Cited1 and Cited2 are differentially expressed in the developing kidney but are not required for nephrogenesis

Abstract

Early kidney development in mammals is characterized by reciprocal tissue interaction between the ureteric bud and the metanephric mesenchyme. The coordinated response to this interaction is regulated largely at the transcriptional level. Here, we investigate the expression and function of Cited1, a transcriptional cofactor that we have previously implicated in kidney development. We show that Cited1 is expressed in the metanephric mesenchyme after invasion of the ureteric bud and that its expression is limited to the cap mesenchyme, those cells that aggregate most tightly around the tip of the ureteric bud and give rise to nephronic epithelium of the adult kidney. Cited1 is down-regulated during the initial stages of epithelial conversion and is not expressed past this progenitor stage. Despite its unique expression pattern, deletion of Cited1 does not disrupt kidney development. We hypothesized that this finding was due to functional redundancy with other members of this gene family. The expression pattern of Cited2 overlaps that of Cited1, but its deletion, either alone or in combination with Cited1, does not disrupt epithelial differentiation of the metanephric mesenchyme. From these studies, we conclude that Cited1 and 2 are dynamically expressed during kidney development, but are not required for nephrogenesis. Developmental Dynamics 236:2321–2330, 2007. © 2007 Wiley-Liss, Inc.

INTRODUCTION

Development of the mammalian kidney is based upon the establishment of a reciprocal tissue interaction between the ureteric bud (UB) and the metanephric mesenchyme (MM; for review, see Dressler,2006). The UB is an epithelial tube that emerges from the nephric duct around embryonic day (E) 10.5 in the mouse, invading the MM, an overlying population of renal progenitor cells in the nephrogenic cord. Invasion of the UB leads to patterning of the MM into the condensed mesenchyme, which surrounds UB tips and branch points, and stromal mesenchyme, a more loosely associated group of cells that encircle the condensed mesenchyme. Cross-talk between the UB, condensed mesenchyme, and stroma drives conversion of progenitor cells into nephronic epithelia and provides for the continued growth and iterative branching of the UB. Much work over the past decade has demonstrated that these events are controlled and integrated at the transcriptional level (for review, see Yu et al.,2004; Boyle and de Caestecker,2006; Schmidt-Ott et al.,2006).

In a screen to identify factors that regulate this process, we previously reported that the transcriptional cofactor Cited1 is expressed in the MM and is down-regulated as these cells undergo epithelial differentiation in response to UB-derived inductive signals (Plisov et al.,2000). Cited1 (Cbp-P300 Interacting Transactivators with E/D rich tails) is the founding member of a family of non-DNA binding transcriptional cofactors that includes Cited2 and Cited4 in mammals and an additional member, Cited3, in lower vertebrates. These factors share a C-terminal transactivation domain and bind core elements of the transcriptional machinery such as P300, in addition to a variety of transcription factors (Shioda et al.,1997; Yahata et al.,2001,2002; Braganca et al.,2002,2003; Plisov et al.,2005). Germline deletion studies implicate these genes in a variety of developmental processes. Cited2 is required for cardiac development as well as neural tube closure and vascularization of the placenta (Bamforth et al.,2001; Weninger et al.,2005; Withington et al.,2006). Cited1 has been implicated in placental trophoblast function (Rodriguez et al.,2004) and maturation of the mammary gland (Howlin et al.,2006), and has been linked to malignancies of the thyroid (Huang et al.,2001; Prasad et al.,2004; Fluge et al.,2006) and skin (Sedghizadeh et al.,2005). The function of Cited family proteins in the developing kidney is unknown.

Previously, we have shown in vitro that Cited1 can act as a bifunctional transcriptional regulator, both activating BMP and repressing WNT-dependent responses (Plisov et al.,2005). Overexpression of Cited1 protein in cultured metanephric kidney explants using a TAT fusion system resulted in the formation of fewer nephronic epithelial structures, suggesting that Cited1 may play a role in maintaining MM cells in an undifferentiated state (Plisov et al.,2005).

