View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Kidney International, Vol. 49 (1996), pp.1615—1623

Structure,regulation and physiological roles of transporters

MATFHIAS A. HEDIGER, ClJG P. SMITH,1 GUOFENG You, WEN-SEN LEE,2 YosHwTsu KANAI,3 and CHAIRAT SHAYAKUL

Renal Division, Department of Medicine, Brigham and Women's Hospital, Harvard Medical School, and Department Biological Chemistry & Molecular Phannacology, Harvard Medical School, Boston, Massachusetts, USA

Structure, regulation and physiological roles of urea transporters. on land [1]. In aquatic animals, the maintenance of low ammonia Urea is the major constituent of the urine and the principal means for concentrations in the body has not been a problem because of the disposal of nitrogen derived from amino acid metabolism. Specialized phioretin-inhibitable urea transporters are expressed in medulla unlimited water supply. For example, teleost fish maintain their and play a central role in urea excretion and water balance. Thesebody osmolality below that of sea water and dispose waste transporters allow accumulation of urea in the medulla and enable the nitrogen as ammonia with little metabolic expenditure [2—41.By kidney to concentrate urine to an osmolality greater than systemic plasma. Recently, expression cloning with Xenopus oocytes has led to the isolation contrast, land animals have overcome the problem of limited of a novel phioretin-inhibitable urea transporter (UT2) from rabbit, and water supply by synthesizing urea as a mechanism to detoxify subsequently from rat kidney. UT2 from both species has the character- ammonia. istics of the -sensitive urea transporter previously defined in Certain species also use urea as an important osmolyte [5—71. kidney by in Vitro perfused tubule studies. Based on these advances, Forexample, marine elasmobranch such as sharks, skates and rays Ripoche and colleagues cloned a homologous urea transporter (HUT11) from eiythrocytes. UT2 and HUTI 1 predict 43 kDa polypeptides andpossess large quantities of urea in the blood, body fluids and exhibit 64% amino acid sequence identity. Since regulation of ureatissues [5]. Their strategy for living in a marine environment is to transport in the kidney plays an important role in the orchestration of the maintain their body osmolality isosmotic with sea water, using antidiuretic response, we have studied the regulation of urea transporter in rat kidney at the mRNA level. On Northern blots probed at highurea as an osmolyte. stringency, rat UT2 hybridized to two transcripts of 2.9 kb and 4.0 kb, Given the vital roles urea plays in vertebrates, an important which have spatially distinct distributions within the kidney. Northern question arises as to how urea moves across biological mem- analysis and in situ hybridization of kidneys from rats maintained at different physiologic states revealed that the 2.9 and 4.0 kb transcripts are branes. Urea movement was originally thought to occur by simple regulated by separate mechanisms. The 4 kb transcript was primarily lipid permeation. Subsequent studies, however, revealed that responsive to changes in the dietary protein content, whereas the 2.9 kb several cell types including erythrocytes and epithelial cells of transcript was highly responsive to changes in the hydration state of the certain nephron segments have membrane permeabilities that are animal. We propose that the two UT2 transcripts are regulated by distinct mechanisms to allow optimal fluid balance and urea excretion. considerably higher than can be explained by simple lipid-phase diffusion, suggesting the existence of specialized urea transporters [8]. In the past few years, carrier mediated urea transport has been Urea is the major end product of nitrogen metabolism in mostextensively characterized in kidney inner medulla [9—11] and red mammals. It is synthesized in the liver by the urea cycle andblood cells [12]. In general, these transporters were found to be excreted in the urine (Fig. 1). Although the synthesis of urea frompassive and phloretin-inhibitable, and to exhibit low affinities for ammonia requires the high metabolic expenditure of four ATPurea (Km > 100 mM). The urea permeability of in vitro perfused molecules, urea formation is necessary to eradicate ammonia,kidney tubule segments has been studied by several investigators which is toxic to animals at relatively low concentrations, From an[reviewed in 11]. Moderate permeabilities were detected in thin evolutionary perspective, the development of urea cycle enzymesdescending limb of short loops of Henle [131, and in the inner and the ability of animals to synthesize urea from ammonia has medullaiy portion of long-loop descending and ascending thin become of particular importance with the movement of animalslimbs (Fig. 2) [14—17]. The highest permeability was detected in the terminal inner niedullary collecting duct (IMCD) [18—221. This permeability was found to increase to high values in response Note: In this paper UT2 refers to the urea transporter encoded by the rabbit 3.0 kb or the rat 2.9 kb mRNA, whereas UT refers to the urea to stimulation by [18—20, 22]. There was no saturation transporters encoded by the 2.9 and 4 kb mRNAs from rat kidney. of radiolabeled urea transport across terminal IMCD segments in Present address: Department of Physiological Sciences, University of zero trans experiments when the bath urea was varied between Manchester, School of Biological Sciences, Manchester M13 9PT, En- 0 and 800 mM [9]. However, when the urea concentration was gland, United Kingdom. varied in both lumen and bath, the lumen to bath urea flux 2Presentaddress: Cardiovascular Biology Laboratory, Harvard School of Public Health, 677 Huntington Avenue, Boston, MA 02115, USA. approached a limiting value at 400 to 500 m urea. Present address, Department of Pharmacology, Kyorin University Toon and Solomon measured the permeability of and School of Medicine, 6-20-2 Shinkawa, Mitaka, Tokyo 181, Japan. amides through the red blood and concluded that © 1996 by the International Society of Nephrology permeation is a saturable process with Km values of around 400

