Plumbing in the embryo: developmental defects of the urinary tracts


Maxime Bouchard, Goodman Cancer Centre, Department of Biochemistry, McGill University, Montreal, 1160 Pine Avenue West, Montreal, Quebec, Canada, H3A 1A3.
Tel.: 1-514-398-3532;
fax: 1-514-398-6769;


Kidney and urinary tract malformations are among the most frequent developmental defects identified in newborns. Ranging from asymptomatic to neonatal lethal, these malformations represent an important clinical challenge. Recent progress in understanding the developmental origin of urinary tract defects in the mouse and other animal models suggests a new framework for the interpretation of these defects in humans. Gene inactivation studies in mice provided invaluable information on the formation of the Wolffian duct, a central component of embryonic renal development, on ureter and kidney induction as well as on distal ureter maturation. All three developmental processes are crucial for normal urinary tract morphogenesis. A failure to complete these developmental steps is responsible for a spectrum of kidney and urinary tract malformations including renal agenesis, renal dysplasia, vesicoureteral reflux, hydroureter, hydronephrosis and ureterocele. Surprisingly, distal ureter maturation, the process by which the ureter is displaced from the Wolffian duct to its final position within the bladder wall, has only recently been characterized at the morphological level. Anomalies in this process are emerging as a major source of urinary tract developmental defects. This review is aimed at bridging the current knowledge on the morphological and molecular events identified in the mouse, together with clinical observations of urinary tract malformation in humans.

Section Editor:
Roderick R. McInnes, email:
Jacques L. Michaud, email:

Developmental defects affecting the urinary tracts are highly heterogeneous, ranging from lethal conditions, such as bilateral renal agenesis (BRA), to asymptomatic abnormalities, as is frequently observed for cases of duplex ureter systems. These malformations are part of the disease group congenital anomalies of the kidney and urinary tract (CAKUT). In general, urinary tract defects occur in as many as 1:100 live births and constitute the most frequent cause of chronic kidney disease in children (1). Their treatment varies widely depending on the type of malformation and ranges from simple prophylactic antibiotic treatment to prevent urinary tract infections to surgical reconstruction of the vesicoureteral junction (1). The most severe cases are, however, not viable and lead to therapeutic termination of pregnancy. The identification of animal models that mimic human urinary tract malformations has helped to elucidate the key morphological and molecular events underlying urinary tract morphogenesis. These can be grouped around three major developmental steps: Wolffian (or nephric) duct formation, ureter budding and distal ureter maturation. Although most attention has been given to the mechanisms of ureteric bud (UB) induction, the process of distal ureter maturation is now emerging as another crucial processes in urogenital system morphogenesis.

Thanks to recent advances, it is now possible to link clinical conditions in infants and adults to precise events during early development and to pinpoint the molecular cascades responsible for these defects. This review focuses on the morphological, molecular and clinical aspects of two central manifestations of urinary tract defects: ureter agenesis and ureter ectopia.

Installing the pipes:

Wolffian duct development and ureter budding

Morphological events underlying ureter formation

The first critical step in genitourinary tract formation is the formation of the pro/mesonephros in the intermediate mesoderm (IM; Fig. 1a). This is a highly coordinated process that involves the formation of a central epithelial duct, the Wolffian (or nephric) duct (WD) from mesodermal precursor cells. The WD elongates along a precise path down the trunk of the embryo until it reaches and fuses with the cloaca, primordium of the bladder and urethra (Fig. 1b–d). Concomitantly, the Wolffian duct induces mesonephric tubule formation in the adjacent mesenchyme (Fig. 1c). In addition to its role as transient embryonic kidney, the pro/mesonephros is central to the development of the genital tracts, as the Wolffian duct is the precursor of the epididymis and vas deferens in males. The Wolffian duct additionally guides the elongation of the Müllerian duct, the precursor of the female oviduct, uterus and upper vagina. The Wolffian and Müllerian ducts will coexist in the embryo until secreted signals from the differentiating gonads promote the survival of only one of the two systems [reviewed in (2)].

Figure 1.

