Integrin-Linked Kinase Controls Renal Branching Morphogenesis Via Dual Speciﬁcity Phosphatase 8

BASIC RESEARCH www.jasn.org

Integrin-linked Kinase Controls Renal Branching Morphogenesis via Dual Speciﬁcity Phosphatase 8

†‡ Joanna Smeeton,* Priya Dhir,* Di Hu,* Meghan M. Feeney,* Lin Chen,* and † Norman D. Rosenblum* §

*Program in Developmental and Stem Cell Biology, and §Division of Nephrology, The Hospital for Sick Children, Toronto, Ontario, Canada; and †Departments of Paediatrics, and ‡Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Ontario, Canada

ABSTRACT Integrin-linked kinase (ILK) is an intracellular scaffold protein with critical cell-specific functions in the embryonic and mature mammalian kidney. Previously, we demonstrated a requirement for Ilk during ureteric branching and cell cycle regulation in collecting duct cells in vivo. Although in vitro data indicate that ILK controls p38 mitogen-activated protein kinase (p38MAPK) activity, the contribution of ILK- p38MAPK signaling to branching morphogenesis in vivo is not defined. Here, we identified genes that are regulated by Ilk in ureteric cells using a whole-genome expression analysis of whole-kidney mRNA in mice with Ilk deficiency in the ureteric cell lineage. Six genes with expression in ureteric tip cells, including Wnt11, were downregulated, whereas the expression of dual-specificity phosphatase 8 (DUSP8) was upregulated. Phosphorylation of p38MAPK was decreased in kidney tissue with Ilk deficiency, but no significant decrease in the phosphorylation of other intracellular effectors previously shown to control renal morphogenesis was observed. Pharmacologic inhibition of p38MAPK activity in murine inner medullary collecting duct 3 (mIMCD3) cells decreased expression of Wnt11, Krt23,andSlo4c1.DUSP8over- expression in mIMCD3 cells significantly inhibited p38MAPK activation and the expression of Wnt11 and Slo4c1. Adenovirus-mediated overexpression of DUSP8 in cultured embryonic murine kidneys decreased ureteric branching and p38MAPK activation. Together, these data demonstrate that Ilk controls branching morphogenesis by regulating the expression of DUSP8, which inhibits p38MAPK activity and decreases branching morphogenesis.

J Am Soc Nephrol 27: 1465–1477, 2016. doi: 10.1681/ASN.2015020139

Renal branching morphogenesis, defined as growth contrast, a limited number and variety of intracellular and branching of the ureteric bud (UB) and its molecules that function within these signaling path- derivatives, is essential to mammalian kidney devel- ways to control ureteric branching have been identi- opment. The UB grows, branches, and differentiates fied. In vitro treatment of embryonic urogenital to form the collecting ducts and pelvis of the mature explants with pharmacologic inhibitors revealed dis- kidney. Further, UB tips induce adjacent metanephric tinct roles for extracellular signal-regulated kinase mesenchyme cells to undergo the process of nephro- genesis.1 Defects in renal branching morphogenesis cause congenital renal hypodysplasia, characterized Received February 6, 2015. Accepted August 11, 2015. by abnormal collecting-system morphology and Published online ahead of print. Publication date available at 2 function and low nephron number. www.jasn.org. TheUBarisesasadirectevaginationofthein- fi Correspondence: Dr. Norman D. Rosenblum, The Hospital for termediate mesoderm-derived Wolf an duct. Ex- Sick Children, Peter Gilgan Centre for Research and Learning, tracellular ligands and cell-surface receptors in mul- 686 Bay Street, Toronto, ON M5G 0A4, Canada. Email: norman. tiple distinct molecular signaling pathways function [email protected] during the induction and early patterning of the UB. In Copyright © 2016 by the American Society of Nephrology

J Am Soc Nephrol 27: 1465–1477, 2016 ISSN : 1046-6673/2705-1465 1465 BASIC RESEARCH www.jasn.org

(ERK), protein kinase B (AKT), and p38 mitogen-activated protein kinase (p38MAPK) during renal development.3 Further- more, ERK and AKT can be activated downstream of the tyrosine kinase receptor, RET, in the UB tip domain, suggesting that intracellular kinase signaling is functionally important during ureteric branching.4,5 Integrin-linked kinase (ILK) is a scaffold protein that was initially identified through its ability to interact with the cytoplasmic domain of b-integrins.6 Investigation of ILK function has revealed a requirement for ILK upstream of intracellular kinases AKT,7 glycogen synthase kinase 3b,8 and p38MAPK.9 While each of AKT, glycogen synthase kinase 3b, and p38MAPK have been shown to play a role in controlling renal branching morphogenesis, their role downstream of ILK in the embryonic kidney has not been defined. Nor is it predicted by their functions in nonrenal tissues because their regulation by ILK is context-dependent.10 Previously, we demonstrated that Ilk is required for renal branching morphogenesis in vivo11 and controls UB branching via p38MAPK in vitro.9,12 However, the role of p38MAPK in vivo and the genes that act downstream of Ilk during the early stages of branching morphogenesis have not been previously defined. In this study we characterize the intracellular signaling pathways and UB transcriptome in the early embryonic Ilk- deficient kidney to determine the primary role of Ilk in the regulation of renal branching. Our results demonstrate that Ilk regulates the expression of a distinct subset of UB branching 2 2 genes through both p38MAPK-independent and -dependent Figure 1. E12.5 Ilk / UB kidneys demonstrate reduced UB pathways. Our data also demonstrate upregulation of a phos- branching and abnormal gene regulation. (A,B) GFP branching 2/2UB phatase, dual-specificity phosphatase 8 (DUSP8), not previ- pattern of (A) GFP+ heterozygote and (B) Ilk kidneys con- fi 2/2UB ously implicated in kidney development. We demonstrate a rmed the branching defect in E12.5 Ilk kidneys selected for microarray analysis. (C) Gene expression differs between WT functional role for DUSP8 in attenuating activation of 2/2UB and Ilk mutant (MUT) kidneys with similar expression pat- p38MAPK and p38MAPK-dependent genes and inhibiting terns clustering within the WT and mutant samples. Expression branching morphogenesis. Together, these results describe a values of differentially regulated genes (P,0.003) are shown in novel DUSP8-mediated mechanism by which ILK controls this heatmap, with red indicating lower expression and yellow renal branching morphogenesis. higher expression.

