS
Molecular Characterization of Two Rat UreaTransporters
The first clone (rUT2), amplified with UT2-specific
oligonucleotides, consists of a 773-base pair [bp]
cDNA coding for 244 amino acids (Figure 1 ). This
sequence exhibits 89.3% identity with rabbit UT2 and
67.6% with human UT1 1. The cDNAfragment (rUT1 1)
amplified with HUT1 1-specific oligonucleotides is 384
bp long and codes for a 1 1 4 amino acid peptide
exhibiting 83.3% with HUT1 1 and 68.4% identity with
rabbit UT2 (Figure 1). These two rat sequences share
66.7% of their amino acids. These homologies suggest
that rUT2 and rUT 1 1 most probably represent the rat
UT2 and UT1 1 , respectively. Recently, the rat UT2
amino acid sequence has been made available in the
GenPro database (accession number RNU 09957).
Our rUT2 sequence is identical to the deposited rat
UT2 sequence.
Distribution in Different Tissues and in theDifferent Zones of the Kidney
Northern blot analysis with rUT1 1 probe revealed a
4.2-kb transcript intensely expressed in the brain,
well-expressed in the spleen, kidney, and testis, and
moderately expressed in the duodenum and colon
(Figure 3). A longer exposure of the autoradiograms
revealed a faint signal in the lungs, but remained
negative for the liver (not shown). rUT2 produced a
positive signal only in the kidney, with two transcripts
of 4.4 and 3.2 kb of different intensities (Figure 3).
Within the kidney, neither probe hybridized with the
mRNA extracted from the cortex, suggesting that
these two urea transporters are not expressed in
gbomeruli. proximal and distal convoluted tubules,
and cortical collecting ducts. Both probes hybridized
heavily in the two deepest layers ofthe medulla (IS and
IM) with a different distribution (Figure 4). For rUT2,
only the short transcript was present in IS, whereas
Q).�
E(I)
-0
�E EC0) a� C
�‘C 0� C
C a _ a � �
C.- 0 0 a a C � - ��B � #{149}a#{149}a�> � n aCl) 0 0 Cl) � s2 � � �- .i
15 S
Figure 3. Tissue localization of rUT2 and rUT1 1 , determined by
Northern analysis of 15 �g total RNA extracted from differentorgans.
28 S
18 S
Figure 4. Intrarenal localization of rUT2 and rUT1 1 , deter-mined by Northern analysis of 15 �g total RNA extractedfrom the four zones of the kidney in three different rats. Cx,cortex; OS and IS, outer stripe and inner stripe of the outermedulla, respectively; lM, Inner medulla.
both the short and the long transcripts were observed
in the IM. Interestingly, the expression of the long
transcript was roughly similar in all rats, whereas that
of the short transcript exhibited a varying intensity in
the different rats. The rUT 1 1 probe revealed a single
transcript that was more heavily expressed in the IM
than in the IS in all three rats, and was very weakly
expressed in the OS (Figure 4).
Study in Hydronephrotic Kidneys
Complete obstruction of the urinary tract and the
resulting increase in hydrostatic pressure are known
to induce a progressive atrophy of the renal paren-
Renal (Vascular and Tubular) Urea Transporters
856 Volume 7 . Number 6 . 1996
Rat
rUT2
rUTh
4
��ua � � � �-- 01 (.)I 01 01 01
Rat 1 2 3 4 5
�‘ � � � � � � �
C.) I 01 0 I 0 I 01
‘y-GT
Figure 5. Northern analysis of mRNA for rUT2, rUT1 1, and y.GT(gamma-glutamyltranspeptidase) in control (CK) and con-tralateral hydronephrotic (HK) kidneys of five different rats.lntrarenal localization of rUT2 and rUT1 1 In normal rat kidneyis shown in the first 4 lanes.
chyma with full preservation of the vascular elements
of the kidney. The main arteries and veins, arteriolesand venules, glomeruli and main postglomerular yes-
sels remain discernible ( 15). This approach was used
to provide further insight into the presence of mRNA
for the two urea transporters in vascular and/or
tubular structures of the kidney.
Nine wk after surgery, the left hydronephrotic kid-
ney (HK) appeared as a widely dilated, well-vascular-
ized, thin-walled bladder filled with fluid (approxi-
mately 9 to 10 mL of fluid per kidney). The right
control kidney (CK) showed some compensatory hy-
pertrophy (2. 14 ± 0.03 g, which is about 50% heavier
than one kidney of a normal rat of the same body
weight). The weight of the emptied HK was 0.65 ±
0.05 g.