Genitourinary tracts formation in the mouse embryo. (a) Pro/mesonephros development is initiated in the intermediate mesoderm (IM). At day 9.0 of development (E9.0), the Wolffian duct (WD) is formed by mesenchymal–epithelial transition and induces mesonephric tubule (MT) formation as it elongates towards the cloaca. (b) Anatomical location of the urogenital system. At the level of the hindlimb bud, the metanephric mesenchyme (MM) (in pink) induces ureteric bud formation at E10.5. Dotted square indicates region shown in E. (c) Schematic view of the pro/mesonephros and early metanephros at E11.5. After this stage, the Wolffian duct gradually becomes part of the genital tract (epididymis, vas deferens and seminal vesicles). (d) The process of ureter formation initially requires that the Wolffian duct reaches and fuses with the cloaca, primordium of the bladder and urethra. (e) Signals emanating from the MM induce the formation of the ureteric bud as a diverticulum of the Wolffian duct. (f) The newly formed ureter (Ur) grows into the MM and branches to form the collecting duct system of the MM. The section of the ureter located between the MM and the Wolffian duct differentiates into ureter. The Wolffian duct segment located between the ureter (red) and the cloaca (gray) constitutes the CND. This short duct is eliminated by apoptosis during the process of distal ureter maturation.

Once the Wolffian duct has fully elongated and fused with the cloaca, ureter budding is induced in the caudal part of the Wolffian duct. This process occurs at the level of the hindlimb where the metanephric mesenchyme (MM), a specialized region of the IM, sends signals to the Wolffian duct to initiate ureteric bud formation (Fig. 1b,e) (3, 4). This process involves major cell shape changes as well as increased cell proliferation. The emerging bud invades the MM and branches to form the ‘T’-stage kidney primordium (Fig. 1f). At this point in development, the mesenchyme located between the Wolffian duct and the MM already initiates the regionalization of the UB stalk into ureter [reviewed in (5)], while the part of the UB located within the MM will branch several times to form the collecting duct system of the metanephric kidney. As the ureter branches, the newly formed ureter tips induce waves of nephron differentiation in the adjacent mesenchyme, thus forming the mature metanephros. Metanephric kidney development and associated malformations have been reviewed extensively elsewhere (3, 4, 6) and will not be discussed in detail here.

Molecular events underlying ureter formation

A number of genes have been identified that play an important role at different steps of Wolffian duct and UB formation. The first renal fate identity in the embryo is provided by the combined action of the transcription factor genes Pax2 and Pax8 in the IM (7). These genes are required and sufficient for the specification of the nephric lineage and for the formation of the pronephros by promoting mesenchymal–epithelial transition and cell survival in the IM (7). The mechanisms of Pax2 and Pax8 activation in the IM are still unclear but seem to involve signals from surrounding tissues (8–11) and may be under the transcriptional control of Osr gene family members (12). One of the key targets of Pax2/8 gene activity at this stage is the transcription factor gene Gata3, which is necessary for proliferation control of the Wolffian duct and for its guidance towards the cloaca (13). In turn, the elongation and maintenance of epithelial integrity of the Wolffian duct is under the control of the transcription factor gene Lim1 (14, 15). Due to their early role in Wolffian duct development, gene inactivation of either Pax2/8, Gata3 or Lim1 lead to a complete lack of genital tracts, ureters and kidneys (BRA). This condition is compatible with embryonic development, but not with post-natal life.

The next step of ureter budding has been extensively studied at the molecular level and is thus relatively well understood. The signaling pathway central to this process consists of the GDNF ligand secreted from the MM, which signals to the Ret/GFRα1 receptor complex expressed in the Wolffian duct epithelium. The importance of this pathway for ureter budding has been clearly demonstrated in both loss-of-function and gain-of-function experiments. Gene inactivation of GDNF, Ret or GFRα1 indeed results in complete failure of ureter budding in over 65% of mutant embryos (16–21). However, ectopic activation of Ret signaling by GDNF, using GDNF-soaked beads or transgenic animals, was sufficient to induce multiple ectopic ureteric buds along the Wolffian duct (22, 23).

The central role of the GDNF-Ret/GFRα1 pathway is further supported by the fact that most genes identified as being necessary for ureter budding are in fact regulators of GDNF or Ret expression (Table 1). Among them, Osr1 was recently identified as a regulator of transcriptional regulators Eya1, Six2, Pax2, Sall1 (12). These genes were previously identified as key regulators of GDNF expression in the MM (Table 1). Interestingly, some of their encoded proteins, namely Pax2 and Eya1, form a complex with Hox11 paralogs that directly bind and activate the Gdnf promoter (24).