RESULTS To determine changes in expression levels for individual genes, the top table of genes based on the t-statistic ranking was gen- Mice with conditional deletion of Ilk in the developing UB erated. Using a statistical cutoff of P,0.003 to identify changes 2 2 (Ilk / UB) have decreased UB branching at E12.5.11 However, in the expression of individual genes, we identified 131 down- the underlying molecular mechanisms are unknown. We in- regulated probe sets and 96 upregulated (Figure 1C, Supple- vestigated these mechanisms using whole-genome–based anal- mental Tables 2 and 3). The magnitude of changes in gene ysis of gene expression in intact kidney tissue isolated at E12.5 expression, particularly for downregulated genes, was twofold (n=6 kidneys per sample, three biologic replicates per geno- or lower. This may be due to the fact that UB gene expression type). At the time of dissection, the presence of a branching was assayed in whole tissue in which UB cells constitute a mi- phenotype in Ilk-deficient kidneys was documented using nority of cells. While a gene expression analysis in isolated UB green fluorescent protein (GFP) expression driven by the cells may have generated gene expression changes of higher Hoxb7 promoter (Figure 1). Gene expression changes were in- magnitude, the quality of RNA isolated from flow-sorted UB vestigated through microarray analysis using Affymetrix cells derived from mutant mice precluded a whole-genome GeneChip Mouse 430 2.0 arrays. Expression levels were nor- gene expression analysis. Notwithstanding these considera- malized to wild-type (WT) and then analyzed for differential tions, consistent with genetic deletion of Ilk, the most statisti- 2 2 expression between Ilk / UB and WT samples using cally significant gene expression change was a downregulation 2 2 the Bioconductor limma package in R statistical program.13 of Ilk in Ilk / UB kidneys.

1466 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 1465–1477, 2016 www.jasn.org BASIC RESEARCH

Gene Ontology Analysis Suggests Abnormal statistical significance for six genes by qRT-PCR (n=6/genotype, Regulation of Tube Morphogenesis Genes two technical replicates) and these were further confirmed by in To define the global state of gene expression in an Ilk-deficient situ hybridization in tissue sections at E12.5 (Figure 2). In con- state, we performed pathway analysis to identify gene ontology trast to UB tip genes Ret and Sox9, the mRNA expression of which (GO) biologic processes that are enriched within the down- was not significantly altered, in situ hybridization analysis con- regulated gene set. Sixty-one biologic process GO terms were firmed decreased expression of Sox8, Wnt11, Myb, CXCR4, Krt23, 2 2 statistically enriched in the downregulated gene list (Supple- and Slco4c1 in the UB of E12.5 Ilk / UB kidneys (Figure 2). mental Table 4). Within the downregulated gene set, GO terms involved in epithelial tube development as well as the regula- Ilk Deficiency Causes a Specific Decrease in the tion of organ morphogenesis were enriched (Table 1). These Activation of p38MAPK terms are highly consistent with the mutant phenotype and a Because previous studies in nonrenal tissues have implicated ILK putative role for Ilk in controlling the expression of genes in- in the control of kinase activity,7,8 we investigated the state of 2 2 volved in branching morphogenesis and tube morphogenesis. kinase signaling in E13.5 Ilk / UB embryonic kidney tissue protein lysates. Consistent with our prior results in vitro,12 phos- A Subset of Ilk-Dependent Genes have a UB-Specific phorylation of p38MAPK, an indicator of p38MAPK activation, 2 2 Expression Pattern was significantly decreased by 61% in Ilk / UB embryonic kidney Because knockdown of Ilk was specific to the UB lineage, next tissue (Figure 3, A and B). In contrast, we observed no statistically we investigated genes expressed in the ureteric lineage and significant difference in the expression of the phosphorylated 2 2 which function in ureteric branching. Accordingly, the forms of AKT and ERK in Ilk / UB embryonic kidney tissue as whole-kidney gene data set was filtered for genes that are compared with WT (Figure 3, A and B), nor was there a signif- known to be expressed in the UB by crossreferencing the icant difference in the expression of the functionally active form downregulated genes with the microarray and in situ hybrid- of b-catenin protein, in which residues Ser33, Ser37, and Thr41 ization expression data available in GUDMAP (www.gudmap. are not phosphorylated (Figure 3, A and B). Consistent with the org).14 This procedure resulted in reduction of the 131 down- analysis of protein expression, expression of the TCF/LEF-LacZ regulated genes to 14 genes with confirmed UB-specificex- reporter allele, a surrogate measure of canonical WNT target gene 2 2 pression, including Wnt11, a UB tip–specific gene essential for activation, was comparable in Ilk / UB and WT kidney tissue UB development15 (Table 2). While a subset of these genes (Figure 3C). (Wnt11, Sox8, CXCR4, Myb, Etv5, Etv4,andSpry1) are com- ponents of the RET signaling pathway or are RET signaling p38MAPK Activation is Required for the Expression targets,16 expression of Ret mRNAwas not significantly altered of a Subset of Ilk-Dependent Genes 2 2 in Ilk / UB kidneys (Figure 2, B and C). Expression of RETwas We investigated the functional role of p38MAPK in regulating further investigated at the level of protein activity by assaying expression of Ilk-dependent genes in the UB by inhibiting the levels of phosphorylated RET-1062, an isoform critical for p38MAPK function. Inhibition of p38MAPK was achieved RET activity.17 Results demonstrated that levels of phosphor- through treatment of explanted kidneys with commercially ylated RET 1062, assayed by Western blot analysis of protein available p38MAPK inhibitors SB203580 and SB202190. Fol- lysates generated from E12.5 kidney tissue, were not signifi- lowing two days of treatment with SB203580 or SB202190, 2 2 cantly different between Ilk / UB and control mice (Supple- cultured E12.5 embryonic kidneys displayed abnormal UB mental Figure 1, A and B). The expression of Ilk-dependent branch morphology (Figure 4, A–C) and 40% or 66% fewer UB genes was validated by quantitative RT-PCR (qRT-PCR). branches, respectively, compared with vehicle-treated litter- AsshowninFigure2A,differentialexpressionreached mate kidneys (Figure 4D). Next, we determined the effect

2 2UB Table 1. Selected GO terms enriched in genes decreased in Ilk / kidneys GO ID Name P Value Genes in Test Set GO:0009887 Organ morphogenesis 6.22E-08 UNCX|HES1|GCNT3|MYO7A|POU3F4|SOX8|GJA1|FGF9|FG F1|CXCR4|BMP2|LAMA1|BCL11B|JAG1|ATIC|HIF1A|SOST DC1|WNT11|CA2|ETV4|ETV5|ALDH1A1|FAM20C|SPRY1|S PRY2|WNT8B GO:0051094 Positive regulation of 1.66E-07 HES1|CAPRIN1|SOX8|GJA1|NRCAM|CHRNA7|FGF9|FGF1|N developmental process PNT|CXCR4|BTC|NELL1|LRP8|BMP2|KITLG|JAG1|MYB|HIF 1A|CA2|CDH4|ETV5|FAM20C|SPRY1|ILK GO:0002009 Morphogenesis of an epithelium 1.39E-06 HES1|SOX8|GJA1|FGF1|NPNT|CXCR4|BMP2|LAMA1|JAG1| HIF1A|WNT11|CA2|ETV4|ETV5|ALDH1A1|FREM2|ILK GO:0035295 Tube development 2.25E-06 HES1|SOX8|GJA1|FGF9|FGF1|NPNT|CXCR4|ARG2|BMP2|LA MA1|JAG1|HIF1A|WNT11|ETV4|ETV5|SPRY1|SPRY2|ILK