Here, we carefully define the temporal and spatial expression pattern of Cited1 in the metanephric kidney, and explore its functional role during nephrogenesis by evaluating Cited1 null mutant mice. As these studies showed that loss of Cited1 expression does not interfere with normal kidney development, we went on to evaluate genetic interactions with another member of the Cited family, Cited2. These studies show that, while Cited1 and Cited2 have overlapping expression domains within the metanephric mesenchyme, both genes are dispensable for nephrogenesis.

RESULTS

Cited1 Is Expressed in the Developing Metanephric Kidney

We created an expression map for Cited1 in the developing kidney using reporter mice in which the Cited1 open reading frame has been replaced with β-galactosidase (Fig. 1; Sado et al.,2000). This insertion can be bred to homozygosity, resulting in mice that are null for Cited1 (Howlin et al.,2006). Previous work has shown that Cited1 mRNA is expressed in the nephric duct at E10.5 (Dunwoodie et al.,1998). Whole-mount analysis confirms expression in the nephric duct at E11 (Fig. 1A) and indicates that as development of the urogenital system progresses, Cited1 expression becomes restricted to the metanephric kidney (Fig. 1B).

Figure 1. Cited1^LacZ expression in the developing kidney. Mapping of Cited1 expression embryonic day (E) 11.0 to adult using Cited1^LacZ/LacZ reporter mice. A: At E11.0; whole-mount staining. Arrowhead, nephric duct. B: Whole-mount of urogenital system from an E15.5 embryo. Inset: Magnification of kidney surface in E17.5, whole-mount preparation. C,D: At E11.0; low-power (C) and high-power (D) images of ureteric bud (UB) invading the metanephric mesenchyme (MM). Arrowhead, UB; arrow, MM. E,F: At E12.0; low-power (E) and high-power (F) images of metanephrogenic region. Arrow, condensing MM; arrowhead, stromal MM. G: At E15.5; Low-power image of metanephric kidney. H: At E15.5; High-power image of branching UB and MM differentiation. arrow, cap mesenchyme; white arrowhead, stromal mesenchyme; black arrowhead, condensed MM and stroma; asterisk, renal vesicle. I: At E17.5; low-power image of metanephric kidney. J: At E17.5; nephrogenic zone; K: Postnatal day (P) 0; metanephric kidney at birth. L: Adult; low-power image of cortical region. FL, forelimb; Ad, adrenal gland; MS, mesonephros; MN, metanephros; Bl, bladder; ND, nephric duct; HG, hindgut; MM, metanephric mesenchyme; UB, ureteric bud.

During the initial stages of metanephric development Cited1 is expressed in the stalk, but not the tip, of the UB (Fig. 1C,D). It is at this time that Cited1 is first detected in the MM, where levels are low and its expression is patchy (Fig. 1C,D). As these cells condense around the branching UB, expression becomes more robust, and by E12, Cited1 is expressed broadly in MM cells (Fig. 1E,F). Beginning at this early stage, its expression becomes limited to a subset of cells within the condensed MM that are most closely aggregated around UB tips, the so called “cap” mesenchyme. The restriction of Cited1 to the cap mesenchyme is striking, with an absence of expression in the condensed MM that lies in the cleft between UB tips (Fig. 1H). This pattern persists throughout nephrogenesis as the UB branches dichotomously and induces the formation of new nephrons (Fig. 1G–J). Cited1 is not expressed in differentiated elements of the mature nephron or collecting system and is absent in the adult kidney (Fig. 1I–L).

To verify the fidelity of the Cited1^LacZ reporter mouse and analyze Cited1 expression at the protein level, we characterized a commercially available anti-Cited1 polyclonal antibody (Neomarkers). The specificity of the antibody was confirmed by the complete lack of staining in Cited1 null kidneys, despite persistence of cap mesenchyme cells (Fig. 2A,B). Using this antibody, we mapped Cited1 protein expression over the same time course shown in the Cited1^LacZ studies. Fidelity of the reporter was confirmed and representative images are shown for E11 (Fig. 2C, compare to Fig. 1D) and E15.5 (Fig. 2D, compare to Fig. 1H). Like βgal reporter expression, Cited1 is restricted to the cap mesenchyme with no expression observed in the clefts between UB tips or in the surrounding stromal mesenchyme.

