The emerging physiological roles of the SLC14A family of urea transporters

Abstract

In mammals, urea is the main nitrogenous breakdown product of protein catabolism and is produced in the liver. In certain tissues, the movement of urea across cell membranes is specifically mediated by a group of proteins known as the SLC14A family of facilitative urea transporters. These proteins are derived from two distinct genes, UT-A (SLC14A2) and UT-B (SLC14A1). Facilitative urea transporters play an important role in two major physiological processes – urinary concentration and urea nitrogen salvaging. Although UT-A and UT-B transporters both have a similar basic structure and mediate the transport of urea in a facilitative manner, there are a number of significant differences between them. UT-A transporters are mainly found in the kidney, are highly specific for urea, have relatively lower transport rates and are highly regulated at both gene expression and cellular localization levels. In contrast, UT-B transporters are more widespread in their tissue location, transport both urea and water, have a relatively high transport rate, are inhibited by mercurial compounds and currently appear to be less acutely regulated. This review details the fundamental research that has so far been performed to investigate the function and physiological significance of these two types of urea transporters.

Overview

Urea is the main breakdown product of protein catabolism in mammals. It is produced in the liver via the ornithine-urea cycle and has classically been viewed simply as a toxic nitrogenous waste product, emitted from the body in both urine and faeces. Urea (H₂NCONH₂) is a water soluble molecule and was originally thought to simply pass slowly across cell membranes via passive diffusion. However, the cloning of the specific proteins responsible for the enhanced trans-epithelial urea transport that is present in certain tissues has revolutionized our understanding of the physiological importance of urea.

Urea transporters (UTs) allow the rapid movement of urea molecules across cell membranes. Although this can occur in either direction through the transporters, net urea movement can only occur down a concentration gradient. This equilibrative movement of urea is also independent of ions such as sodium and chloride (You et al., 1993). These proteins are generally referred to as ‘facilitative urea transporters’, and are distinct from the undefined class of proteins responsible for the ‘active’ uptake of urea against concentration gradients.

Urea transporters are derived from two separate genes: SLC14A1 (UT-B) and SLC14A2 (UT-A). These UT-A and UT-B genes are on the same chromosome (18q12.1-q21.1), suggesting gene duplication from a common ancestor (Olives et al., 1996). The Kidd (Jk) blood group, previously known to be located on this chromosome, does in fact represent the human UT-B urea transporter (Olives et al., 1995).The human UT-A gene consists of 26 exons covering 476 kb, with similarly large genes being present in mouse and rat (Smith and Fenton, 2006). In contrast, the human UT-B gene is smaller, consisting of 11 exons spread over 30 kb (Lucien et al., 1998). Again, the UT-B gene is of a similar size in both mouse and rat (Yang and Bankir, 2005). [Note: a third urea transporter gene, UT-C, has so far only been identified in two species of fish (Mistry et al., 2005).]

The most studied physiological role for urea transporters is in the urinary concentration mechanism. Renal UT-A transporters within the kidney nephron help facilitate the reabsorption and recycling of urinary urea, hence increasing medullary urea concentration – see Figure 1. This prevents the problematic osmotic effect that urinary urea would otherwise have and hence prevents excess water loss in the urine (Fenton et al., 2004). During reasonable levels of dietary nitrogen intake, urea transporters are therefore vital in the production of concentrated urine and hence in maintaining body fluid balance. Detailed discussion of the precise renal function of urea transporters will not be presented here, as this topic has been the subject of an excellent recent review (see Fenton, 2009).

Figure 1.

Schematic model of the kidney nephron showing role of Slc14a transporters in the renal urinary concentrating mechanism. Blood urea is freely filtered at the glomerulus and ∼40% is constitutively reabsorbed in the proximal tubule. The remaining urea can be reabsorbed across the epithelial cells of the inner medullary collecting duct, mainly via apical UT-A1 transporters and basolateral UT-A3 transporters that are both regulated by vasopressin. This has the effect of increasing urea concentration in the kidney medulla and hence preventing the excessive water loss that would otherwise occur due to the osmotic effect of the urinary urea. In order to maintain this high medullary concentration, it is necessary for the urea to be recycled and the concentration gradient not to be dissipated. This is achieved by the facilitated movement of urea across the apical and basolateral of both the thin descending limbs (via UT-A2 transporters) and the descending vasa recta blood vessels (via UT-B1 transporters).

A second important physiological role for urea transporters is now emerging in respect to its role in the process of urea nitrogen salvaging (UNS) in the mammalian intestinal tract. This process supplies intestinal bacteria with a source of nitrogen that they utilize for their growth and is hence vital in maintaining the symbiotic relationship between mammals and their bacterial populations, particularly in ruminant species (for detailed review – see Stewart and Smith, 2005). The crucial first step in UNS is the movement of urea from the blood into the intestinal tract, via UT-B urea transporters in the epithelial layers – see Figure 2. UT-B proteins have now been identified in various intestinal tissues and species, such as bovine rumen (Stewart et al., 2005), rat caecum (Collins et al., 2011) and, most interestingly, human colon (Inoue et al., 2004; Collins et al., 2010). Urea transporters may therefore indirectly play a significant role in both nutritional balance and intestinal health.