28 S The disappearance of tubular structures in the HKwas attested by the absence of transcript for y-glu-
tamyltranspeptidase, an enzyme of the brush border
membrane of the proximal tubule (Figure 5). With
18 S regard to urea transporters, the rUT 1 1 probe revealeda much higher expression of the 4.2-kb transcript in
28 S HK than in CK. In contrast, the rUT2 probe, which
revealed the usual two transcripts in CK, did not
detect any mRNA in HK (Figure 5).18 S
Influence of Urine Concentrating Activity
Chronic addition of water to the food or infusion ofdDAVP did not significantly influence the growth rate
of the rats. However, the rats that received dDAVP28 S tended to be somewhat lighter than those of the three
other groups (Table 1). As expected, the experimental
protocol used to alter the UCA influenced the drinking
18 S behavior of the rats and their urine flow rate andosmolality without influencing their urea excretion
rate. In HWI and MWI, in spite of the water supplied
with the food (exceeding, in HWI, the spontaneouswater intake of normal rats), rats drank additional
water from the bottle, as already described (16,19).
Their total fluid intake was thus higher than that of
NWI rats. In contrast, rats infused with dDAVP re-
duced their water intake (Table 1 ). Plasma osmolality
did not differ between groups. Mean 24-h urine osmo-lality ranged from 586 ± 53 for the HWI group to 2575
± 1 19 mosmol/kg H20 for the dDAVP group. dDAVP
infusion led to an increase in kidney weight relative to
body weight (Table 1) as reported in several previous
studies (17; also see the review in Reference 20).
Increasing UCA did not influence the abundance of
the long transcript revealed by rUT2 in the rat kidney
but greatly enhanced that of the short transcript. This
effect was particularly intense in rats infused withdDAVP (Figure 6). Quantification of the signals by
densitometry shows that dDAVP rats express short
rUT2 about fourfold more than NWI rats although
their urine osmolality is only modestly higher (Figure
TABLE 1 . Body weight, kidney weight, and urine parameters in rats with different urine concentrating activity
Statistical analysis was one-way analysis of variance. HWI. high water intake; MWI, moderately high water Intake; NWI. normal water; dDAVP,treated with 1-desamino 8-D-arginine vasopressin; NS, not significant.
HWI MWI NWI dDAVP
Rat I 2 �3 4 5 E3 7 8 9 101112
rUT2
rUTh
Figure 6. Northern analysis of mRNA for rUT2 and for rUT1 1 inrat kidneys after chronic alterations of urinary concentratingactivity. HWI, MWI, and NWI, rats with high, moderate, ornormal water Intake, respectively; dDAVP, rats treated withdDAVP.
rUT2 UU
I
. Sr = 0.907
� � � L
rUTh
a
a ar=0.592
200�
150’
100
50�
0#{149}
600
400
200
.�
‘�
I
0 500 1000 1500 2000 2500 3000
0
‘J -�
Promeneur et al
Journal of the American Society of Nephrology 857
7). A significant correlation was observed between
short rUT2 and urine osmolality for the 12 rats (r =
0.685, P < 0.01). However, a higher correlation coef-ficient was observed when linear regression concerned
only the nine rats in which endogenous AVP plasmalevels were altered (dDAVP excluded) (r = 0.907, P <
0.001) (Figure 7).The UCA also influenced the rUT1 1 message (Figure
6). A significant positive correlation was found be-tween rUT! 1 mRNA expression and urine osmolality
(r = 0.592, P < 0.05). Treatment with dDAVP did notinduce a marked overexpression of rUT 1 1 , as it did for
rUT2 (Figure 7). These changes in mRNA expression
were specific of urea transporters because mRNAexpression of G3PDH in the same kidneys was unal-
tered by changes in UCA (Figure 7).
DISCUSSION
The existence of facilitated urea transport in themammalian collecting ducts and erythrocytes hasbeen known for a long time. Two urea transportershave been recently cloned in two different species: aninner medullary-derived urea transporter (UT2 [7] inthe rabbit), and an erythrocyte urea transporter(HUT1 1 [81 in humans). In the investigation presented
here, we studied the messages for these two transport-ers in a single species, the rat. In this species, the
isolation of two different cDNA fragments correspond-ing, respectively, to UT2 and to UT1 1 , and the study of
gene expression of these two transporters confirm thatthey correspond to two different genes. We observed
that their tissue localization is different. Althoughthey are both expressed in the kidney, they are most
probably not present in the same structures, and areregulated differently by changes in urine concentrat-
ing activity.