J Am Soc Nephrol 27: 1465–1477, 2016 Ilk Controls Ureteric Branching via DUSP8 1467 BASIC RESEARCH www.jasn.org

2 2UB Table 2. Selected UB-expressed mRNA transcripts downregulated in Ilk / kidneys (P,0.003) Probe ID Gene Name Gene Symbol Fold Change P Value 1435438_at SRY-box containing gene 8 Sox8 22.037065358 8.71E-06 1437870_at Solute carrier organic anion transporter family, member 4C1 Slco4c1 21.963658776 2.48E-05 1418213_at Keratin 23 Krt23 21.835321044 7.02E-04 1448710_at Chemokine (C-X-C motif) receptor 4 Cxcr4 21.613128354 4.71E-05 1450772_at Wingless-related MMTV integration site 11 Wnt11 21.572346139 3.12E-04 1420998_at Ets variant gene 5 Etv5 21.459696529 3.89E-04 1435303_at TAF4B RNA polymerase II, TATA box binding protein-associated factor Taf4b 21.413147086 1.05E-03 1422165_at POU domain, class 3, transcription factor 4 Pou3f4 21.379100185 2.44E-03 1423232_at Ets variant gene 4 (E1A enhancer binding protein, E1AF) Etv4 21.368048482 8.53E-04 1424367_a_at Homer homolog 2 (Drosophila) Homer2 21.344712561 1.14E-03 1450194_a_at Myeloblastosis oncogene Myb 21.307002458 6.57E-04 1415874_at Sprouty homolog 1 (Drosophila); similar to sprouty 1 Spry1 21.304973111 4.38E-04 1418153_at Laminin, a 1 Lama1 21.291465414 1.62E-03 1415855_at Kit ligand Kitl 21.283595843 1.33E-03

of p38MAPK inhibition on the expression of the six Ilk- mIMCD3 cells. Our data demonstrate that overexpression of dependent UB genes using qRT-PCT. While levels of Sox8, DUSP8 results in a significant decrease in the expression of CXCR4,andMyb were not significantly affected by inhibition Wnt11 and Slco4c1 mRNA (Figure 5E). Krt23 was also de- of p38MAPK in kidney explant culture (Figure 4E), expression creased, but the change in expression was not statistically sig- of Wnt11, Krt23,andSlco4c1 were each significantly decreased nificant. mRNA levels of Myb and Sox8 were unchanged. (Figure 4F). These data indicate that Ilk acts via p38MAPK- Because CXCR4 mRNA was undetected in both mIMCD3- dependent and -independent mechanisms to regulate genes and DUSP8-overexpressing mIMCD3 cells, we could not eval- expressed in the developing UB. uate its regulation by DUSP8. Together, these data show that Wnt11 and Slco4c1, both of which are regulated by p38MAPK, Overexpression of DUSP8 Decreases Both p38MAPK are also regulated by DUSP8. Activation and Expression of UB Tip–Specific Genes GO analysis of the complete transcriptome identified altered Adenovirus-Mediated DUSP8 Overexpression in 2 2 regulation of MAPK phosphatases in Ilk / UB kidneys Embryonic Kidney Explants Decreases UB Branching (Table 3). At the individual gene level, the p38MAPK phos- Weinvestigated the functional contribution ofDUSP8 in renal phatase, Dusp8, was among the most significantly upregulated branching morphogenesis in intact kidney tissue. E12.5 WT 2 2 genes in Ilk / UB kidneys (Table 4). We confirmed upregula- kidney explants were treated with either GFP-expressing 2 2 tion of DUSP8 in Ilk / UB embryonic kidneys by qRT-PCR recombinant adenovirus (Ad-Gfp)oradenoviruscarrying analysis (n=6/genotype, two technical replicates; Figure 5A). Gfp and Dusp8 (Ad-Gfp-Dusp8)for5daysat107 plaque- While our published work demonstrated that ILK activates forming units/ml. Transduction efficiency of Dusp8 tran- p38MAPK, the mechanism underlying their interaction is not script was analyzed in three independent experiments by understood. Accordingly, we investigated the functional contri- qRT-PCR analysis (n=9/group, three technical replicates; bution of DUSP8 to p38MAPK activity in the UB lineage using Figure 6A). A minus-reverse transcription (RT) negative- inner medullary collecting duct cells (mIMCD3) in which we control sample was included in qRT-PCR experiments to es- varied DUSP8 expression and stimulated p38MAPK signaling tablish the effective removal of adenoviral DNA in the RNA using EGF12 (Figure 5B). mIMCD3 cells respond to EGF within samples. The mRNA of Dusp8 was 68-fold more abundant in 15 minutes with a 52.2% increase in p38MAPK phosphoryla- the Ad-Gfp-Dusp8–treated explants compared with control tion (Figure 5, C and D). In contrast, no increase in p38MAPK explants. Levels of DUSP8 protein were remarkably higher in phosphorylation was observed in DUSP8-overexpressing Ad-Gfp-Dusp8–treated explants compared with control ex- mIMCD3 cells. While a further increase in p38MAPK was ob- plants (Figure 6B). Next, we measured the effect of DUSP8 served by 30 minutes following administration of EGF in overexpression on ureteric branching. Following 5 days of mIMCD3 cells, no change in p38MAPK phosphorylation treatment with Ad-Gfp-Dusp8, ureteric branches were ana- was observed in DUSP8-overexpressing mIMCD3 cells. lyzed by whole-mount immunofluorescence analysis of cyto- Next, we investigated whether DUSP8 regulates the genes keratin expression (Figure 6, C and D). Quantification of whose expression is regulated by ILK. Expression of p38MAPK- ureteric tips demonstrated a significant branching phenotype regulated genes (Wnt11, Krt23,andSlco4c1) and p38MAPK- in the DUSP8-overexpressing kidneys, with an average of 45 independent genes (Sox8, CXCR4, and Myb) was analyzed after UB branch tips compared with the 74 branch tips present in EGF-stimulation of mIMCD3 cells and DUSP8-overexpressing the control explants (n=14/group) (Figure 6E).

