AJP - Renal Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Physiol Renal Physiol 283: F150-F163, 2002. First published February 20, 2002; doi:10.1152/ajprenal.00333.2001
0363-6127/02 $5.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
283/1/F150    most recent
00333.2001v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (10)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Leal-Pinto, E.
Right arrow Articles by Abramson, R. G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Leal-Pinto, E.
Right arrow Articles by Abramson, R. G.
Vol. 283, Issue 1, F150-F163, July 2002

Functional analysis and molecular model of the human urate transporter/channel, hUAT

Edgar Leal-Pinto1, B. Eleazar Cohen2, Michael S. Lipkowitz1, and Ruth G. Abramson1

1 Division of Nephrology, Department of Medicine, Mount Sinai School of Medicine, New York, New York, 10029; and 2 Division of Extramural Activities, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland 20892


    ABSTRACT
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

Recombinant protein, designated hUAT, the human homologue of the rat urate transporter/channel (UAT), functions as a highly selective urate channel in lipid bilayers. Functional analysis indicates that hUAT activity, like UAT, is selectively blocked by oxonate from its cytosolic side, whereas pyrazinoate and adenosine selectively block from the channel's extracellular face. Importantly, hUAT is a galectin, a protein with two beta -galactoside binding domains that bind lactose. Lactose significantly increased hUAT open probability but only when added to the channel's extracellular side. This effect on open probability was mimicked by glucose, but not ribose, suggesting a role for extracellular glucose in regulating hUAT channel activity. These functional observations support a four-transmembrane-domain structural model of hUAT, as previously predicted from the primary structure of UAT. hUAT and UAT, however, are not functionally identical: hUAT has a significantly lower single-channel conductance and open probability is voltage independent. These differences suggest that evolutionary changes in specific amino acids in these highly homologous proteins are functionally relevant in defining these biophysical properties.

pyrazinoate; lactose; oxonate; glucose; adenosine; glycophorin A


    INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

HUMANS, AS WELL AS BIRDS, reptiles, and some nonhuman primates lack functional uricase and, as a consequence, urate is the end product of the intracellular degradation of the purines adenine and guanine (3). Subsequent to its metabolic production by the enzyme xanthine oxidase (5, 40, 78), urate effluxes from cells by an unknown mechanism to enter the extracellular compartment. In the absence of degradation of urate to allantoin by hepatic uricase in these species, maintenance of urate homeostasis is entirely dependent on elimination of urate from the body by the kidneys (3) and, to a much lesser extent, the gastrointestinal tract (70, 71). Considering the very limited solubility of urate (80) within cells, plasma, and urine, it is apparent that the avoidance of urate crystallization in any of these compartments in humans is even more critically dependent on the efflux and excretory transporters than in species that have hepatic uricase to metabolize urate to the water-soluble compound allantoin (12, 25, 50). Although there are limited data on the mechanism(s) responsible for elimination of urate by the intestinal epithelium (70, 71), the handling (filtration, reabsorption, and secretion) and mechanisms of urate transport have been extensively evaluated in the kidney (3). Renal transport has been ascribed to both an electroneutral urate-anion exchanger (9, 20, 21, 28-30, 61) and an electrogenic urate uniporter (1, 2, 33, 61) in a number of species, including humans (61).

We recently cloned a cDNA from a rat renal expression library that encodes a 322-amino acid protein, prepared recombinant protein from the cDNA, and demonstrated that this protein functions as a highly selective, voltage-sensitive 10-pS urate transporter/channel in planar lipid bilayers (38). This protein, designated UAT, displays a number of characteristics (36) that suggest that this channel is the transporter responsible for urate efflux from systemic cells as well as the molecular representation of the rat renal electrogenic urate transporter. Moreover, recent studies in which UAT has been expressed as a chimeric protein documented that UAT is an integral plasma membrane protein with intracellular termini in a variety of renal and nonrenal epithelial cells derived from a number of species (59). However, it is noteworthy that UAT belongs to a family of proteins, the galectins, that have all been presumed to be soluble, cytoplasmic, or secreted proteins (4, 7, 10, 15-17, 22, 26, 48, 56). Equally importantly, galectins have been assigned multiple functions (4, 7, 10, 15-17, 22, 26, 48, 56), but none have ever been proposed to serve a transport function. The demonstration that UAT functions as a channel in synthetic lipid bilayers (36, 38) and resides as an integral plasma membrane protein in living cells (59) thus represents both a unique function and previously undescribed subcellular localization for a galectin.

Subsequent to our publication on the cloning of UAT (38), galectin 9 was reported in rats (76, 77), mice (76, 77) and humans (51, 52, 75). It is of note that the cDNAs for rat, mouse and human galectin 9 are, respectively, 99, 89, and 73% identical to UAT, and the translated proteins, like other members of the galectin family, are considered to be soluble, cytoplasmic, or secreted proteins (51, 75-77). Although a functional role has not been assigned to rat galectin 9 (76, 77), mouse galectin 9 has been proposed to serve a role in thymocyte-epithelial interactions (76, 77), whereas human galectin 9 is believed to participate in cellular interactions of the immune system (75) and in eosinophil chemoattraction (51). In view of the high degree of homology between UAT (accession no. U67958) and human galectin 9/ecalectin (accession nos. Z49107, AB006782, and AB005894), we recently generated galectin 9 cDNA by RT-PCR from RNA of human white blood cells, prepared recombinant protein, and performed studies to evaluate the possibility that the apparent human homologue of UAT might serve a transport function (44). These studies demonstrated that recombinant human galectin 9, a 323-amino acid protein that is identical to accession no. AB006782 (minus the 32-amino acid insertion specific to the intestinal isoform of galectin 9), both functions as a highly selective urate channel in synthetic lipid bilayers and represents an integral plasma membrane protein with cytoplasmic NH2 and COOH termini in epithelium-derived cells (44). These observations led us to propose that the human homologue of UAT is also likely to represent the urate channel in plasma membranes of systemic cells and the electrogenic renal urate transporter in humans (44).

The present studies were conducted to evaluate the functional characteristics of the human urate transporter/channel, designated hUAT, to model its transmembrane organization and to assess the potential role in channel function of the two beta -galactoside binding sites within hUAT, the signature amino acid sequence of a galectin. These studies demonstrate that hUAT channel activity displays a number of characteristics that suggest that the topologies of hUAT and UAT are quite similar. In contrast, single-channel conductance and voltage sensitivity of open probability of hUAT differ significantly from that of UAT, presumably as a consequence of some evolutionary divergence in critical amino acids in the respective sequences. Furthermore, these studies provide evidence to suggest that binding of the beta -galactoside alpha -lactose to hUAT is not simply a confirmation that this protein contains the signature sequences for galectins but rather that such binding significantly influences hUAT channel activity. Finally, evidence is provided that glucose, the physiologically more relevant sugar, similarly has a significant modulating effect on hUAT channel activity.


