A Mutant Ahr Allele Protects the Embryonic Kidney from Hydrocarbon-Induced Deficits in Fetal Programming

Background: The use of experimental model systems has expedited the elucidation of pathogenetic mechanisms of renal developmental disease in humans and the identification of genes that orchestrate developmental programming during nephrogenesis. Objectives: We conducted studies to evaluate the role of AHR polymorphisms in the disruption of renal developmental programming by benzo(a)pyrene (BaP). Methods: We used metanephric cultures of C57BL/6J (C57) mice expressing the Ahrb-1 allele and B6.D2N-Ahrd/J (D2N) mice expressing a mutant allele deficient in ligand binding (Ahrd) to investigate molecular mechanisms of renal development. Deficits in fetal programming were evaluated in the offspring of pregnant mice treated with BaP during nephrogenesis. Results: Hydrocarbon challenge of metanephri from C57 mice altered Wilms’ tumor suppressor gene (Wt1) mRNA splice variant ratios and reduced mRNAs of the Wt1 transcriptional targets syndecan-1 (Sdc1) paired box gene 2 (Pax2), epidermal growth factor receptor (Egfr), and retinoic acid receptor, alpha (Rarα). These changes correlated with down-regulation of effectors of differentiation [secreted frizzled-related sequence protein 1 (Sfrp1), insulin-like growth factor 1 receptor (Igf1r), wingless-related MMTV-integration site 4 (Wnt4), Lim homeobox protein 1 (Lhx1), E-cadherin]. In contrast, metanephri from D2N mice were spared hydrocarbon-induced changes in Wt1 splice variant ratios and deficits of differentiation. We observed similar patterns of dysmorphogenesis and progressive loss of renal function at postnatal weeks 7 and 52 in the offspring of pregnant C57 but not D2N mice gavaged with 0.1 or 0.5 mg/kg BaP on gestation days 10–13. Conclusions: These findings support a functional link between AHR and WT1 in the regulation of renal morphogenesis and raise important questions about the contribution of human AHR polymorphisms to the fetal origins of adult-onset kidney disease.


Research
Embryonic develop ment requires the orches tration of temporally precise genetic events that culminate in formation of a complete organism and that respond to diverse environ mental and somatic signals in utero. Previous studies identified the aryl hydrocarbon recep tor (AHR) as an important nuclear transcrip tion factor during mammalian embryogenesis and throughout maturity. Ligandactivated AHR interacts with hypoxiainducible fac tor 1β (HIF1β) to form transcriptional com plexes that bind specific DNA sequences to regulate genetic targets (Denison and Nagy 2003). Nuclear translocation triggers proteo lytic degradation of AHR protein, a signaling event that defines the biologi cal responsive ness to AHR ligands (Roberts and Whitelaw 1999). AHR protein is highly polymorphic in the core ligandbinding and trans activa tion domains, and this variability dictates differ ences in susceptibility to environ mental injury in mice (Moriguchi et al. 2003) and humans (Sasaki et al. 2006).