Figure 2.

Model of the urea nitrogen salvaging process. Urea is produced in the liver via the ornithine-urea cycle and passed into the blood. Urea can then pass (a) into the kidney, where it is freely filtered and either reabsorbed or excreted, or (b) via UT-B urea transporters into specific regions of the gastrointestinal tract that contain large bacterial populations (e.g. rumen, caecum, colon). Within these regions urea is broken down by the bacterial enzyme urease into ammonia and carbon dioxide. The ammonia can either be reabsorbed directly back into the blood or be utilized as a nitrogen source by the bacteria to produce amino acids and peptides, which themselves can then be reabsorbed. The return of the nitrogen to the mammalian host in these different forms represents the ‘salvaging’ of the original urea nitrogen.

Isoforms

Both the Slc14a1 and Slc14a2 genes produce multiple isoforms, via the process of alternative splicing (for review of genomic organization – see Smith and Fenton, 2006). There are six known Slc14a2 (UT-A) transporters and two Slc14a1 (UT-B) transporters. Figure 3 shows a schematic representation of these eight urea transporter proteins. Some of these isoforms have yet to be fully characterized in more than one species at present. For example, cDNA sequences for UT-B2 have been reported in human caudate nucleus (GenBank Acc. No. AK091064) and mouse thymus (GenBank Acc. No. AK153891), but the proteins have yet to be investigated. Evidence is also emerging of the existence of further novel isoforms, particularly for UT-B transporters. For example, a cDNA clone from human thalamus appears to encode a novel 281-amino acid UT-B protein (GenBank Acc. No. AK127452) that has a truncated N-terminus compared with the UT-B1 transporter. Because the current nomenclature for Slc14a urea transporters was not originally utilized, the previous aliases used in the literature are listed in Table 1. This table also details the small variations in amino acid length between species that occurs for certain transporters and includes a basic guide to tissue distribution (for further details see distribution section below).

Figure 3.

Schematic representation of the different isoforms of UT-A and UT-B urea transporters. The different boxes represent regions of hydrophobic amino acids. The black lines show coding sequences which are common, while the red lines show coding sequences that are unique to that particular isoform (i.e. derived from novel exons) (adapted from Smith, 2009).

Table 1.

The nomenclature of all the currently identified members of the SLC14A urea transporter family

Biochemistry and genetics

The proposed basic structure for both UT-A and UT-B urea transporters consists of 10 transmembrane spanning domains (TMDs), a large extracellular loop containing an N-glycosylation site, plus intracellular amino and carboxy terminals (Olives et al., 1994) – see Figure 4. The main exception to this 10 TMDs structure is UT-A1, which is proposed to have 20 TMDs, and is basically UT-A2 and UT-A3 combined by a 73-amino acid central linking loop. This central loop region of UT-A1 contains serine 486, which is responsible for protein kinase A (PKA) activation of this isoform (Mistry et al., 2010). A recent paper also reports that the N-terminal 81-amino acids in rat UT-A1 are required for transport activity (Huang et al., 2010a), although further investigation is still required to fully understand its precise role. Interestingly, bacterial homologues of kidney urea transporters have recently been used to show that urea transporters function in the plasma membrane as multimers, for example, as dimers in Actinobacillus pleuropnemoniae (Raunser et al., 2009) or as trimers in Desulfovibrio vulgaris (Levin et al., 2009).

Figure 4.

Schematic diagram representing topology of two Slc14a facilitative urea transporters: UT-A1 and UT-B1. Various important residues are highlighted in each transporter: the asparagine (Asn) residues known to be important in glycosylation; the serine residues (Ser) known to be involved in the phosphorylation events that regulate transporter function; the cysteine (Cys) residues important in targeting the protein to the plasma membrane.

As mentioned, urea transporters are N-linked glycosylated proteins that have a unique pattern of hydrophobicity, as shown originally for rabbit UT-A2 (You et al., 1993). Rat UT-A1 has two glycosylated versions (97 & 117 kDa) that are both deglycosylated to an 88 kDa core 929-amino acid protein (Bradford et al., 2001). We now know that glycosylation of UT-A1 at two sites (Asn 279 and Asn 742) is required in order for correct trafficking of the protein to the plasma membrane to occur in response to vasopressin (Chen et al., 2006). The other main UT-A isoforms are also known to have N-linked glycosylations, such as the 55 kDa glycosylated rat UT-A2 protein (Wade et al., 2000). The exact nature of the glycosylations can vary between species however. For example, rat UT-A3 has 44 & 67 kDa glycosylated forms that are deglycosylated down to 40 kDa (Terris et al., 2001), while mouse glycosylated UT-A3 is a 45–65 kDa continuous smear that again shifts to a 40 kDa core protein (Stewart et al., 2004).

