Introduction

Hepatocyte nuclear factor-1β (HNF-1β, vHNF1) is a developmentally regulated transcription factor that is required for tissue-specific gene expression in the epithelial cells of many organs, including kidney, pancreas, liver, and genitourinary tract [1, 2]. In the developing kidney, HNF-1β is expressed in developing nephrons and the branching ureteric bud that gives rise to the renal collecting system [3,4,5]. In the mature kidney, expression of HNF-1β persists in the epithelial cells of renal tubules but not in the glomeruli, blood vessels, or interstitial cells. The importance of HNF-1β in renal tubules is underscored by the finding that combined expression of HNF-1β with the transcription factors Emx-2, HNF-4α, and Pax-8 is sufficient to reprogram fibroblasts into functional renal tubular epithelial cells [6].

HNF-1β activates or represses transcription of target genes through binding of its POU (Pit-1, OCT1/2, UNC-86; POUS)-specific domain and atypical POU homeodomain (POUH) to the DNA consensus sequence 5′-GTTAATNATTAAC-3′. The dimerization domain located at the N-terminus of the protein mediates the formation of HNF-1β homodimers or heterodimers with the related protein HNF-1α. The protein complex also includes two molecules of the dimerization cofactor of HNF1 known as pterin-4 alpha-carbinolamine dehydratase (PCBD1) [7]. The C-terminal region of HNF-1β contains a transactivation domain(s) that is responsible for the recruitment of coactivators and regulation of transcription [8]. To date, the DNA-binding domain is the only HNF-1β motif whose crystal structure has been determined (PDB file: 2H8R) [9]. HNF-1β and HNF-1α recognize the same DNA consensus sequence but can regulate gene transcription with different outcomes depending on the cellular context [10, 11].

In humans, heterozygous mutations of the HNF1B gene cause a spectrum of inherited and sporadic malformations of the kidney and genitourinary tract. Molecular genetic defects include whole-gene deletions in about 50% of patients; point mutations are detected in the remaining cases [12]. De novo mutations of HNF1B are encountered in up to 30–50% of cases [12, 13]. The first HNF1B-related disorder that was described was Renal Cysts and Diabetes (RCAD; MIM:137920), a syndrome characterized by autosomal dominant inheritance, renal cystic abnormalities, and maturity-onset diabetes of the young type 5 (MODY5) [12, 14]. Additional extrarenal manifestations include exocrine pancreatic failure, abnormal liver function tests, and genital tract malformations. HNF1B mutations also cause congenital anomalies of the kidney and urinary tract (CAKUT) [15, 16]. HNF1B mutations are found in 20–30% of fetuses with renal abnormalities and represent the most common prenatal cause of hyperechogenic kidneys with or without cysts [17]. Studies on large cohorts of fetuses and children showed that genetic variants in HNF1B are the most common cause of isolated renal hypodysplasia in humans [13, 18]. Other renal manifestations include multicystic dysplastic kidneys, glomerulocystic kidney disease, oligomeganephronia, renal agenesis, renal hypoplasia, familial juvenile hyperuricemic nephropathy, and renal interstitial fibrosis associated with a highly variable rate of decline of estimated glomerular filtration rate (eGFR) [12, 19, 20]. Humans with HNF1B mutations may also suffer from electrolyte disturbances, such as hypomagnesemia and hypokalemia [21, 22]. In addition, heterozygous deletions of HNF1B have been associated with prune belly syndrome (PBS, MIM: 100100), characterized by deficiency or absence of abdominal wall musculature, dilatation of the urinary tract and bilateral undescended testes [23, 24]. The renal phenotypes observed in patients with HNF1B mutations are highly heterogeneous and may manifest during the antenatal period, in childhood, or in adulthood [21]. This review summarizes recent insights into the roles of HNF-1β in the developing and mature kidney and reflects on future perspectives.