AHR participates in develop mental regula tion of vascular structures in liver (Fernandez Salguero et al. 1995;Lahvis et al. 2005;Schmidt et al. 1996), as well as morpho genesis of heart (Lund et al. 2003(Lund et al. , 2006 and kidney (Falahatpisheh and Ramos 2003;McMillan and Bradfield 2007). A role for AHR in renal develop ment was firmly established in experi ments showing that Ahrnull mice exhibit deficits in renal condensation, appearance of differentiated structures, and cellular pro liferation (Falahatpisheh and Ramos 2003). The kidney derives from the nephric ridge of the inter mediate meso derm where the pro nephric duct elongates to form the Wolffian duct, which in turn gives rise to the ureteric bud and sets the stage for condensation of metanephric mesenchyme and mesynchemal toepithelial cell transition. Wilms' tumor suppressor gene (Wt1) functions as a criti cal regulator of nephro genesis by encoding a Cis 2 His 2 zincfinger protein that functions via transcriptional regulation of insulinlike growth factor 2 (Igf2r), syndecan1 (Sdc1), epidermal growth factor receptor (Egfr), and retinoic acid receptor, alpha (Rarα), among other targets (Avner 1993;Hosono et al. 1999). Homozygous Wt1 -/mice do not undergo differentiation from pro nephros to metanephros and die in utero (Kreidberg et al. 1993). Wt1 activity can be regulated in cis by the different ratios of its own splice variants and in trans by proteins such as bone marrow zinc finger 2 (BMZF2) (Lee et al. 2002). The most studied regulatory mechanism of Wt1 involves the formation of +KTS and -KTS splice variants. KTS splice variants originate from the insertion of a lysinethreonineser ine between the third and fourth zinc fingers, and this change regulates Wt1 DNA binding specificity (Menke et al. 1998). In humans, reduced WT1 +KTS mRNA isoforms result in severe kidney and gonad develop mental deficits, known as Frasier syndrome (Barbaux et al. 1997). Changes in exon 5 splice vari ants are also associated with deficits in renal differentia tion (Iben and RoyerPokora 1999). Addition of 17 amino acids in exon 5 creates an mRNA isoform (17aa) that regulates trans activation (Wang et al. 1995). Nterminal residues 1-182 encode a dimerization region implicated in the regulatory mecha nism exerted by dominant negative mutants (Englert et al. 1995).
The precise mechanism by which AHR regulates genetic elements during nephro genesis is not known. Given the requisite activation of AHR by endogenous or exoge nous ligands, it is likely that develop mental regulatory functions of AHR involve tran scriptional regulation of genes during early kidney morpho genesis. The polymorphic nature of the AHR locus suggests that the inherent susceptibility of mice and humans to develop mental interference by AHR ligands is variable. Such relationships can be studied taking advantage of murine models express ing variant AHR proteins that structurally and functionally resemble those in humans. The C57BL/6J (C57) mouse is perhaps the most widely used mouse model to evaluate the biology of AHR (Moriguchi et al. 2003). This strain is naturally sensitive to exogenous AHR ligands, as evidenced by trans regulation of AHRregulated genes (Moriguchi et al. 2003). In contrast, B6.D2NAhr d /J (D2N) mice are highly resistant to ligandmediated AHR acti vation and gene transactivation. The C57 and Background: The use of experimental model systems has expedited the elucidation of pathogenetic mechanisms of renal developmental disease in humans and the identification of genes that orchestrate developmental programming during nephro genesis. oBjectives: We conducted studies to evaluate the role of AHR polymorphisms in the disruption of renal developmental programming by benzo(a)pyrene (BaP). Methods: We used metanephric cultures of C57BL/6J (C57) mice expressing the Ahr b-1 allele and B6.D2N-Ahr d /J (D2N) mice expressing a mutant allele deficient in ligand binding (Ahr d ) to investigate molecular mechanisms of renal development. Deficits in fetal programming were evaluated in the offspring of pregnant mice treated with BaP during nephro genesis. results: Hydrocarbon challenge of metanephri from C57 mice altered Wilms' tumor suppressor gene (Wt1) mRNA splice variant ratios and reduced mRNAs of the Wt1 transcriptional targets syndecan-1 (Sdc1) paired box gene 2 (Pax2), epidermal growth factor receptor (Egfr), and retinoic acid receptor, alpha (Rarα). These changes correlated with down-regulation of effectors of differentiation [secreted frizzled-related sequence protein 1 (Sfrp1), insulin-like growth factor 1 receptor (Igf1r), wingless-related MMTV-integration site 4 (Wnt4), Lim homeobox protein 1 (Lhx1), E-cadherin]. In contrast, metanephri from D2N mice were spared hydrocarbon-induced changes in Wt1 splice variant ratios and deficits of differentiation. We observed similar patterns of dys morpho genesis and progressive loss of renal function at postnatal weeks 7 and 52 in the offspring of pregnant C57 but not D2N mice gavaged with 0.1 or 0.5 mg/kg BaP on gestation days 10-13. conclusions: These findings support a functional link between AHR and WT1 in the regulation of renal morpho genesis and raise important questions about the contribution of human AHR polymorphisms to the fetal origins of adult-onset kidney disease. key words: aryl hydrocarbon receptor, benzo(a)pyrene, fetal programming, nephro genesis, WT1.  mice are isogenic strains that differ in AHR ligandbinding affinity due to a single nucleo tide substitution that replaces valine for alanine at codon 375 in the ligandbinding domain, resulting in a 10fold reduction in ligandbinding affinity (Poland et al. 1994). Humans predominantly express an AHR protein of reduced ligandbinding affinity similar to that in D2N mice, with several poly morphic variants identified to date that exhibit increased ligandbinding affinities and trans activation potentials (Sasaki et al. 2006). Some of these variants afford heightened suscepti bility to cancer, and possibly develop mental interference. We conducted the present study to evaluate the impact of polymorphic variants of the AHR on hydrocarboninduced deficits in nephro genesis. We present evidence that a mutant Ahr allele deficient in ligandbinding affinity and nuclear transactivation protects the developing murine kidney from hydrocarbon induced deficits of fetal genetic programming and loss of renal function in adult life. These findings implicate AHR in the regulation of renal develop mental programming and the fetal basis of adultonset kidney disease.

Materials and Methods
Metanephric cultures. On gestation day (GD) 11.5, mouse embryos were dissected from C57BL/6J Ahr b-1/b-1 wildtype and B6.D2NAhr d/d /J mice (Jackson Laboratories, Bar Harbor, ME). All animals were treated humanely and with regard for allevia tion of suffering. Metanephri were cultured on 0.45mm polyethylene terephthalate cyclopore cell culture inserts (Fisher Scientific, Pittsburgh, PA) for 1-4 days. Kidney explants were maintained at the liquid-gas interface in a solution consisting of a 1:1 mixture of Dulbecco's modified Eagle's medium and F12 medium supplemented with 10% fetal bovine serum and a 5× concentration of MITO serum extender (Becton Dickenson, Bedford, MA). Explants were equilibrated for 1 day before start of experi ments. Seven or more kid ney explants from four dams were placed on individual inserts and exposed daily to 3 μM benzo(a)pyrene (BaP) or an equivalent volume of dimethyl sulfoxide (DMSO), with or with out 20 nM αnapthoflavone (αNF), for 1, 3, or 4 days. αNF is a ligand of AHR that fails to induce conformational changes for efficient transactivation of target genes and thus func tions as a competitive antagonist. The dose of BaP tested in vitro represents an environ mentally relevant dose that affords optimal activation of AHR (Bowes and Ramos 1994). Explants were fixed in situ and processed for further evaluation.
Intrauterine exposures to BaP. Timed pregnant C57BL/6J Ahr b-1/b-1 and B6.D2N Ahr d/d /J mice were gavaged on GDs 10-13 with 0.1 or 0.5 mg/kg BaP or medium chain triglyceride oil (MCT; Mead Johnson Nutritionals, Evansville, IN). MCT is a well characterized emulsion vehicle optimal for oral BaP formulations. The doses examined are considerably lower than those used previously in studies of teratogenicity (Lummus and Henningsen 1995;MacKenzie and Angevine 1981;Rodriquez et al. 1999;Wells et al. 1997;Winn and Wells 1997) and approximate human exposures among atrisk populations when corrected for differences in ontogenic profiles (Menzie et al. 1992;Rebagliato et al. 1995). One week after birth, some pups were euthanized and kidney, heart, liver, testis, and aorta were fixed in situ under physiological pressure; others were kept for studies of renal structure and function 52 weeks after birth.