Table 1.  Mouse models of renal agenesis
GeneType of proteinGenotypeDefects in humansReferences
  • ECM, extracellular matrix; HDR, hypoparathyroidism–deafness–renal anomalies syndrome.

  • a

    blebbed (bl).

  • b

    It has been suggested from transgenic mouse phenotypes.

Involved in Wolffian duct formation
 Pax2 and Pax8Transcription factorPax2−/−, Pax2−/−;Pax8−/−Renal coloboma syndrome(7, 36, 62)
 Gata3Transcription factorGata3−/−HDR syndrome(13, 39)
 Lim1Transcription factorLim1−/−HoxB7-Cre;Lim1lz/flox (14, 15, 63)
Involved in ureter budding
GDNF-Ret/GFRα1 components
 GDNFSecreted molecule, neurotrophin familyGDNF−/−Hirschprung disease(18–20, 64)
 RetReceptor tyrosine kinaseRet−/−Hirschprung disease(21, 65)
 GFRα1GPI-linked neurotrophin receptorGFRa1−/− (16, 17)
Factors required for GDNF expression
 Eya1Transcription factorEya1−/−Branchio-oto-renal(37, 66)
 Hox11 paralogousTranscription factorHoxa11−/−;Hoxc11−/−;Hoxd11−/− (67)
 Sall1Transcription factorSall1−/−Townes–Brocks syndrome(41, 68)
 Six2Transcription factorSix2−/− (69)
 Pax2Transcription factorPax2−/−Renal coloboma syndrome(70)
 Osr1Transcription factorOsr1−/− (12)
 Gdf11TGFb family proteinGdf11−/− (71)
 Fras1ECM proteinbl/blaFraser syndrome(40, 72, 73)
 NpntECM proteinβ-actin-Cre; NpntDex1/Dex1 (74)
 Integrinα8Cell surface ECM receptorIntegrin α8−/− (75)
Factors required for Ret expression
 Gata3Transcription factorGata3−/−, HoxB7-Cre;Gata3flox/floxHDR syndrome(25, 39, 76)
 β-cateninTranscriptional coactivator, Wnt pathwayHoxB7-Cre;β-catenin−/flox (26)
 Emx2Transcription factorEmx2−/− (28)
 Hs2stHeparan sulfate biosynthetic enzymeHs2st−/− (29)
 WT1Transcription factor, tumor suppresserWT1−/−Denys–Drash syndrome(31)
 Grip1Cytoplasmic multi-PDZ scaffolding proteinGrip1−/−Fraser syndromeb(32)

Gene inactivation studies have further identified a key role for Gata3 in regulating Ret expression in the Wolffian duct. Early on, Pax2/8-regulated Gata3 expression is indeed necessary for Ret activation (13). At a later stage, Gata3 expression in the Wolffian duct becomes dependent on β-catenin signaling. Accordingly, Ret expression is lost in both β-catenin and Gata3 conditional mutant mice (25, 26). These genetic experiments also showed that β-catenin expression and signaling are unaffected in Gata3 mutant embryos, thereby establishing that β-catenin acts upstream of Gata3. Together, the data therefore point to a β-catenin-Gata3-Ret regulatory cascade in the Wolffian duct (25). Ret expression in the Wolffian duct was also shown to depend on retinoic acid signaling from the surrounding mesenchyme (27). The relationship between retinoic acid signaling and the β-catenin-Gata3 cascade is, however, still elusive. Following ureter budding, the maintenance of Ret expression in the Wolffian duct and ureter seems to additionally require the action of the transcription factors Lim1 and Emx2 (14, 28). In accordance with a later role in Ret maintenance, the initial step of ureter budding is not affected in Emx2 mutant embryos. Instead, the ureter buds out and invades the MM but fails to branch, leading to a complete arrest in ureter and metanephric kidney development.

In addition to the molecules described above, a number of other signals have been identified as being critical for UB formation. These are the heparan sulfate synthesis protein Hs2st (29), the transcription factor Wt1 (30, 31), as well as the PDZ-domain protein Grip1 (32). Grip1 interacts with the extracellular matrix protein Fras1, involved in Fraser syndrome (Table 1) and is additionally necessary for Wt1 expression in the MM (32). The failure of UB induction in these gene inactivation mutants supports an important role of the extracellular matrix for the proper induction of ureter and metanephric kidney development.