Figure 2. Cited1 protein expression the developing kidney. A,B: Characterization of anti-Cited1 antibody. Kidneys from embryonic day (E) 15.5 embryos colabeled with rabbit αCited1 (red) and mouse αE-Cadherin (green). A: Wild-type. B:Cited1 null. C,D: Immnohistochemistry using rabbit αCited1 antibody. C: At E11.0. Arrowhead, ureteric bud (UB); arrow, metanephric mesenchyme (MM). Compare with Figure 1D. D: At E15.5. Arrow, cap mesenchyme; white arrowhead, stromal mesenchyme; black arrowhead, condensed MM and stroma (compare with Fig. 1H).

This antibody was also a useful tool to examine the subcellular localization of Cited1 in the cap mesenchyme, as well as its regulation as these cells undergo epithelial differentiation. In the cap mesenchyme, Cited1 is localized primarily in the cytoplasmic compartment (Fig. 3A,B). As these cells begin to differentiate, they migrate around the tip of the UB as the pretubular aggregate (PTA) and form the renal vesicle (RV), marked by expression of E-Cadherin. Dual labeling demonstrates a clear gradient of expression during this transition, with Cited1 levels highest in the cap mesenchyme, reduced in the region of the PTA and completely absent in the RV (Fig. 3B; also see Figs. 1H,J, 2D). By staining sequential sections from E15.5 kidneys, we compared this expression pattern with that of two well-characterized transcriptional regulators within the MM, Pax2 and WT1 (Dressler et al.,1990; Kreidberg et al.,1993; Torres et al.,1995). The expression patterns of Cited1 and Pax2 are largely overlapping in the condensed mesenchyme; however, Pax2 is also expressed in the cleft between UB tips, as well as in the UB itself (Fig. 3C,D). While Cited1 is down-regulated as cap cells begin to differentiate, Pax2 expression is robust in the forming RV. Similarly, expression of Cited1 coincides with that of WT1 in the cap mesenchyme, although WT1 expression increases as MM cells differentiate to form early epithelial structures. Furthermore, WT1 is present in primitive glomerular podocytes, which do not express Cited1 (Fig. 3E,F).

Figure 3. Cited1 is down-regulated as cap mesenchyme cells differentiate and has an expression pattern distinct from Pax2 and WT1.A,B: Colabeling for Cited1 (red) and E-Cadherin (green). A: At embryonic day (E) 12.0. B: At E15.5. Arrow, cap metanephric mesenchyme (MM); arrowhead, pretubular aggregate (PTA); asterisk, renal vesicle (RV). C–F: At E15.5; sequential sections colabeled for either Cited1 (red, C,E), Pax2 (red, D), or WT1 (red, F), and E-Cadherin (green). DAPI (4′,6-diamidine-2-phenylidole-dihydrochloride) stains nuclei blue. C,D: Cited1 and Pax 2 sequential sections. Arrow, cap MM; asterisk, RV; arrowhead, stromal MM. E,F: Cited1 and WT1 sequential sections. arrow, cap MM; asterisk, RV arrowhead, glomerular podocytes; dagger, S-shaped body.

Cited1 Is Not Required for Nephrogenesis

To evaluate the function of Cited1 in the developing kidney, we analyzed Cited1^LacZ/LacZ mice as this targeting strategy results in a complete loss of Cited1 protein (Howlin et al.,2006). This analysis was carried out on the 129/Svj background to avoid the lethal placental defect that is observed in null mice on the C57Bl/6 background (Rodriguez et al.,2004). Cited1 null animals are born in the expected proportions and display no overt physical abnormalities (data not shown). Examination of the kidney, both during development and in the adult, reveals that deletion of Cited1 does not disrupt proper organ formation (Fig. 4). Tetragonolobus purpureas lectin staining demonstrates that proximal tubules form normally in Cited1 null mice and in numbers comparable to wild-type (Fig. 4C–F, data not shown). Formation of glomeruli is also unaffected in these animals. As a measure of total nephrogenic capacity, we counted glomeruli in adult wild-type and Cited1 null animals. No reduction in total glomerular number was observed in the absence of Cited1 (Table 1).