Cloning of Two Urea Transporters in the RatThe isolation of a cDNA fragment of UT2 and UT1 1
from the same species provides us with a valuable tool
for investigating their respective contribution in 5ev-
2000
150028 S
� 1000
185
28S � 400
185 :�‘ 300
200
100
0
0
G3PDH
��_n 0
0 500 1000 1500 2000 2500 3000
Uosm (mOsm I kg H20)
Figure 7. Relationship between urine osmolality (Uosm) andrenal abundance of mRNA In 12 rats with different urInaryconcentrating activity. The mRNA abundance was quanti-fled by densitometry. Unear regressions are shown. Upperpanel: 5, short transcript (closed symbols); 1, long transcript(open symbols); the three closed squares correspond to theshort transcript In the three rats treated with dDAVP, whichwere excluded from the regression (see Text).
eral urea transport-dependent physiopathologicalconditions. The sequence analysis of rabbit UT2 andHUT1 1 already demonstrates that the duplication of
this gene in the genome occured after the already
described internal duplication (2 1). Moreover, the sig-nificant homology between UT2 and UT1 1 suggests
either that the function of the protein is highly se-
quence-dependent, or that the duplication is relatively
Renal (Vascular and Tubular) Urea Transporters
858 Volume 7 . Number 6 . 1996
recent. By comparison, there is only 45% identity
between the human red blood cell water channel
aquaporin 1 and the human vasopressin-sensitiverenal water channel aquaporin 2.
Localization of the Two Urea Transporters
In the rat, UT2 mRNA was found exclusively in the
kidney, whereas UT1 1 mRNA was present in several
organs. The signals detected in the brain, spleen.kidney, testis, and lung cannot be accounted for by
reticulocytes trapped in these organs because other
highly vascularized organs like liver or muscle did not
express this message. In intact rat kidney, rUT1 1
detected only one transcript, whereas HUT1 1 revealedtwo transcripts in the human tumoral kidney (8). The
very strong signal observed for rUT1 1 in the brain
suggests the existence of facilitated urea transport in
this organ. Such a transport has previously been
described in the frog brain (22). Regarding the mam-
malian brain, several reports indicate that the blood-brain barrier has a low permeability to urea (23,24).Further studies are required to investigate the func-tional significance of this transport. as well as the
brain areas where it takes place. The existence of
facilitated urea transport in the testis and lung was
not suspected and also deserves to be investigated. A
urea transporter might be necessary in cells that
possess a significant arginase activity (as known for
brain and testis). This enzymatic activity cleaves argi-
nine into ornithine (the precursor of the polyaminepathway), and urea (which needs to exit from the
cells).Neither rUT2 nor rUT 1 1 probes hybridized with
mRNA extracted from the rat liver. Hepatocytes syn-
thesize urea at a high rate and have been shown toexhibit facilitated urea transport. which accelerates
the secretion of newly synthesized urea into blood(4,25). Accordingly, a third urea transporter, with noor weak homology with UT2 and UT 1 1 , is likely
present in the liver. At variance with the rabbit (7), aUT2 message was not detected in the rat colon. This
might be accounted for by the different dietary regi-men of the two species and the possibility, in herbi-vores, to reuse urea nitrogen after urea is broken
down by the urease-rich bacterial microflora residing
in their digestive tract (26).In the kidney, a facilitated transport of urea has
been demonstrated in several tubule segments andvessels (2). This Includes the IMCD (6), the thin de-
scending limb of short-looped nephrons and thin as-
cending limb oflong-looped nephrons (27,28), and the
descending vasa recta in the inner stripe of the outermedulla ( 10,29). The existence ofan active urea trans-port in the pars recta of the proximal tubule in the
mammalian kidney remains controversial. Kawamura
and Kokko reported a modest but significant activesecretion of urea in the rabbit pars recta (30), butKnepper did not confirm this finding and concluded
that urea transport in this segment could be explained
by simple passive diffusion through the lipid bilayer of
the plasma membranes (3 1).Some information has already been obtained in
different species regarding the intrarenal localization
of UT2 and UT 1 1 . In the rat and rabbit kidney, the
UT2 message has been observed in the outer and
inner medulla but not in the cortex (7,25,32,33).
These studies suggest that the long UT2 transcript
resides in the IMCD and the short UT2 transcript in
(7,32). In the human kidney, HUT! 1 has been local-
ized in vasa recta of the other and inner medulla byimmunofluorescence and in situ hybridization (10).