    MATERIALS AND METHODS
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

Preparation of Recombinant Protein

As previously reported, recombinant protein was made from cDNA prepared by RT-PCR of RNA that was harvested from human white blood cells (44). In brief, the full length of the coding sequence of hUAT in pBluescript was amplified by PCR using a sense primer with an XhoI site immediately 5' to the start codon (5'-GCCTCGAGATGGCCTTCAGCGGTTCCCAG-3') and an antisense primer with a HindIII site just 3' to the stop codon (5'-GCAAGCTTCTATGTCTGCACATGGGTCGC-3'). The purified PCR product was subcloned into XhoI- and HindIII-digested pRSETA (Invitrogen, San Diego, CA) for subsequent production of a fusion protein with a six-histidine metal-chelating domain 5' to the coding region of UAT. pRSETA-hUAT was isolated (Qiagen Plasmid Maxi kit, Qiagen, Chatsworth, CA) and used to transform BL21(DE3)pLysE cells (Novagen, Madison, WI). Colonies of BL21(DE3)pLysE cells containing pRSETA-hUAT were grown until the optical density reached 0.6-0.7. Thereafter, isopropyl-1-thio-beta -D-galactopyranoside was added to a final concentration of 0.4 mM, and the culture was grown for an additional 4 h and then centrifuged at 5,000 g for 20 min in a Sorvall RC-5B refrigerated centrifuge (DuPont). Cell pellets were stored at -70°C until recombinant protein was isolated. After cell lysis, the recombinant protein was harvested by metal-affinity chromatography on a nickel-chelating resin (Ni-NTA, Qiagen) in the presence of denaturants (6 M guanidine, 6 M urea), detergent (0.1% Triton X-100), a reducing agent (1 mM beta -mercaptoethanol), and glycerol (10%) using a modification of a single-step purification/solubilization technique, in which denatured recombinant protein is solubilized in Tris-buffered saline and eluted in the same solution with EDTA (24). Eluate fractions containing hUAT were aliquoted and stored at -70°C until used in the lipid bilayer experiments.

Functional Evaluation of Recombinant hUAT

Formation of proteoliposomes. A 1:1 (wt/wt) mixture of bovine brain phosphatidylethanolamine (PE) and phosphatidylserine (PS; Avanti Polar Lipids, Birmingham, AL), each at a concentration of 10 mg/ml, were evaporated to dryness under a stream of nitrogen. The resultant pellet was suspended in 25 µl of 220 mM Cs2SO4 and 10 mM HEPES-NaOH at pH 7.4, after which 2 µl of recombinant hUAT protein were added. Proteoliposomes were formed by sonicating the suspension for 30 s at 80 kHz in a bath sonicator (Laboratory Supplies, Hicksville, NY) (36-38, 44). Fresh proteoliposomes were prepared for each experiment.

Lipid bilayer chamber, formation of lipid bilayer, and channel reconstitution. The lipid bilayer system was identical to that previously reported (36-38, 44). In all experiments, both chambers of the Plexiglas apparatus were filled with 1 ml of a solution containing 2.5 mM urate, 220 mM Cs2SO4, and 0.25 mM CaCl2 that was buffered to pH 7.4 with 10 mM HEPES-NaOH. Subsequently, a 50-µm hole in a Teflon film (type C-20, 12.5 µm thick, DuPont Electronics, Wilmington, DE) that had been tightly fitted between the two wells of the chamber was painted with lipids using a club-shaped glass rod. The lipids used to paint the bilayer were identical to those used to make the proteoliposomes (a 1:1 mixture of PE and PS, each at 10 mg/ml) but, after drying under nitrogen, the lipids were dissolved in n-decane (Sigma, St. Louis, MO) at a concentration approximating 50 mg lipid/ml. Junction potentials were corrected with the zero-adjust system of the patch-clamp amplifier (Axopatch 200B, Axon Instruments, Burlingame, CA). The cis chamber is defined as the chamber connected to the voltage-holding electrode; all voltages are referenced to the trans (ground) chamber. Voltage was generated, clamped at different voltages (-100 to +100 mV), and controlled with the patch-clamp amplifier. When a stable resistance of at least 100 GOmega and a noise level of <0.1 pA were maintained, the experiments were initiated by addition of 5 µl of the hUAT-containing proteoliposomes to the trans chamber. The solution in the trans chamber was stirred until the proteoliposomes fused with the bilayer.

Functional analysis of the channel. In each experiment, the activity of the channel was initially assessed in the presence of symmetrical solutions of 2.5 mM urate in 220 mM Cs2SO4, 0.25 mM CaCl2, and 10 mM HEPES-NaOH at pH 7.4 in the cis and trans chambers. Thereafter, the channel was reexamined in the symmetrical 2.5 mM urate solutions, but after the cis or trans chamber was pulsed with microliter volumes of one of the following reagents to achieve progressively increased concentrations in the bath: 2.5 mM alpha -lactose monohydrate (Sigma), 1.0 M D(+)-glucose (Sigma), 1.0 M D(-)-ribose (Sigma), 2.5 mM oxonate (Sigma), 2.5 mM pyrazinoate (PZA; Aldrich Chemical, Milwaukee, WI), or 1 mM adenosine (Sigma). All reagents were prepared in 220 mM Cs2SO4 and 10 mM HEPES-NaOH buffered to pH 7.4. In some experiments, channel activity was reexamined after the solution in the trans and/or cis chamber was replaced with reagent-free fresh urate solution.

Data collection and analysis. Current output of the patch clamp was filtered at 10 kHz through an eight-pole filter (Bessel filter, model 902, Frequency Devices, Haverhill, MA) that was digitized at 5 kHz (Digi Data 1200 series Interface, Axon Instruments). Data were analyzed with commercial software (pCLAMP, version 8.0, Axon Instruments) after additional digitized filtering at not less than 1 kHz.


    RESULTS
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

Characteristics of the Reconstituted Channel

Figure 1 demonstrates single-channel activity of hUAT (evidenced by clear transitions between open and closed states) in the presence of symmetrical urate solutions after fusion of the hUAT-containing proteoliposomes with the lipid bilayer. Single-channel activity was evident in most experiments; however, multiple channels and apparent substate conductances were also observed. As is evident from the traces (Fig. 1, A and B), the open probability of hUAT is independent of voltage and, in this experiment, the slope conductance was 2 pS (Fig. 1C). As previously reported, the mean single-channel slope conductance of hUAT, calculated by linear regression analysis, was 4.0 ± 0.4 pS (n = 11), rectification was not obvious in symmetrical urate solutions, and the channel was highly selective to urate (44). These combined findings indicate that two important properties of hUAT, its single-channel conductance and its voltage insensitivity, are distinctly different from those previously described for rat UAT (36, 38). Moreover, in contrast to rat UAT, hUAT channel activity generally displayed run-down.