Clinical considerations

Because Wolffian duct formation and ureter budding are essential steps in metanephric kidney development, the animal models described above often lead to a complete absence of ureter and kidney. This condition occurs in about 1:10,000 births and represents a subset of Potter syndrome, which is characterized by severe developmental malformations of the lungs, limbs and craniofacial features (33). This characteristic phenotype results from a lack of amniotic fluid, which is mostly derived from fetal urine during late gestation. With the increasing use of antenatal ultrasounds, these severe cases with Potter syndrome are readily identified and often lead to therapeutic termination of pregnancy (1). Noteworthy, as these severe conditions are likely caused by homozygous loss-of-function mutations in important developmental regulator genes, bilateral kidney agenesis in the fetus is a strong indication for urogenital system anomalies in the siblings (250 times higher) and parents (15 times higher) (34). These anomalies, expected to derive from heterozygous mutations, can have serious consequences during adult life (35). Several human syndromes are indeed caused by genetic haploinsufficiency. It is the case for some of the developmental regulators described above, which are associated with human syndromes including urinary tract and kidney defects (Table 1). Among them are PAX2 in renal coloboma syndrome (36), EYA1 and SIX1 in branchio-oto-renal syndrome (37, 38), GATA3 in hypoparathyroidism–deafness–renal anomalies (HDR) syndrome (39), FRAS1 in Fraser syndrome (40) and SALLI in Townes–Brocks syndrome (41). Recently, mutations in the RET gene were identified specifically in BRA fetuses (42). Hence, identifying the molecular cause of BRA in the fetus may provide important information for an eventual diagnosis and intervention in parents. It is indeed conceivable that milder variants of these syndromes would not produce the complete phenotype such that parents and siblings may be relatively asymptomatic. In support of this, a common haplotype affecting PAX2 expression levels has been identified in patients with hypoplastic kidneys without renal coloboma syndrome (43).

The fact that the genital tracts are derived from the Wolffian and Müllerian ducts provides an initial indication as to which genes may be affected (as heterozygous) in the parents of BRA fetuses. Based on the animal models available, BRA associated with absent genital tracts would point to the central regulators of pro/mesonephros development such as PAX2, GATA3 and LIM1. Alternatively, BRA in the presence of genital tracts would rather involve regulators of ureter budding.

In principle, the possibility of affected parents and siblings also applies to newborns or infants diagnosed with unilateral renal agenesis, as this condition is expected to be caused by the same genes. Unilateral renal agenesis is a relatively frequent condition (1:5000) that is usually asymptomatic early in life but may lead to hypertension and proteinuria later in life due to an overall nephron deficit (44).

Reconnecting the main pipe to the waste tank: distal ureter maturation

Morphological events underlying ureter ectopia

The process of UB induction is crucial for kidney development and has accordingly received much attention over the years. In addition to ureter and kidney agenesis, other defects at the level of ureter budding can have dramatic consequences on renal function. These are multiple ureter inductions, where two or more ureter and associated renal apparatuses are formed, and ectopic single kidney induction, where only one ureter is formed but is located either too rostrally (direction of the head) or too caudally (direction of the tail). UB formation at an ectopic location will likely not find sufficient surrounding mesenchyme to support normal growth, leading to kidney hypoplasia or dysplasia (45). However, the clinical consequences of abnormal budding on ureter functionality only make sense in light of the subsequent step of urinary tract morphogenesis: the maturation of the distal ureter. This process consists of moving the distal end of the newly formed ureter from the Wolffian duct to its final position within the bladder wall. This process has been the focus of recent attention with the identification of mouse mutants specifically affecting this morphogenetic step of development without influencing ureter budding per se. Hence, another major source of lower urinary tract malformation lies in the process of distal ureter maturation itself.