1615 1616 Hedigeret al: Mammalian urea transporters

Fig. 1. Urea cycle in the liver. Most of the urea excreted by the mammalian kidney is synthesized in the liver. Two ammonium ions are consumed by the urea cycle: One is consumed in the formation of carbamoyl phosphate (the carbamoyl moiety is then transferred to ornithine to form citrulline). The second is consumed in the formation of Urea Kidney aspartate from oxaloacetate (in this reaction, an Arginine cs-amino nitrogen from glutamate is transferred to oxaloacetate by transamination to form aspartate and a-ketoglutarate; this reaction can be considered to consume one ammonium ion NH44-e Urea although it does not directly utilize free ammonium). Citrulline and aspartate then form Liver OrnithinJ Colon argininosuccinate, and subsequent cleavage of argininosuccinate yields arginine and fumarate. Arginine is then hydrolyzed into urea and NR4+CO2 (NH4j ornithine. An equivalent of four ATP molecules are utilized during the formation of each urea 4itl/$1__• molecule. The Figure also highlights the urea Citrulline recycling pathway which exists in some species between the liver and the colon.

a Fig. 2. Role of urea transporters in the urinaty 5 concentrating mechanism, and possible sites of V Ot expression of urea transporters in the rat kidney. a, Abbreviations are: DVR, descending vasa recta; 2.9 kb AVR, ascending vasa recta; tDL, thin descending limb; TAL, thick ascending limb; o CD, collecting duct. The 2.9 kb signal is expressed in the inner stripe of the outer medulla (probably in thin descending limbs of short ioops of Henle). The 4.0 kb signal corresponds to the transporter in the IMCD. The Figure shows the thin descending limbs of short ioops of Henle to cross over the long ioops and to the peripheiy of the vascular bundles which include descending and 4kb UT2 ascending vasa recta. This morphological phenomenon was observed in some rodents such as rat and mouse and has implications for the urine concentrating ability.

mM [12]. Physiologically, it was thought that facilitated ureaconcentrate urine [25]. Based on their study, Jk(a—b—) patients transport in red cells minimizes volume perturbations when theyhad a maximal urine osmolality of 819 mOsm/kg H20, whereas circulate through the renal medulla. Studies of urea transport inthe maximal urine osmolality in control individuals was —1,000 human erythrocytes of the Kidd blood type Jk(a—b—) revealedmOsm/kg H20. Because of the slightly impaired urine concen- that these cells lack the facilitated urea transporter. Experimen-trating ability in Jk(a—b—) patients, it was proposed that the tally, it was demonstrated that erythrocytes from individuals whoerythrocyte and the kidney urea transporters are encoded by a lack the Kidd antigen exhibit an increased resistance to lysis tosingle that is deficient in patients with K.idd blood type. Our aqueous 2 m urea [23]. Froehlich et al also showed that theserecent insights into the molecular identity of urea transporters, erythrocytes have a specific defect in urea transport [24]. Al-however, revealed that there are different urea transporters in though patients with the Jk(a—b—) blood type have no obviouskidney and erythrocytes and that the Kidd blood group is associ- medical problems, recent studies by Sands and colleagues indi-ated with a defect of the urea transporter expressed in erythro- cated that these individuals have a minor defect in the ability tocytes. Hediger et al: Mammalian urea transporters 1617