View larger version (20K):
[in this window]
[in a new window]
  		Fig. 1.   Human urate transporter/channel (hUAT) activity, open probability of the channel, and current-voltage relationship in symmetrical urate solutions. Channel activity was recorded in symmetrical solutions of 2.5 mM urate, 220 mM Cs2SO4, and 10 mM HEPES-NaOH, pH 7.4, after fusion of hUAT-containing proteoliposomes with the lipid bilayer. A: 1-min traces of channel activity obtained at various holding potentials. Vertical arrows, time at onset of 1-s traces; solid horizontal lines, closed state. B: 1-s traces recorded during the 1-min traces depicted in A. Solid horizontal lines, closed state. C: current-voltage relationship of the channel depicted in A. Solid line, best fit by linear regression analysis. G and R, slope conductance and correlation coefficient, respectively.

Effect of alpha -Lactose on Activity of hUAT

The amino acid sequence of hUAT contains two highly conserved beta -galactoside binding domains, H x N P R 7x V x N 6x W 2x E x R 5x F 2x G and H x N P R 6x V x N 6x W 2x E x R 7x F 2x G, where x represents any amino acid and the number indicates the number of variable amino acids (45). The initial beta -galactoside binding domain is located within the first predicted extracellular domain (amino acids 61-96), and the other within the second predicted extracellular domain (amino acids 235-271) of hUAT (Fig. 2). Although these domains represent signature sequences for the galectins, and are known to bind selective sugars (6, 39, 45), the functional role, if any, of these domains has not been ascertained. To assess the possibility that these sites participate in the function of the urate channel, increasing concentrations of alpha -lactose, a well-known substrate for these binding sites (6), was added to the chambers bathing the cis or trans side of the channel. In the absence of alpha -lactose, the mean single-channel conductance approximated 2 pS (Fig. 3, A and B). However, as noted above and depicted in Fig. 3, several higher conductance levels were observed. Of note, simultaneous openings and/or closing to the higher conductance levels were seen intermittently, suggesting cooperativity between a number of subunits (31, 35). In the absence of alpha -lactose, the open probability of the channel was quite low (Fig. 3C), independent of the voltage applied (not depicted), and there was a rather rapid run down of channel activity over time. Addition of alpha -lactose to the cis side of the bilayer did not influence channel activity (Fig. 3, A and B). In distinct contrast, after addition of 70 ± 14.6 µM alpha -lactose to the trans chamber, the conductance of the channel increased significantly (Fig. 3, A and B), reaching a mean value of 8 ± 1.9 pS (n = 7). The increase in conductance to this level occurred in a progressive manner in association with increments in the concentration of lactose in the trans chamber (Fig. 3, A and B). Additionally, as in the control state, in the presence of alpha -lactose in the trans chamber simultaneous openings and closings to the higher conductance state were evident (Fig. 3, A and B). Finally, the presence of alpha -lactose resulted in a significant increase in the open probability of the channel (Fig. 3C) from 10.6 ± 5.1 to 58.3 ± 12.9% (n = 7) and reduced the likelihood of channel run down. This stabilization of channel activity at the higher conductance level (Fig. 3, A and B) could be consequent to a lactose-induced sustained cooperativity between hUAT subunits (multimerization) and/or modification in the conformation of the pore of the channel that results in a higher conductance state. On the basis of the assumption that alpha -lactose binds with the same or similar affinities to the two beta -galactoside binding sites within hUAT, the observed unilateral effect of this substrate requires that the topology of hUAT is such that both binding sites must be located on the same side of the channel (Fig. 2). Moreover, in view of the consistency in the unilateral (trans) effect of alpha -lactose, it appears that hUAT must insert in the lipid bilayer in a specific orientation. A similar uniformity in the direction of lipid insertion was observed for UAT and, as previously noted, the consistent orientation of the channel in the bilayer likely reflects the nonsymmetrical distribution of electrical charges on the bilayer lipids (36).


View larger version (21K):
[in this window]
[in a new window]
  		Fig. 2.   Topological model of hUAT. Numbers 1-4 designate the NH2-to-COOH terminal transmembrane domains. Numbers adjacent to the transmembrane domains indicate the amino acid residues at the beginning and end of each domain. The approximate location of the 2 beta -galactoside binding sites within the 2 extracellular domains of hUAT, the site with homology to the A1/A3 adenosine receptors on the extracellular side of transmembrane domain 2, and the urate/oxonate binding site with homology to uricase in the cytoplasmic loop between transmembrane domains 2 and 3 are indicated by arrows. Brackets in the loop into the membrane from the extracellular face of the channel, pointed to by 2 arrows from the cytoplasmic side of the model, represent the location of the 2 beta -sheets, which are connected by 6 amino acids that carry a net neutral charge in hUAT.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 3.   Channel activity in the absence and presence of alpha -lactose at a holding potential of 30 mV. A: 10-s traces of channel activity in symmetrical solutions of 2.5 mM urate, 220 mM Cs2SO4, and 10 mM HEPES-NaOH, pH 7.4, in the absence (top trace) and presence of the designated concentrations of alpha -lactose in the cis (2nd trace) and trans chambers (3rd and 4th traces) Solid horizontal lines, closed state. B: 1-s traces recorded during the 10-s traces depicted in A. The 1st, 2nd, and 4th traces represent the first second of the 10-s traces depicted in A; the 3rd trace was recorded at the time indicated by the arrow in A. Solid horizontal lines, closed state. C: open probability (%O.P.) of the channel during the traces depicted in A.

Effect of D(+)-Glucose, But Not D(-)-Ribose, on Activity of hUAT

It has been presumed that the galactose moiety of alpha -lactose forms the major interaction with the beta -galactoside binding sites in galectins (7, 39). The observation that there is at least a 100-fold higher affinity for alpha -lactose than galactose, however, has suggested that an interaction between the glucose moiety of alpha -lactose and the beta -galactoside binding sites is also important (7, 39). To assess the possibility that glucose per se may interact with hUAT, presumably via the beta -galactoside binding sites, hUAT channel activity was examined in the presence of increasing concentrations of glucose (n = 6). In three of these studies, hUAT channel activity was first assessed in the presence of increasing concentrations of D(-)-ribose, used as a control for a nonspecific sugar effect. As depicted in Fig. 4, A and B, addition of up to 50 mM D(-)-ribose to the trans side of the chamber failed to activate hUAT: open probability remained at < 1.0% for as long as 2 h after exposure of the channel to ribose. In distinct contrast, within minutes of addition of 5 mM D(+)-glucose to the trans chamber (Fig. 4, A and B) channel activity increased such that the open probability of the channel increased significantly to 11.7 ± 7.5% (n = 6). Of note, a further increase in glucose concentration in the trans chamber to 20 mM (Fig. 4, A and B) was associated with a further increase in the channel's open probability to 25.8 ± 14.7% (n = 6). Although these studies demonstrate that hUAT has a much higher affinity for alpha -lactose (µM) than for glucose (mM) (Figs. 3 and 4) under these experimental conditions, it is apparent that glucose, like alpha -lactose, significantly modulates hUAT channel activity (Fig. 4), presumably via conformational changes secondary to an interaction with the beta -galactoside binding domains in hUAT.