The first model of distal ureter maturation was presented in a seminal report by Mackie and Stevens (45). In this paper, the authors analyzed 51 cases of duplex kidneys resulting from bifid or duplex ureters. The careful analysis of these samples led them to postulate that ureter budding at a position too rostral (cranial) would lead to ectopic ureter insertion within the bladder neck, urethra or sex ducts. Such ectopic ureters would become dilated as a result of vesicoureteral junction obstruction. They also proposed that, abnormal budding of the ureter from the caudal portion of the Wolffian duct would lead to a positioning of the ureter too far lateral within the bladder wall. Such cases are usually associated with vesicoureteral reflux. These conclusions were supported by clinical observations associating the upper pole of duplex kidneys to a lower ureter orifice positioning in the bladder, while the lower pole connected higher (Weigert–Meyer law). The displacement of the ureter towards the bladder was assumed to result mostly from the enlargement of the bladder/urethra primordium that would engulf the Wolffian duct segment between the ureter and the bladder wall (common nephric duct, CND) (45). Hence, the CND tissue was thought to contribute to the bladder trigone, which is the bladder region located between the ureters and the sex ducts.

However, recent studies in the mouse have shed a new light on the process of distal ureter maturation. It has now become clear that the displacement of the ureter from the Wolffian duct to the bladder wall does not involve an engulfment of the CND into the bladder trigone but rather results from the elimination of the CND by programmed cell death (46). This observation was followed by another report further characterizing the morphological events occurring during this maturation process (47). Based on these reports we propose the following model of distal ureter maturation: (i) Following ureter budding, the CND is progressively eliminated (chewed up) by apoptosis in the region immediately adjacent to the cloaca epithelium (Fig. 2a). The elimination of the CND brings the distal ureter in immediate contact with the cloaca epithelium (Fig. 2b). Concomitantly to this process, the ureter undergoes a 180 degrees rotation around the axis of the Wolffian duct (Fig. 2a,b), (ii) the distal segment of the ureter (dUr) then lies down against the cloaca epithelium and is also eliminated by apoptosis (Fig. 2c). Ureter rotation is crucial as it allows the distal ureter to orient itself towards the bladder primordium before being eliminated, (iii) this process generates a new ureter connection point in the cloaca region that will give rise to the bladder, while the Wolffian duct remains in the region giving rise to the urethra (Fig. 2d), and (iv) the separation between the Wolffian duct and ureter is further increased by the growth of the bladder and urethra.

Figure 2.

Revised model of distal ureter maturation and associated lower urinary tract malformations. (a) Following the invasion of the metanephric mesenchyme (MM) by the newly formed ureteric bud (UB), a segment of the Wolffian duct (WD) located between the ureter and the cloaca (common nephric duct; CND; brown) is eliminated by apoptosis. (b) CND elimination brings the ureter in direct contact with the cloaca. Concomitantly with this process, the ureter rotates around the axis of the Wolffian duct. (c) Ureter rotation orients the distal segment of the ureter towards the bladder primordium of the cloaca. The distal-most segment of the ureter (dUr) lies down on the cloaca epithelium and is also eliminated by apoptosis. (d) This process generates a new connection point between the ureter and the bladder primordium. After this stage, the distance between the ureter and Wolffian duct is further increased by expansion of the urethra and bladder tissues. (e) In duplex systems, a second ureter is induced along the Wolffian duct. This additional ureter (blue) is typically forming at a rostral position in reference to the primary ureter (red). (e′) Following distal ureter maturation, the rostral ectopic ureter connects below the primary ureter, either in the bladder neck, urethra or remains attached to Wolffian duct. (f) In some cases, a rostral ectopic ureter (blue) forms in the absence of primary ureter. (f′) Following distal ureter maturation, these ectopic ureters behave similarly to the rostral ureters of duplex system. (g) Caudal ureter budding places the connection point between the ureter and the Wolffian duct very close to the cloaca. (g′) The near absence of CND is expected to accelerate the elimination of the distal ureter and thus position the vesicoureteral junction in a ectopic lateral position. (h) When the process if distal ureter maturation is defective, the ureter (dark purple) is induced at the right location along the Wolffian duct. (h′) The defective elimination of the CND prevents or slows down distal ureter elimination, resulting in an aberrant positioning of the ureter on the Wolffian duct, urethra or bladder neck. When urine starts to be produced, the pressure exerted on the urethra or bladder neck is likely to cause ureterocele (not shown on diagram).

