Epigenetics, Development, and the Kidney

SPECIAL ARTICLE www.jasn.org

Gregory R. Dressler Department of Pathology, University of Michigan, Ann Arbor, Michigan

ABSTRACT How cells partition the genome into active and inactive genes and how that informa- gant series of nuclear transplant experi- tion is established and propagated during embryonic development are fundamental to ments by Solter (for review see refer- maintaining the normal differentiated state. The molecular mechanisms of epigenetic ence2). Mouse zygotes containing two action and cellular memory are increasingly amenable to study primarily as a result of maternal genomes developed to midges- the rapid progress in the area of chromatin biology. Methylation of DNA and modi- tation but lacked extraembryonic tissues, fication of histones are critical epigenetic marks that establish active and silent chro- whereas zygotes with two paternal ge- matin domains. During development of the kidney, DNA-binding factors such as nomes generated mostly extraembryonic Pax2/8, which are essential for the intermediate mesoderm and the renal epithelial tissues. These classic experiments dem- lineage, could provide the locus and tissue specificity for histone methylation and onstrate the need for both maternal and chromatin remodeling and thus establish a kidney-specific fate. The role of epigenetic paternal genomes in mammalian devel- modifications in development and disease is under intense investigation and has opment. Furthermore, the early embryo already affected our view of cancer and aging. must be able to distinguish two different haploid genomes through some epige- J Am Soc Nephrol 19: 2060–2067, 2008. doi: 10.1681/ASN.2008010119 netic mechanism and subsequently erase or reprogram these epigenetic recogni- tion marks once the genomes pass Implicit in the structure of the DNA dou- WHAT IS EPIGENETICS? through the new germline. ble helix proposed by Watson and Crick More recently, the study of epigenet- was the heritable nature of the genome The term epigenetics, most broadly de- ics has concerned itself with inheritance during cell division, for one complemen- fined, refers to a type of inheritance that not just through the germline but also in tary strand served as a template to replicate is not encoded directly within the DNA somatic cells during embryonic develop- the other. Once the genetic code was re- sequences of genes. Such heritability can ment. Development can be thought of as vealed as triplets of nucleotides transcribed be through the germ line, as from parents a series of decisions that restrict the fate into messenger RNA and translated into to offspring, or from a single mother cell of rapidly dividing cells in response to proteins, the central dogma of molecular to its progeny during mitosis. In hu- positional information. Although em- biology was complete. Thus, genetics, mans, classic examples of epigenetic phe- bryonic stem cells and cells of the epiblast which predated the discovery of DNA and nomena include maternal and paternal are truly pluripotent, once gastrulation concerned itself with the abstract concepts imprinting, in which a disease phenotype occurs, the three germ layers become re- of heritability and phenotype, was explain- is differentially expressed depending on stricted in their eventual fates. Despite able in more biochemical terms as genes whether the allele is inherited from the that the cells are still proliferating rapidly were assigned to blocks of DNA sequence mother or the father. For example, dele- and do not express many differentiated and alleles became variants of such se- tions near chromosome 15q11 are asso- markers, such restriction is generally sta- quences. Despite the remarkable progress ciated with Prader-Willi syndrome, ble and inherited in all progeny cells. As made in deciphering the meaning of the when inherited from the father, yet man- cells of the neural ectoderm, mesoderm, genome, certain phenomena remained ifest a very different spectrum of pheno- outside the realm of classical genetics and types, Angelman syndrome, when inher- Published online ahead of print. Publication date the DNA centric view of inheritance. These ited through the mother. These types of available at www.jasn.org. epigenetic phenomena have received epigenetic imprinting effects likely arose Correspondence: Dr. Gregory R. Dressler, Depart- much attention in recent years because with the evolution of placental mam- ment of Pathology, 2049 BSRB 2200, University of they have a direct impact on our under- mals1 and underscore a fundamental dif- Michigan, Ann Arbor, MI 48109. Phone: 734-764- 6490; Fax: 734-763-6640; E-mail: dressler@umich. standing of genome regulation, the differ- ference between maternal and paternal edu entiated state, and the stability of cellular genomes. That haploid genomes are not Copyright ᮊ 2008 by the American Society of identity in many biologic systems. equivalent was also addressed in an ele- Nephrology