Figure 4. Deletion of Cited1 does not disrupt nephrogenesis. A,B: Low-power hematoxylin and eosin–stained section from adult wild-type (A) and Cited1^LacZ/LacZ (B) mice. C–F:Teragonoglobus purpureas lectin staining of proximal tubules (arrow) in embryonic day (E), 18.5 wild-type (C,E), and Cited1^LacZ/LacZ (D,F) kidneys. Gl, glomeruli.

Table 1. Quantification of Embryonic and Adult Kidney Structures in Wild-Type and Cited Family Mutant Micea

Genotype	n	Ureteric bud tips	Nephronic epithelia	n	Adult glomeruli/kidney
a Metanephric kidneys for culture were isolated at embryonic day 12 and grown for 5 days. Adult mice were >8 weeks of age. n/a, not applicable. *P < 0.01.
Wild-type	11	70 ± 9	52 ± 9	5	7251 ± 182
Cited1^−/−	10	79 ± 7	51 ± 5	4	7192 ± 169
Cited2^+/−	13	78 ± 5	52 ± 6	4	7228 ± 188
Cited2^−/−	13	67 ± 12	43 ± 7*	–	n/a
Cited1^−/−; Cited2^+/−	12	72 ± 6	49 ± 5	4	6987 ± 141
Cited1^−/−; Cited2^−/−	2	48 ± 6	38 ± 7	–	n/a

Cited2 and Cited4 Are Expressed in the Developing Kidney

It is possible that the lack of a kidney phenotype in the Cited1 knockout animal is the result of functional redundancy through other members of this gene family. Reverse transcriptase-polymerase chain reaction (RT-PCR) using total RNA from E15.5 kidneys demonstrates that Cited2 and Cited4 are also expressed in the developing kidney (Fig. 5A). We compared the expression patterns of Cited2 and Cited4 to Cited1 using LacZ reporter animals and in situ hybridization. Whole-mount in situ hybridization analysis of E15.5 kidneys reveals that like Cited1, Cited2 is expressed in the characteristic arc shape marking the MM (Fig. 5C). Analysis of sections from E15.5 kidneys by in situ hybridization and LacZ staining shows Cited2 localized to the condensed mesenchyme, but also persisting in differentiated epithelial structures and glomeruli (Fig. 5E–H). In contrast, Cited4 is expressed only in the UB, most prominently in the tips, in a pattern reciprocal to that of Cited1 (Fig. 5D,E).

Figure 5. Cited2 and Cited4 are expressed in the developing kidney. A: Reverse transcriptase-polymerase chain reaction using RNA extracted from embryonic day (E) 15.5 wild-type kidneys. B–D: Whole-mount staining of E15.5 kidneys showing expression of Cited1 (B, Cited1^LacZ), Cited2 (C, in situ hybridization), and Cited4 (D, Cited4^LacZ). E–H: Expression of Cited2 in E15.5 metanephric kidney. E,F: In situ hybridization (E) confirms fidelity of Cited2^LacZ reporter (F). G,H: High-power images of branching ureteric bud (UB, G) and primitive glomeruli (H) from Cited2^LacZ mice. I: Section of D demonstrating Cited4^LacZ expression in the ureteric bud at E15.5. Arrowhead, UB tips; arrow, metanephric mesenchyme.

Deletion of Cited2 Does Not Disrupt Branching Morphogenesis or Induction of New Nephronic Epithelia

Based on its overlapping expression pattern with Cited1, we focused on Cited2 as a candidate for redundant function in the cap mesenchyme. The role of Cited2 during kidney development has not been previously evaluated. Cited2 null mice die in utero beginning at E13.5 due to cardiac defects (Bamforth et al.,2001; Weninger et al.,2005), complicating embryonic studies and making analysis of adult kidneys impossible. Because of this finding, we used an ex vivo system in which metanephric rudiments were isolated at E12 and grown in culture for a period of 5 days. Dual labeling with the UB marker CalbindinD_28K (red) and the pan epithelial marker E-Cadherin (green) allows us to evaluate branching of the UB and the induction of nephronic epithelia. When these images are merged, branched UB tips appear yellow, expressing both markers, while nephronic epithelial structures express only E-Cadherin and will be green. UB tips and new epithelial structures can then be counted (Supplementary Figure S1, which can be viewed at http://www.interscience.wiley.com/jpages/1058-8388/suppmat).