Role of urea transport in urine concentration suitable for screening cDNA libraries. A powerful expression An important function of the mammalian kidney is to conservecloning technique for studying membrane proteins at the molec- water. Urea transport in the mammalian kidney medulla isular level has subsequently emerged which involves screening of eDNA libraries by functional expression in Xenopus laevis intimately involved in the urinary concentrating mechanism [re- viewed in 11]. Vasopressin increases the water permeability inoocytes. We have adapted this strategy to the cloning of mem- brane transport proteins, and the first transporter to be cloned by collecting ducts [26]resultingin a progressive increase in luminal urea concentration (Fig. 2). Vasopressin also increases the per-this method was the intestinal Na4-coupled glucose meability of the terminal part of the collecting duct to urea,SGLT1 [33, 34]. Further use of this technique in our laboratory resulting in movement of urea into the medullary interstitiumled to the cloning of several transporters cDNAs, each of which encoded the first member of a novel protein family. The trans- [19]. Increasing urea transport in the IMCD, in conjunction with Na-K-2C1 cotransport in the medullary thick ascending limbs [27], porters encoded by these cDNAs include the Na and K- enables the establishment and maintenance of the hypertonicdependent epithelial and neuronal high affinity glutamate trans- medulla which provides the osmotic gradient required for waterporter EAAC1 [35], the intestinal H-coupled oligopeptide reabsorption. This allows urine to be concentrated efficiently andtransporter PepTi [36, 37], and the renal dibasic and neutral urea to be disposed of in a small volume of water. amino acid transport protein D2, defects of which cause the Previous studies indicated that urea absorbed in the IMCD isformation of kidney stones in patients with cystinuria [38, 39]. To isolate a cDNA encoding a mammalian urea transporter we also secreted into thin descending and ascending limbs and into screened a rabbit kidney inner medulla cDNA library for the descending vasa recta in a process called urea recycling [11]. Urea recycling is thought to limit urea dissipation from the inneruptake of 14C-labeled urea [40] (a general description of the medulla, allowing the maintenance of the corticomedullary os-expression cloning technique is in [34]). Poly(A) RNA from motic gradient. This process is at least in part provided by urearabbit kidney inner medulla was first micro-injected into collage- transporters since both the IMCD and thin descending limbs ofnase-treated and manually defolliculated Xenopus oocytes. Three short loops of Henle have been shown to contain urea transport-days after injection, the uptake of '4C-labeled urea (1 mM) into oocytes was measured during 30 minutes. A fivefold increase in ers (see above). uptake of urea was observed compared to water-injected control Additional urea recycling pathways are provided by the unique RNAwas size-fraction- anatomical organization of the vascular bundles in the inner stripeoocytes. Rabbit kidney medulla poly(A) + of the outer medulla. In this region, the vessels heading for theated using a preparative agarose gel electrophoresis apparatus, and the size-fractions were analyzed for their ability to induce inner medulla (descending vasa recta) or returning from the innerurea uptake in oocytes. Fractions within the 3.0 to 4.4 kb medulla (ascending vasa recta) are incorporated into tightlysize-range showed peak stimulation of urea transport in oocytes packed vascular bundles [28]. Recent studies by Pallone et al using and were used to construct a directional cDNA library, using the dissected and perfused outer medullary vasa recta from vascular bundles revealed that the endothelia of descending vasa rectaexpression vector pSPORT1 (Gibco BRL). Plasmid DNA was then extracted from pools of 300 clones by a mini-prep technique contain a phloretin-inhibitable urea transporter [29, 30]. It was and in vitro transcribed using T7 RNA polymerase. The cRNA proposed that this transporter serves to prevent the escape of urea was micro-injected into oocyte and the uptake of '4C-labeled urea from the renal medulla by efficiently trapping it in the vascular counter current system and by delivering urea back to the deeperwas measured three days after injection. Analysis of —20 poois led portion of the medulla. Since the concentration of urea returningto the identification of one positive pool, and its subsequent in ascending vasa recta from the inner medulla approaches thedivision into progressively smaller subpools, followed by func- concentration in the descending vasa recta, the close associationtional analysis in Xenopus oocyte, resulted in the isolation of a 3,1 kb cDNA which encodes the rabbit urea transporter UT2 [40]. of descending and ascending vasa recta in vascular bundles allows The cDNA contains an open reading frame from nucleotides efficient countercurrent exchange of urea between the two struc- 1025 to 2215 that predicts a 397 residue protein (UT2; Fig. 3A). tures. Such an exchange would involve exit of urea from ascending vasa recta via the fenestrated endothelial structure of theseA search of the translated sequence databases revealed that UT2 vessels, and carrier mediated entry of urea into descending vasais not homologous to any known sequence. Hydropathy analysis of UT2 revealed an unusual pattern of hydrophobicity different from recta. In some species, descending thin limbs of short loops of Henlethat found in previously studied transporters (Fig. 3B). Overall, cross over the long loops of Henle and are either attached to theUT2 is extremely hydrophobic and, in contrast to the character- istic pattern observed in most plasma membrane transport pro- periphery of the bundles (rat) or become incorporated into the teins which includes several distinct hydrophobic transmembrane vessels of the bundles (mouse) (Fig. 2) [31, 32]. These species- specific anatomical arrangements of the bundles in the innersegments interspersed with hydrophilic regions, UT2 has two stripe of the outer medulla likely provide efficient recycling andlarge extended hydrophobic domains (Fig. 3B). This suggests that a major portion of the protein may be entirely embedded in the trapping pathways for urea, and further prevent its escape frommembrane, as indicated in Figure 2A. the renal medulla. High stringency Northern analysis of rabbit tissues was used to study the tissue distribution of UT2 (Fig. 4A) [40]. UT2 mRNA is Expression cloning and characterization of the urea strongly expressed in kidney and colon. Strong bands at 4.0 kb transporter UT2 from rabbit kidney were observed in samples from rabbit renal papillary tip, renal In the past decade, the analysis of mammalian transporters atmedulla and colon. A band at 3.0 kb was present in the outer the molecular level has been initially limited by the ability tomedulla of the kidney. purify membrane proteins and to prepare probes and antibodies Urea transport in colon may be involved in nitrogen salvaging. 1618 Hediger et al: Mammalian urea transporters

A B PKA Ser3l Ser 386 397 PKC COOH Thr 13 NH2 B +3 123 45 67 8 9 10

a) Fig. 3. A. Hypothetical representation of the 0 topology of rabbit UT2inthe membrane. O00 Putative PKA and PKC phospholylation sites are indicated. Asn 210 is a putative N- glycosylation site. Potential membrane spanning A B C regions are numbered 1 to 10. B. Hydropathy analysis of the UT2 amino acid sequence using —3 a window of 21. A and C are extended hydrophobic regions. B is a more hydrophilic 0 100 200 300 400 region predicted to be extracellular.