HNF-1β in kidney development

Mouse kidney development begins approximately at embryonic day E10.5 with the outgrowth of the ureteric bud (UB) from the Wolffian duct (WD) into the adjacent metanephric mesenchyme (MM, Fig. 1) [25]. Subsequently, the UB undergoes branching morphogenesis to generate the entire collecting system and ureter, whereas signals from the UB tips will induce an epithelial transition of surrounding MM cells to form renal vesicles (RV) [25]. Renal vesicles differentiate into comma- and S-shaped bodies, and subsequently into Bowman’s capsule of the glomerulus and nephron tubules that connect to the collecting system [25]. Functional nephrons are present by E16.5 in mice, but elongation and remodeling of tubular segments continue for 2 weeks after birth [25, 26].

Fig. 1
figure 1

HNF1-β is required for UB branching, initiation of nephrogenesis, and nephron segmentation. Schematic showing the major stages of kidney development. HNF1-β-expressing cells are depicted in light pink. At E10.5, the UB emerges from the WD upon signaling from the MM. Starting from E12.5, HNF1-β is essential for UB branching. In developing nephrons, HNF1-β is expressed in the distal region of RVs, in the distal CSB, and in the distal to proximal SSB, but it is absent in the most proximal region that will form the glomerulus. By E17.5, HNF1-β is present in all tubular epithelial cells of the mature nephrons but not in the glomeruli. MM, metanephric mesenchyme; UB, ureteric bud; WD, Wolffian duct; RV, renal vesicle; CSB, comma-shaped bodies; SSB, S-shaped bodies; D, distal; I, intermediate; P, proximal; DCT, distal convoluted tubule; PT, proximal tubule; TAL, thick ascending limb of Henle; CD, collecting duct

Full size image

HNF-1β is an essential transcriptional regulator of renal epithelial organization and differentiation [3,4,5, 27]. HNF-1β is expressed in the WD, UB, RV, comma-shaped bodies, S-shaped bodies, and developing renal tubules but not in MM (Fig. 1). Within comma- and S-shaped bodies, HNF-1β is expressed in the proximal and distal segments that are the anlagen of the renal tubules, but not in the most-proximal regions that will give rise to glomeruli [28]. Using constitutive inactivation of Hnf1b in the epiblast by tetraploid aggregation, Lokmane et al. found that HNF-1β is involved in the timing of UB outgrowth and its early branching. It is also required for maintenance of the WD epithelium and for early nephrogenesis [3]. HNF-1β exerts these functions, at least in part, through the direct control of several key regulatory genes [27]. HNF-1β acts directly upstream of Wnt9b, an essential factor that acts as a paracrine signal to orchestrate the mesenchymal-epithelial transitions underlying the initiation of nephrogenesis [29, 30]. HNF-1β also maintains the expression of Pax2 and Lim1, which have essential functions in multiple steps of renal epithelial tubular morphogenesis and are required for WD formation [3]. Recently, Desgrange et al. conditionally inactivated Hnf1b in the WD and UB using a HoxB7-Cre line [27]. They found that HNF-1β is required for the formation of a specialized UB tip domain through the maintenance of cell-cell adhesion, Pax2 expression, and Ret signaling. HNF-1β is subsequently required in the UB trunks for the establishment of normal apico-basal polarity and differentiation of the collecting system [27].

Inactivation of Hnf1b in MM using a Wnt4-Cre line or a Six2-Cre line leads to defective S-shaped bodies lacking an intermediate limb that gives rise to the loop of Henle loop and proximal tubule [4, 5]. HNF-1β regulates the acquisition of a proximal-intermediate nephron segment fate by controlling the expression of Irx1, Osr2, Pou3f3, and Notch signaling components [4, 5]. The importance of HNF-1β in nephron segmentation has also been demonstrated in zebrafish embryos [31].