Importantly, this revised model is fully compatible with the initial clinical observations reported by Mackie and Stevens (45). According to this recent model, supernumerary ureters induced in the rostral region of the Wolffian duct cannot reach the cloaca epithelium and therefore fail to lie down and connect properly to the bladder primordium (Fig. 2e,e′). A similar situation is expected for single ectopic ureters forming too rostrally (Fig. 2f,f′). These ureters therefore remain attached either to the Wolffian duct, the urethra region or the bladder neck, depending on how far from the normal ureter budding site they have been initially induced. Once urine starts to be produced in the affected kidney, these malformations cause vesicoureteral junction obstruction and thus lead to hydroureter/hydronephrosis and ureterocele (a urine filled ureter-derived cyst formed within the urethra or protruding within the bladder). Conversely, when ureter budding occurs too caudally along the Wolffian duct, the ureter is expected to reach the cloaca prematurely (and possibly before complete ureter rotation) and consequently generate a ureter–bladder connection point that is more lateral than normally expected (Fig. 2g,g′). In this ectopic location, the vesicoureteral valve does not form adequately, leading to vesicoureteral reflux. The fourth possibility is that ureter budding occurs at the right location, but the process of ureter maturation itself is deficient (Fig. 2h,h′). Although we know relatively little about this process, it is already evident that the regulation of apoptotic cell death is central to ureter maturation (27, 47). A failure or delay in the elimination of the CND and distal ureter leads to vesicoureteral junction obstruction equivalent to that observed for ectopic rostral ureter budding (Fig. 2h′). Similarly, one would expect defects in the connection between the Wolffian duct and the cloaca or between the ureter and the bladder primordium to also lead to hydroureter and hydronephrosis.

Molecular events underlying ureter ectopia

At the molecular level, ectopic rostral budding (usually duplex/multiplex), ectopic caudal budding and ureter maturation defects are relatively distinct. Rostral budding can be seen as the phenotypical opposite of kidney agenesis described above. While agenesis is typically due to a loss of function in GDNF-Ret signaling, the excessive activation of this pathway leads to either duplex/multiplex systems or single ureter formation in an ectopic rostral location. As GDNF expression in the MM determines the site of primary ureter budding, it is crucial to restrict its expression domain and signaling potential to the appropriate region. Proteins such as Slit2, its receptor Robo2 and the transcription factors FoxC1 and FoxC2 are necessary to repress the expression of GDNF in the rostral mesenchyme adjacent to the Wolffian duct (48, 49). In support of their importance in GDNF regulation, all these gene mutations lead to duplex/multiplex systems (Table 2). The signaling molecule BMP4 also counteracts GDNF activity, but does so without affecting GDNF expression levels (50). The negative action of BMP4 on GDNF is itself inhibited by the secreted protein Gremlin at the ureter budding site, thereby allowing UB formation to occur (51). Interestingly, in contrast to most other GDNF regulators, BMP4 heterozygous mutant embryos do not harbor multiple UBs but rather shifts the ureter budding site more rostrally (50). More recently, Gata2 hypomorphic mutant mice were shown to have a similar phenotype, which may come as a result of BMP4 deregulation (52). The detailed mechanisms leading to single vs multiple ectopic bud formation as a result of GDNF deregulation remain unclear at the moment.

Table 2.  Mouse models of ureter ectopia
GeneType of proteinGenotypeDefects in humanReference
  1. CAKUT, congenital anomalies of the kidneys and urinary tracts; HDR, hypoparathyroidism–deafness–renal anomalies syndrome; VUR, vesicoureteral reflux.

Duplex/Multiplex ureter budding
 Foxc1 (Mf1)  and Foxc2  (Mfh1)Transcription factorFoxc1−/−Foxc1+/−;Foxc2+/−CAKUT(49, 77)
 Slit2Robo2 ligand, secreted glycoproteinSlit2−/− (48)
 Robo2Slit2 receptor, Ig superfamilyRobo2−/−VUR(48, 57)
 Gata3Transcription factorHoxB7-Cre;Gata3flox/floxHDR Syndrome(25)
 β-cateninTranscriptional coactivator, Wnt pathwayHoxB7-Cre; β-catenin−/flox (26)
 Agtr2G-protein coupled angiotensin II receptorAgtr2−/−CAKUT(78)
 Spry1Receptor tyrosine kinase antagonistSpry1−/− (53)
Rostral ureter budding
 Gata2Transcription factorGata2fGN/fGN (52)
 BMP4Secreted molecule, TGF familyBMP4+/−Renal Hypodysplasia(50)
Caudal ureter budding
 RetReceptor tyrosine kinaseHoxB7/Ret+/−Hirschsprung disease(54)
 Pax2Transcription factorPax21Neu/+Renal coloboma syndrome(55)
Distal ureter maturation
 RetReceptor tyrosine kinaseRet−/−Hirschsprung disease(27)
 Rara and  Rarb2Retinoic acid receptorRara−/−;Rarb2−/− (27)
 Aldh1a2Retinaldehyde dehydrogenaseAldh1a2−/− (46)
 Ptprs and  Ptprf (LAR)Receptor tyrosine phosphatasePtprs−/−;Ptprf−/− (47)