Histology and morphometric analysis. Metanephri were fixed in 4% paraformal dehyde at 4°C for 16 hr, immobilized in Histogel (Richard Allan Scientific, Kalamazoo, MI), and embedded in paraffin. Serial sections (4 μm) were cut and stained with hematoxylin and eosin (H&E) for visualization of differen tiated structures. Images of at least five meta nephri per treatment group were captured with an Axiovert 200 inverted microscope (Carl Zeiss Microscopy, Thornwood, NY) and stored as ZVI files. Glomeruli and Sshaped and commashaped bodies were quanti fied using manual functions in AxioVision (release 4.1; Carl Zeiss Microscopy). For in vivo measure ments, kidneys were fixed in 4% paraformaldehyde at 4°C for 12 hr and embedded in paraffin. Sections (5 μm) were cut and stained with H&E. Images of entire kidney crosssections from five different renal planes were captured and analyzed using AxioVision (release 4.3) image analysis soft ware (Carl Zeiss Microscopy). All values were normalized to renal area.
Immunohistochemistry. Slides were exposed under pressure to Antigen Unmasking Solution® (Vector Labratories, Burlingame, CA). Sections were incubated with Wilms tumor suppressor (WT1) antibody (180 amino acids in length; Santa Cruz Biotechnology, Santa Cruz, CA) or AHR rabbit polyclonal antibodies (Biomol International, Plymouth Meeting, PA) overnight at 4°C in a solution of 0.3% TritonX and 5-10% goat serum. Primary antibodies were bound to a goat antirabbit biotinylated secon dary antibody (InvitrogenMolecular Probes, Carlsbad, CA), amplified with the the Vectastain Elite ABC Kit; Vector Laboratories), developed with diamino benzidine (DAB), and counter stained with Mayer's hematoxylin (Vector Laboratories). Threshold optimization was completed relative to negative controls, and indices of protein expression were expressed as sum density normalized to total area. Podocyte numbers were quantified using WT1 signal filtered for color, intensity, and size. All values were normalized to glomerular density.
Quantitative polymerase chain reaction (PCR). Total RNA was extracted using TRIzol® (Invitrogen, Carlsbad, CA) and cDNA synthesized using Super Script II (Invitrogen) per manufacturer's instructions. Quantitative PCR was performed to detect differences in the ratio of Wt1 splice variants (±KTS) in response to AHR ligand treatment (Falahatpisheh and Ramos 2003). All primers were designed using Beacon Designer (ver sion 5.1; PREMIER Biosoft, Palo Alto, CA) to create amplicons from 150 to 300 base pairs with an average melting temperature of 55°C. All primers used are listed in Table 1.
Urinary albumin. Individual urine sam ples were stored at -80°C in stabilizing buffer (Biotrin International, Dublin, Ireland). One microliter of sample was loaded onto 4-12% NuPAGE BisTris gel (Invitrogen) under reducing conditions and processed for silver staining per the manufacturer's specifications. Sum density values were calibrated to mouse serum albumin standards ranging from 10 to 0.001 μg/μL. Urinary renal papillary antigen 1 (RPA1) and glutathione S-transferase Yb1 (GSTYb1) measurements. Ninetysix-well microtiter plates were conjugated with antiRPA1 or GSTYb1 IgG. Urine was diluted 1:25 and equilibrated for 1 hr at room temperature before addition of antibody-enzyme conju gate. After substrate develop ment, absorbance was read at 450 nm using 630 nm as a refer ence. Absorbance was normalized to internal controls and expressed as relative units.
Western blot analysis. Protein was extracted using TPER reagent (Pierce, Rockford, IL) per the manufacturer's specifications. Samples were run on 4-12% NuPAGE BisTris gels under reducing conditions, transferred to a polyvinyl difluoride membrane, and probed with WT1(180) rabbit polyclonal antibody (Santa Cruz Biotechnology) and horseradish peroxidase-conjugated secondary antibody.
Statistical analysis. Statistical significance was determined as noted using Student's ttest, analysis of variance (ANOVA), Wilcoxon rank sums, least significant difference (LSD), and Tukey post hoc tests at the p < 0.05 level.