View larger version (25K):
[in this window]
[in a new window]
  		Fig. 4.   Channel activity in the presence of ribose and glucose at a holding potential of 75 mV. A: 10-s traces of channel activity in symmetrical solutions of 2.5 mM urate, 220 mM Cs2SO4, and 10 mM HEPES-NaOH, pH 7.4, in the presence of 50 mM ribose (top trace) and subsequent presence of the designated concentrations of glucose (2nd and 3rd traces) in the trans chamber. Solid horizontal lines, closed state. B: open probability (%O.P.) of the channel during the traces depicted in A.

Local Block of Homology to Glycophorin A Within hUAT

On the basis of the data obtained after addition of alpha -lactose to the trans chamber that suggest that hUAT may multimerize (Fig. 3), the amino acid sequence of hUAT was assessed with the multiple protein sequence alignment program MACAW (65) to search for a local block of homology between hUAT (and rat UAT) and the extensively characterized dimerization motif within the single transmembrane domain of glycophorin A (GpA) (41-43, 46, 47, 62, 63, 66). As depicted in Fig. 5, the dimerization motif of GpA is formed by seven amino acids, Leu75, Ile76, Gly79, Val80, Gly83, Val84, and Thr87 (42, 43), with the G79xxxG83 sequence being described as the motif that is likely to be involved in high-affinity association of transmembrane alpha -helices (62, 66). Alignment of amino acids 18-33 of both rat UAT and hUAT [the block of residues that was previously proposed to represent the first transmembrane domain of UAT (36)] reveals significant homology to GpA (Fig. 5). In both UAT and hUAT, four residues are identical to the seven residues of the dimerization motif in GpA, including G24xxxG28 (Fig. 5). In UAT and hUAT, additional residues are homologous to those within the dimerization motif of GpA: three in UAT and two in hUAT (Fig. 5). Of interest, a second GxxxG sequence exists in UAT and hUAT (residues 19-23) that may be relevant to dimerization; however, alignment of G19xxxG23 with G79xxxG83 of GpA yields a lower overall homology between the 16-amino acid block of GpA (Fig. 5) and comparably sized blocks of UAT and hUAT.


View larger version (9K):
[in this window]
[in a new window]
  		Fig. 5.   Local blocks of homology to glycophorin A in hUAT and rat UAT. The residue no. of the individual proteins is indicated at the beginning and end of each line. The arrows above the amino acid sequences indicate the amino acids that form the dimerization motif, i.e., Leu75, Ile76, Gly79, Val80, Gly83, Val84, and Thr87, in glycophorin A. Double lines between amino acids indicate identical residues; single lines indicate homologous residues. Homology is defined according to the Swiss-Prot data bank in which amino acids in each of the following groups are homologous: [S, T, A, G, P]; [N. D., E, Q]; [R, K, H]; [M, L, I, V]; and [F, Y. W].

Effect of Oxonate on the Activity of hUAT

Oxonate, a specific inhibitor of the enzyme uricase (14), both inhibits electrogenic urate transport in rat and rabbit renal cortical membrane vesicles (1, 2, 33) and blocks the activity of recombinant rat UAT that is reconstituted in the lipid bilayer system (36). As in other studies, multiple conductance states were detected in the absence of oxonate (Fig. 6, A and B). Oxonate concentrations up to 188 µM in the trans chamber failed to influence hUAT activity (Fig. 6, A and B). In contrast, addition of increasing concentrations of oxonate to the cis chamber, to a concentration of 138 ± 30.5 µM (n = 7), progressively decreased the number of conductance states and ultimately virtually abolished channel activity (Fig. 6, A and B). As depicted (Fig. 6C), the effect of oxonate on the open probability of hUAT (in the presence or absence of lactose) was quite similar to its effect on the activity of UAT (36). This oxonate-induced block of hUAT was reversible in that channel activity was fully restored after the oxonate-containing solution in the cis chamber was replaced with a fresh oxonate-free urate solution (not depicted). Of note, the effect of oxonate on both rat UAT (36) and hUAT activity is restricted in that the oxonate-induced block is only observed when the cytoplasmic face of the channel is exposed to the reagent. As previously proposed with UAT (36), the distinct asymmetrical effect of oxonate implies that this compound interacts with a specific domain in hUAT that is consistently localized on the cis face of the channel. It is of note that in our bilayer system the cis side of the chamber is exposed to changes in voltage, simulating an intracellular compartment; the cis chamber is connected to the voltage-holding electrode with all voltages referenced to the trans (ground) side. Insofar as the cis chamber represents an intracellular compartment, the domain in hUAT that interacts with oxonate, like the domain in UAT (36), must then reside on the cytoplasmic face of the channel. In this context, because lactose only influences hUAT when added to the opposite chamber of the bilayer (Fig. 3), the two beta -galactoside binding sites must reside on the extracellular face of hUAT.


View larger version (23K):
[in this window]
[in a new window]
  		Fig. 6.   Channel activity in the presence of alpha -lactose and in the absence and presence of oxonate at a holding potential of 100 mV. A: 10-s traces of channel activity in symmetrical solutions of 2.5 mM urate, 220 mM Cs2SO4, and 10 mM HEPES-NaOH, pH 7.4, in the absence (top trace) and presence of the designated concentrations of oxonate in the trans (2nd trace) and the cis chambers (3rd and 4th traces) Solid horizontal lines, closed state. B: 1-s traces recorded during the 10-s traces depicted in A. The 1st, 2nd, and 4th traces represent the first second of the 10-s traces depicted in A; the 3rd trace was recorded at the time indicated by the arrow in A. Solid horizontal lines, closed state. C: open probability (%O.P.) of the channel during the traces depicted in A.

Local Block of Homology to Uricase Within hUAT

Because oxonate is a competitive inhibitor of uricase (14), and oxonate blocks hUAT channel activity (Fig. 6), the amino acid sequence of the human homologue was evaluated to determine whether it contains a block of homology to the substrate binding site in uricase. Importantly, the Q228 of Aspergillus uricase, which is critical to substrate binding (11) (presumably to oxonate as well as urate), is conserved within a 12-amino acid domain of porcine uricase, Aspergillus uricase, hUAT, rat UAT, and the intestinal isoform of galectin 9 (Fig. 7A). As depicted, alignment of a 12-amino acid block of porcine uricase (residues 231-242) and Aspergillus uricase (residues 224-235) with residues 158-169 of hUAT reveals that hUAT has 50% homology to both porcine and Aspergillus uricase (Fig. 7A). Alignment of residues 157-168 of rat UAT with residues 158-169 of hUAT reveals that this block of amino acids is highly conserved with 92% homology between the rat and human sequences (Fig. 7B). It is of note that a sequence for a gastrointestinal isoform of human galectin 9, which is inserted immediately after residue 148 of galectin 9, has been deposited in GenBank (accession no. AB006782). This domain in hUAT appears to be in part duplicated insofar as alignment of amino acids 4-15 of the 32-amino acid isoform sequence also has a high degree of homology to uricase, having 50 and 67% homology to porcine and Aspergillus uricase, respectively (Fig. 7).