After 5 days in culture, there were no gross differences in overall morphology of metanephric cultures from wild-type, Cited1 null, or Cited2 null animals (Fig. 6). Quantitative analysis revealed no difference in total number of UB tips or nephronic epithelial structures in Cited1 null cultures as compared with wild-type (Table 1). Cultured kidneys from Cited2 null mice displayed an 18% reduction in nephronic epithelia compared with wild-type (Table 1; mean of 43 per kidney compared with 52 in wild-type). While this effect is statistically significant, the overall growth and UB branching of the rudiment was not significantly impaired and the MM was able to undergo epithelialization.

Figure 6. Deletion of Cited2 does not perturb ureteric bud (UB) branching or induction of new nephronic epithelia. A–F: Representative images of metanephric kidneys isolated at embryonic day (E) 12 and grown for 5 days in culture from: wild-type (WT, A–C), Cited1^LacZ/LacZ (D–F), or Cited2^−/− mice. G–I: Left column stained with UB marker CalbindinD28_K; middle column stained with pan-epithelial marker E-Cadherin; right column merge of two images differentiating UB epithelia from nephronic epithelia.

Compound Deletion of Cited1 and Cited2 Does Not Disrupt Inductive Events in the Developing Kidney

Having established that Cited1 and Cited2 are not individually required for branching morphogenesis or nephron induction during kidney development, we generated compound mutant mice to test our hypothesis that these two proteins could substitute for one another in the MM. During the course of these studies, we were able to generate Cited1^LacZ/LacZ / Cited2^+/− adult animals that had normal kidneys both in terms of morphology and glomerular numbers (data not shown, Table 1). This finding indicates that, if Cited2 compensates for loss of Cited1 in the developing kidney, one copy is sufficient. An organ culture system was again used to evaluate early nephrogenic events owing to the embryonic lethality of Cited2 homozygous deletion. Generation of Cited1/2 double homozygous null embryos for organ culture however, was problematic. Mendelian analysis of our crosses (totaling 14 litters) dictates that, of the 94 embryos recovered, 15 should be null for both Cited1 and Cited2. We recovered, however, only four Cited1/2 double homozygous null animals that survived until E12 (Table 2). We were able to recover tissue from an additional eight reabsorbed embryos, seven of which were double null. This finding indicates that deleting both of these genes accelerates the embryonic lethality observed in Cited2 null animals, increasing the proportion of animals that die before the onset of kidney development. Among those embryos that did reach E12, we observed fewer branch tips and new epithelial structures in cultured kidneys (Fig. 7G–I; Table 1), although their overall morphology appears normal. We were not able to recover a sufficient number of Cited1^LacZ/LacZ / Cited2^−/− animals to demonstrate this to be a statistically significant difference. On this basis we conclude that the key branching and inductive signaling events are intact in Cited1/2 double null kidneys.

Table 2. Distribution of Expected and Observed Survival to Embryonic Day 12 of Relevant Genotypes During Generation of Cited1/2 Compound Mutant Mice

Genotype	Expected	Recovered	Reabsorbed
Wild-type	4–5	5	0
Cited1^−/−	15–16	17	0
Cited2^+/−	7–8	8	0
Cited2^−/−	3–4	5	1
Cited1^−/−// Cited2^+/−	31–32	37	0
Cited1^−/−// Cited2^−/−	16–17	4	7

Figure 7. Compound deletion of Cited1 and Cited2 does not perturb ureteric bud (UB) branching or induction of new nephronic epithelia. A–I: Representative images of metanephric kidneys isolated at embryonic day (E) 12 and grown for 5 days in culture from wild-type (WT, A–C), Cited1^LacZ/LacZ//Cited2^+/− (D–F), or Cited1^LacZLacZ//Cited2^−/− (G–I) mice. Left column stained with UB marker CalbindinD28_K; middle column stained with pan-epithelial marker E-Cadherin; right column merge of two images differentiating UB epithelia from nephronic epithelia.