The vital role of the human UT-B1 N-terminal in protein targeting to the plasma membrane has been clearly shown (Lucien et al., 2002). Double mutation of the cysteine residues Cys 25 and Cys 30 prevented successful localization of hUT-B1 to the plasma membrane (Lucien et al., 2002). UT-B urea transporters are also known to have N-linked glycosylations. Human UT-B1 gives a 46–60 kDa glycosylated protein that deglycosylates to 36 kDa in red blood cells (Olives et al., 1995), while UT-B1 in human kidney gives a 41–54 kDa signal (Timmer et al., 2001). This tissue-specific variation in the extent of glycosylation is characteristic of urea transporter proteins and occurs in other species, for example, rat glycosylated UT-B1 protein is 32 kDa in brain but 45–55 kDa in kidney (Trinh-Trang-Tan et al., 2002). The glycosylation of human UT-B1 all occurs at the Asn 211 N-linked glycosylation site in the extracellular loop (Sidoux-Walter et al., 2000). Interestingly, evidence from oocyte expression studies suggests that mutating this site surprisingly does not affect either membrane trafficking or function (Lucien et al., 2002). However, it should be noted that UT-B glycosylation function may differ in a mammalian system, hence the precise role it plays has yet to be fully determined.

Single-amino acid mutations in human UT-A2 (Val/Ile 227 or Ala/Thr 357) have been associated with a decrease in blood pressure (Ranade et al., 2001). These UT-A2 mutations have been linked with increased efficacy of the anti-hypertensive drug nifedipine (Hong et al., 2007) and, more recently, shown to be associated with metabolic syndrome (Tsai et al., 2010). Jk null individuals have mutations in UT-B1 and therefore have red blood cells that lack functional urea transporters. For example, the first recorded mutations were shown to cause the skipping of exon 6 or 7 during the transcription process, producing non-functional UT-B1 proteins that did not reach the plasma membrane (Lucien et al., 1998). However, various Jk null polymorphisms occur in different populations around the world – see Table 2. Many of these mutations have functional consequences, such as the S291P mutation, whose failure to be expressed in red blood cells explains the Finnish Jk null phenotype (Sidoux-Walter et al., 2000). However, some mutations have no functional effect. For example, the G838A mutation (Asp280 to Asn 280) is a common polymorphism (i.e. A or B Kidd allele) in human UT-B1, but does not affect transport function (Lucien et al., 2002).

Table 2.

A list of the various mutations in SLC14A1 (UT-B) urea transporters found in different populations around the world

Population	SNP/mutation	Consequence	Reference
Polynesian	Splice site mutation	Exon 6 missing	(Irshaid et al., 2000)
Finnish	S291P	Prevents efficient trafficking to membrane	(Sidoux-Walter et al., 2000)
English	Genomic deletion	Exons 4, 5 missing	(Irshaid et al., 2002)
Swiss	T194Stop	Truncated exon 7	(Irshaid et al., 2002)
Chinese	Splice site mutation	Exon 6 missing	(Meng et al., 2005)
American
Caucasians	G68Stop	Nonsense mutation	(Wester et al., 2008)
Spanish	I262Stop	Nonsense mutation
African	T319M	Missense mutation

SNP, single-nucleotide polymorphism.

Pharmacology

Using a fluorescent based enzyme-linked immunosorbent assay, it was estimated that UT-A1 transporters in the rat renal inner medullary collecting duct (IMCD) had a turnover number of 100 000 urea molecules per second and that there were ∼5 million copies of UT-A1 per IMCD cell (Kishore et al., 1997). More recently, stopped flow fluorometry measurements of mouse UT-A transporters expressed in Xenopus oocyte plasma membranes showed transport rates of 46 000 and 59 000 urea molecules per second per protein for UT-A2 and UT-A3 respectively (MacIver et al., 2008).

Two reports have independently estimated that there are 14 000 copies of UT-B1 per human red blood cell (Masouredis et al., 1980; Mannuzzu et al., 1993). It has therefore been calculated that human UT-B1 transporters have a turnover of up to 6 000 000 urea molecules per second (i.e. at least ∼60-fold greater than UT-A transporters) (Mannuzzu et al., 1993). Although not as impressive as the Aquaporin-1 (AQP1) water channel, which transports 3 billion water molecules per second (Gade and Robinson, 2006), one might argue that with this channel-like transport rate the correct terminology for these proteins is ‘UT-B channels’. Interestingly, the structure of a bacterial homologue urea transporter has recently been shown to consist of two oppositely orientated homologous halves, known as ‘an inverted repeat motif’ and operate by a channel-like mechanism (Levin et al., 2009). Furthermore, this ‘inverted repeat motif’ is also found in channels that transport other neutral solutes, such as ammonia and water, for example, AQP1 itself (Murata et al., 2000).

Urea transporters have a low affinity for urea, for example, rabbit UT-A2 Km > 200 mM (You et al., 1993), and so are not saturated by the levels of between 2 and 10 mM of urea generally found in mammalian blood. Initial reports suggested that UT-A proteins are highly selective for urea and do not transport water (Hill et al., 2005). Recent stopped flow fluorometry measurements of transporters expressed in Xenopus oocyte plasma membranes confirmed that mouse UT-A2 and UT-A3 did not transport water, ammonia or urea analogues, such as formamide, acetamide, methylurea and dimethylurea (MacIver et al., 2008).