View larger version (32K):
[in this window]
[in a new window]
  		Fig. 7.   Local blocks of homology to uricase. A: alignment of hUAT and the human galectin 9 intestinal isoform with pig/rat uricase and Aspergillus uricase. The residue numbers of the individual proteins are indicated at the beginning and end of each line. Double lines between amino acids indicate identical residues; single lines indicate homologous residues as defined in Fig. 4. B: alignment of the rat, human, pig, and mouse homologues of UAT and their intestinal isoforms with pig/rat uricase and Aspergillus uricase. The residue numbers of the individual proteins are indicated at the beginning and end of each line. gal, Galectin. The arrows in A and B demonstrate conservation in all of these sequences of the amino acid Q 228, which is critical to substrate binding in Aspergillus uricase.

Effect of PZA on Activity of hUAT

PZA, a potent inhibitor of urate transport in intact kidneys of multiple species (3) and an inhibitor of electrogenic urate transport in rat and rabbit membrane vesicles (1, 2, 33), also blocks channel activity of recombinant rat UAT (36). Comparable to observations made with recombinant rat UAT (36), despite the raising of the PZA concentration to 150 µM in the cis chamber, PZA failed to alter channel activity of hUAT (Fig. 8, A and B). Similar to the effect of PZA on rat UAT (36), PZA induced a dose-dependent block of hUAT activity (Fig. 8, A and B) and a reduction in open probability (Fig. 8C) when added to the trans chamber (in the presence or absence of lactose). The open probability of the channel was profoundly reduced at a concentration of 87.5 ± 43.3 µM (n = 5). This block was completely reversed after the PZA-containing solution was replaced with a fresh PZA-free urate solution (not shown). Because PZA and oxonate only effectively block channel activity when in contact with the trans and cis faces of the channel, respectively, the specific domains that bind these substrates in hUAT, as in UAT (36), must be located on opposite faces of the channel. Insofar as oxonate binds to a site on the cytoplasmic face of the channel, then the domain that binds PZA in hUAT, as in the case of the domain in UAT (36), must reside within an extracellular portion of the channel.


View larger version (18K):
[in this window]
[in a new window]
  		Fig. 8.   Channel activity in the presence of alpha -lactose and in the absence and presence of pyrazinoate (PZA) at a holding potential of 75 mV. A: 10-s traces of channel activity in symmetrical solutions of 2.5 mM urate, 220 mM Cs2SO4, and 10 mM HEPES-NaOH, pH 7.4, in the absence (top trace) and presence of the designated concentrations of PZA in the cis (2nd trace) and trans chambers (3rd and 4th traces). Solid horizontal lines, closed state. B: 1-s traces recorded during the 10-s traces depicted in A. All 4 traces represent the first second of the 10-s traces depicted in A. Solid horizontal lines, closed state. C: open probability (%O.P.) of the channel during the traces depicted in A.

Effect of Adenosine on Activity of hUAT

As demonstrated in Fig. 9, hUAT, like UAT, contains a local block of amino acids that has 73 and 45% homology to the adenosine A1 (60, 73) and A3 (64) receptors, respectively. Importantly, this block of amino acids is identical to or homologous with the specific residues in the A1 receptor, P249, H251, and N254, that bind adenosine and xanthine (32, 57). These amino acids align with amino acids P127, H129, and D132 of hUAT. Because functional studies with rat recombinant protein documented that adenosine is a potent blocker of UAT channel activity (36), the possibility was evaluated that adenosine would also interact with hUAT. As depicted in Fig. 10, adenosine essentially eliminated channel activity when the concentration of adenosine in the trans chamber reached 14.2 ± 5.8 µM, producing a profound decrease in open probability (n = 3). As was the case with rat UAT, adenosine failed to block channel activity when a comparable concentration was achieved in the cis chamber (Fig. 10, A and B). This unilateral effect of adenosine on hUAT clearly implies that the block of homology to the adenosine receptors in hUAT is functional and, as in the case of UAT (36), is only exposed when adenosine is applied to the extracellular (trans) face of the channel.


View larger version (8K):
[in this window]
[in a new window]
  		Fig. 9.   Local block of homology in hUAT to the adenosine A1/A3 receptors. The residue nos. of the individual proteins are indicated at the beginning and end of each line. Double lines between amino acids indicate identical residues; single lines indicate homologous residues as defined in Fig. 4.



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 10.   Channel activity in the absence and presence of adenosine at a holding potential of 50 mV. A: 10-s traces of channel activity in symmetrical solutions of 2.5 mM urate, 220 mM Cs2SO4, and 10 mM HEPES-NaOH, pH 7.4, in the absence (top trace) and presence of the designated concentrations of adenosine in the cis (2nd trace) and trans chambers (3rd and 4th traces). Solid horizontal lines, closed state. B: 1-s traces recorded during the 10-s traces depicted in A. All 4 traces represent the first second of the 10-s traces depicted in A. Solid horizontal lines, closed state. C: open probability (% O.P.) of the channel during the traces depicted in A.


    DISCUSSION
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

The present studies demonstrate that the human homologue of the urate transporter/channel, hUAT, (44) has characteristics that are both similar to and different from the highly homologous rat protein, rat UAT (36, 38). Three reagents that were previously documented to block rat UAT channel activity, oxonate, PZA, and adenosine (36), have been shown to similarly block hUAT channel activity (Figs. 5, 7, and 9). Moreover, with both the human and rat channels, each of these substrates only blocks channel activity when a specific side of the channel is exposed to the compound (Figs. 5, 7 and 9). There are, however, two important differences in the biophysical properties of these channels. First, the mean single-channel conductance of hUAT approximates one-half that of rat UAT (4.0 ± 0.4 vs. 9.5 ± 0.47 pS). Second, the open probability of hUAT is voltage independent (Fig. 1), whereas that of rat UAT is consistently voltage dependent (36, 38).

The concordance of findings observed with recombinant rat (36) and human homologues of UAT relative to the inhibitory effects of oxonate, PZA, and adenosine on channel activity (Figs. 5, 7 and 9), in conjunction with the identical sidedness of effects of the respective substrates, implies that the topologies of the human and rat transporters are similar. We previously proposed a molecular model for rat UAT that incorporated intracellular NH2 and COOH termini and four transmembrane domains (36). This model, including the specific amino acid residues that represent the four transmembrane alpha -helices (36), was based on electrophysiological studies in lipid bilayers that revealed the sidedness of effects of these same three reagents, the location of local blocks of homology to the A1/A3 receptors and uricase within UAT, the hydrophobicity profile of UAT, and the detection of hydrophobic segments (long enough to span the membrane) with significant homology to transmembrane domain 2 in urate/xanthine permease (19), the alpha -helix documented to form transmembrane domain E in bacterial rhodopsin (58), and a portion of the alpha -helix reported to form transmembrane domain IX of subunit 1 of cytochrome c oxidase (74). By incorporating all of this information, the hydrophobic segments with homology to urate/xanthine permease, bacterial rhodopsin, and cytochrome c oxidase were modeled as transmembrane domains 1, 2, and 3, respectively, in rat UAT (36).