Deletion of HNF-1β in elongating renal tubules using the Ksp-Cre line results in renal cyst formation [32]. The mitotic spindles of dividing tubular cells from HNF-1β-mutant kidneys are significantly misoriented compared with controls, a finding that suggests dysregulation of planar cell polarity (PCP) [33]. In elongating renal tubules, PCP signaling coordinates tubular narrowing and lengthening via convergent extension (CE) and the spatial distribution of daughter cells via oriented cell division (OCD) [34]. These processes control tubule lumen diameter and ensure that cell proliferation results in tubular elongation rather than dilation [34]. In HNF-1β-mutant kidneys, abnormalities in OCD are observed in precystic tubules, suggesting that loss of PCP may be responsible, at least in part, for cyst formation [35].

Role of HNF-1β in polycystic kidney disease

In experimental mice, kidney-specific deletion of Hnf1b by Cre/loxP recombination or transgenic expression of dominant-negative mutant HNF-1β leads to kidney cysts and renal failure, similar to the phenotype of humans with HNF1B mutations [32, 36]. Molecular characterization of the cystic kidneys has revealed that HNF-1β regulates the transcription of Pkhd1, the gene mutated in autosomal recessive polycystic kidney disease (ARPKD; MIM:263200); Pkd2, the gene mutated in autosomal dominant polycystic kidney disease (ADPKD; MIM:613295); Umod, associated with medullary cystic kidney disease (MCKD; MIM:174000); and Glis2, associated with nephronophthisis, and the cpk modifier gene Kif12 [32, 37, 38]. These findings suggest that mutations of HNF-1β produce kidney cysts by downregulating the expression of multiple cystic disease genes.

Most of the proteins encoded by cystic disease genes are located in the primary cilium, a sensory organelle on the cell surface, or the basal body, which anchors the cilium in the cell body. Primary cilia play essential roles during development and tissue homeostasis by regulating cell proliferation, migration, apoptosis, PCP, and differentiation through the activation of multiple signaling pathways, including canonical and non-canonical Wnt signaling and hedgehog, cAMP, and mTOR signaling [39,40,41]. Ciliary dysfunction or cilia loss in the kidney has been associated with renal cyst formation and renal failure [42]; however, the precise mechanism whereby the loss of renal cilia produces kidney cysts remains poorly understood.

Although HNF-1β is not directly involved in primary cilium function, it regulates the expression of proteins that localize to the cilium or that belong to biological processes mediated by cilia [43]. Polycystin-2 (PC2), encoded by the Pkd2 gene, is a Ca2+-permeable cation channel that interacts with polycystin-1 (PC1) in the primary cilium. Deletion of PC2 or mutations in HNF-1β are associated with decreased Ca2+ entry, activation of the Ca2+-inhibitable adenylyl cyclases AC5 and AC6 and elevated cAMP levels [40]. cAMP in turn promotes cyst growth by stimulating cell proliferation and fluid secretion. In addition, HNF-1β regulates the expression of phosphodiesterase 4C (PDE4C), which catabolizes cAMP in the primary cilium. In Hnf1b mutant kidney cells and mice, PDE4C is downregulated and cAMP levels are increased [40]. Inhibition of cAMP synthesis may be beneficial in the treatment of PKD as knockout of AC5 reduces renal cAMP levels and reduces cyst burden in Pkd2 mutant mice [44, 45].

The protein encoded by Pkhd1, named polyductin or fibrocystin, is an integral membrane protein that localizes to the basal body of primary cilia as well as the centrosomes and mitotic spindles of dividing cells [46]. Pkhd1 knockout mice develop kidney cysts, pancreatic cysts, and biliary ductal plate malformations, which confirms that downregulation of Pkhd1 is sufficient to produce renal cystic disease [47]. Expression of dominant-negative mutant HNF-1β or kidney-specific knockout of Hnf1b in mice reduces the levels of Pkhd1 mRNA transcripts in the kidney via direct downregulation of Pkhd1 promoter activity in renal collecting ducts (CDs) [32, 36, 48, 49].

Umod encodes uromodulin, a protein produced in the thick ascending limb of the loop of Henle (TAL) [50]. Uromodulin is found in the mitotic spindles of dividing cells. In addition, the decreased number of uromodulin-positive cilia in the kidney biopsies of patients with known UMOD mutations suggests a specific function of the protein in cilia [50].