In the study presented here, the expression of the
messengers for the two transporters was investigated
in parallel in the same species. Results show that the
UT2 message disappears completely in the hydrone-
phrotic kidney, suggesting that it is localized only in
tubular structures. According to the conclusions of
Smith et al. (32) and to present results, the short UT2signal seen in outer and inner medulla likely resides in
thin limbs (34). The message of UT! 1 found in therenal medulla, and still present in the hydronephrotickidney. is compatible with the immunodetection of the
UT! 1 protein in vasa recta reported by Xu et at. (10).
The faint UT! 1 signal observed in OS likely originates
from branches of the efferent arterioles of deep gb-meruli and early vascular bundles (Figure 2). In sum-
mary, the information available so far regarding the
localization of urea transporters ( 10,32; present
study), suggests that UT! 1 is located in vasa recta,
short UT2 in thin limbs, and long UT2 in IMCD.
Influence of UCA on the Two Urea Transporters
Urea accumulation in the inner medulla is known to
be essential for the efficient operation of the urinary
concentrating mechanism (2). This accumulation re-sults (1 ) from the permanent delivery of concentrated
urea in the inner medulla, permitted by the action of
vasopressin on urea permeability in the terminal cob-
becting duct (28), and (2) from efficient countercurrent
exchanges that continuously return urea carried awayfrom the IM in the ascending venous blood to the inner
medulla. This intrarenal urea recycling involves the
exit of urea from the terminal IMCD and reentry of
urea in thin limbs and descending vasa recta (1,35-
37).
Smith et at. have investigated the influence of hy-dration state on UT2 expression (32). Their experi-
mental protocols induced complex changes in fluid
and nutrient balance because they involved simulta-
neous and/or successive deprivation of food and fluid
and/or addition of glucose in the drinking water.
These protocols most likely induced differences in
total urea excretion, in addition to differences in urine
concentration. In the study presented here, UCA was
altered in both directions either by increasing fluidintake or infusing dDAVP. Rats in the different groups
were given a limited amount of food per day. These
Promeneur et al
Journal of the American Society of Nephrology 859
protocols modified only the water intake/vasopressin
axis and ensured stable and equal food intake among
groups during the entire study, resulting in similarurea excretion in all rats (Table 1). This study, bike
that of Smith et at. , revealed that the message for thelong UT2 transcript (coding for the collecting duct
urea transporter) did not increase with increasingUCA. In contrast, the message for the short UT2
transcript (probably corresponding to thin limb urea
transporter) was upregulated in both studies. The
results presented here also reveal that UT 1 1 (which is
likely to be the vasa recta transporter) is also upregu-
bated with increasing UCA. Taken together, these ob-servations suggest that chronic changes in urinaryconcentrating activity influence the abundance of
urea transporters in thin limbs and in vasa recta, but
not in terminal CD. Further studies are required to
determine whether these changes result from a directeffect of vasopressin (34) on gene expression or aresecondary to the operation of the urinary concentrat-
ing mechanism and the associated intrarenal urea
recycling.
The KIdd antigen has recently been shown to becarried by UT! 1 protein (9). Homozygous individuals
[Jk(a-b-)J lack this antigen, and their red blood cells
do not exhibit facilitated urea transport. BecauseUT! 1 protein is present in vasa recta ( 10), the ureatransporter is most likely also missing in the vasa
recta of Jk(a-b-) individuals. In two of these sub-
jects, Sands et at. observed a moderate concentratingdefect (their maximum urine osmobality was 818 and767 mosmol/kg H2O versus approximately 1000mosmol/kg in control subjects) (38). This suggests
that in humans, urea transport by red cells and urearecycling in vasa recta play a modest role in the
urinary concentrating process.Among rats that exhibited the highest urine osmo-
balities (>2000 mosmob/kg H20), those of the dDAVP
group had about a threefold higher expression of theshort UT2 transcript than those of the NW! group.Such a marked overexpression did not occur for UT 1 1.In the first three groups, the experimental increases in
water intake (HWI and MWI) or spontaneous intern-
dividual variations (NWI) resulted in differences in the
secretion of endogenous AVP. This natural hormoneexerts different influences in its target cells throughtwo different types of receptors, V1 and V2. In thefourth group, the infusion of exogenous dDAVP mim-
icked only the V2 effects of AVP and likely decreased
endogenous AVP secretion, thus minimizing possible
vi effects. It is well established that collecting ductspossess both V1 and V2 receptors and that the V1
action of vasopressin weakens the V2-dependent in-crease in water (and probably also urea) permeabilityby stimulating the synthesis of prostaglandmns, which
act as negative feedback modulators of the antidi-
uretic effect of AVP (39,40). It may thus be assumed
that dDAVP exerted a much stronger influence than
AVP on intrarenal urea recycling and urea accumula-
tion in the inner medulla. It is conceivable that the
UT2 message was influenced to a much greater extent
by dDAVP than by AVP because the V2 effects of thisdrug were not counterbalanced by a V1 influence.