Confirmation that rat and human UAT are transmembrane proteins has been obtained in surface biotinylation studies of renal and nonrenal epithelia-derived cells transfected with the cDNA of rat and human UAT (44, 59). Moreover, recent evidence in support of the above-described model was obtained with immunofluorescent and confocal microscopy of nonpermeabilized and permeabilized epithelial cells subsequent to transfection with NH2 or COOH FLAG-tagged UAT cDNAs; the NH2 and COOH termini of both rat and human UAT were observed to reside on the intracellular side of the plasma membrane (44, 59). Additional strong support for this model is provided by the present studies. First, the high degree of homology that has been detected within amino acids 18-33 of both hUAT and UAT to the dimerization domain within the single transmembrane alpha -helix of GpA (Fig. 5) (41-43, 46, 47, 62, 63, 66) supports our previous molecular model in which amino acids 15-35 of UAT were designated as transmembrane domain 1 (36). Second, we previously proposed that UAT contains two large extracellular domains, one located between transmembrane domains 1 and 2 and the second located between transmembrane domains 3 and 4 (Fig. 2). Importantly, hUAT contains two beta -galactoside binding sites, one encompassed by residues 61-96 within the first putative extracellular domain and the second incorporated by residues 235-271 within the second putative extracellular domain. Of note, the specific amino acids involved in each of these sites are 100% conserved in human and rat UAT (the latter within residues 60-95 and 234-270). The present finding that alpha -lactose only influences hUAT channel activity when in contact with the extracellular face of the channel (Fig. 3) is thus consistent with our model. The combination of findings of intracellular locations of the NH2 and COOH termini of hUAT and UAT (44, 59) and extracellular locations of the two beta -galactoside binding sites could be consistent with two transmembrane alpha -helices. However, the unilateral intracellular block of channel activity induced by the uricase inhibitor oxonate (Fig. 6) and the high degree of likelihood that this substrate interacts with amino acids 158-169 of the uricase-like domain in hUAT (Fig. 7) require that this site is exposed to the intracellular face of the channel. Because the uricase-like domain is located between the two beta -galactoside binding sites, hUAT therefore must contain at least four rather than two transmembrane alpha -helices (Fig. 2), a model entirely compatible with our previously proposed molecular model of the rat urate transporter/channel (36).

We previously suggested that a local block of homology to uricase within UAT (36) is most likely responsible for the functional similarities between the electrogenic urate transporter and uricase (1, 2, 33, 34, 37) and the ability of our polyclonal antibody to porcine uricase to select the UAT clone from the rat cDNA library (38), react with recombinant UAT (38), block electrogenic urate transport in membrane vesicles prepared from rat kidney (34), and selectively block UAT channel activity from the cytoplasmic side of the channel (36). It was also suggested that the oxonate-induced block of UAT activity was most likely consequent to its interaction with the uricase-like domain in UAT (36). In previously aligning a local block of homology in rat UAT to the substrate binding site in uricase, Q156 in rat UAT was aligned with Q228 in Aspergillus uricase (36), the amino acid that has been identified by X-ray crystallography as being critically important in formation of the substrate-uricase complex (11). However, on the basis of additional sequence data, specifically hUAT (44), the human (51, 75), mouse, and rat sequences of galectin 9 (76), the pig sequence for a urate transporter/channel (72), and the intestinal isoforms of human (accession no. AB006782), mouse, and rat galectin 9 (76) and pig urate transporter/channel (72), an adjustment has been made in this alignment. We now assign the substrate binding glutamine as Q161 in rat UAT as it is this glutamine, rather than Q156, that is conserved in the rat, human, pig, and mouse sequences (Fig. 7, A and B). In addition to conservation of this glutamine in the four mammalian species, it is evident that there is also an extremely high degree of homology within the block of amino acids that is located just proximal to the uricase-like domain in the intestinal isoform of these same species (Fig. 7B).

Despite the high degree of homology (Fig. 11) between and apparent similarities in the topology of the human and rat homologues of UAT, evidence has been obtained that two important biophysical properties of these proteins differ (voltage sensitivity of the open probability and single-channel conductance). The voltage sensitivity of channels has been ascribed to the presence of charged residues, specifically, basic residues localized in critical points in the structure of the channel (8). These residues are presumed to sense the electrical field across the membrane (8) and produce a significant change in the conformation of the protein that affects the open probability of the channel. We previously reported (36) that rat UAT contains two putative beta -sheets (residues 96-104 and 111-119) linked by six amino acids (105-110) that carry a net positive charge (R106, E108, and K110) in a domain located between transmembrane alpha -helices 1 and 2 (Fig. 2). We proposed that this segment of the protein could act as a mobile domain (69), interact with the amphiphilic alpha -helices forming the UAT pore, result in presentation of this charged region to the cytoplasmic side of the channel (23), and thereby function as the voltage sensor. It is of note that in contrast to rat UAT, the six amino acids (106-111) that link the two putative beta -sheets in hUAT (residues 97-105 and 112-120) carry a net neutral charge (S107, D109, and K111) (Fig. 2). Insofar as this region does contain the voltage sensor in the channel, the evolutionary modification in the first of these three amino acids may then be responsible for the lack of voltage sensitivity of the open probability of hUAT. Clearly, confirmation of this proposal will require an analysis of the voltage sensitivity of recombinant rat UAT subsequent to an experimentally induced mutation of amino acid 106 from arginine to serine.


View larger version (69K):
[in this window]
[in a new window]
  		Fig. 11.   Comparison of amino acid sequences of the rat and human homologues of the urate transporter/channel. R and H, rat and human UAT sequences, respectively; shaded amino acids, those residues that are identical or homologous in the rat and human sequences.

The second significant difference between the biophysical properties of the rat and human homologues of UAT is the conductance of the respective channels. The single-channel conductance of hUAT that has been measured in prior (44) and present studies (Figs. 1 and 3) in the absence of alpha -lactose approximates one-half of that observed with rat UAT (36, 38). Although the actual basis for this difference is not known, we assume that the conductance difference reflects evolutionary changes in some amino acids in the region of the rat and human channels that are critical to the conformation of the pore. It is of interest that there appears to be a clustering of nonhomologous amino acids in the human and rat sequences within putative transmembrane alpha -helices 2 and 3 and, probably most significantly, in the block of cytoplasmic amino acids that connect these alpha -helices (Figs. 2 and 11). Importantly, within the hairpin turn between these transmembrane helices, the human channel contains three prolines (P150, P154, and P157), whereas rat UAT has only one (P153) (Fig. 11). Because prolines are known to induce turns in transmembrane segments and result in the formation of helical hairpins (53, 55), the increased number of prolines in the hairpin turn of hUAT may alter the conformation of this domain and result in increased packing of the transmembrane alpha -helices. Insofar as transmembrane alpha -helices 2 and 3 participate in formation of the channel pore, closer packing of these alpha -helices may decrease the size of the channel pore and thereby reduce channel conductance in hUAT relative to that of rat UAT. Alternatively, the difference between the initial amino acid of the dimerization motif in the first transmembrane alpha -helix of human and rat UAT (Fig. 5) may influence packing of the transmembrane domains and thereby affect conductance. Induced mutations in hUAT to recapitulate the amino acids found in rat UAT in the first transmembrane alpha -helix and/or the domain between transmembrane alpha -helices 2 and 3 will be required to assess the possibility that the difference in conductance in the rat and human homologues is consequent to naturally evolved divergences in their primary structure.