1468 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 1465–1477, 2016 www.jasn.org BASIC RESEARCH

Figure 2. Validation of Ilk-dependent genes confirmed differential expression in the UB. (A) Results of qRT-PCR analysis of UB- 2 2 expressed genes downregulated in Ilk / UB kidneys. Significantly decreased genes Ilk, Sox8, Wnt11, CXCR4, Slco4c1, Myb,andKrt23 are shown (P,0.05). (B) In situ hybridization analysis of gene expression in tissue sections demonstrated normal expression of (B and C) Ret and (R and S) Sox9 in UB tips. (D–Q) Decreased UB-specific expression of Ilk-target genes was confirmed. Arrows indicate UBs.

In parallel experiments we investigated the effect of DUSP8 different mice and treated individually and then pooled to overexpression on p38MAPK activity in embryonic kidney generate sufﬁcient protein for Western blot analysis. The levels explants. Experiments were designed in a manner parallel to of phosphorylated p38MAPK, controlled for total p38MAPK, those using mIMCD3 cells, using EGF to stimulate p38MAPK in each treatment group were compared with Ad-GFP plus activity (Figure 5). Three independent experiments were per- EGF (no DUSP8). Treatment of explants with Ad-GFP- formed, each consisting of three explants isolated from DUSP8 or Ad-GFP did not signiﬁcantly change p38MAPK

J Am Soc Nephrol 27: 1465–1477, 2016 Ilk Controls Ureteric Branching via DUSP8 1469 BASIC RESEARCH www.jasn.org

DISCUSSION

A requirement for Ilk expression in ureteric cells during branching morphogenesis has been demonstrated in the murine kidney.11 Yet, the molecular mechanisms that under- lie its actions have not been previously defined in vivo. Our previous analysis of intracellular signaling downstream of ILK in cultured collecting duct cells with gain or loss of ILK expression suggested that ILK controls tubule morphogenesis via p38MAPK.9 In this study, we demonstrate that Ilk controls p38MAPK-dependent expression of genes involved in renal branching morphogenesis by regulating the expression of DUSP8, which inhibits p38MAPK activity. We first identified genes that are regulated by Ilk in ureteric cells using a whole-genome mRNA expression analysis of whole-kidney mRNA in mice with Ilk deficiency targeted to the UB. The expression of six genes with expression in ureteric tip cells – Wnt11, Krt23, Slo4c1, Myb, Sox8,andCxcr4 – was downregulated in Ilk-deficient tissue. Next, we demonstrated that kidney tissue with Ilk deficiency is characterized by decreased phosphorylated p38MAPK but no significant decrease in the phosphorylation of other intracellular effectors (ERK, AKT, or b-catenin) previously shown to control renal morphogenesis. We then showed that inhibition of p38MAPK activity controls the expression of Wnt11, Krt23,and Slo4c1.UpregulationofDUSP8,adual- specificity phosphatase, in Ilk-deficient kidney tissue suggested a mechanism by which ILK controls p38MAPK and ureteric tip cell gene expression. Our data show that DUSP8 overexpression in murine-collecting duct cells inhibited p38MAPK activation 2/2UB Figure 3. Embryonic Ilk kidneys have decreased activation of p38MAPK. (A) and expression of Wnt11 and Slo4c1 and de- Western blot analysis of E13.5 kidney lysates for expression of phosphorylated (phos- creased ureteric branching and p38MAPK pho) p38MAPK, p38MAPK, phospho-ERK, phospho-AKT, pS33/S37/741 b-catenin, and activation in intact embryonic kidney tissue. total b-catenin. (B) Quantification of protein bands relative to total protein demon- 2 2 strated significantly decreased expression of phospho-p38MAPK in Ilk / UB kidneys Our data support a model of ILK signaling 2 2 (P,0.001). (C) TCF-LacZ expression in the UB lineage is unchanged in Ilk / UB kidneys. during UB branching (Figure 7). Our model suggests that ILK acts upstream of p38MAPK to regulate Wnt11, Krt23,and activation compared with Ad-GFP plus EGF. In contrast, Slco4c1 expression in the UB tip to stimulate UB branching. treatment with Ad-GFP-DUSP8 plus EGF decreased ILK also regulates expression of Sox8, CXCR4,andMyb in the p38MAPK activation by 62% (P,0.02) compared with the UB. Our model further proposes that ILK negatively regulates Ad-GFP plus EGF control (Figure 6F). Together, these data the expression of DUSP8, a protein phosphatase, which can act show that DUSP8 decreases branching morphogenesis and to modulate p38MAPK activation and downstream target gene p38MAPK activation in intact murine kidney tissue. expression.

1470 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 1465–1477, 2016 www.jasn.org BASIC RESEARCH

Figure 4. p38MAPK inhibition in kidney culture causes decreased branching and decreased expression of a subset of Ilk-dependent genes. (A–C) E12.5 Hoxb7Cre-EGFP kidneys were cultured in the presence of DMSO, SB203580, or SB202190 for 48 hours. Branching was visualized by GFP expression in the UB lineage. Insets show high magnification of UB tip regions. (D) Quantification demonstrated decreased tip number in p38MAPK-inhibitor–treated kidneys. (E) qRT-PCR analysis of gene expression demonstrated that Sox8, Myb, and CXCR4 were not significantly altered following inhibition of p38MAPK activity (P.0.05). (F) qRT-PCR analysis demonstrated that expression of Wnt11, Krt23,andSlco4c1 were decreased following inhibition of p38MAPK activity in kidney culture (P,0.05).

Ilk-Dependent Genes and Control of Renal Branching characterized by renal hypoplasia and decreased ureteric branch- 2 2 Morphogenesis ing.15 This phenotype is similar to that observed in Ilk / UB mice The expression of genes (Wnt11, Cxcr4,andSox9) expressed in and suggests that Wnt11 deﬁciency in the absence of Ilk is critical the ureteric tip and with demonstrated functions in ureteric branch- to the genesis of renal hypoplasia. We also observed decreased 2 2 ing are decreased in Ilk / UB mice. Wnt11-deﬁcient mice are expression of Cxcr4, a chemokine encoding gene that is highly

J Am Soc Nephrol 27: 1465–1477, 2016 Ilk Controls Ureteric Branching via DUSP8 1471 BASIC RESEARCH www.jasn.org