Studies by Jackson and coworkers indicated that, in some species,ration of transport with increasing substrate concentration. The urea formed in the liver is disposed of not only by excretion in themaximal urea concentration that can be used in experiments with urine, but also by secretion into the colon where it is hydrolyzedXenopus oocytes is —200 mM. However, as discussed above, urea by gut microflora (Fig. 1) [411. Ammonia formed in this way maytransporters in kidney and red blood cells have Km values be salvaged and transported to the liver via the portal vein wheresignificantly above 200 m. Consistent with this, UT2-mediated it is made available for further metabolic interaction. This recy-transport did not exhibit saturation in the concentration range cling of nitrogen between the liver and the gut may be ofbetween 1 and 200 m urea. particular importance in herbivores, which have a relatively low Urea transport in the kidney is also characterized by inhibition protein intake and therefore a limited nitrogen supply. Interest-with phloretin and urea analogues. With phloretin concentrations ingly, in rat, UT2 mRNA does not appear to be expressed in colonof 0.35 and 0.7 m rabbit UT2-mediated uptake of urea (1 mM) [42]. Whether dietaiy protein restriction induces expression ofwas inhibited by 48 and 78%, respectively [40]. Studies addressing UT2 in rat colon remains to be elucidated. the inhibition of UT2 mediated urea uptake (1 mM) by urea Low stringency Northern analysis was used to test for theanalogues (150 mM) revealed that the pattern of inhibition existence of UT2-related sequences in rabbit (Fig. 4B). Under lowresembles that reported for the apical vasopressin-sensitive urea stringency condition, the rabbit UT2 probe hybridized to distincttransport in in vitro perfused rat kidney IMCDs [40]. 4.0 kb bands in liver and lung [37]. A urea transporter in the Also consistent with the predicted distribution of the vasopres- sinusoidal plasma membrane of hepatocytes would allow exit ofsin-regulated urea transporter are our data from in situ hybrid- urea that is produced by the urea cycle. ization (Fig. 6A) [40]. Using 35S-labeled UT2 antisense cRNA, a UT2 expressed in Xenopus oocytes mediated passive movementstrong signal was observed in epithelial cells lining the IMCD. of urea across the oocyte plasma membrane and increased its ureaThis signal corresponds to the 4.0 kb message detected on the permeability 23-fold (Fig. 5). The calculated urea permeability ofNorthern blot in lanes corresponding to inner medulla and UT2-injected oocytes based on initial rates of uptake was 4.5 xpapillary tip (Fig. 4A). There is also a striking pattern of hybrid- i0 cm/sec [40]. This is in agreement with the published value forization in the inner stripe of the outer medulla (Fig. 6A) that in vitro perfused terminal IMCD segments which is between 11.6corresponds to the 3.0 kb UT2 transcript in the outer medulla as and 13.1 x i0 cm/sec [21]. In contrast, the urea permeability ofshown by Northern analysis (Fig. 4A). The 3.0 kb transcript water-injected control oocytes is 1.9 X 10 cm/sec, a value whichappears to encode a urea transporter which is located in thin is characteristic of lipid-phase permeation of urea. descending limbs of short loops of Henle. UT2 in this structure is A characteristic feature of carrier-mediated transport is satu-most likely involved in urea recyling. Hediger et al: Mammalian urea transporters 1619

ci) 0 - Ci, 0 E 0 ci) (C) = A : E C .0.0 C ci) a ci) 0 E U) ci) U) S o 2 o E — 5 (C) CD a 0 U) U) - ci) ,0. ci) ID-,=0 co 0 5 ocO I coi -J0 C.) 0

4 kb 3 kb I B Fig. 4. Tissue distribution of rabbit UT based on low and high-stringency Northern analysis. A. High stringency Northern analysis. The filter fr,. was hybridized at 42°C in 50% formaldehyde 4 kb and washed at 42°C in 0.1 XSSC-0.1%SDS. Each lane contains 3 g poly(A) RNA. B. 3 kb Low stringency Northern analysis. The filter was hybridized at 35°C in 50% formaldehyde and washed at 42°C in 0.1 X SSC-0.1% SDS. 32P-labeled full-length UT2 cDNA was used for hybridization.