Lactose, a beta -galactoside, has previously only been utilized as a tool to isolate and purify galectins (4, 7, 10, 15-17, 22, 26, 48, 51, 56, 76). The ability to use this reagent for purification purposes is consequent to the fact that lactose binds to the highly conserved beta -galactoside binding domains within galectins (7). Because various beta -galactosides are found on glycolipids and glycoproteins on cell surfaces and extracellular matrix, some of which have also been shown to bind to the beta -galactoside binding sites in galectins (39), it has been suggested that secreted galectins could function as biologically significant ligands that play a role in cell migration, cell proliferation, immune function, and adhesion (4, 7, 10, 15-17, 22, 26, 48, 56). To date, however, there is no direct evidence to support these proposals.

In contrast to the absence of functional data relative to lactose or the beta -galactoside binding sites in galectins, the present study indicates that lactose, presumably by binding to the beta -galactoside binding domains in hUAT, regulates the activity of hUAT by significantly augmenting its conductance and open probability (Fig. 3). Although alpha -lactose per se would not be relevant to physiological function in vivo, it is important to note that both of lactose's component sugars, galactose and glucose, interact with the beta -galactoside binding domain in galectins (7), and therefore these sugars could regulate channel activity in vivo. The functional consequence of an increase in conductance and open probability of hUAT would be an increase in urate flux. On the basis of membrane potential and the likely prevailing electrochemical gradient for urate, this would represent an increase in urate efflux from systemic cells (inducing hyperuricemia) and an increase in urate excretion consequent to an increase in the rate of urate secretion in the renal proximal tubule and intestine (inducing hypouricemia). In this context, it is of interest that elevation of blood galactose levels in galactosemic patients (13) and rather modest elevations in blood glucose levels in diabetic patients (79) are associated with hyperuricemia. However, during periods of poor metabolic control hypouricemia and hyperuricosuria are evident in diabetic patients (18, 49). Of note, both are corrected when blood glucose is normalized (18). Consistent with the likelihood that increased proximal tubular fluid glucose concentration also has a direct effect on urate flux is the observation that the infusion of glucose induces an increase in urate clearance in humans that significantly exceeds that induced by mannitol at comparable osmolar clearances (54, 67, 68). Reports of this (18, 49, 54, 67, 68, 79), in conjunction with the observed effect of glucose on hUAT channel activity at concentrations comparable to physiological (5 mM) and pathological (20 mM) plasma levels in humans (Fig. 4), suggest that extracellular glucose may interact with urate transporter/channels that reside in systemic cells and the renal proximal tubule (27) and thereby exert a direct regulatory effect on the activity of this channel.

In summary, the present studies have provided evidence that recombinant protein that was prepared from the cDNA of hUAT has topological characteristics that are comparable to those of the rat homologue UAT. However, these proteins are not functionally identical: differences in their biophysical properties suggest that evolutionary changes in specific amino acids in these two highly homologous proteins are functionally relevant in defining these properties. Finally, the present data suggest that an interaction of selective sugars with the beta -galactoside binding domains in hUAT may be responsible, at least in part, for regulating hUAT channel activity in vivo.


    ACKNOWLEDGEMENTS

This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants DK-52785 (R. G. Abramson) and DK-57867 (M. S. Lipkowitz).


    FOOTNOTES

Address for reprint requests and other correspondence: R. G. Abramson, Div. of Nephrology, Box 1243, Mount Sinai School of Medicine, One Gustave L. Levy Place, New York, NY 10029 (E-mail: ruth.abramson{at}mssm.edu).

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

First published February 20, 2002;10.1152/ajprenal.00333.2001

Received 1 November 2001; accepted in final form 12 February 2002.


    REFERENCES
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

1.   Abramson, RG, King VF, Reif MC, Leal-Pinto E, and Baruch SB. Urate uptake in membrane vesicles of rat renal cortex: effect of copper. Am J Physiol Renal Fluid Electrolyte Physiol 242: F158-F170, 1982[Abstract/Free Full Text].

2.   Abramson, RG, and Lipkowitz MS. Carrier-mediated concentrative urate transport in rat renal membrane vesicles. Am J Physiol Renal Fluid Electrolyte Physiol 248: F574-F584, 1985[Abstract/Free Full Text].

3.   Abramson, RG, and Lipkowitz MS. Evolution of the uric acid transport mechanisms in vertebrate kidney. In: Basic Principles in Transport, edited by Kinne RKH. Basel: Karger, 1990, p. 115-153.

4.   Albrandt, K, Orida NK, and Liu FT. An IgE-binding protein with a distinctive repetitive sequence and homology with an IgG receptor. Proc Natl Acad Sci USA 84: 6859-6863, 1987[Abstract/Free Full Text].

5.   Al-Khalidi, UAS, and Chaglassian TH. The species distribution of xanthine oxidase. Biochem J 97: 318-320, 1965[ISI][Medline].

6.   Barondes, SH, Castronovo V, Cooper DN, Cummings RD, Drickamer K, Feizi T, Gitt MA, Hirabayashi J, Hughes C, Kasai K, Barondes SH, Castronovo V, Cooper DNW, Cummings RD, Drickamer K, Feizi T, Gitt MA, Hirabayashi J, Hughes C, Kasai K, Leffler H, Liu F-T, Lotan R, Mercurio AM, Monsigny M, Pillai S, Poirer F, Raz A, Rigby PWJ, Rini JM, and Wang JL. Galectins: a family of animal beta-galactoside-binding lectins. Cell 76: 597-598, 1994[ISI][Medline].

7.   Barondes, SH, Cooper DN, Gitt MA, Leffler H., and Galectins Structure and function of a large family of animal lectins. J Biol Chem 269: 20807-20810, 1994[Free Full Text].

8.   Bezanilla, F. The voltage sensor in voltage-dependent ion channels. Physiol Rev 80: 555-592, 2000[Abstract/Free Full Text].

9.   Blomstedt, JW, and Aronson PS. pH gradient-stimulated transport of urate and p-aminohippurate in dog renal microvillus membrane vesicles. J Clin Invest 65: 931-934, 1980[ISI][Medline].