Genome-wide chromatin immunoprecipitation combined with DNA microarray (ChIP-chip) and transcription profiling identified the PKD modifier gene Kif12 and the suppressor of cytokine signaling Socs3 as novel HNF-1β target genes [37, 51]. KIF12 participates in mitotic spindle formation, and variation in KIF12 copy number has been associated with CAKUT in humans [52]. HNF-1β promotes tubulogenesis of renal cells by repressing the transcription of Socs3, which enables activation of hepatocyte growth factor (HGF) signaling [51]. Conversely, upregulation of Socs3 in dominant-negative HNF-1β-expressing cells inhibits HGF signaling and impairs tubule formation. Consistent with these findings, HNF1B is downregulated and SOCS3 is highly upregulated in human polycystic kidneys [53]. Moreover, humans with heterozygous mutations of both PKD1 and HNF1B have a more severe phenotype, suggesting that HNF1B functions as a modifier gene in PKD [54].

Recently, it was shown that HNF-1β regulates the expression of a long noncoding RNA (lncRNA) containing the miR-200 cluster of microRNAs [55]. Kidney tubule-specific deletion of Hnf1b results in decreased expression of miR-200, and in turn, knockdown of miR-200 in cultured renal epithelial cells inhibits tubulogenesis and produces cyst-like structures [55]. The molecular mechanism whereby miR-200 regulates cyst pathogenesis may involve modulation of PKD1 expression via binding to the 3′UTR of the PKD1 transcript. miR-200 also regulates epithelial-to-mesenchymal transition (EMT), a process in which epithelial cells acquire mesenchymal properties, such as increased migratory capacity [56, 57]. EMT is integral to development and is reactivated during wound healing and tissue fibrosis. miR-200 and its transcriptional regulator HNF-1β inhibit EMT [58, 59]. Downregulation of HNF-1β is associated with EMT [58], consistent with the observed decrease in miR-200 levels, and inactivation of HNF-1β in CDs using a Pkhd1-Cre line causes interstitial fibrosis [55, 60]. Clinically, some patients with HNF1B mutations present primarily with renal fibrosis, a syndrome called autosomal dominant tubulointerstitial kidney disease (ADTKD) [22]. The role of EMT in the pathogenesis of ADTKD requires further investigation.

HNF-1β regulates ion transport in kidney

Almost 50% of patients with HNF-1β mutations develop hypomagnesemia due to renal Mg2+ wasting [21]. In the kidney, the major sites of Mg2+ reabsorption are the thick ascending limb of the loop of Henle (TAL) and distal convoluted tubule (DCT). Whereas defects in Mg2+ handling in the TAL lead to concomitant renal wasting of Ca2+, defects in the DCT cause hypermagnesuria associated with normo- or hypocalciuria [61]. In patients with HNF1B mutations, hypomagnesemia is usually accompanied by hypocalciuria, although a few affected adults have normocalciuria suggesting impaired reabsorption of Mg2+ in DCT [21, 62]. Hypomagnesemia is attributed to decreased expression of FXYD2, which encodes the γ subunit of Na+-K+-ATPase. Expression of FXYD2 is directly regulated by HNF-1β (Fig. 2) [62, 63], and mutations of FXYD2 cause autosomal dominant renal hypomagnesemia with hypocalciuria (MIM 601814) [64]. Recently, it was demonstrated that HNF-1β is a transcriptional activator of CaSR, encoding the calcium-sensing receptor, in the TAL [65]. Consequently, patients with HNF1B mutations may have reduced CaSR activity in the kidney, which could contribute to the hypocalciuria.