2 2 expressed in the UB tip. CXCR4 / mice are characterized by a mild which could play a structural role specifictotheureterictip.The degree of renal hypoplasia.18 Pharmacologic inhibition of CXCR4 function of Slco4c1, a unique marker of the ureteric tip domain22 in embryonic kidney explants inhibits branching morphogenesis.19 and a uremic toxin transporter protein,23 remains to be defined. Moreover, the overall state of chemokine signaling is misregulated in Growth factor signaling via the RET receptor is crucial for 2 2 Ilk / UB kidneys (Supplemental Table 5, Table 3). Together, these renal branching morphogenesis. Ret expression and phos- 2 2 observations suggest that loss of Cxcr4 may also contribute to the phorylation is not decreased in Ilk / UB kidney tissue. Inter- 2 2 hypoplastic phenotype in Ilk / UB mice. Sox8 was markedly de- estingly, UB isolated from Ilk loss-of-function (K220A or 2 2 creased in Ilk / UB mice. Sox8 acts in combination with Sox9 and K220M) mutant mice is unresponsive to the RET ligand, alone to control branching morphogenesis.20 However, the effects of and GDNF24 and GDNF-responsive genes are downregulated 2 2 Sox9 deficiency on branching morphogenesis are far less severe than in Ilk / UB kidneys. Phosphorylation of specifictyrosinesin those observed in mice with deficiency of both Sox8 and Sox9,sug- the cytoplasmic domain of RET causes phosphorylation of gesting that loss of Sox8 may act in combination with the loss of multiple distinct pathways downstream of RET.25 Because Wnt11 and Cscr4 to disrupt branching morphogenesis. We iden- loss of Ilk results in alterations in a subset of GDNF-dependent tified loss of expression of other genes in ureteric tip cells, genes, including Myb and CXCR4, it is possible that ILK may but without known functions during branching morphogenesis, be acting downstream of a distinct phosphorylation site of 2 2 in Ilk / UB mice. Keratin 23 is an intermediate filament protein,21 RET to regulate a subset of signals downstream of RET.

2/2UB Table 4. Selected mRNA transcripts upregulated in Ilk kidneys (P,0.003) Probe ID Gene Name Gene Symbol Fold Change P Value 1455625_at Predicted gene 4799; Taf10 2.43524092 3.45E-06 TAF10 RNA polymerase II, TATA box binding protein-associated factor 1418930_at Chemokine (C-X-C motif) ligand 10; Cxcl10 2.09356283 5.34E-06 similar to small inducible cytokine B10 precursor (CXCL10) (interferon-g-induced protein CRG-2) (g-IP10) (IP-10) (C7) 1418714_at Dual-speciﬁcity phosphatase 8 Dusp8 1.73978671 8.55E-05 1427975_at RAS-like, family 10, member A Rasl10a 1.72559525 4.78E-06 1422029_at Chemokine (C-C motif) ligand 20 Ccl20 1.46633699 5.64E-05 1449195_s_at Chemokine (C-X-C motif) ligand 16 Cxcl16 1.4264928 5.05E-05

1472 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 1465–1477, 2016 www.jasn.org BASIC RESEARCH

Figure 5. DUSP8 overexpression results in altered p38MAPK phosphorylation and decreased UB gene expression. (A) Dusp8 over- 2 2 expression in Ilk / UB kidneys was confirmed by qRT-PCR (P,0.05). (B) Following Dusp8 cDNA transfection and stable colony selection, qRT-PCR analysis demonstrated successful overexpression of Dusp8 in mIMCD3 cells (DUSP8 o/e) (P,0.05). (C and D) After 15 minutes of EGF treatment in culture, levels of phosphorylated (phospho) p38MAPK are comparable between WT and DUSP8-overexpressing cells. After 30 minutes of EGF treatment, phospho-p38MAPK levels in DUSP8-overexpressing mIMCD cells are significantly decreased compared with WT cells (P,0.001). (E) In an EGF-stimulated manner, DUSP8-overexpressing mIMCD cells have significantly decreased the expression of Wnt11 and Slco4c1 (*Wnt11, P,0.05; **Slco4c1, P,0.001). Inhibition of Krt23 and Myb in DUSP8- overexpressing cells was observed, but not statistically significant.

DUSP8 is a Novel Modulator of p38MAPK Activity in morphogenesis. In fact, the body of evidence related to the the Murine Embryonic Kidney role of phosphatases during kidney development has previ- We identiﬁed DUSP8 as a novel ILK target that modulates ously been limited to proteins that interact with RET. The p38MAPK activity. To our knowledge, this is the ﬁrst reported tyrosine phosphatase SHP2 promotes signaling downstream functional role of DUSP8 in regulating branching of RET; loss of Shp2 in the developing UB decreases both ERK

J Am Soc Nephrol 27: 1465–1477, 2016 Ilk Controls Ureteric Branching via DUSP8 1473 BASIC RESEARCH www.jasn.org

unchanged in contrast to a significant decrease in phosphorylation of p38MAPK. Three members of the DUSP family – DUSP8, DUSP10, and DUSP16 – have been identified in nonrenal tissues and shown to modulate p38MAPK activity.28–30 These observations are consistent with our data, which demonstrate that DUSP8 decreases p38MAPK phosphorylation and expression of genes that are modulated in their expression by p38MAPK in the embryonic kidney. Interestingly, attenuation of p38MAPK activation by DUSPs can lead to positive regulation of ERK phosphorylation and activity.31 Our finding that phosphorylation of ERK is unchanged suggests that abnormal ERK activity does not contribute to 2 2 the phenotype in Ilk / UB kidneys. The regulation of DUSP8 expression is incompletelyunderstood.Givenourprevious demonstration of a BMP7-ILK-p38MAPK signaling axis and our observation that DUSP8 modulates p38MAPK activation, an autoregulatory feedback loop to regulate its expression is likely. If so, our model (Figure 7) suggests that p38MAPK activation leads to expression of DUSP8, but ILK is required directly or indirectly to negatively modulate DUSP8 and maintain the levels of p38MAPK within a precise range to facilitate gene expression and branching. Loss of neg- 2 2 Figure 6. Adenovirus-mediated DUSP8 overexpression in WT E12.5 embryonic kidney ative inhibition in Ilk / UB kidneys would explants cultured for 5 days results in impaired UB branching and decreased activation cause excessive DUSP8 expression leading of p38MAPK. (A) Overexpression of Dusp8 in explant cultures treated with Ad-GFP- to decreased phosphorylated p38MAPK acti- Dusp8 compared with Ad-GFP control was confirmed by qRT-PCR (P,0.001). (B) DUSP8 protein is expressed in explants treated with Ad-GFP-Dusp8 compared with vationintheUB,alteredgeneexpression,and explants treated with Ad-GFP. (C and D) Immunofluorescence analysis of cytokeratin decreased branching morphogenesis. expression in whole-mount preparations of E12.5 kidney explants at 5 days post- In summary, we demonstrated that Ilk infection demonstrate markedly decreased branching. (D) The number of tips was is crucial for the sustained activation of quantified from Ad-GFP and Ad-GFP-Dusp8 kidneys stained with cytokeratin. Differ- p38MAPK and downstream gene expression ences between the two treatment groups were statistically significant (n=14 kidneys/ in the developing ureteric cell in lineage. De- 2 2 group; P,0.001). (E and F) Western blot analysis of p38MAPK activation in EGF- creased branching morphogenesis in Ilk / UB stimulated Dusp8-overexpressing explants cultured for 48 hours shows significantly mice is due, in part, to increased expression decreased phosphorylation of p38MAPK in Ad-GFP-Dusp8 compared with Ad-GFP of DUSP8, which attenuates p38MAPK , (n=3 groups, P 0.02) and relevant controls. signaling. Ilk is required in the UB for the expression of specificmarkersofthetipcelldo- activation in UB tips as well as ureteric branching.26 The phos- main, as well as expression of renal branching genes crucial for phatase PTEN antagonizes PI3K activity downstream of the morphogenesis. RET receptor affecting branching morphogenesis through the regulation of the localization and generation of phosphoinositol 3–5 triphosphates.27 CONCISE METHODS The DUSP family of proteins are specifictoMAPKs.DUSPs dephosphorylate the serine-threonine and tyrosine residues in Mice the kinase domain, thereby inactivating the kinases. In Ilk- Mice with conditional ILK deficiency targeted to the ureteric lineage deficient kidney tissue, the effect of ILK deficiency was specific were generated using Hoxb7-Cre-EGFP32 and IlkloxP mice33 to gener- 2 2 to p38MAPK because phosphorylation of ERK and AKT was ate Hoxb7-Cre-EGFP;Ilk-/loxP (Ilk / UB) progeny with a specific