Urea transport across the apical membrane of terminal IMCDscreening a eDNA library from rat kidney medulla with the cells is thought to be the rate-limiting step and the control site forfull-length rabbit UT2-cDNA probe [42]. A 2,9 kb cDNA was regulation of trans-epithelial urea permeability by vasopressin.isolated that has an open reading frame from nucleotides 928 to Studies by Star et al [22]usingisolated perfused rat terminal2118 and encodes a 397 residue protein (rat UT2). The rat UT2 IMCD segments suggested that stimulation of urea transport byamino acid sequence is 88.1% identical to that of rabbit UT2. vasopressin results from the occupation of V2-type vasopressinExpression studies of rat UT2 in Xenopus oocytes showed that it receptors with subsequent activation of adenylate cyclase, andis a phloretin-sensitive urea transporter. that the apical urea permeability increases predominantly as a High stringency Northern analysis of rat kidney mRNA re- result of an increase in the number of functional transporters invealed that rat UT is encoded by two transcripts, a 2.9 and a the apical membrane. Activation is thought to involve recruitment4.0 kb transcript (Fig. 7), analogous to the 3.0 and 4.0 kb of urea transporter molecules from intracellular vesicle pools, astranscripts found in rabbit [42]. Based on in situ hybridization, is the case for the vasopressin-regulated water channel AQP2 [27], using rat UT2 antisense cRNA as a probe (Fig. 6B), in combina- and/or activation of urea transporter molecules by direct phos- phorylation via protein kinase A. UT2 has two putative proteintion with Northern analysis, these two transcripts exhibit spatially kinase A phosphorylation sites (Fig. 3). However, the cAMPdistinct distributions within the kidney [42], again in complete agreement with the findings in rabbit. Figure 6B shows a promi- analogues Sp-cAMP-S, dibutyryl-cAMP and 8-bromo-cAMP (30 mm preincubation, 1 m analogues) were found to have nonent signal in epithelial cells lining the IMCD. Northern analysis significant effect on UT2-mediated urea transport in oocytes [40].(Fig. 7) revealed that this signal corresponds to the 4.0 kb Thus, further studies involving the subcellular localization of thetranscript, its localization is consistent with the predicted distri- urea transporter in IMCD cells and analysis of the molecularbution of the vasopressin-regulated urea transporter. An in situ nature of the 3.0 and 4.0 kb transcripts will be necessary tohybridization signal is also present in the lower part of the inner determine the mechanisms by which vosopressin activates ureastripe of the outer medulla (Fig. 6B). This signal corresponds to transport in the IMCD. the 2.9 kb transcript detected by Northern analysis in RNA from the inner stripe of the outer medulla (Fig. 7). As alluded to above, Cloning and characterization of UT2 from rat kidney UT2 in this region may participate in urea recycling. Studies are UT2 from rat kidney was cloned to further study the regulationnow in progress to determine the exact location of UT2 in the of expression of this transporter. A rat UT2 cDNA was isolated bykidney medulla. 1620 Hedigeret al: Mammalian urea transporters

300 expression of a passive vasopressin-inducable urea transporter in the initial inner medullary collecting duct [44]. Interestingly, these investigators also observed that feeding rats a low protein diet for 1250 x 10 cm-1 three weeks induces a second, Nat-coupled urea transporter that 0 is not sensitive to phloretin or vasopressin [45]. Determination of 0 the molecular basis of this second, "active" urea transporter, 200 however, requires further studies. To study the effect of dehydration on the expression of UT mRNA levels, male Sprague-Dawley rats had limited access to 150 water (10 mI/day) for three days [42]. Based on Northern analysis and in situ hybridization, the 2.9 kb transcript was highly respon- sive to changes in the hydration state. Dehydration of the animals 100 strongly increased the expression of this transcript (Fig. 8, right). Since the 2.9 kb transcript appears to be expressed in thin descending limbs of short ioops of Henle, this response may be 1.9 x10cms beneficial during dehydration because it limits dissipation of urea from the inner medulla by stimulating urea recycling. Up-regula- 0 tion of the 2.9 kb transcript may therefore enhance the cortico- 0 20 40 60 80 100 120 papillary osmolality gradient and osmotic water reabsorption. Dehydration also appears to induce a spreading of the 2.9 kb Incubation time signal towards the papilla (Fig. 8, right). Whether the spreading Fig. 5. Expression of rabbit UT2 in Xenopus oocytes. The Figure shows the involves expression of the 2.9 kb transcript in inner medullary time course of the uptake of 14C-labeled urea (1 mM) into UT2-cRNA- portions of long-loop descending and/or ascending thin limbs injected oocytes and water-injected control oocytes. remains to be determined. Taken together, our studies on the long-term regulation of UT reveal that the expression levels of the 2.9 and 4.0 kb transcripts Differential regulation of the 2.9 and 4.0 kb transcripts in are regulated by separate mechanisms. Given the predicted roles rat kidney and localization of the two transcripts in kidney it seems plausible Using Northern analysis and in situ hybridization, we havethat distinct regulatory pathways of these transcripts are necessary investigated the expression of UT mRNA in rat kidney into allow optimal fluid balance in response to different physiologic response to various physiologic conditions [42]. The studiesstates. Whether the transcripts are controlled by separate promo- revealed that the 2.9 and 4.0 kb transcripts are regulated bytors remains to be elucidated. separate mechanisms (Fig. 8). The reponse of UT mRNA expression to dietary protein Differences in the distribution of expression of UT mRNA in restriction was determined in Sprague-Dawley rats that were rabbit and rat kidney given for four weeks either a normal diet containing 18% protein A comparison of the UT in situ hybridization signals in rabbit (NIH-31 diet) or a low protein diet containing 8% proteinand rat is shown in Figure 6. The different patterns of the UT (NIH-31LP diet) [421. The low protein diet was the same as thatsignals in the inner medulla of these species are in agreement with used by Isozaki et al [43, 44] and has been shown to cause aprevious observations by Knepper et al [46]. These investigators significant decrease in blood urea nitrogen and in the urinary ureademonstrated that in the inner medulla, the collecting ducts fuse excretion [43]. The experiments revealed that the 4.0 kb transcriptwith a different frequency in rabbits compared to rats. In rabbits, which is expressed in the inner medullary collecting ducts isthe number of collecting ducts per unit cross-sectional surface strongly up-regulated in response to protein dietary restrictionarea was found to decrease continuously from the outer-inner while the 2.9 kb transcript is much less affected (Fig. 8, left) [421.medullaiy border to the papillary tip. By contrast, in rat, this This finding is consistent with the observation by Isozaki et al thatnumber was found to be constant throughout most of the inner dietary protein restriction causes a decrease in the fraction ofmedulla and to diminish only close to the papillary tip. The filtered urea excreted and an associated increase of phloretin-different branching patterns in rabbit and rat are entirely consis- inhibitable urea transport in the initial inner medullary collectingtent with the UT in situ hybridization staining patterns obtained ducts [43]. Up-regulation of UT2 mRNA in collecting ducts mayfor the inner medulla of these two species (Fig. 6). help to preserve blood urea so that urine can be concentrated Further differences in the distribution of the two UT2 tran- maximally when the need arises. scripts in rabbit and rat kidney may parallel adaptation in Previous studies revealed that in rats fed a normal diet theresponse to different amounts of dietary protein consumed by initial inner medullary collecting duct has a very low urea perme-these species, given that the herbivorous rabbits have a lower ability that cannot be stimulated by vasopressin [19]. Our in situprotein intake than the omnivorous rats. The strong expression of hybridization data indicate that feeding a low protein diet inducesUT mRNA in rabbit medulla (Fig. 6A) appears to resemble that a spreading of the 4.0 kb hybridization signal to the upper third ofobserved in the kidney of rats fed a low protein diet (Fig. 8A). the inner medullary collecting duct (Fig. 8, left) [42]. The spread- ing of the UT signal induced by protein restriction is consistent The urea transporter from erythrocytes is the Kidd antigen with recent studies of Sands and colleagues which showed that Ripoche and colleagues recently isolated a eDNA clone from a feeding rats a low protein (8%) diet for two weeks triggers thebone marrow eDNA library which corresponds to the red blood Hediger et a!: Mammalian urea transporters 1621