To determine whether ligand binding to AHRmediated deficits in renal cell differen tiation, we examined effects of BaP exposure on D2N mice expressing the Ahr d/d allele and C57 mice cotreated with αNF, a competi tive inhibitor at the AHR ligandbinding site. Activation of the lowaffinity Ahr d/d recep tor was not associated with morphologic or genetic deficits in developing kidneys. Although αNF slightly decreased glomerular density, we observed no reciprocal changes in the abundance of undifferentiated structures or alterations in gene expression ( Figure 1B). As expected, αNF efficiently neutralized the actions of BaP on metanephric differentia tion, as evidenced by reversal of dedifferentia tion deficits. Immunohistochemical analysis showed a decrease in AHR protein after 4 days of BaP exposure in Ahr b-1/b-1 mice (Figure 2), whereas the abundance of AHR d/d protein was not influenced by BaP treatment (Figure 2A).
Instead, expression of Ahr d/d , or cotreatment with BaP and αNF, prevented loss of AHR protein. These findings are consistent with the hypothesis that AHR is required for renal Figure 1. BaP inhibits nephro genesis via an Ahr allele-specific mechanism, as shown by metanephric cultures treated with 3 µM BaP or DMSO without or with α-NF (see "Materials and Methods" for details). (A) Photomicrographs of C57 and D2N metanephri stained with H&E. BaP-exposed C57 metanephri display less morphologically distinct differentiated structures compared with the C57 DMSO control. Expression of the Ahr d allele in D2N mice abrogates BaP-induced deficits; co-treatment with the competitive inhibitor α-NF also inhibited BaP effects. Abbreviations: C, comma-shaped bodies; g, glomeruli; s, S-shaped bodies. Bars = 100 µm. (B and C) Quantification (mean ± SD) of glomeruli (B) and comma and S-shaped bodies (C) normalized to area from serial sections (n ≥ 6 metanephri/group). develop mental signaling and that disruption of nephro genesis by BaP requires AHR ligand binding, signaling, and protein degradation. Next, we monitored the expression of mesenchymal [Sfrp1 (secreted frizzled related sequence protein 1)]and epithelial [Igf1r ( insulinlike growth factor 1 receptor), Wnt4 (winglessrelated MMTVintegration site 4), Lhx1 (Lim homeobox protein 1), and Ecadherin] markers of differentiation by qRTPCR to determine if morphologic deficits correlated with modulation of genetic targets. Markers of renal epithelial cell differ entiation (Igf1r, Wnt4, Lhx1, and Ecadherin) were modulated to variable degrees relative to vehicle control after BaP exposure of C57 metanephri for 1, 3, or 4 days ( Figure 3A). Igf1r, Wnt4, Lhx1, and Ecadherin were also significantly downregulated by day 4 of BaP treatment in C57 cultures ( Figure 3A), whereas Sfrp1 was downregulated at all time points ( Figure 3A). Metanephri expressing the Ahr d allele showed variable degrees of induc tion in Sfrp1, Igf1r, and Wnt4 throughout the exposure period and were completely spared deficits in differentiation, except for changes in Lhx1, which showed decreased levels at all time points ( Figure 3B). Collectively, these data indicate that nephro genic deficits induced by BaP require integrity of AHR signaling and involve inter ference with coordinated renal cell differentiation programming.
Ahr allele mediates disruption of Wt1 mRNA splice variants by BaP. BaP exposure of C57 metanephri for 4 consecutive days resulted in 3 and 8fold induction of +KTS and -KTS variants, respectively ( Figure 4A,B). Expression of +17aa or -17aa was not altered by BaP ( Figure 4C,D). Consistent with the known transcriptional repressive activity of WT1, significant reductions in the relative expression of several WT1 targets, includ ing Sdc1, paired box gene 2 (Pax2), Egfr, and Rarα, were observed by day 4 of BaP treatment ( Figure 4E). D2NAhr d/d meta nephri exposed to 3 μM BaP did not exhibit changes in any of the Wt1 mRNA splice variants ( Figure 5A-D) or Wt1 target genes ( Figure 5E), except for Rarα, where the pat tern of regulation was reversed compared with the C57 strain. C57 but not D2N metanephri exhibited decreased taurine transporter (TauT) mRNA (compare Figures 4E and 5E).

Intrauterine exposures to BaP alter renal develop ment and function of the offspring.