1474 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 1465–1477, 2016 www.jasn.org BASIC RESEARCH

Microarray Data Set Analysis The six samples were analyzed and compared using the Bioconductor limma package in the R statistical program. Brieﬂy, the limma package was used to perform background correction, robust multiarray (GC-RMA) normalization, expression calculation, and comparative 2 2 analysis using a contrast matrix between WTand Ilk / UB samples to determine log-fold change and differential expression for each probe. The limma package calculates the expression values for each probe and ranks the probe list by statistical signiﬁcance.35 To investigate the contributions to the mutant phenotype of genes with common attri- butes, the relative enrichment of GO terms in the downregulated gene list was investigated using the ToppGene server (false discovery rate, P,0.05).36 The complete data set was analyzed by the program ErmineJ37 to identify globally altered molecular pathways.

qRT-PCR RNA was isolated using Qiagen RNeasy Micro Kit and cDNA was generated using First Strand cDNA Synthesis (Invitrogen) from total RNA. Real-time PCR was performed using the Applied Biosystems 7900 HT Fast RT-PCR system with a PCR reaction mix containing a cDNA sample, SYBR green, and gene-specific primers (Supplemental Table 1). Relative mRNA expression levels were determined using the standard curve method and individual expression values were nor- Figure 7. Model of the role of ILK in UB branching. ILK acts malized by comparison to b2-microglobulin. In embryonic kidney upstream of p38MAPK to regulate Wnt11, Krt23,andSlco4c1 explants treated with adenovirus, relative mRNA expression levels expression in the UB tip to stimulate UB branching. Wnt11 is were determined using the comparative CT method and individual required for maintenance of Gdnf expression in the adjacent expression values were normalized by comparison to glyceraldehyde metanephric mesenchyme. ILK also regulates expression of Sox8, 3-phosphate dehydrogenase. CXCR4,andMyb in the UB. ILK negatively regulates the expression of the phosphatase DUSP8, which can act to modulate p38MAPK activation and downstream target gene expression. In Situ mRNA Hybridization, Western Analysis, and Immunocytochemistry Whole embryos were fixed in 4% paraformaldehyde in PBS for 16 2 2 deficiency of Ilk in the UB lineage. Ilk / UB mice were crossed with hours at 4°C. In situ hybridization was performed as described,38 on the TCF/LEF-b-galactosidase reporter strain.34 All mice were housed paraffin-embedded sections (4 mm) using dioxygenin-labeled cDNA at the Toronto Centre for Phenogenomics according to the guidelines probes for Ret, Wnt11, Sox8, Sox9, Slco4c1, Krt23, Myb,andCXCR4. of the Canadian Council for Animal Care. For Western blot analysis, kidneys or cultured cells were lysed in radioimmunoprecipitation assay buffer and subjected to immuno- Global Gene Expression Analysis in Mouse Kidney blotting as described.11 The following primary antibodies were used Tissue at a dilution of 1:1000: p38MAPK, 9212; P-p38MAPK, 9211; ERK, E12.5 embryos were dissected and kidneys removed. GFP fluorescence 4695; P-ERK, 9101; AKT, 9272; P-AKT, 9275; nonphospho (Active) and branching pattern were used to identify mutant and WT kidneys b-catenin (Ser33/37/Thr41), 8814; b-catenin, 8480 (Cell Signaling prior to RNA stabilization in RNAlater (Invitrogen). Due to the absence Technology), DUSP8, NBP1–58302 (from Novus Biologicals), P-RET of GFP expression in both heterozygote and WT littermates, the (Tyr-1062)-R, sc-20252-R (Santa Cruz Biotechnology), RET (C-19), genotypes were confirmed by PCR prior to RNA isolation. Six kidneys sc-167 (Santa Cruz Biotechnology), and Actin, A5441 (Sigma- were pooled in each biologic replicate prior to RNA isolation by Qiagen Aldrich). E12.5 urogenital ridges were stained for b-galactosidase RNeasy Micro Kit according to the manufacturer’s instructions. Prior to expression as previously described.39 microarray sample submission, total RNA quality and quantity was determined by Nanodrop analysis. RNA quality was confirmed by Agilent Cell Culture 2100 Bioanalyzer nano chip analysis prior to linear RNA amplification Both mIMCD3- and DUSP8-overexpressing mIMCD3 cells were and microarray profiling on the Affymetrix GeneChip Mouse Genome grown in monolayer in DMEM-Ham’s F12 medium (DMEM-F12), 430 2.0 Array. All samples were of sufficient RNA quality (RNA integrity supplemented with 10% FBS (Gibco), 1% penicillin-streptomycin . fi number 9) and 400 ng of each sample was ampli ed for array analysis in an incubator with 5% CO2 at 37°C. The medium for DUSP8- using the Ambion IVT kit. Each sample was amplified, hybridized, and overexpressing mIMCD3 cells was supplemented with 8 mg/ml G418. scanned according to the standard protocol by The Centre for Applied DUSP8 (NM_008748) was overexpressed in mIMCD3 cells using a Genomics, The Hospital for Sick Children, Toronto, Canada. TrueClone full-length cDNA clone from Origene (MC203407). Cells