Fig. 2
figure 2

HNF-1β regulates the expression of solute transporters along the nephron. Defects in ion homeostasis reported in patients with HNF1B mutations include hypomagnesemia, hypokalemia, hyperuricemia, and rarely Fanconi syndrome. Whereas HNF-1α is restricted to the PT, HNF-1β is expressed in all epithelial cells of the nephron. HNF-1β target genes in the mature kidney encode OAT1/3/4, URAT1, NPT1, collectrin, uromodulin, NKCC2, the γ-subunit of Na+-K+-ATPase, and FXR. Refer to the text for detailed information. PT, proximal tubule; TAL, thick ascending limb of Henle; DCT, distal convoluted tubule; CD, collecting duct; HNF-1α/β, hepatocyte nuclear factor-1α/β; OAT1/3/4, organic anion transporter 1/3/4; NPT1, Na+-Pi cotransporter 1; URAT1, urate anion exchanger 1; OA−, organic anion; UA, uric acid/urate; ROMK, renal outer medullary K+ channel; NKCC2, Na+-K+-Cl− transporter; CaSR, calcium-sensing receptor; TRPM6, transient receptor potential melastatin 6; FXR, farnesoid X nuclear receptor

Full size image

Faguer et al. found that 46% of adults with HNF1B mutations had hypokalemia [21]. Since hypokalemia is frequently associated with Mg2+ deficiency [66], it was hypothesized that a decrease in intracellular Mg2+ concentrations leads to the release of inhibition of renal outer medullary K+ channel (ROMK) causing urinary K+ wasting [67]. In addition, HNF-1β can directly regulate the transcription of genes involved in renal K+ handling, such as UMOD and SLC12A1 encoding respectively uromodulin and the Na+-K+-Cl− transporter (NKCC2) in the TAL (Fig. 2). Under physiological conditions, NKCC2 mediates Na+ and Cl− transport across the apical membrane, which is maintained by K+ recycling through ROMK. Uromodulin regulates both NKCC2 and ROMK [68, 69]. Mutations in UMOD result in urinary salt wasting, gout, and hypertension [70], whereas mutations in SLC12A1 and in KCNJ1, encoding ROMK, are associated with antenatal Bartter’s syndrome (MIM 601668, MIM 241200), a severe salt wasting syndrome that leads to a pronounced urinary concentrating defect and metabolic alkalosis [71, 72]. HNF-1β also directly regulates the transcription of KCNJ10 that encodes the K+ channel Kir5.1 [73]. Kir5.1 is a major contributor to the basolateral K+ conductance in the DCT, and mutations in KCNJ10 cause a salt-wasting syndrome that includes hypomagnesemia and hypokalemia (MIM 612780) [73]. Hence, dysregulation of multiple transporters may contribute to the K+ and Mg2+ wasting observed in patients with HNF1B mutations.

In the study by Faguer et al., two HNF1B mutant patients had PT dysfunction presenting as Fanconi syndrome, which is more typically associated with HNF1A mutations. This finding is in line with the evidence that HNF-1α and HNF-1β are co-expressed in the PT and co-regulate expression of renal organic anion transporters such as OAT1 (Slc22A6), OAT3 (Slc22A83), and OAT4 (Slc22A11), the renal urate transporter (URAT1/Slc22A12) and renal Na+-phosphate (Pi) transporter 1 (NPT1/Slc17A1; Fig. 2) [10, 74,75,76]. Another HNF1-1β target gene involved in PT transport is Tmem27 encoding collectrin [77, 78]. Collectrin increases the surface expression of amino acid transporters, and disruption of Tmem27 in mice results in a severe defect in renal amino acid uptake [79, 80]. In general, PT dysfunction in humans with HNF1B mutations is rare, probably because HNF-1α is able to compensate for the loss of HNF-1β. Hyperuricemia with reduced fractional excretion of uric acid is the only widespread finding in HNF1B mutant patients that can be ascribed to an effect in the PT [81]. This phenotype is probably due to inhibition of the OAT1/3/4 transporters and/or uromodulin. Mutation of the latter also causes hyperuricemia [70], attributed to inhibition of NKCC2 in the TAL with a consequent increase in proximal Na+ reabsorption, which drives uric acid reabsorption (Fig. 2) [81].