Another consequence of this stronger influence of
dDAVP over AVP is revealed by the kidney hypertro-
phy induced by dDAVP in this study (Table 1) and a
preceding study (17). Notably, dDAVP did not amplifythe UT! 1 message more than AVP did. This is consis-tent with the fact that UT ! 1 is likely located in vasarecta, which possess V1 but not V2 receptors.
In conclusion, the study presented here shows thatUT2 and UT! 1 are indeed two different urea trans-porters. UT2 is probably expressed only in the kidneyand, within this organ, only in epithelial structures.The two transcripts of UT2 are unequally expressed inthe different regions of the rat renal medulla and only
the short transcript is influenced by chronic changesin the vasopressin/water intake axis. The UT! 1 mes-
sage is expressed in renal vessels (most likely in thearterial vasa recta) but not in renal tubules, and is
influenced by changes in the vasopressin/water in-take axis. The expression of this transporter in otherorgans (testis and brain) deserves further studies.
ACKNOWLEDGMENTThis work was supported in part by the Groupe DANONE. Paris,France.
REFERENCES
1 . Bankir 1.: Urea and the kidney. In: Brenner BM, RectorFC. Eds. The Kidney. 5th Ed. Philadelphia: W.B. Saun-ders Company; 1995:571-606.
2. Marsh DJ, Knepper MA: Renal handling of urea. In:Windhager EE, Ed. Handbook of Physiology. New York:Oxford University Press; 1992:1317-1348.
3. Macey R: Transport ofwater and urea in red blood cells.Am J Physiol 1984;246[Cell Physiol 151:C195-C208.
4. Effros R, Jacobs E, Hacker A, Ozker K, Murphy C:Reversible inhibition of urea exchange in rat hepato-cytes. J Cliii Invest 1993;91:2822-2828.
5. Chou C, Knepper M: Inhibition ofurea transport in innermedullary collecting duct by phloretin and urea ana-logues. Am J Physiol 1989;257(Renal Fluid ElectrolytePhysiol 261:F359-F365.
6. Sands JM, Nonoguchi H, Knepper MA: Vasopressineffects on urea and H2O transport in inner medullarycollecting duct subsegments. Am J Physiol 1987;253[Renal Fluid Electrolyte Physiol 221:F823-F832.
7. You G, Smith G, Kanal Y, Lee W, Stelzner M, Hediger M:
Cloning and characterization of the vasopressin-regulated urea transporter. Nature (Lond) l993;365:844-847.
8. Olives B, Neau P, Bailly P. Hediger M, Rousselet G,Cartron J-P, Ripoche P: Cloning and functional expres-sion of a urea transporter from bone marrow cells. J BiolChem 1994;269:31649-31652.
9. Olives B, Mattei M, Huet M, Neau P. MartIal S. Cartron J,Bailly P: Kidd blood group and urea transport function ofhuman erythrocytes are carried by the same protein.J Biol Chem 1995;270:15607-15610.
10. Xu Y, Olives B, Bailly P. Rlpoche P, Ronco P, Cartron J,Rondeau E: Endothelial expression of a new urea trans-porter in human kidney [Abstracti. J Am Soc Nephrol1995;6:33 1A.
1 1 . Abrami L, Simon M, Rousselet G, Berthonaud V. BuhlerJ, Ripoche P: Sequence and functional expression of anamphibian water channel, Fa- CHIP: A new member ofthe MIP family. Biochim Biophys Acta 1994;1192:147-151.
Renal (Vascular and Tubular) Urea Transporters
860 Volume 7 ‘ Number 6 . 1996
12. Sambrook J, Fritsch E, Maniatis T: Molecular Cloning,A Laboratory Manual. Cold Spring Harbor, NY: ColdSpring Harbor Laboratory Press; 1989.
13. Chomczynski P. Sacchi N: Single-step method of RNAisolation by acid guanidinium thiocyanate-phenol-choroform extraction. Anal Biochem 1987; 162:156-159.