The observation that rat UT-B1, as well as facilitating the movement of urea, also transports water (1.4 × 10⁻¹⁴ cm³·s⁻¹ per channel cf. 0.1 × 10⁻¹⁴ cm³·s⁻¹ per channel for rat UT-A2) (Yang and Verkman, 1998) was initially controversial (Sidoux-Walter et al., 1999). However, stopped flow light scattering measurements using mouse erythrocytes from knockout (KO) models confirmed UT-B1 has a water permeability (Pf) of 7.5 × 10⁻¹⁴ cm³·s⁻¹ per channel (Yang and Verkman, 2002). This promiscuous nature of the UT-B channel has now been confirmed by the fact it also transports formamide, acetamide, methylurea, methylformamide, ammonium carbonate and acrylamide (Zhao et al., 2007).

In humans, this relatively unselective UT-B mediated transport is shown by the fact that red blood cells from Jk null individuals (i.e. ones lacking UT-B1) have a decreased permeability to both urea and water (Meng et al., 2005). Indeed, it has been reported that a known urea transporter inhibitor, the analogue thiourea, can actually go through human UT-B1 (Km = 40 mM), but not through human UT-A2 (Martial et al., 1996).

Urea transport in perfused rat IMCD was originally shown to be inhibited by 250 µM phloretin and 200 mM of the urea analogues thiourea, methylurea and acetamide (Chou and Knepper, 1989). The K_1/2-value for thiourea inhibition was 27 mM, which was unaffected by vasopressin (see regulation section below) even though transport increased fourfold (Chou and Knepper, 1989). Rat IMCD urea transport was also inhibited by dimethylurea and phenylurea (Zhang and Verkman, 1990). Rat UT-A1, located in the IMCD, was later confirmed to indeed be inhibited by phloretin and urea analogues (Shayakul et al., 1996), as well as by thionicotinamide (Frohlich et al., 2004). Importantly, all other known UT-A isoforms are also inhibited by phloretin – including UT-A2 (You et al., 1993), UT-A3 and UT-A4 (Karakashian et al., 1999), UT-A5 (Fenton et al., 2000) and UT-A6 (Smith et al., 2004).

Human UT-B1 expressed in oocytes was initially confirmed to be inhibited by phloretin and thiourea (Olives et al., 1994). In agreement with this, rat UT-B1 function in vasa recta was also inhibited by thiourea, methylurea, acetamide and phloretin (Pallone, 1994), with a K_1/2-value for thiourea of 19 mM, as well as dimethylurea (Yang and Verkman, 1998). More recently, mouse UT-B1 mediated transport in red blood cells has been shown to be inhibited by more than 60% by dimethylurea, acrylamide, thiourea and methylformamide (Zhao et al., 2007). As expected, both bovine UT-B1 and UT-B2 isoforms were also inhibited by both phloretin and thionicotinamide (Stewart et al., 2005). Interestingly, it has been suggested that UT-B proteins are more sensitive to phloretin inhibition than UT-A transporters – for example, IC₅₀ phloretin: human UT-B1 = 75 µM versus human UT-A2 230 µM (Martial et al., 1996). In addition, human red blood cell urea transport is also inhibited by mercurial compounds, such as p-chloromercuribenzenesulphonate (pCMBS) (Mannuzzu et al., 1993), because human UT-B1 is pCMBS-sensitive (Olives et al., 1994; Lucien et al., 2002) with an IC₅₀ of 150 µM (Martial et al., 1996). In direct contrast, human UT-A2 is not inhibited by pCMBS (Olives et al., 1996), while rat kidney UT-B1 is indeed pCMBS-sensitive (Pallone, 1994).

Finally, a human red blood lysis assay has been used to investigate the potential inhibitory effects of over 50 000 compounds on UT-B1 facilitated acetamide transport (Levin et al., 2007). This research discovered 30 specific inhibitors that selectively inhibited UT-B but not UT-A transporters, while also having no effect on AQP1 (Levin et al., 2007). These compounds were from the phenylsulphoxyoxazole, benzensulphonanilide, phthalazinamine and aminobenzimidazole classes (Levin et al., 2007).

Some members of the aquaporin water channel group are also permeable to urea. For example, AQP7 is urea permeable and was originally localized in the testis (Ishibashi et al., 1997). AQP7 has also been localized to the proximal tubule in rat kidney (Ishibashi et al., 2000) and is hence located in a different nephron segment to the UT-A urea transporters. Although this implies that AQP7 could play a role in renal urea transport, a mouse AQP7 KO model does not actually display a urinary concentrating defect and shows the renal function of AQP7 to be the reabsorption of water and glycerol (Sohara et al., 2005). AQP9 is another urea permeable aquaporin, which was originally located in human leukocytes (Ishibashi et al., 1998). However, AQP9 KO mice have no change in their plasma urea levels and results showed AQP9 to be involved in glycerol metabolism in the liver (Rojek et al., 2007). These results strongly suggest aquaporins do not play a significant physiological role in urea transport.