Control C57 and D2N mice showed similar numbers of glomeruli ( Figure 6A). In utero exposure to both 0.1 and 0.5 mg/kg BaP dur ing GDs 10-13 caused signifi cant reductions in glomerular numbers ( Figure 6). One week after birth, the kidneys of offspring of C57 dams exposed to BaP exhibited significant reductions in glomerular size and increased numbers of undifferentiated cells compared with controls ( Figure 6B). In contrast, D2N mice expressing the lowaffinity Ahr allele were spared BaP induced glomerular deficits ( Figure 6). C57 and D2N mice exhibited similar amounts of urinary albumin, demonstrating that expression of the Ahr d allele in itself does not compromise renal function in unstressed animals ( Figure 7A). Measurements of urinary albumin 52 weeks after intra uterine exposure to BaP showed dosedependent increases in urinary albumin only in C57 mice express ing the responsive Ahr b-1 allele, whereas Figure 2. AHR expression correlates with nephro genesis, as shown by immunohistochemical analysis of metanephric cultures treated with 3 µM BaP or DMSO without or with α-NF (see "Materials and Methods" for details). Exposure to BaP decreases AHR protein levels, as indicated by DAB staining (A) and density (mean ± SD) of AHR normalized to total area (B). Bars = 100 µm. Protein expression was similar in DMSOtreated C57 and D2N metanephri, but D2N and α-NF-co-treated C57 metanephri were not sensitive to BaP-induced deficits in AHR protein expression.  Figure 7A,B). Immunohistochemical quantification of podocyte numbers revealed no differences between C57 and D2N mice ( Figure 7C).
In utero exposure to 0.1 or 0.5 mg/kg BaP was associated with decreased podocyte numbers in 52weekold C57 but not D2N mice ( Figure 7C). Consistent with this find ing, both doses of BaP decreased WT1 protein expression in an Ahr allele-specific manner ( Figure 7D). BaP exposure did not alter uri nary RPA1 or GSTYb1 levels in either C57 or D2N mice (Figure 8), suggesting that in utero BaP exposure was selectively associated with glomerular pathology and did not involve col lecting duct (RPA1) or distal tubular injury (GSTYb1) (Falkenberg et al. 1996;Hildebrand et al. 1999;Kilty et al. 1998).

Discussion
Significant progress has been made in under standing the genetic basis of renal develop mental disease. Of note is the discovery that mice overexpressing the -KTS isoform of Wt1 have severely compromised renal develop ment, increased stromal tissue, decreased tubular epithelium and glomeruli, and altered podo cyte structure (Hammes et al. 2001). Although the exact molecu lar conse quences resulting from alterations of the four predominant and biologically rele vant iso forms of WT1 remain to be defined, it is well established that the -KTS mRNA encodes a protein that participates in transcriptional regulation, whereas the protein encoded by +KTS mRNA associates with splicing factors (Hastie 2001). Here we present evidence that BaP induces shifts in -KTS that correlate with downregulation of differentiation markers. The actions of BaP are likely mediated by unregulated activation of AHR signaling dur ing nephro genesis and involve the classical AHR-HIF1β macro molecular complex or molecu lar inter actions with NFκB (nuclear factor κB), activator protein 1 (AP1), and/or glucocorticoid receptor (Nanez and Ramos 2010). Of relevance are reports show ing that the glucocorticoid receptor associ ates with peroxisome proliferatoractivated receptor gamma coactivator1 (PGC1), a known regulator of cotranscriptional splic ing (Knutti et al. 2001). Another possibil ity involves direct binding of AHR to the Wt1 promoter, which contains two consensus AHRresponsive elements within a regula tory region known to regulate transcription coupled splicing events (Cohen et al. 1997). Because AHR associates with proteasomal complexes consisting of damagedDNA binding protein 1, aryl hydrocarbon receptor nuclear translocator, transducinβlike 3, and cullin 4B and itself possesses E3 ligase activity (Ohtake et al. 2007), AHR may regulate pro teasomal degradation of factors required for Wt1 splicing. Although the E3 ligase activ ity of AHR has not been fully characterized, AHR ubiquiti na tion occurs in both cyto plasm (Song and Pollenz 2002) and nucleus (Marlowe and Puga 2010). A model detailing a constitutive function for AHR in the regulation of nephro genesis via WT1 is further supported by our finding that use of the genetically resistant D2N strain, which expresses the Ahr d allele or competitive antagonists