J Am Soc Nephrol 27: 1465–1477, 2016 Ilk Controls Ureteric Branching via DUSP8 1475 BASIC RESEARCH www.jasn.org

were transfected using Turbofect reagent (R0531; Thermo Fisher Sci- 4. Jijiwa M, Fukuda T, Kawai K, Nakamura A, Kurokawa K, Murakumo Y, entific), according to the manufacturer’s instructions, and stable clones Ichihara M, Takahashi M: A targeting mutation of tyrosine 1062 in Ret were generated following G418 selection over one week. EGF stimula- causes a marked decrease of enteric neurons and renal hypoplasia. Mol Cell Biol 24: 8026–8036, 2004 tion experiments were performed in triplicate after 18 hours of serum 5. Jain S, Encinas M, Johnson EM Jr, Milbrandt J: Critical and distinct roles starvation in culture using 10 ng/ml EGF (E4127; Sigma-Aldrich). for key RET tyrosine docking sites in renal development. Genes Dev 20: 321–333, 2006 Cultured Embryonic Kidney Explants 6. Hannigan GE, Leung-Hagesteijn C, Fitz-Gibbon L, Coppolino MG, Kidneys were dissected from E12.5 Hoxb7-Cre-EGFP mice and grown Radeva G, Filmus J, Bell JC, Dedhar S: Regulation of cell adhesion and anchorage-dependent growth by a new beta 1-integrin-linked protein in culture on filter membrane inserts for 2 days in the presence of kinase. Nature 379: 91–96, 1996 DMSO vehicle control or chemical inhibitors SB203580 (20 mM, 7. Persad S, Attwell S, Gray V, Delcommenne M, Troussard A, Sanghera J, V1161; Promega) and SB202190 (10 mM, S7067; Sigma-Aldrich) Dedhar S: Inhibition of integrin-linked kinase (ILK) suppresses activation (n=3 kidneys/treatment). For adenovirus treatment, embryonic kid- of protein kinase B/Akt and induces cell cycle arrest and apoptosis of neys dissected at E12.5 were grown in culture on filter membrane PTEN-mutant prostate cancer cells. Proc Natl Acad Sci U S A 97: 3207– inserts for 5 days. Ad-Gfp and Ad-Gfp-Dusp8 were purchased from 3212, 2000 7 8. Delcommenne M, Tan C, Gray V, Rue L, Woodgett J, Dedhar S: Phos- Vector Biolabs and added daily to culture medium at 10 plaque-forming phoinositide-3-OH kinase-dependent regulation of glycogen synthase units/ml for 5 days. For imaging of ureteric branches, explants were kinase 3 and protein kinase B/AKT by the integrin-linked kinase. Proc fixed for 1 hour in methanol at –20°C, permeabilized with 0.1% Triton Natl Acad Sci U S A 95: 11211–11216, 1998 X-100 for 10 minutes, blocked with Protein Block Serum Free (Dako) 9. Leung-Hagesteijn C, Hu MC, Mahendra AS, Hartwig S, Klamut HJ, for 1 hour and incubated overnight with cytokeratin mAb (1:200; Rosenblum ND, Hannigan GE: Integrin-linked kinase mediates bone morphogenetic protein 7-dependent renal epithelial cell morphogen- Sigma-Aldrich). After washes in PBS, staining was visualized by incu- esis. Mol Cell Biol 25: 3648–3657, 2005 bation with Alexa Fluor 488 goat anti-mouse antibody (1:2000 dilu- 10. McDonald PC, Fielding AB, Dedhar S: Integrin-linked kinase–essential tion; Invitrogen). The number of UB tips was quantified using the cell roles in physiology and cancer biology. JCellSci121: 3121–3132, 2008 counter plugin in ImageJ. For assays of p38MAPK activation, 12.5 11. Smeeton J, Zhang X, Bulus N, Mernaugh G, Lange A, Karner CM, embryonic kidney explants were treated for 2 days and then analyzed. Carroll TJ, Fässler R, Pozzi A, Rosenblum ND, Zent R: Integrin-linked kinase regulates p38 MAPK-dependent cell cycle arrest in ureteric bud development. Development 137: 3233–3243, 2010 Statistical Analyses 12. Hu MC, Wasserman D, Hartwig S, Rosenblum ND: p38MAPK acts in the Results were analyzed for statistical differences between two groups BMP7-dependent stimulatory pathway during epithelial cell morpho- using the t test, on GraphPad Prism software (version 6.0). P,0.05 genesis and is regulated by Smad1. J Biol Chem 279: 12051–12059, 2004 was considered statistically significant. Values are given as mean6SEM. 13. Smyth GK: Limma: linear models for microarray data. In: Bioinformatics and Computational Biology Solutions using R and Bioconductor,ed- ited by Gentleman VCR, Dudoit S, Irizarry R, Huber W, New York, Springer, 2005, pp 397–420 ACKNOWLEDGMENTS 14. Harding SD, Armit C, Armstrong J, Brennan J, Cheng Y, Haggarty B, Houghton D, Lloyd-MacGilp S, Pi X, Roochun Y, Sharghi M, Tindal C, McMahon AP, Gottesman B, Little MH, Georgas K, Aronow BJ, Potter We thank Shoukat Dedhar for kindly providing IlkloxP mice and SS, Brunskill EW, Southard-Smith EM, Mendelsohn C, Baldock RA, Carlton Bates for Hoxb7GFPCre mice. Davies JA, Davidson D: The GUDMAP database–an online resource for This work was supported by the Hospital for Sick Children genitourinary research. Development 138: 2845–2853, 2011 Foundation Graduate Training Award (to J.S.) and by grants from the 15. Majumdar A, Vainio S, Kispert A, McMahon J, McMahon A: Wnt11 and Canadian Institutes of Health Research and Canada Research Chairs Ret/Gdnf pathways cooperate in regulating ureteric branching during – Program (to N.D.R.). metanephric kidney development. Development 130: 3175 3185, 2003 16. Lu BC, Cebrian C, Chi X, Kuure S, Kuo R, Bates CM, Arber S, Hassell J, MacNeil L, Hoshi M, Jain S, Asai N, Takahashi M, Schmidt-Ott KM, Barasch J, D’Agati V, Costantini F: Etv4 and Etv5 are required down- DISCLOSURES stream of GDNF and Ret for kidney branching morphogenesis. Nat None. Genet 41: 1295–1302, 2009 17. Wong A, Bogni S, Kotka P, de Graaff E, D’Agati V, Costantini F, Pachnis V: Phosphotyrosine 1062 is critical for the in vivo activity of the Ret9 receptor tyrosine kinase isoform. Mol Cell Biol 25: 9661–9673, 2005 REFERENCES 18. Takabatake Y, Sugiyama T, Kohara H, Matsusaka T, Kurihara H, Koni PA, Nagasawa Y, Hamano T, Matsui I, Kawada N, Imai E, Nagasawa T, Rakugi 1. Saxen L: Organogenesis of the kidney, Cambridge, Cambridge Uni- H, Isaka Y: The CXCL12 (SDF-1)/CXCR4 axis is essential for the devel- versity Press, 1987 opment of renal vasculature. J Am Soc Nephrol 20: 1714–1723, 2009 2. Cain JE, Di Giovanni V, Smeeton J, Rosenblum ND: Genetics of renal 19. Ueland J, Yuan A, Marlier A, Gallagher AR, Karihaloo A: A novel role for hypoplasia: insights into the mechanisms controlling nephron endow- the chemokine receptor Cxcr4 in kidney morphogenesis: an in vitro study. Dev Dyn 238: 1083–1091, 2009 ment. Pediatr Res 68: 91–98, 2010 20. Reginensi A, Clarkson M, Neirijnck Y, Lu B, Ohyama T, Groves AK, Sock 3. Tang MJ, Cai Y, Tsai S-J, Wang Y-K, Dressler GR: Ureteric bud out- E, Wegner M, Costantini F, Chaboissier MC, Schedl A: SOX9 controls growth in response to RET activation is mediated by phosphatidylinositol epithelial branching by activating RET effector genes during kidney 3-kinase. Dev Biol 243: 128–136, 2002 development. Hum Mol Genet 20: 1143–1153, 2011