Urea has also been reported to pass through other co-transporters, such as the sodium-dependent glucose transporter (SGLT1) (Leung et al., 2000). However, the actual urea transport rate through SGLT1 was P_urea= 1.2 × 10⁻⁷ cm·s⁻¹ (Leung et al., 2000), which is several orders of magnitude lower than P_urea values for urea transporters, for example, UT-A2 = 4.5 × 10⁻⁵ cm·s⁻¹ (You et al., 1993). It therefore again seems unlikely that these co-transporters have a significant physiological role in transporting urea.

Distribution

In the original research that cloned the first urea transporter, rabbit UT-A2 was discovered in the kidney, but also detected in the colon (You et al., 1993). Human UT-A1 was detected in the kidney, as it was for other species such as rat (Nielsen et al., 1996) and mouse (Fenton et al., 2002b). Similarly, UT-A2 (Olives et al., 1996), UT-A3 (Stewart et al., 2004) and UT-A4 (Karakashian et al., 1999) are also all found in the kidney of various species. Human UT-A3 protein has yet to be investigated, although a cDNA clone has been identified (Smith and Fenton, 2006) and an appropriately sized 2.4 kb transcript detected in human kidney medullary RNA (Bagnasco et al., 2001). In contrast to these renal locations, mouse UT-A5 is found in the testis (Fenton et al., 2000) and human UT-A6, a 235-amino acid protein including a novel 5a exon and unique 19-amino acid C-terminal, has been located in the colon (Smith et al., 2004). Other tissues in which UT-A proteins have been found include heart (Duchesne et al., 2001), cochlea (Kwun et al., 2003), placenta (Damiano et al., 2006), brain and liver (Fenton et al., 2002b).

Human UT-A1 was located in the renal IMCD (Bagnasco et al., 2001). In rats, UT-A1 was present in the terminal portions of the IMCD (Nielsen et al., 1996), while UT-A2 was found in the late part of descending thin limbs of short loops of Henle and the inner medullary part of descending thin limbs of long loops of Henle (Shayakul et al., 1997). In mice, UT-A1 was again found in terminal IMCD, and UT-A2 located in short (type I) and long (type 3) thin descending limbs of the loops of Henle (Fenton et al., 2002b). UT-A3 was also only located in the IMCD (Stewart et al., 2004). Interestingly, mouse UT-A1 and UT-A3 co-localized with AQP2 in principal cells in IMCD (Fenton et al., 2006). Mouse UT-A5 protein is localized to the peritubular myoid cells of the seminiferous tubules in the testis (Fenton et al., 2000).

In IMCD cells, UT-A1 was mainly located at the apical membrane region (Nielsen et al., 1996), while UT-A3 was mainly located at the basolateral membrane region (Shayakul et al., 2001; Stewart et al., 2004). However, it should be noted that UT-A1 is capable of going to the basolateral membrane (Frohlich et al., 2004) and UT-A3 has been reported in one study at the apical membrane after vasopressin treatment (Blount et al., 2007). UT-A2 was located on both apical and basolateral membranes in rats (Lim et al., 2006) and mice (Fenton et al., 2002b; Potter et al., 2006). The subcellular location of UT-A4, UT-A5 and UT-A6 remain unclear at present.

In contrast to the mainly renal location of UT-A transporters, UT-B protein has been detected in numerous tissues. Initially, a triated urea analogue 1-(3-azido-4-chlorophenyl)-3methyl-2-thiourea ([³H]MeACPTU) was used as a probe to photolabel human red blood cell urea transporters (Neau et al., 1993). This technique detected a 40 kDa polypeptide in red blood cells, later called UT-B1, which was absent in Jk null red blood cells (i.e. showing that the Kidd antigen was a urea transporter) (Neau et al., 1993). This UT-B1 transporter was initially detected in human bone marrow, erythrocytes and also the kidney, as 2.5 kb and 4.7 kb transcripts (Olives et al., 1994). Rat UT-B1 was found in brain, spleen, kidney and testis, compared with rat UT-A2 which was only located in the kidney (Promeneur et al., 1996). Rat UT-B1 was also found in urinary tract epithelia, thymus and lung (Tsukaguchi et al., 1997). UT-B1 proteins were also detected in the heart (Meng et al., 2009), bladder and gastrointestinal tract in mice (Lucien et al., 2005), human colon (Inoue et al., 2004), rat colon (Inoue et al., 2005) and throughout sheep gastrointestinal tract, including the salivary glands (Ludden et al., 2009). Bovine UT-B1 and UT-B2 isoforms are both present in the bovine rumen (Stewart et al., 2005), while UT-B1 has also been reported in the cochlea of rats (Kwun et al., 2003).