Within the Wolffian duct, the signaling response of Ret to GDNF further needs to be modulated. This is performed in part by the negative regulation of Ret signaling by members of the Sprouty family, as evidenced by the duplex/multiplex systems observed in Sprouty1 mutant embryos (53). In addition, β-catenin and Gata3 were also shown to affect ureter budding in the Wolffian duct (25, 26). In both models, specific gene inactivation in the Wolffian duct using the Cre-lox system leads to a mosaic loss of Ret expression. The mutant Wolffian duct cells undergo premature differentiation, while those maintaining Ret expression initiate ectopic ureter budding as a consequence of an inappropriate cellular response to both GDNF and fibroblast growth factor signals (25). These mice harbor severe urinary tract defects such as hydronephrosis, hydroureter and kidney dysplasia, that resemble the malformations observed in GATA3 heterozygous mutations in humans (HDR syndrome). It is thus possible that the mosaic deletion of Gata3 in mice mimics a threshold effect generated by GATA3 haploinsufficiency in humans.

Although most mouse models of urinary tract anomalies result in rostral ectopic budding, cases of single caudal ectopic ureter have also been described. These are the transgenic overexpression of Ret (54) as well as Pax2 heterozygous mice (55). In accordance with the revised model of distal ureter maturation, these mice exhibit vesicoureteral reflux as a result of lateral positioning of the ureter within the bladder wall. In this respect, Pax2 heterozygous mice reflect the renal phenotype observed in humans (renal coloboma syndrome) (36).

It is only recently that the process of distal ureter maturation has emerged as a primary cause of lower urinary tract malformations in mouse models. In these cases, the site of ureter budding is normal but the sequence of events displacing the ureter from its site of induction on the Wolffian duct to its final position within the bladder wall is deficient. Such defects are observed in Rara;Rarb2 and Aldh1a2 mutant mice (46), which fail to generate retinoic acid, a key morphogenic molecule during embryonic development. Similar defects were observed in mice inactivated for both PTPRF and PTPRS, two receptor-type tyrosine phosphatase genes of the leucocyte antigen-related (LAR) family (47). In all three mouse models, the CND failed to be normally eliminated by programmed cell death. As a result, the single ureter often remains connected to the urethra or bladder neck, which prevents normal urine flow from the affected kidney. In PTPRF;PTPRS double mutants, the resulting obstruction often led to the formation of ureterocele associated with megaureter and hydronephrosis. Although ureterocele was not reported for Aldh1a2-deficient mice, megaureter and hydronephrosis were also present in these mice. Importantly, both retinoic acid and PTPRF;PTPRS deficient mice were found to regulate Ret activity at the transcriptional and protein levels, respectively (46, 47). Noteworthy, Ret mutant mice showed similar lower urinary tract malformations (27), although these defects were not characterized in details at the morphological level. Ret is unlikely to be the only important player in distal ureter maturation as LAR-family phosphatases were identified as negative regulators of Ret signaling, presumably inhibiting its pro-survival activity (47), while retinoic acid signaling was found to positively regulate Ret expression in the CND region (46). These results underscore the fact that much remains to be clarified about the molecular mechanisms of distal ureter maturation.

Clinical considerations

A key aspect of the new model of distal ureter maturation presented above is that it provides an integrated understanding of most developmental anomalies affecting the lower urinary tract. Although derived from mouse models, the close resemblance between mouse and human urinary tract development and malformations leads us to believe that this model is also suitable to understand urinary tract anomalies in humans. Importantly, this model is compatible with previous clinical observations and identifies three major causes of lower urinary tract defects (in addition to ureter agenesis): rostral ureter budding, caudal ureter budding and distal ureter maturation defects.