Finally, hydronephrosis and polyuria have been observed in humans with mutations of HNF1B [62]. Proposed etiologies of polyuria include glycosuria due to early-onset diabetes mellitus and renal Na+ and K+ wasting causing intrauterine polyuria and polyhydramnios. Another possible mechanism was recently suggested by Aboudehen et al. [60] who found that inactivation of Hnf1b in mouse CDs produces polyuria and defects in urinary concentration by: (i) direct inhibition of osmosensitive genes, such as FXR; (ii) indirect downregulation of the urea transporter UT-A1; and (iii) impaired apical trafficking of aquaporin 2 (AQP2) [60].

HNF-1β and intrarenal metabolism

A combination of chromatin immunoprecipitation/next generation sequencing (Chip-seq) and gene expression profiling in renal epithelial cells uncovered a novel role of HNF-1β in a transcriptional network that regulates intrarenal cholesterol metabolism [82]. Studies in Hnf1b mutant kidney cells showed that HNF-1β stimulates cholesterol synthesis and inhibits cholesterol uptake, a mechanism that is reminiscent of the function of SREBP-2, which also activates transcription of cholesterol biosynthetic genes [82]. Indeed, ChIP-seq and reporter gene assays demonstrated that HNF-1β directly regulates Srebf2 (SREBP-2) transcription as well as the expression of Pcsk9, which encodes a serine protease that plays a crucial role in regulating cholesterol influx by internalizing the LDL receptor [83]. The physiological consequences of the regulation of intrarenal cholesterol metabolism, without affecting circulating cholesterol levels, are unknown. It is possible that HNF-1β may regulate metabolic changes in response to renal injury. Renal expression of HNF-1β shows an early downregulation followed by transient over-expression after ischemic acute kidney injury (AKI) [84, 85], a condition that is also associated with stimulation of renal cholesterol synthesis [86]. The increase in renal cholesterol, which seems to have a cytoprotective effect, may be mediated by the transient upregulation of HNF-1β expression. In sepsis-induced AKI, HNF-1β levels decrease within 24 h after injury and quickly normalize by 36 h [87]. HNF-1β downregulation during sepsis in vivo is associated with a decrease in the expression of PPARGC1A, which is involved in mitochondrial biogenesis and function. In vitro experiments have shown that HNF-1β directly regulates PPARGC1A and mitochondrial respiration, whereas ablation of Hnf1b in proximal tubule cells leads to a shift from oxidative phosphorylation to glycolysis [87]. Kang et al. reported that favoring glycolysis in renal tubular cells worsens the development of fibrosis [88], a common finding in HNF1B mutant humans. The contribution of the metabolic events affected by HNF-1β to the pathogenesis of cystic kidney disease and other mutant phenotypes remains to be further defined.

Mechanisms of HNF-1β regulation

Many intracellular events that occur at multiple levels regulate the abundance and activity of HNF-1β (Fig. 3). Whereas the upstream effectors that are involved in HNF1B gene transcription are largely unknown, with the exception of SNAIL and C/EBP, microRNAs (miRNAs) have been found to control Hnf1b mRNA abundance in kidney. Patel et al. reported that the miR-17 ∼ 92 miRNA cluster regulates the posttranscriptional expression of PKD genes, including Pkd1, Pkd2, and Hnf1b [89]. Within the cluster, miR-92a repressed the 3′ UTR of Hnf1b, whereas mutations of the miR-92a binding site abrogated this repression. Since miR-17 ∼ 92 is upregulated in PKD, miR-17 ∼ 92-dependent post-transcriptional silencing of PKD genes may represent a mechanism underlying cyst growth.