1476 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 1465–1477, 2016 www.jasn.org BASIC RESEARCH

21. Zhang S, Lin Y, Itäranta P, Yagi A, Vainio S: Expression of Sprouty genes 30. Tanoue T, Yamamoto T, Maeda R, Nishida E: A Novel MAPK phos- 1,2and4duringmouseorganogenesis.Mech Dev 109: 367–370, 2001 phatase MKP-7 acts preferentially on JNK/SAPK and p38 alpha and 22. Thiagarajan RD, Georgas KM, Rumballe BA, Lesieur E, Chiu HS, Taylor beta MAPKs. JBiolChem276: 26629–26639, 2001 D, Tang DT, Grimmond SM, Little MH: Identification of anchor genes 31. Finch AR, Caunt CJ, Perrett RM, Tsaneva-Atanasova K, McArdle CA: during kidney development defines ontological relationships, molec- Dual specificity phosphatases 10 and 16 are positive regulators of EGF- ular subcompartments and regulatory pathways. PLoS One 6: e17286, stimulated ERK activity: indirect regulation of ERK signals by JNK/p38 2011 selective MAPK phosphatases. Cell Signal 24: 1002–1011, 2012 23. Toyohara T, Suzuki T, Morimoto R, Akiyama Y, Souma T, Shiwaku HO, 32. Zhao H, Kegg H, Grady S, Truong HT, Robinson ML, Baum M, Bates Takeuchi Y, Mishima E, Abe M, Tanemoto M, Masuda S, Kawano H, CM: Role of fibroblast growth factor receptors 1 and 2 in the ureteric Maemura K, Nakayama M, Sato H, Mikkaichi T, Yamaguchi H, Fukui S, bud. Dev Biol 276: 403–415, 2004 Fukumoto Y, Shimokawa H, Inui K, Terasaki T, Goto J, Ito S, Hishinuma 33. Terpstra L, Prud’homme J, Arabian A, Takeda S, Karsenty G, Dedhar S, T, Rubera I, Tauc M, Fujii-Kuriyama Y, Yabuuchi H, Moriyama Y, Soga T, St-Arnaud R: Reduced chondrocyte proliferation and chondrodysplasia Abe T: SLCO4C1 transporter eliminates uremic toxins and attenuates in mice lacking the integrin-linked kinase in chondrocytes. JCellBiol hypertension and renal inflammation. J Am Soc Nephrol 20: 2546– 162: 139–148, 2003 2555, 2009 34. Mohamed OA, Clarke HJ, Dufort D: Beta-catenin signaling marks the 24. Lange A, Wickström SA, Jakobson M, Zent R, Sainio K, Fässler R: prospective site of primitive streak formation in the mouse embryo. Dev Integrin-linked kinase is an adaptor with essential functions during Dyn 231: 416–424, 2004 mouse development. Nature 461: 1002–1006, 2009 35. Smyth GK: Linear models and empirical bayes methods for assessing 25. Kawamoto Y, Takeda K, Okuno Y, Yamakawa Y, Ito Y, Taguchi R, Kato differential expression in microarray experiments. Stat Appl Genet Mol M, Suzuki H, Takahashi M, Nakashima I: Identification of RET auto- Biol 3: 1–25, 2004 phosphorylation sites by mass spectrometry. J Biol Chem 279: 14213– 36. Chen J, Bardes EE, Aronow BJ, Jegga AG: ToppGene Suite for gene list 14224, 2004 enrichment analysis and candidate gene prioritization. Nucleic Acids 26. Willecke R, Heuberger J, Grossmann K, Michos O, Schmidt-Ott K, Res 37: W305–311, 2009 Walentin K, Costantini F, Birchmeier W: The tyrosine phosphatase Shp2 37. Lee HK, Braynen W, Keshav K, Pavlidis P: ErmineJ: tool for functional acts downstream of GDNF/Ret in branching morphogenesis of the analysis of gene expression data sets. BMC Bioinformatics 6: 269, 2005 developing mouse kidney. Dev Biol 360: 310–317, 2011 38. Bridgewater D, Di Giovanni V, Cain JE, Cox B, Jakobson M, Sainio K, 27. Kim D, Dressler GR: PTEN modulates GDNF/RET mediated chemotaxis Rosenblum ND: b-catenin causes renal dysplasia via upregulation of and branching morphogenesis in the developing kidney. Dev Biol 307: Tgfb2andDkk1.JAmSocNephrol22: 718–731, 2011 290–299, 2007 39. Bridgewater D, Cox B, Cain J, Lau A, Athaide V, Gill PS, Kuure S, Sainio 28. Muda M, Theodosiou A, Rodrigues N, Boschert U, Camps M, Gillieron K, Rosenblum ND: Canonical WNT/beta-catenin signaling is required C, Davies K, Ashworth A, Arkinstall S: The dual specificity phosphatases for ureteric branching. Dev Biol 317: 83–94, 2008 M3/6 and MKP-3 are highly selective for inactivation of distinct mitogen-activated protein kinases. JBiolChem271: 27205–27208, 1996 29. Masuda K, Shima H, Watanabe M, Kikuchi K: MKP-7, a novel mitogen- activated protein kinase phosphatase, functions as a shuttle protein. J This article contains supplemental material online at http://jasn.asnjournals. Biol Chem 276: 39002–39011, 2001 org/lookup/suppl/doi:10.1681/ASN.2015020139/-/DCSupplemental.

J Am Soc Nephrol 27: 1465–1477, 2016 Ilk Controls Ureteric Branching via DUSP8 1477