DISCUSSION

In these studies, we have evaluated the expression and role of the Cited family of transcriptional cofactors during kidney development in the mouse. We have shown that Cited1 expression is initiated in the MM at E11 and is subsequently restricted to a subset of cells in the condensed mesenchyme closely aggregated around UB tips. Its expression is down-regulated as these cells undergo epithelial conversion and is absent in all differentiated structures. Despite its dynamic expression pattern, deletion of Cited1 does not disrupt kidney development. We hypothesized that this finding was due to functional redundancy through other Cited family members. Indeed, Cited2 and Cited4 are also expressed in the developing kidney. Cited2 is expressed in the MM, although unlike Cited1 its expression persists in differentiated epithelial elements. In contrast, Cited4 is expressed only in the tips of the UB, in those cells directly opposed to Cited1-expressing cells. Owing to its overlapping expression pattern, we focused on Cited2 as a candidate to compensate for loss of Cited1. Neither deletion of Cited2 alone, nor in combination with Cited1 overtly disrupted branching morphogenesis or induction of new epithelia in cultured kidneys.

The striking expression pattern of Cited1 is unique among other characterized transcriptional regulators of kidney development. Unlike factors such as Pax2 (Dressler et al.,1990), WT1(Kreidberg et al.,1993), and Six1 (Xu et al.,2003), Cited1 is not present in the MM before UB invasion. After invasion, expression in the MM is patchy, but strong in the nephric duct. As the UB undergoes initial branching events Cited1 is up-regulated in the cap mesenchyme. These are renal progenitor cells that are directly responding to inductive cues from the UB and will soon undergo epithelial differentiation en route to becoming nephrons. It is also thought that a group of cells within the cap are responsible for self-renewal of the MM, although this has yet to be clearly demonstrated and no markers have been identified that distinguish this niche. Expression of Cited1 within the cap mesenchyme persists throughout nephrogenesis, and overlaps with that of Six2, which is expressed in the condensed mesenchyme and is down-regulated as these cells undergo differentiation (Self et al.,2006). However, unlike Six2, Cited1 is restricted to the cap mesenchyme and is not expressed in cells in the cleft between UB tips.

Cited1 is down-regulated as cap cells form the PTA and begin to undergo epithelial differentiation. This finding contrasts Pax2 and WT1, which are expressed in the MM, but are up-regulated as early epithelial structures form. Furthermore, WT1 is expressed in glomerular podocytes, both in the embryo and the adult. Like Pax2, Cited1 is not expressed in the adult kidney.

Our studies show that Cited1 protein is primarily localized within the cytoplasmic compartment of cells within the cap mesenchyme. This is consistent with data demonstrating the existence of a strong nuclear export signal in the C-terminal domain of Cited1 (Shi et al.,2006). As the primary role of Cited family proteins relates to their ability to interact with and modify CBP and p300-dependent transcriptional responses (Shioda et al.,1997; Yahata et al.,2001,2002; Braganca et al.,2002,2003), these findings suggest that nuclear export may serve to limit transcriptional effects of Cited1 in the cap mesenchyme. Interestingly, Cited1 expression in Wilms' tumors, a pediatric malignancy of renal origin characterized by the persistence of undifferentiated MM cells (Rivera and Haber,2005), is primarily confined to the nucleus (unpublished data, Mark de Caestecker). This difference from normal nephrogenesis raises questions about the regulation and functional impact of changes in Cited1 localization during abnormal MM differentiation, as well as the possibility that Cited1 may serve a nontranscriptional function in the cytoplasm of cap mesenchyme cells.

We relied on in situ hybridization and Cited2 and Cited4 LacZ reporter mice to analyze gene expression during kidney development as we were unable to detect Cited2 and Cited4 in the developing kidney using antibody staining (despite published reports using the same antibodies in different tissues; Bhattacharya et al.,1999; Yahata et al.,2002). Results from these experiments show that Cited1 and Cited2 expression overlaps in condensed MM cells, but that Cited2 persists in early epithelial structures and is present in glomeruli, as previously reported (Takemoto et al.,2006). In contrast, Cited4 is not present in the MM, but is expressed most prominently in the UB tips. This expression pattern is reminiscent of that of the tyrosine kinase receptor Ret, which is responsible for growth of the UB in response to MM-derived signals (Pachnis et al.,1993; Schuchardt et al.,1996). The function of Cited4 during development is unknown; however, homozygous null animals are viable, and have morphologically normal kidneys (unpublished data, Mark de Caestecker).