UT-B1 localized to non-fenestrated endothelial cells in descending vasa recta of human kidney (Xu et al., 1997; Timmer et al., 2001) and rat kidney (Pallone, 1994). In addition to its location in descending vasa recta, mouse UT-B1 has been reported in the renal proximal tubule and papillary surface epithelium (Jung et al., 2003). UT-B1 was prevalent in the colonic epithelial cells in the human ascending colon (Collins et al., 2010), while bovine UT-B protein was present in rumen epithelial layers (Stewart et al., 2005). UT-B1 was also found in the Sertoli cells of the seminiferous tubules in the testis (Tsukaguchi et al., 1997).

UT-B1 was located on both apical and basolateral membranes of descending vasa recta in both rats (Lim et al., 2006) and mice (Jung et al., 2003). UT-B1 was also located on both apical and basolateral membranes in rat testis Sertoli cells (Fenton et al., 2002c). In agreement with these findings, bovine ruminal UT-B protein has been localized to both apical and basolateral membranes of epithelial cells (Stewart et al., 2005; Simmons et al., 2009). In addition, functional evidence for bovine UT-B2, when over-expressed in a Madin–Darby canine kidney (MDCK) cell line, also showed that this isoform is capable to going to both apical and basolateral membranes (Tickle et al., 2009).

Regulation

IMCD plasma membranes limit the rate of trans-epithelial urea transport (Star, 1990) and it is known that the antidiuretic hormone vasopressin stimulates an increase in membrane urea permeability (Nielsen and Knepper, 1993). Vasopressin binds to V2 receptors on the basolateral membrane of IMCD cells, increasing cAMP levels and stimulating PKA – hence vasopressin stimulation of urea transport was prevented by PKA inhibition (Zhang and Verkman, 1990).

Vasopressin regulates rat UT-A1 function in the short term (Terris et al., 1998), and its action is prevented by PKA inhibition (Frohlich et al., 2006). Vasopressin activates PKA rapidly (within 5 to 10 min.) and stimulates phosphorylation of UT-A1 protein (Zhang et al., 2002). This causes an increase in urea transport by increasing UT-A1 abundance at the apical plasma membrane in IMCD cells (Klein et al., 2006a). This regulation of UT-A1 activity and membrane accumulation involves rapid phosphorylation of serine 486 (Klein et al., 2010). Other UT-A1 amino acids reported to be phosphorylated in response to vasopressin include serine 84 (Hwang et al., 2010) and serine 499 (Blount et al., 2008) (see Figure 4).

Vasopressin also stimulates UT-A3 function (Stewart et al., 2007), via a process again involving the stimulation of both phosphorylation and membrane accumulation (Blount et al., 2007). Vasopressin firstly produces PKA-dependent stimulation of UT-A3 transporters in the basolateral membrane (Stewart et al., 2009). Secondly, vasopressin also then stimulates casein kinase II-dependent trafficking of additional UT-A3 transporters to the basolateral membrane, via a process that is dependent on both protein kinase C and calmodulin (Stewart et al., 2009).

Both UT-A1 (Shayakul et al., 1996) and UT-A3 (Karakashian et al., 1999; Stewart et al., 2007) are stimulated by increased levels of cAMP, although the increase in human UT-A1 function is modest (Bagnasco et al., 2001). PKA also stimulates the transport function of human UT-A6 (Smith et al., 2004), but not UT-A2 (Fenton et al., 2002b). In contrast, when mouse UT-A2 was expressed in a stable MDCK cell line it was acutely regulated by vasopressin, cAMP and calcium (Potter et al., 2006). Increased tonicity in rat IMCD suspensions increased plasma membrane localization of UT-A1 and UT-A3 (Blessing et al., 2008). Indeed, hypertonicity stimulates urea transport through a protein kinase C-mediated phosphorylation event (Wang et al., 2010). Angiotensin II also increases vasopressin-stimulated urea transport in rat IMCD via a protein kinase C-dependent effect on UT-A1 (Kato et al., 2000).

As well as insertion, regulation of urea transporter removal from the plasma membrane also occurs. For example, ubiquitination regulates the plasma membrane expression of the three main renal UT-A urea transporters: UT-A1, UT-A2 and UT-A3 (Stewart et al., 2008). Ubiquitination and subsequent degradation of UT-A1 by the proteasome pathway involves MDM2 (murine double minute) E3 ubiquitin ligase (Chen et al., 2008). UT-A1 is internalized by a dynamin-dependent mechanism, which is mediated by both caveolae and clathrin coated pit pathways (Huang et al., 2010b). In contrast to all this, UT-B proteins have not been reported to be acutely regulated. For example, bovine bUT-B2 is constitutively activated when over-expressed in an MDCK cell line and is not stimulated by cAMP, calcium, vasopressin or protein kinases (Tickle et al., 2009).