Caudal ureter budding is believed to be the primary cause of vesicoureteral reflux (VUR) [reviewed in (56)]. It should be noted, however, that in addition to caudal budding, a premature or accelerated maturation of the distal ureter (e.g. by excessive cell death in the CND) is also predicted to result in VUR by lateral ureter positioning within the bladder wall. As suggested by the revised model, excessive apoptosis of the CND is indeed expected to bring the distal ureter against the bladder wall before complete ureter rotation. Such mechanism will have to be explored further. The high occurrence of VUR in humans (about 1:100 live births) suggests that caudal budding is a common defect in humans, which contrasts with the rare report of such defects in the mouse. The reason for this may simply be that reflux in mice does not lead to major phenotypical consequences and may thus have been missed. Nonetheless, the functional analysis of genes involved in mouse urinary tract morphogenesis has already led to the identification of vesicoureteral reflux genes in humans (57, 58).

Ectopic positioning of the ureter within the bladder neck, urethra or sex ducts can occur in duplex systems or as single ectopic ureter. Both conditions lead to similar obstruction nephropathies such as hydroureter (megaureter), hydronephrosis and ureterocele. Importantly, however, these malformations result from distinct developmental anomalies. Based on the mouse models available, duplex systems typically come as a result of supernumerary ureter inductions due to a failure to repress GDNF rostral signaling activity. Hydroureter, hydronephrosis and/or ureterocele associated with a single urinary tract can be caused by rostral ureter budding or ureter maturation defects. At the moment, it is difficult to differentiate between both underlying causes. However, ectopic rostral budding is expected to affect metanephric kidney development due to insufficient mesenchymal tissue to support normal kidney growth, leading to hypodysplasia (45), whereas distal ureter maturation defects are not necessarily linked to metanephric kidney development. In support of this, Ptprf;Ptprs double-mutant embryos exhibit severe defects in distal ureter maturation without showing apparent kidney malformations (47). This observation also shows that distal ureter maturation and ureter branching morphogenesis are two distinct processes at the molecular level.

In Caucasians, the proportion of duplex system to single rostral ectopic ureter is estimated to be about 4 to 1. Interestingly, this proportion is reverted (1:4) in the Indian and Asian populations (59, 60). The explanation for this regional difference is unclear but it may be, at least in part, due to a vitamin A-deficient diet observed in some populations, notably in India. Such deficiency would indeed affect the production of retinoic acid, which is crucial for distal ureter maturation and bladder development (46, 61).


Urinary tract anomalies are an important component of the disease group CAKUT. CAKUT is characterized by a high degree of phenotypic variability and can include hydroureter, hydronephrosis, ureterocele, duplex systems and kidney adysplasia. A number of questions remain on how these different conditions can be found in association. The revised model of distal ureter maturation presented here suggests that both, the site of ureter budding and the subsequent process of distal ureter maturation determine the final position of the ureter. If either step is abnormal, lower urinary tract functionality is affected accordingly, leading to complete or partial obstruction or reflux. Importantly, the continuum of possible ureter positions observed in diseased models is submitted to a threshold effect. The resulting phenotypic outcomes are distinct and depend on whether the ureter reaches the urethra, the bladder neck or the lower bladder, or connects too far in the lateral portion of the bladder wall. This threshold effect is probably part of the great variability in lower urinary tract deficiencies. Another crucial aspect is that lower urinary tract malformations often occur in association with additional anomalies of the genitourinary and other organ systems. The characterization of new mouse models with known molecular alterations and their association to known phenotypes in humans will be necessary to resolve the complexity of urinary tract-related diseases. Although many questions remained to be answered, a more complete and integrated view of kidney and urinary tract development has emerged in the recent years, both at the morphological and molecular levels. Constant efforts should be made so that these discoveries, mostly made in animal models, are being translated towards the treatment and diagnosis of urinary tract-related diseases.


We wish to thank David Grote, Indra Gupta and Susanne Kaitna for the critical review of the manuscript. M. B. is supported by operating grants from the Kidney Foundation of Canada and the Canadian Institutes for Health Research and holds a Canada Research Chair on Developmental Genetics of the Urogenital System.