Based on these findings and the observation that Cited4 expression did not overlap with that of Cited1, we went on to evaluate whether Cited2 might be compensating for loss of Cited1 in the condensed MM of Cited1 null mice. Initial studies evaluated Cited2 null mutant mice. In examining cultured kidneys from Cited2 null animals, we did see a decrease in the formation of new epithelial structures, although the overall morphology of the kidney was not perturbed. This modest effect is likely due to a general underdevelopment of Cited2 null animals resulting from the associated cardiac and/or placental defects at this stage of development (Bamforth et al.,2001; Weninger et al.,2005; Withington et al.,2006). Having established that Cited2 null mice did not have a major defect in nephrogenesis, we went on to ascertain whether loss of Cited2 expression in a Cited1 null background would unmask a defect in nephrogenesis. We were only able to recover a small number of Cited1/2 double null embryos during the course of our studies, as deletion of both genes resulted in earlier embryonic lethality than is observed in mice lacking only Cited2 (Bamforth,2001; Weninger,2005). This effect may result from exacerbation of cardiac and/or placental defects in double null animals as Cited1 is also expressed in the developing heart and placental trophoblasts (Dunwoodie et al.,1998; Rodriguez et al.,2004). However, despite some reduction in overall growth compared with wild-type kidneys at the same gestational age, we did observe induction of nephronic epithelia and relatively normal UB branching in cultured metanephric kidneys that we were able to isolate from two of the Cited1/2 double null embryos. Furthermore, we were able to generate Cited1 null / Cited2 heterozygous compound mutant mice and show that these mice lived into adulthood had normal kidneys. Taken together, these findings indicate that Cited2 is not required for nephrogenesis and that its expression does not compensate for loss of Cited1 in the condensed mesenchyme.

In summary, these studies show that Cited1 is not required for kidney development and that deletion of Cited2 does not disrupt branching of the UB or the ability of the MM to undergo epithelial differentiation. Furthermore, while compound deletion of Cited1 and Cited2 hastens the onset of embryonic lethality observed in Cited2 null mice, it does not interfere with inductive events in the kidney. Based on these findings, we conclude that, despite their dynamic expression patterns within the MM, both Cited1 and Cited2 are dispensable for nephrogenesis.

EXPERIMENTAL PROCEDURES

Animals

Generation and genotyping of the Cited1^LacZ (Sado et al.,2000; Howlin et al.,2006), Cited2^+/− (Bamforth et al.,2001), and Cited2^LacZ (Barbera et al.,2002) mice has been previously described. Cited4^LacZ mice were made in the T. Shioda lab in a manner similar to the Cited1^LacZ animals, with a β-galactosidase/neomycin cassette replacing the open reading frame (Supplementary Figure S2). Primers used to detect the wild-type Cited4 allele were F-tcaaggtagtgtctagccca; R-tgagctgttgagagccacca. Primers used to detect the Cited4^LacZ insertion were F-ccgggtgagacagtgaacgaatccgatttattc R-gcgagtaacaacccgtcggattctccgtgg. Cited1^LacZ mice are maintained on the 129/SvJ background. Cited2^+/− and Cited4^LacZ mice are maintained on the C57Bl/6 background, and Cited2^+/LacZ mice on a mixed background. Timed pregnancies were counted with the morning of vaginal plug appearance as day 0.5.

LacZ Staining

Whole embryos or kidneys were isolated in cold phosphate buffered saline (PBS) and fixed in 0.2% glutaraldehyde in PBS with 2 mM MgCl₂ and 5 mM ethyleneglycoltetraacetic acid (EGTA). Fixation times were determined empirically and varied from 1 to 4 hr at 4°C according to age. For whole-mount staining, embryos were equilibrated in βGal wash (0.1 M phosphate buffer pH 7.3 containing 2 mM MgCl₂, 5 mM EGTA, 0.02% NP40, 0.01% Na deoxycholate) and stained overnight at 37°C in βGal wash containing 1 mg/ml XGal and 5 mM K ferro- and ferricyanate. Embryos were cleared for photography using a glycerol gradient in PBS. For frozen sections, fixed tissues were cryoprotected in 30% sucrose in PBS overnight, sectioned at 10 μm, and stained as described above. Sections were counterstained with eosin, dehydrated, and mounted.