• b •

a ••#aaI/lap Mt 'i?..''' Vd,'-;•'2 / — ,•z"•• 2*ZT

•

Fig.6. In situ hybridization of kidney from rabbit (A) and rat (B). Cryosections (7 .rm) of paraformaldehyde-fixed kidney were probed with 35S-labeled antisense UT2 cRNA synthesized from the full-length rabbit or rat UT2 cDNAs. The cRNA probes were hydrolyzed for 50 minutes to generate probes of --100 nucleotides. Hybridization signals are evident in the terminal part of the TMCD and in the inner stripe of the outer medulla in bothrabbit and rat.

kidney [471.However,its exact tissue distribution within the kidney has not yet been determined. Since the erythrocyte urea transporter does not appear to be under hormonal control, HUT1I in the kidney may represent a constitutive transporter.

Summary and future perspectives kb Our data show that UT2 is a phloretin-inhibitable passive urea transporter that is strongly expressed in the kidney medulla. UT2 is the first member of a growing family of novel transporters that are unusually hydrophobic. HUTI 1 was recently identified as the erythrocyte homologue of UT2 and there is 64% amino acid sequence identity between the two isoforms. It appears likely that this transporter family includes additional transporters such as the basolateral urea transporter from kidney IMCD and putative UT2 d isoforms from liver, lung and possibly brain. Our data strongly reveal that UT plays an important role in the urinary concentrating mechanism and in the regulation of nitro- gen balance. In rat, UT mRNA is extremely responsive to changes Fig. 7. Renal localization of rat UT2 by high stringency Northern analysis of in the physiological state of the animal and the pattern of the dissected kidney. Full length 32P-labeled UT2 eDNA was used as a probe. responses are specific to the functional state of the kidney. The 2.9 Filters were hybridized at 42°C and final washing was in 0.1 XSSC-0.1% SDS at 65°C. kb and 4.0 kb transcripts appear to be independently regulated and their message levels are most likely controlled by multiple factors, possibly by multiple hormones and/or due to the presence of receptors or sensors expressed in different cell types of the cell urea transporter (HUT11) [47]. At the amino acid sequencekidney. Studies to characterize the 4.0 kb UT transcript in the level, HUTI 1 is 64% identical to rat UT2. When expressed inkidney are now in progress. In addition to the regulation at the Xenopus oocytes, its inhibition pattern by phloretin, the sulfhydrylmessage level, UT expression in the terminal IMCD is expected to specific agent pCMBS, and urea analogues was as described forbe highly responsive to vasopressin. Studies using immunocyto- the mammalian erythrocyte urea transporter. In oocytes express-chemistry are currently in progress to determine whether activa- ing HUT11 the urea permability was increased from 1.2 to 2.0 Xtion of UT by vasopressin involves recruitment of transporter 10'cm/secondto 2.4 xiOcm/second. This gives approxi-molecules from intracellular vesicle pools, in analogy to the water mately a 20-fold increase in the urea permeability. The investiga-channel AQP-2 [26]. Further localization of rat UT and HUT1 1 in tors also demonstrated that HUTiland UT2 do not affect thekidney inner and outer medulla, combined with the identification permeability to water. Recent studies by Olives et al revelaed thatand localization of other renal urea transporters such as the HUT11 corresponds to the Kidd blood group and that JK(a—b—)transporter in basolateral membranes of terminal IMCD cells, will red cells have a defect in this urea transporter [48]. lead to a better understanding of the specific roles of urea Based on Northern analysis, HUTI1 is also expressed in thetransporters in the renal urea concentrating mechanism. 1622 Hedigeret at: Mammalian urea transporters

Protein deprivation Water restriction Fig. 8. Alteration of rat UT mRNA expression in response to changes of the dietary protein content and the hydration state. Left. In situ hybridizationof kidneys from rats fed a low protein diet containing 8% protein (NIH-31 diet) for four weeks. Right. In situ hybridization of kidneys from dehydrated rats (rats had free access to food but water was restricted to 10 mI/day for three days). The Figure shows that protein deprivation strongly increasesthe expression of the 4 kb transcript in the IMCD whereas water restriction increases the expression of the 2.9 kb transcript in the inner stripe ofthe outer medulla and in the inner medulla (see Fig. 2).