10.   Clerch, LB, Whitney P, Hass M, Brew K, Miller T, Werner R, and Massaro D. Sequence of a full-length cDNA for rat lung beta-galactoside-binding protein: primary and secondary structure of the lectin. Biochem 27: 692-699, 1988[Medline].

11.   Colloc'h, N, El Hajji M, Bachet B, L'Hermite G, Schiltz M, Prangé T, Castro B, and Mornon JP. Crystal structure of the protein drug urate oxidase-inhibitor complex at 2.05 Å resolution. Nature Structural Biol 4: 947-952, 1997[ISI][Medline].

12.   De Duve, C, and Baudhuin P. Peroxisomes (microbodies and related particles). Physiol Rev 46: 323-357, 1966[Free Full Text].

13.   Forster, J, Schuchmann L, Hans C, Niederhoff H, Kunzer W, and Keppler D. Increased serum urate in galactosemia patients after a galactose load: a possible role of nucleotide deficiency in galactosemic liver injury. Klin Wochenschr 53: 1169-1170, 1975[ISI][Medline].

14.   Fridovich, I. The competitive inhibition of uricase by oxonate and by related derivatives of s-triazines. J Biol Chem 240: 2491-2494, 1965[Free Full Text].

15.   Gitt, MA, Colnot C, Poirier F, Nani KJ, Barondes SH, and Leffler H. Galectin-4 and galectin-6 are two closely related lectins expressed in mouse gastrointestinal tract. J Biol Chem 273: 2954-2960, 1998[Abstract/Free Full Text].

16.   Gitt, MA, Massa SM, Leffler H, and Barondes SH. Isolation and expression of a gene encoding L-14-II, a new human soluble lactose-binding lectin. J Biol Chem 267: 10601-10606, 1992[Abstract/Free Full Text].

17.   Gitt, MA, Wiser MF, Leffler H, Herrmann J, Xia YR, Massa SM, Cooper DN, Lusis AJ, and Barondes SH. Sequence and mapping of galectin-5, a beta-galactoside-binding lectin, found in rat erythrocytes. J Biol Chem 270: 5032-5038, 1995[Abstract/Free Full Text].

18.   Gonzalez-Sicilia, L, Garcia-Estan J, Martinez-Blazquez A, Fernandez-Pardo J, Quiles JL, and Hernandez J. Renal metabolism of uric acid in type I insulin-dependent diabetic patients: relation to metabolic compensation. Horm Metab Res 29: 520-523, 1997[ISI][Medline].

19.   Gorfinkiel, L, Diallinas G, and Scazzocchio C. Sequence and regulation of the uapA gene encoding a uric acid-xanthine permease in the fungus Aspergillus nidulans. J Biol Chem 268: 23376-23381, 1993[Abstract/Free Full Text].

20.   Guggino, SE, and Aronson PS. Paradoxical effects of pyrazinoate and nicotinate on urate transport in dog renal microvillus membranes. J Clin Invest 76: 543-547, 1985[ISI][Medline].

21.   Guggino, SE, Martin GJ, and Aronson PS. Specificity and modes of the anion exchanger in dog renal microvillus membranes. Am J Physiol Renal Fluid Electrolyte Physiol 244: F612-F621, 1983[Abstract/Free Full Text].

22.   Hadari, YR, Paz K, Dekel R, Mestrovic T, Accili D, and Zick Y. Galectin-8. A new rat lectin, related to galectin-4. J Biol Chem 270: 3447-3453, 1995[Abstract/Free Full Text].

23.   Hille, B. Ionic Channels of Excitable Membranes. Sunderland, MA: Sinauer, 1992.

24.   Holzinger, A, Phillips KS, and Weaver TE. Single-step purification/solubilization of recombinant proteins: application to surfactant protein B. Biotechniques 20: 804-808, 1996[ISI][Medline].

25.   Hruban, Z, and Recheigl M. Microbodies and related particles: morphology, biochemistry, and physiology. In: International Review of Cytology, edited by Bourne GH, and Danielli JF.. New York: Academic, 1969, suppl. 1, p. 20-39.

26.   Hughes, RC. Secretion of the galectin family of mammalian carbohydrate-binding proteins. Biochim Biophys Acta 1473: 172-185, 1999[Medline].

27.   Hyink, DP, Rappoport JZ, Wilson PD, and Abramson RG. Expression of the urate transporter/channel is developmentally regulated in human kidneys. Am J Physiol Renal Physiol 281: F875-F886, 2001[Abstract/Free Full Text].

28.   Kahn, AM, and Aronson PS. Urate transport via anion exchange in dog renal microvillus membrane vesicles. Am J Physiol Renal Fluid Electrolyte Physiol 244: F56-F63, 1983[Abstract/Free Full Text].

29.   Kahn, AM, Branham S, and Weinman EJ. Mechanism of urate and p-aminohippurate transport in rat renal microvillus membrane vesicles. Am J Physiol Renal Fluid Electrolyte Physiol 245: F151-F158, 1983[Abstract/Free Full Text].

30.   Kahn, AM, Shelat H, and Weinman EJ. Urate and p-aminohippurate transport in rat renal basolateral vesicles. Am J Physiol Renal Fluid Electrolyte Physiol 249: F654-F661, 1985[Abstract/Free Full Text].

31.   Keleshian, AM, Edeson RO, Liu GJ, and Madsen BW. Evidence for cooperativity between nicotinic acetylcholine receptors in patch clamp records. Biophys J 78: 1-12, 2000[Abstract/Free Full Text].

32.   Kim, J, Wess J, van Rhee AM, Shöneberg T, and Jacobson KA. Site-directed mutagenesis identifies residues involved in ligand recognition in the human A2A adenosine receptor. J Biol Chem 270: 13987-13997, 1995[Abstract/Free Full Text].

33.   Knorr, BA, Beck JC, and Abramson RG. Classical and channel-like urate transporters in rabbit renal brush border membranes. Kidney Int 45: 727-736, 1994[ISI][Medline].

34.   Knorr, BA, Lipkowitz MS, Potter BJ, Masur SK, and Abramson RG. Isolation and immunolocalization of a rat renal cortical membrane urate transporter. J Biol Chem 269: 6759-6764, 1994[Abstract/Free Full Text].

35.   Krouse, ME, and Wine JJ. Evidence that CFTR channels can regulate the open duration of other CFTR channels: cooperativity. J Membr Biol 182: 223-232, 2001[ISI][Medline].

36.   Leal-Pinto, E, Cohen BE, and Abramson RG. Functional analysis and molecular modeling of a cloned urate transporter/channel. J Membr Biol 169: 13-27, 1999[ISI][Medline].

37.   Leal-Pinto, E, London RD, Knorr BA, and Abramson RG. Reconstitution of hepatic uricase in planar lipid bilayer reveals a functional organic anion channel. J Membr Biol 146: 123-132, 1995[ISI][Medline].

38.   Leal-Pinto, E, Tao W, Rappaport J, Richardson M, Knorr BA, and Abramson RG