of AHR, prevented renal develop mental deficits in BaPtreated metanephri and restored normal ratios of Wt1 mRNA iso forms. The modulation of renal differentiation markers in organ culture induced by BaP is consistent with the deficits in kidney morpho genesis observed after intr auterine exposure to BaP and with mounting evidence impli cating AHR in the regulation of prolifera tion, develop ment, adhesion, migration, and proteasomal degradation of several organ sys tems (Nanez and Ramos 2010). Mutation of the ligandbinding domain in the Ahr d allele decreases ligandbinding affinity by 10fold compared with Ahr b-1 and decreases nuclear translocation efficiency, transcriptional activa tion, and degradation upon ligand binding (Poland et al. 1994). Likewise, αNF does not elicit the conforma tional changes necessary for efficient trans loca tion and/or protein degrada tion (Henry and Gasiewicz 2003). Ahrnull mice exhibit delayed nephro genesis and com promised renal develop ment (Falahatpisheh and Ramos 2003) and a hyper tensive pheno type (Lund et al. 2003(Lund et al. , 2006, so the renal pheno type of Ahrnull mice, as well as that of mice treated in utero with BaP, may be mediated by interference with develop mental programming at the meta nephric stage of renal cell differentiation. Together, our find ings implicate AHR as a key regulator of renal morpho genesis and differentiation and suggest that exogenous ligands of AHR alter genetic programming of the kidney during critical periods of development. The disruption of nephro genesis by BaP involved changes in expression of epithelial markers of differentiation. Sfrp1 is a mesen chymal marker expressed in the metanephric medullary and cortical stroma that acts as a regulator of branching morpho genesis and tubule formation (Yoshino et al. 2001). Igf2r is expressed during early nephro genesis, whereas Igf1r is expressed ubiquitously throughout renal maturation (Duong Van Huyen et al. 2003), and loss of either recep tor compromises renal growth and differen tiation (Feld and Hirschberg 1996). Wnt4 is expressed in condensed mesenchyme and pre tubular aggregates (Vainio et al. 1999), and Wnt4 deficiency inhibits tubule develop ment (Stark et al. 1994). Lhx1 is chiefly expressed in the ureteric bud and induced in mesenchy mal aggregates and differentiating comma and Sshaped bodies (Barnes et al. 1994;Karavanov et al. 1998), and murine models of Lhx1 deficiency are born headless and without kidneys despite the presence of other organs (Shawlot and Behringer 1995). Lhx1null mice do not progress past mesonephric development (Shawlot and Behringer 1995). Ecadherin is a general marker of mesynchemaltoepithelial cell transition that is expressed in the ureteric bud epithelium, distal tubule progenitor cells, and most differentiated tubular epithelium (Cho et al. 1998). Ecadherin-null mice show decreased nephrons due to a failure of proper fusion of metanephric mesenchyme to the ureteric bud (Mah et al. 2000). The criti cal functions executed by these genes during metanephric differentiation and the disrup tion of coordinated patterns of gene expression induced by BaP in an AHRspecific manner further implicate the AHR as a critical regula tor of renal cell differentiation.
Fetal programming is a process whereby during a critical window of develop ment a stimulus induces lasting effects on the struc ture or function of the organism (Barker 2004). The evidence we present here shows that intra uterine exposure to BaP during nephro genesis is associated with sustained deficits of renal structure and function that compromise organ function long after birth. These observations are consistent with pre vious correlations linking low birth weight (LBW) and smoking with the disruption of renal develop mental programming. LBW humans have reduced nephron numbers and glomerular hypertrophy and are prone to micro albuminuria, proteinuria, and decreased glomerular filtration (Celsi et al. 1998;Nwagwu et al. 2000;Sanders et al. 2005). Reduced renal capacity leads to glomeru lar hyper tension and compensatory hypertro phy, which in turn is associated with disrup tion of the glomerular basement membrane and glomerulo sclerosis. Sustained glomeru lar injury exacerbates nephron loss, further reducing renal glomerular filtration rate. A reduction in renal capacity further increases blood pressure and completes a futile cycle of glomerulo sclerosis and nephron loss that may ultimately result in progressive renal failure (ZandiNejad et al. 2006).