Acknowledgments 13. IMAI M, HAYASHI M, ARAKI M: Functional heterogeneity of the descending limbs of Henle's loop. I. Jnternephron heterogeneity of It is a pleasure to acknowledge Drs. Mark Knepper, Jeff Sands, Pierre the hamster kidney. Pftugers Arch 402:385—392, 1984 Ripoche and Jean-Pierre Cartron for their contribution to this work. We 14. MARSH Di: Solute and water flows in thin limbs of Henle's loop in the also thank Dr. Lise Bankir for stimulating discussion. This work was hamster kidney. Am J Physiol 218:824—83 1, 1970 supported by the National Institutes of Diabetes and Digestive and Kidney 15. 1MM M, Koioo J: Sodium chloride, urea and water transport in the Disease Grants DK43632, 43171 and 46289 to M.A. Hediger. thin ascending limb of Henle. J Clin Invest 53:393—402, 1974 16. CHOU CL, KNEPPEP. MA: In vitro perfusion of chinchilla thin limb Reprint requests to Dr. Matthias A. Hediger, Renal Division, Brigham and segments: Urea and NaCI permeabilities. Am J Physiol 264:F337— Women's Hospital, 75 Francis Street, Boston, Massachusetts 02115, USA. F343, 1993 17. IMAI M: Function of the thin ascending limb of Henle of rats and References hamsters perfused in vitro. Am J Physiol 232:F201—F209, 1977 18. MORGAN T, BERLINER RW: Permeability of the loop of Henle, vasa 1. COHEN PP: Evolutionary and comparative aspects of urea biosynthe- recta, and collecting duct to water, urea, and sodium. Am J Physiol sis, in The Urea Cycle, edited by GRISOLIA 5, ET AL, New York, Wiley, 215:108—115, 1968 1976, pp 21—38 19. SANDS JM, NONOGUCHI H, KNEPPER MA: Vasopressin effects on urea 2. SMITH 11W: The excretion of ammonia and urea by the gills of fish. J and H20 transport in inner medullary collecting duct subsegments. Biol Chem 8 1:727—742 1929 Am J Physiol 253:F823—F832, 1987 3. CAMERON JN: Responses to reversed NH3 and NH4 gradients in a 20. ROCHA AS, KUDO LH: Water, urea, sodium, chloride and potassium teleost (Ictalurus punctatus), an elasmobranch (Raja erinacea), and a transport in the in vitro isolated perfused papillary collecting duct. crustacean (Callinectes sapidus): Evidence for NH4/H exchange in Kidney lot 22:485—491, 1982 the teleost and elasmobranch. J Exp Zool 239:183—195, 1986 21. SANDS JM, KNEPI'ER MA: Urea permeability of mammalian inner 4. MOMMSEN T, WALSH PJ: Evolution of urea synthesis in vertebrates: medullary and papillary surface epithelium. J The piscine connection. Science 243:72—75, 1989 Clin Invest 79:138—147, 1987 5. SMITH HW: The retention and physiological role of urea in the 22. STAR RA, NONOGUCHI H, BALABAN R, KNEPPER MA: Calcium and elasmobranchii. Biol Rev 11:49—82, 1936 cyclic adenosine monophosphate as second messengers for vasopres- 6. BROWN OW JR, BROWN SG: Urea and its formation in coelacanth sin in the rat inner medullary collecting duct. J Clin Invest 81:1879— liver. Science 155:570—573, 1967 1888, 1988 7. Piciu'oai GE, GRANT FB: Serum osmolality in the coelacanth,23. HEATON DC, MCLOUGHLIN K: Jk(a-b-) red blood cells resist urea Latimeria chalumnae: Urea retention and ion regulation. Science lysis. Transfusion 22:70—71, 1982 155:568—570, 1967 24. FROEHLICH 0, MACEY RI, EDWARDS-MOULDS J, GARGUS ii, GUNN 8. MARSH DJ, KNEPPER MA: Renal handling of urea, in Handbook of RB: Urea transport deficiency in Jk(a-b-) erythrocytes. Am J Physiol Physiology section 8, Renal Physiology (vol 2), edited by WINDHAGER 260:C778—C783, 1991 EE, Oxford, Oxford University Press, 1992, pp 1317—1347 25. SANDS JM, GARAGUS JJ, FROEHLICH 0, GUNN RB, Koiuo JP: 9. CHOU CL, SANDS JM, NONOGUCHI H, KNEPPER MA: Concentration Urinary concentrating ability in patients with JK(a-b-) blood type who dependence of urea and thiourea transport in rat inner medullary lack carrier-mediated urea transport. JAm Soc Nephrol 2:1689—1696, collecting duct. Am J Physiol 258:F486—F494, 1990 1992 10. CHOU CL, KNEPPER MA: Inhibition of urea transport in inner 26. FUSHIMI K, UCHIDA 5, HARA Y, HIRATA Y, MARUMO F, SASAKI 5: medullary collecting duct by phloretin and urea analogues. Am J Cloning and expression of apical membrane water channel of rat Physiol 257:F359—F365, 1989 kidney collecting tubule. Science 361:549—552, 1993 11. KNEPPER MA, RECTOR FC JR: Urine concentration and dilution, in 27. GAMBA G, MIYANOSHITA A, LOMBARDI M, LYTTON 3, LEE W-S, The Kidney (5th ed), edited by BRENNER BM, Philadelphia, W.B. HEDIGER MA, HEBERT SC: Molecular characterization of the mam- Saunders Company, 1995, pp 532—570 malian renal electroneutral sodium-potassium-chloride cotransporter 12. TooN MR, SOLOMON AK: Transport parameters in the human red cell family. J Biol Chem 269:17713—17722, 1994 membrane: Solute-membrane interactions of amides and ureas. Bio- 28. BANKIR L, DE ROUFFIGNAC C: Urinary concentrating ability: Insights chim BiophysActa 1063:179—190, 1991 from comparative anatomy. Am J Physiol 249:R643—R666, 1985 Hediger et a!: Mammalian urea transporters 1623