A number of factors have been shown to influence UT-A gene expression – including hydration state, dietary protein, glucocorticoids and mineralocorticoids. There are two promoters in the UT-A gene: one controlling UT-A1/UT-A3 expression (called UT-Aα or UT-A promoter I) and one controlling UT-A2 (called UT-Aβ or UT-A promoter II). In mice, the UT-Aα promoter is cAMP and tonicity sensitive, while the UT-Aβ promoter is only cAMP sensitive (Fenton et al., 2002a), hence both UT-A2 and UT-A3 mRNA levels are increased in water-deprived mice (Fenton et al., 2002a). Glucocorticoids have been reported to inhibit transcription and expression of UT-A1 and UT-A3 by decreasing UT-A promoter I activity in rats (Peng et al., 2002), with similar findings for the UT-Aα promoter in mice (Fenton et al., 2006). Interestingly, rat kidney UT-A1 has also been shown to be down-regulated by aldosterone via the mineralocorticoid receptor (Gertner et al., 2004). In contrast, the down-regulation of UT-A1 by glucocorticoids in the same study was not via the mineralocorticoid receptor, but most likely through the glucocorticoid receptor (Gertner et al., 2004). The implications of these findings are that both glucocorticoid agonists used for inflammatory conditions and mineralocorticoid antagonists used for diuresis are likely to alter SLC14A2 transporter expression in the human kidney.

Glucocorticoids have been shown to have no effect on the UT-A promoter II (i.e. UT-Aβ) and so did not alter UT-A2 expression levels (Peng et al., 2002). However, UT-A2 expression in the rat kidney can be influenced by other factors. UT-A2 has a 4.0 kb RNA transcript that is regulated by dietary protein content, while a 2.9 kb transcript is responsive to hydration state (Smith et al., 1995), thus meaning fluid and nitrogen balance can be regulated independently. It has also been reported that tonicity responsive regulation of rat UT-A1 and UT-A3 is mediated by the TonE/TonEBP pathway (Nakayama et al., 2000).

Investigation of the regulation of UT-A protein level expression also confirms many of these findings. For example, osmolality and urea concentration regulate UT-A1 expression (Terris et al., 1998), glucocorticoids down-regulate UT-A1 in rat terminal IMCD (Naruse et al., 1997) and rat UT-A2 abundance is regulated by the antidiuretic hormone vasopressin (Wade et al., 2000). Interestingly, one study has shown that water deprivation increased UT-A3 expression, but decreased UT-A1, in the rat IMCD (Lim et al., 2006). The mechanisms for this differential effect on UT-A1 and UT-A3 are not yet clearly understood.

To date, little is known about the factors that influence expression of the UT-B gene. However, research at the protein level suggests a number of factors may be involved. Long-term vasopressin infusion has been shown to greatly decrease UT-B1 in rat kidney (Trinh-Trang-Tan et al., 2002), while water derivation increased UT-B staining intensity of descending vasa recta (Lim et al., 2006). Rat UT-B1 renal and intestinal protein expression is also regulated by low protein and urea diets (Inoue et al., 2005). Furthermore, dietary intake has shown to alter abundance and localization of UT-B protein in the rumen of two species – cows (Simmons et al., 2009) and sheep (Ludden et al., 2009). Further research is now required to clarify how these dietary effects on UT-B proteins are actually regulated.

Pathology and clinical significance

Major dysfunction of UT-A protein has not been reported in humans. However, specific KO mouse models have been investigated and the lack of major dysfunction may well be due to the compensatory effects of up-regulating other urea transporter isoforms. For example, it has been reported that there are increased levels of UT-A2 in mice lacking UT-B (Klein et al., 2004). The UT-A1/3 KO mice, lacking both the collecting duct urea transporters UT-A1 and UT-A3, have a severe concentrating defect when fed a normal 20% protein diet (Fenton et al., 2004) – with increased fluid consumption, increased urine flow and decreased urine osmolality. More recently, these UT-A1/3 KO mice have also been shown to have increased blood pressure, plus increased chance of hydronephrosis and renal pelvic reflux compared with wild-type controls (Jacob et al., 2008). UT-A2 KO mice have a much milder concentrating defect that is only detectable when on a low protein diet – suggesting UT-A2 mainly plays a role in maintaining inner medulla urea concentration when urea production is low (Uchida et al., 2005). While these findings confirm that UT-A function is predominantly concerned with the renal urinary concentrating mechanism, the implications of UT-A mutations to the health of the human population is not yet fully understood.

As previously mentioned, humans lacking red blood cell UT-B1 protein do exist (i.e. Jk null individuals). These patients displayed moderate maximal concentrating ability (∼800 mOsm vs. ∼1000–1100 mOsm in controls), that is, they have only a mild concentrating defect (Sands et al., 1992). As expected, Jk null individuals have red blood cells that have a selective defect in urea transport (Olives et al., 1995). Jk null individuals are rare within a population, for example, five out of 20 163 Thai individuals (i.e. 0.0002%) (Deelert et al., 2010). However, after immunization anti-Jk3 forms and therefore it can be difficult to find donors suitable for these Jk null individuals (Deelert et al., 2010).