Fig. 3
figure 3

Schematic model of the molecular mechanisms involved in the regulation of HNF-1β expression and activity. Control of HNF1-β expression and activity in the cell nucleus can occur at multiple levels: (1) HNF1B gene transcription; (2) HNF1B mRNA stability; and (3) HNF1-β dimerization and interaction with regulatory proteins. EGF, epidermal growth factor; EGFR, epidermal growth factor receptor; PCBD1, pterin-4 alpha-carbinolamine dehydratase; P, phosphorylation site

Full size image

After translation in the cytosol, HNF-1β is targeted to the nucleus by a nuclear localization signal. Within the nucleus, HNF-1β binds DNA target sites as homodimers or heterodimers with its homolog HNF-1α (Fig. 3). This complex is able to recruit a number of cofactors necessary for transcription. The dimerization domain of HNF-1β interacts with the transcriptional coactivator PCBD1, a 12-kD protein that shuttles between cytosol and nucleus (Fig. 3) [7]. A few disease mutations have been mapped to the dimerization domain of HNF-1β in patients [8, 90]. These mutations are primarily premature termination codons that render the HNF1B mRNA susceptible to nonsense-mediated decay. Therefore, the phenotype in these patients is driven by HNF-1β haploinsufficiency rather than lack of interaction with PCBD1. Interestingly, homozygous or compound heterozygous PCBD1 mutations in humans are associated with MODY diabetes and renal Mg2+ wasting, although with variable penetrance [91]. These manifestations are in line with a defective function of PCBD1 as a dimerization cofactor for HNF-1β.

The transactivation domain at the C-terminus of HNF-1β interacts with CBP, P/CAF, and HDAC1 [8, 49, 92]. Disease-causing mutations in HNF-1β can lead to defective transactivation function through impaired CBP and P/CAF recruitment in vitro [8]. Accordingly, expression of an HNF-1β C-terminal deletion mutant in transgenic mice causes downregulation of Pkhd1 transcriptions and development of renal cysts, possibly due to a defect in recruiting CBP and P/CAF [49].

Zyxin is a novel HNF-1β interacting protein that binds to the POU domain and the C-terminal transactivation domain of HNF-1β, forming a transcriptional complex that also includes CBP [93]. Zyxin localizes in focal adhesions at sites of cell-matrix interaction but can translocate to the nucleus to enhance HNF-1β-mediated transcription upon stimulation by epidermal growth factor (EGF) (Fig. 3) [93]. Since signaling from the extracellular matrix is important for epithelial polarity and tubular morphogenesis in kidney tubules [94], the regulation of HNF-1β by zyxin may play an important role in kidney development and cystic kidney diseases [93].

In conclusion, lack of interaction between mutated HNF-1β and its partners may contribute to the onset of the HNF-1β-related disease in vivo [8, 49, 92]. In addition, HNF-1β-like renal abnormalities may be attributable to mutations in HNF-1β partners.

Conclusion and future perspectives

During the last decade, the identification of novel HNF-1β target genes in kidney by use of genome-wide techniques has provided new insights into the roles of HNF-1β in the developing and adult kidney [3, 32]. HNF-1β coordinates transcriptional networks involved in nephrogenesis, epithelial differentiation, tubular transport, and intrarenal metabolism. Therefore, patients who carry heterozygous mutations in HNF1B may present with complex renal phenotypes. To date, genotype-phenotype studies have not found a correlation between the type of mutations and the type and/or severity of renal disease, and patients who harbor the same HNF1B mutation show inter- and intrafamilial variability [12]. The multiple steps that influence HNF-1β-mediated gene transcription could explain this diversity. First, genetic variations in HNF-1β interacting proteins may affect the expression of target genes. Depending on the developmental stage, the ability of a regulatory protein to bind HNF-1β could give rise to either hypoplasia or cysts. Second, variations in non-coding regions may affect HNF1B gene expression and/or HNF1B mRNA stability. Finally, additional germline or de novo mutations in other genes that belong to the same biological pathways as HNF-1β may contribute to the phenotypic variability observed in HNF1B mutant patients. To test the hypothesis of modifier effects, a screen of HNF-1β interacting genes in patients should be performed. Future studies should also address the upstream signaling pathways that regulate HNF1B transcription and HNF-1β activity. Dissecting the molecular basis of HNF-1β regulation may offer new targets for the treatment of renal anomalies associated with HNF1B mutations.