14. Bankir L, Hollenberg NK: In vivo staining of the kidneywith Alcian Blue: an adjunct to morphological and phys-iological studies. Renal Physiol 1983;6: 151-155.
15. Stelnhausen M, Snoei H, Parekh N, Baker R, JohnsonP: Hydronephrosis: A new method to visualize va affer-ens, efferens, and glomerular network. Kidney mt 1983;23:794-806.
1 6. Bouby N, Bachman 5, Blchet D, Bankir L: Effect of waterintake on the progression of chronic renal failure in 5 I 6nephrectomized rat. Am J Physiol 1990;258IRenal FluidElectrolyte Physiol 271:F973-F979.
17. Bankir L, Fischer C, Fischer 5, Jukkala K, Specht HC,Kriz W: Adaptation of the rat kidney to altered waterintake and urine concentration. Pflugers Arch 1 988;4 12:42-53.
18. Laperche Y, Bulle F, Aissani T, Chobert MN, AggerbeckM, Hanoune J, Guella#{235}nG: Molecular cloning and flu-cleotide sequence of rat kidney -y-glutamyl transpepti-dase cDNA. Proc Nail Acad Sci USA 1986;83:937-94 1.
19. Fltzimons J, Magnen JL: Eating as a regulatory controlof drinking in the rat. J Comp Physiol Psychol 1969;67:273-283.
20. Banklr L, Kriz W: Adaptation of the kidney to proteinintake and to urine concentrating activity: Similar con-sequences in health and CRF. Kidney mt 1995;47:7-24.
2 1 . Rousselet G, Bailly P, Ripoche P: Tandem sequencerepeats in urea transporters: Identification of an UTsignature sequence. Am J Physiol l996;270[Renal FluidElectrolyte Physiol 391:F554-F555.
29. Pallone T: Characterization of the urea transporter inouter medullary descending vasa recta. Am J Physiol1 994;267lRegulatory Integrative Comp Physiol 361:R260-R267.
30. Kawamura 5, Kokko J: Urea secretion by the straightsegment of the proximal tubule. J Clin Invest 1976;58:604-6 12.
3 1 . Knepper MA: Urea transport in nephron segments frommedullary rays of rabbits. Am J Physiol 1983;244[RenalFluid Electrolyte Physiol 131:F622-F627.
32. Smith C. Lee W, Martial 5, Knepper M. You G, Sands J,Hediger M: Cloning and regulation of expression of therat kidney urea transporter (rUT2). J Clin Invest 1995;96:1556-1563.
33. Ashkar Z, MartIal 5, Isozaki T, Price 5, Sands J: Ureatransport in initial IMCD of rats fed a low-protein diet:Functional properties and mRNA abundance. Am JPhysiol 1995;268lRenal Fluid Electrolyte Physiol 371:F12 18-F1223.
34. Bankir L, Promeneur D, Trinh-Trang-Tan M: The 2.9 kbmRNA of the urea transporter UT2 expressed in theupper part of the rat inner medulla is selectively in-creased by acute ADH administration. FASEB (Fed AmSoc Exp Biol) J 1996;1O:273A.
35. Valtin H: Structural and functional heterogeneity ofmammalian nephrons. Am J Physiol 1977;233[RenalFluid Electrolyte Physiol 21:F491-F50 1.
36. Knepper MA, Roch-Ramel F: Pathways of urea transportin the mammalian kidney. Kidney Int 1987;3 1:629-633.
37. Bankir L, Bouby N. Trinh-Trang-Tan MM: The role of thekidney in the maintenance of water balance. Baifii#{232}resChin Endocrinol Metab 1989;3:249-31 1.
38. Sands J, Gargus J, Fr#{246}hlich 0, Gunn R, Kokko J:Urinary concentrating ability in patients with Jk(a-b-)blood type who lack carrier-mediated urea transport. JAm Soc Nephrol 1992;2:1689-1696.
39. Beck T, Dunn M: The relationship of antidiuretic hor-mone and renal prostaglandins. Miner Electrolyte Metab1 98 1 ;6:46-59.
40. Zipser R, Myers 5, Needleman P: Stimulation of renalprostaglandins synthesis by the pressor activity of vaso-pressin. Endocrinology 198 1 ; 108:495-499.
![1-[3-(4-Nitrophenyl)propanoyl]urea acetic acid monosolvate](https://static.documents.page/doc/80x56/6361b653a514f501cd0ce6ae/1-3-4-nitrophenylpropanoylurea-acetic-acid-monosolvate.jpg)