Immunohistochemistry

Paraffin-embedded tissues were sectioned at 6 μm, and subject to antigen retrieval using buffered citrate. After allowing slides to cool, tissue was blocked in 10% goat serum in PBS for 1 hr. Sections were incubated with primary antibodies overnight at 4°C. After washing 3 x 5 min in PBS, sections were incubated with secondary antibodies for 1 hr at room temperature, washed again 3 × 5 min in PBS, and mounted with Vectashield containing DAPI (4′,6-diamidine-2-phenylidole-dihydrochloride; Vector Labs).

Antibodies

Polyclonal rabbit anti-Cited1 (NeoMarkers no. RB-9219) was used at 1:250. Monoclonal mouse anti–E-Cadherin (BD Biosciences catalog no. 610181) was used at 1:300. Polyclonal rabbit anti-calbindin D28_K (Calbiochem no. PC253L) was used at 1:1,000. Polyclonal rabbit anti-Pax2 (Zymed/Invitrogen no. 71-6000) and WT1 (Santa Cruz no. SC192) were used at 1:100. Horseradish peroxidase– conjugated Tetragonolobus purpureas lectin, (Sigma no. L5759) was used at 0.3 mg/ml. Conjugated secondary antibodies included horse anti-mouse fluoresine (Vector Labs no. FI-2000) and goat anti-rabbit Rhodamine X (Jackson Immunoresearch no. 111-295-144), both used at 1:250.

Organ Culture Isolation and Staining

Metanephric rudiments were isolated at E12, and grown for 5 days on clear 0.4 μm transwell filters (Corning Costar no. 3460) in DMEM containing 10% fetal bovine serum. For immunostaining, filters were fixed in cold methanol for 10 min, and incubated with primary antibodies diluted in 5% goat serum in PBST (PBS + 0.1% Tween 20) overnight at 4°C. Filters were washed extensively the next day at room temperature in PBST, with the final wash extending overnight at 4μC. Cultures were incubated with secondary antibodies for 1 hr at room temperature, washed 3 times 1 hr in PBST and mounted with Vectashield (Vector Labs).

Glomerular Counts

Glomeruli were counted as previously described (Godley et al.,1996). Briefly, individual kidneys were isolated from adult mice (> 8 weeks of age) and minced into 2-mm cubes. Fragments were incubated in 5 ml of 6 M HCl at 37°C for 90 min. Tissue was homogenized through repeated pipetting and 25 ml of H₂O added. After overnight incubation at 4°C, glomeruli in 5 × 1 ml of this solution were counted in a 35-mm counting dish. Total glomerular number per kidney was extrapolated mathematically from the mean of these five counts.

RT-PCR

RNA was extracted from E15.5 kidneys using absolutely RNA microprep kit (Stratagene no. 400805). RT reactions were carried out using 1 μg of total RNA and primed with random hexamers. PCR primer sets were designed to cross intron/exon boundaries (except for Cited4 which only has 1 exon) to distinguish between cDNA and genomic DNA in case of contamination. Annealing temperature was 60°C. Primer sets were as follows: mCited1 F, gcacttgatgtcaagggtgg; mCited1 R, gagagacagatcccggagac; mCited2 F, gagcagaaatcgcaaagacg; mCited2 R, tgttgagcttctgcagctcg; mCited4 F, tcaaggtagtgtctagccca; mCited4 R, tgagctgttgagagccacca.

In Situ Hybridization

In situ hybridization for Cited2 was preformed as previously described. (Hogan,1994; Dunwoodie et al.,1998). Digoxigenin-labeled riboprobes were generated from a plasmid containing full-length mouse Cited2. Tissue was treated with PK for 7 min, and probes were hybridized overnight at 55°C. After extensive washing, tissues were incubated with alkaline phosphatase–conjugated α-digoxigenin antibody (Roche no. 1093274) at 1:1,000, overnight at 4°C. Tissue was developed with BM purple AP substrate (Roche no. 1442074).

Acknowledgements

We thank Sally Dunwoodie (Cited2^LacZ) and Shoumo Bhattacharya (Cited2^+/−) for their generous gifts of mice and David Frank and Melissa Langworthy for their valuable technical assistance.