Our model in which maternal BaP insult results in decreases in podocyte numbers that impair glomerular filtration is strikingly similar to clinical manifestations seen in conditions such as focal segmental glomerulo sclerosis or diabetic nephropathy (Hara et al. 2001;Hayden et al. 2005;Kim et al. 2001;Wolf et al. 2005). Podocytes are terminally differentiated cells that line the glomeru lar basement membrane and act as a sizeselective Figure 7. BaP exposure induces glomerular-specific injury, as shown by albumin (A,B), podocyte number (C), and total WT1 signal (D) in urine collected from 52-week-old C57 b1/b1 and D2N-AHR d/d mice exposed to 0.1 or 0.5 mg/kg BaP or MCT oil in utero (see "Materials and Methods" for details). (A) Expression of the Ahr d/d allele abrogates BaP-induced alterations in albumin urinary levels in C57 mice compared with D2N mice. (B) Silver stain visualization of mouse urinary albumin in C57 mice exposed in utero to 0.5 mg/kg MCT or BaP. (C) Immunohistochemical analysis of podocyte numbers quantified using WT1 signal filtered for intensity, color, and size normalized to glomerular density. (D) Immunohistochemical analysis of total WT1 signal quantified using WT1 normalized to podocyte numbers. For A, C, and D, data are mean ± SD.  Figure 8. BaP exposure does not induce collecting duct or distal tubular injury, as shown by absorbances from enzyme immunoassay detection of RPA1 (A) and GSTYb1 (B) in urine collected from 52-week-old C57 b1/b1 and D2N-AHR d/d mice exposed to 0.1 or 0.5 mg/kg BaP or MCT oil in utero (see "Materials and Methods" for details). Absorbances were normalized to internal controls and expressed as relative units (mean ± SD). *p < 0.05 compared with the corresponding MCT control, by ANOVA and LSD post hoc tests. filter that must structurally encompass the entire glomerular surface with foot processes to maintain proper filtration (Wiggins 2007). Podocytopenia results in denudation of the basement membrane that compromises glom erular function to initiate a cycle of injury and continued podocyte loss, leading to progres sive renal failure. In fact, glomerulopathies are the most common causes of endstage renal disease worldwide (Hricik et al. 1998).

Conclusions
The discovery of novel functions of AHR during nephro genesis highlights a potential mechanism for compromised renal develop ment after disruption of AHR signaling in pregnant women exposed to tobacco smoke or babies exposed post natally to secondhand smoke. Such bioavailability of poly cyclic aro matic hydro carbons provides a ready source for agents that modify early determinants of renal cell differentiation. Of relevance are studies linking AHR genetic poly morphisms to tobac cosmoke-induced decreases in birth weight (Sasaki et al. 2006), along with associated microalbuminuria, proteinuria, and height ened risk of endstage renal disease (Zandi Nejad et al. 2006). Other stressors, such as diet and environmental exposures, are associ ated with LBW and deficits in renal develop ment (Everson et al. 1988;Nelson et al. 1999;Roquer et al. 1995;ZandiNejad et al. 2006). In humans, strong associations have been reported between smoking and albuminuria in non diabetic patients (PintoSietsma et al. 2000) and a higher risk of renal dys function and renal cancer (Bertazzi et al. 1989). The protection afforded in mice by Ahr d suggests that the risk of AHRmediated renal develop mental deficits is most likely relevant in sub jects expressing poly morphic variants of AHR with increased ligand affinity. Such deficits are not severe but instead may compromise renal reserve capacity, leading to increased suscepti bility to lateonset renal disease.