UT-B KO mice have a 45-fold reduction in red blood cell urea permeability, 50% increase in urine output and 30% decrease in urine osmolality (i.e. mild, urea-selective concentrating defect (Yang et al., 2002). Interestingly, the UT-B KO mice gain less weight than wild-type littermates, suggesting intestinal UT-B plays important role in weight gain (Yang and Bankir, 2005). Although it cannot be ruled out that this reduced weight gain was simply due to adverse effects on general health, the KO mice displayed no obvious signs of ill-health compared with the wild-type littermates. This proposed nutritional role for intestinal UT-B protein is linked to its involvement in the UNS process that helps maintain the symbiotic relationship between mammals and their intestinal bacteria (Stewart and Smith, 2005) – see Figure 2. Recent advances in our understanding of the importance of colonic bacteria populations to human health and nutrition strongly indicate that the UNS process may play a vital role in our well-being. Urea hydrolysis in the human gastrointestinal tract is regulated by diet (Fouillet et al., 2008) and one may predict that this involves regulation of the UT-B1 proteins mediating trans-epithelial urea transport in the human ascending colon (Collins et al., 2010) – in a manner similar to the dietary effects observed on ruminal UT-B transporters (Simmons et al., 2009). Although the precise role of human colonic urea transporters still remain to be determined, it is intriguing to speculate that alteration of their normal function could alter bacterial populations and potentially contribute to disease states of the human colon.

Numerous factors have been shown to influence renal UT-A transporter expression. Excess of glucocorticoids produces a decrease in rat renal UT-A1 and UT-A3 abundance, and may explain impaired urinary concentrating capacity in human patients suffering from Cushing syndrome (Li et al., 2008). Renal urea transporters are down-regulated by severe inflammation, such as occurs with sepsis-induced acute renal failure (Schmidt et al., 2007), while UT-A1 is down-regulated in adriamycin-induced nephritic syndrome (Fernandez-Llama et al., 1998). It has also been suggested that expressional changes in kidney UT-A protein may be responsible for reduced concentrating ability of mammalian kidney as ageing occurs (Combet et al., 2003).

Rat UT-A1 in kidney medulla is down-regulated in angiotensin II-induced hypertension (Klein et al., 2006b), while UT-A protein was increased in kidney of streptozotocin-induced diabetic rats compared with control rats (Bardoux et al., 2001). Importantly, increases in UT-A1 (but not UT-B1) in renal medulla during diabetes mellitus may help limit fluid loss during this disease (Kim et al., 2003). Lastly, there is increased UT-A1 and UT-A3 expression, and resulting function, in the IMCD of salt-sensitive rats (cf. salt-resistant rat) (Fenton et al., 2003). There are also disease and environmental-related changes in non-renal UT-A transporters. For example, pre-eclampsia appears to increase phloretin-sensitive urea transport and UT-A urea transporter abundance in the placenta (Damiano et al., 2006), while UT-A protein abundance in the heart is increased during uraemia, hypertension and heart failure (Duchesne et al., 2001).

There are a limited number of reports concerning changes in both UT-A and UT-B proteins. Decreases in UT-A1, UT-A3 and UT-B1 in rats with ureteral obstruction may explain reduction in the urinary concentrating ability in these animals (Li et al., 2004). Chronic renal failure significantly decreases UT-A1, UT-A2 and UT-B1 in the rat kidney, and also decreased UT-B1 in rat brain (Hu et al., 2000). A decrease in UT-A1 and UT-B1 abundance in the renal inner medulla has been reported in lithium-fed rats (Klein et al., 2002). However, UT-A1 recovered to normal levels 14 days after cessation of lithium administration (Blount et al., 2010). Treatment with the immunosuppressant drug cyclosporine reduces UT-A2, UT-A3 and UT-B1 levels in the kidney, explaining the impaired urine concentrating ability that is a main feature of cyclosporine-induced nephropathy (Lim et al., 2004). The diuretic furosemide also moderately decreases UT-B1 abundance in rat kidney (Trinh-Trang-Tan et al., 2002), which may have implications for its long-term use. Finally, there has also been a reported increase in UT-B1 abundance in the ageing rat brain (Trinh-Trang-Tan et al., 2003). It is not yet known what might cause this change or whether it leads to dysregulation in cerebral function (Trinh-Trang-Tan et al., 2003).

Conclusion

The family of SLC14A facilitative urea transporters play an important role in two major physiological processes, namely the urinary concentration mechanism and UNS. These facilitative transporters are found in specific locations within different tissues and are derived from two distinct genes: UT-A and UT-B. Generally, the UT-A and UT-B classes of urea transporters have a similar function, topology and basic structure. However, they do display marked differences in terms of substrate specificity, transport rates, inhibition, gene regulation, functional regulation and tissue localization. Although there are no major pathologies linked with urea transporter dysfunction, understanding these proteins has important clinical implications, especially within the context of renal disease.

Glossary

Abbreviations

AQP

aquaporin

IMCD

inner medullary collecting duct

KO

knockout

MDM2

murine double minute

MeACPTU

1-(3-azido-4-chlorophenyl)-3 methyl-2-thiourea

pCMBS

p-chloromercuribenzenesulphonate

PKA

protein kinase A

SGLT1

sodium-dependent glucose transporter

tDL

thin descending limbs

TMD

transmembrane spanning domain

UNS

urea nitrogen salvaging

UT

urea transporter

Conflicts of interest

There are no conflicts of interest to report regarding this article.