29. PALLONETL,WORK J, MAYERS RL, JAMISON RL: Transport of Expression cloning and characterization of the vasopressin-regulated sodium and ureainoutermedullary descendingvasa recta.JClin urea transporter. Nature365:844—847,1993 Invest 93:212—222,1994 41. LM4GRAN M, MoR.&r BJ, MURPHY JL, JACKSON AA: Adaptation to a 30. PALLONE TL: Characterization of the urea transporter in outer diet low in protein: Effect of complex carbohydrate upon urea kinetics medullaiy descending vasa recta. AmJ Physiol 267:R267—R269, 1994 in normal man. ClinSd 82:191—198, 1992 31. KRIZ W: Structural organization of the renal medulla: Comparative 42. SMITH CP, LEE W-S, MARTIAL 5, KNEPPER MA, You G, SANDS JM, and functional aspects. AmJ Physiol 241:R3—R16, 1981 HEDIGER MA: Cloning and regulation of expression of the rat kidney 32. KRIz W: Structural organization of the renal medullary counter urea transporter (rUT2). JClin Invest (inpress) current flow system. FedProc 42:2379—2385,1983 43. IsozI T, GILLAN AG, SWANSON CF, SANDS JM: Protein restriction 33. HEDIGER MA, COADYMJ,IKEDA TS, WRIGHT EM: Expression sequentially induces new urea transport processes in rat initial IMCD. cloning and eDNA sequencing of the Na7glucose cotransporter. AmJ Physiol 266:F756—F761, 1994 Nature 1987 330:379—381, 44.Isoz.iu T, VERLANDER, SANDS JM: Low protein diet alters urea 34.HEDIGER MA, RHOADSD: Molecular physiology of Na/gIucose transport and cell structure in rat intital inner medullary collecting Rev 74:993—1026,1994 . Physiol duct. JClin Invest 92:2448—2457,1993 35.KANAL Y,HEDIGERMA:Primary structure and functional character- 45.IsozAxiT, LEA JP, TUMLIN JA, SANDS JM: Sodium-dependent net ization of a high-affinity . Nature360:467—471, Clin 1992 urea transport in rat initial inner medullary collecting ducts. J 1994 36.FEI Y-J, KANAIYK,NUSSBERGER 5, GANAPATHY V, LEIBACH FH, Invest 94:1513—1517, 46. KNEPPER MA, DANIELSON RA, SAIDEL GM, POST RS: Quantitative ROMERO MF, SINGHSK,BORON WF,HEDIGERMA:Expression analysis of renal medullary anatomy in rats and rabbits. Kidneymt cloningofamammalian proton-coupled oligopeptide transporter. Nature368:563—566, 1994 12:313—323,1977 37. NUSSBERGERS,HEDIGERMA:How peptides cross biological mem- 47. OLIVES B, NEAU P, BAILLY P, HEDIGER MA, ROUSSELET G, CARTRON branes. EpNephro! 3:211—218, 1995 J-P,RIPOCHEP: Cloning and functional expression of a urea trans- 38.WELLS RG, HEDIGERMA:Cloning of a rat kidney eDNA that porter from human bone marrow cells. fBio!Chem 269:31649—31652, stimulates dibasic and neutral amino acid transport and has sequence 1994 similarity to glucosidases. ProcNatlAcad Sci USA 89:5596—5600, 1992 48. OLIvEs B, MATFEI M-G, HUET M, NEAU P, MARATIAL S, CARTRON 39.HEDIGER MA: Structure, function and evolution of solute transporters JP,BAILLYF: Kidd blood group and urea transport function of human in prokaryotes and eukaryotes. JExp Biol 196:15—49, 1994 erythrocytes are carried by the same protein. I BiolChem 270:15607— 40. You G, SMITH CP, KANAI Y, LEE W-S, STELZNERM,HEDIGER MA: 15610, 1995