2060 ISSN : 1046-6673/1911-2060 J Am Soc Nephrol 19: 2060–2067, 2008 www.jasn.org SPECIAL ARTICLE and endoderm continue their differenti- Beads-on-a-string CH CH CH CH ation, cell fates become increasingly 3 3 CH 3 3 more restricted. The stability of the ter- 3 11 nm minally differentiated genome is best il- Ac Ac lustrated by the difficulty of cloning by Solenoid somatic cell nuclear transfer. When a nucleus from a specialized adult cell type is 30 nm introduced into the denucleated zygote, the epigenetic marks accumulated within that differentiated nucleus are not easily Fiber erased. Although Dolly the sheep3 first ~500 nm proved that epigenetic marks can be erased and the somatic genome repro- grammed within the cytoplasmic envi- Mitotic chromosome ronment of a fertilized egg, it is a rare and inefficient process. The loss of pluripotency and the heritability of a differentiated state imply the genome is modified to remember pat- Figure 1. The dynamic structure of chromatin. The basic unit of chromatin is the terns of gene expression in all progeny nucleosome, a histone octamer represented by the tan-colored balls on the 11-nm string. The DNA helix is wrapped around the nucleosome, and a spacer region, bound cells. Dosage compensation by inactiva- with histone H1 or H5, separates adjacent nucleosomes. The histone tails extend out of tion of the X chromosome in female cells the nucleosome and can be modified by methylation or acetylation. The beads-on-a- is a well-studied example of a mitotically string can condense into a solenoid, a theoretical structure thought to be approximately inherited epigenetic phenomenon that 30 nm in diameter. Increased condensation leads to a chromatin fiber that is several controls gene expression patterns.4 Strik- hundred nm in diameter. The mitotic chromosomes is the most compact form of ingly, inactivation of the X chromosome chromatin in the cell. Chromatin remodeling refers to the sliding of nucleosomes along is coincident with the loss of pluripo- the string, the compaction of nucleosomes into higher order structures, and the unfold- tency as the mammalian epiblast under- ing of higher order structures into accessible chromatin. goes gastrulation. Although the initial decision to inactivate the maternal or pa- chemistry and structure, these questions consists of approximately 147 bp of the ternal X seems to be random, once the and more are now being addressed in a DNA helix wrapped around a histone oc- decision is made, all daughter cells from variety of experimental systems. tamer containing two copies of histone a single founder will have the same X H2A, H2B, H3, and H4 (Figure 1). Indi- chromosome inactivated. The inactive X vidual nucleosomes are separated by a chromosome is tightly packaged and eas- BIOCHEMISTRY OF EPIGENETICS spacer, or linker, region that are bound ily identifiable as a Barr body, yet in sub- by histone H1 or H5. The primary struc- sequent rounds of mitosis, it must be un- The study of epigenetic phenomenon has ture of chromatin is often referred to as packaged, replicated, and repackaged presupposed that genes can be marked “beads on a string” because of the ap- once again, suggesting that the silent X is somehow by modifications that are inde- pearance of the nucleosome and spacer marked through an epigenetic mecha- pendent of the primary nucleotide cod- sequences under the electron micro- nisms that distinguishes it from its active ing sequence. Such marks include the co- scope; however, this relatively unfolded counterpart. valent modification of DNA nucleotides form of chromatin can assume higher or- Because the pattern of gene expres- directly, specifically the methylation of der structures to compact the genome sion defines the differentiated state, the cysteine in CpG di-nucleotides. Modifi- even further (Figure 1). The folding and heritability of such patterns must be well cations are also found in the proteins unfolding of chromatin is a dynamic and regulated. On the autosomes, active and tightly associated with DNA. Eukaryotic regulated process that is necessary for inactive genes are present, sometimes in chromatin consists of the double- DNA replication but can also have a di- close proximity, and must be recognized stranded DNA helix and associated pro- rect impact on the accessibility of genes by positive and negative transacting fac- teins. Histones are small, highly basic to transcription machinery. tors. Does the complement of regulatory proteins that are tightly bound to the The methylation of DNA at CpG di- proteins define the differentiated state, DNA. As a family, the histones are ex- nucleotides was among the first genome or are these active and inactive genes tremely well conserved from yeast to hu- modifications described and remains an somehow marked, like the X chromo- mans and are among the most common attractive mechanism for epigenetic in- somes? As a result of many recent break- proteins in the cell. The basic unit of heritance of cell identity.5–7 During em- throughs in the fields of chromatin bio- chromatin is the nucleosome, which bryogenesis, the genome undergoes dra-

J Am Soc Nephrol 19: 2060–2067, 2008 Epigenetics, Development, and the Kidney 2061 SPECIAL ARTICLE www.jasn.org

matic changes in CpG methylation and P M P requires the function of de novo DNA M M M A AAM A M M M ARTKQTARKSTGGKAPRKQLATKAARKSAPATGGVKK--K- Histone methyltransferases Dnmt3a and Dnmt3b H3 for postimplantation development. Hy- 2 4 9 14 17 23 27 permethylation is found on the inactive P M A A A A M X chromosome and generally correlates SGRGKGGKGLGKGGAKRHRKVLRDNIQGITKPAIRRLAR-- Histone H4 with inactive genes on the autosomes. 1 3 5 8 12 16 20 Many regulated promoters have CG-rich sequences upstream of and around the P A A SGRGKQGGKARAKAKSRSSRAGLQFPVGRVHRLLRKGNY-- Histone transcription start site. In general, these H2A 1 5 9 so-called CpG islands show hypomethy- lation when genes express yet can con- A A P vert to a hypermethylated state upon in- PEPAKSAPAPKKGSKKAVTKAQKKDSKKRKRSRKESYSV-- Histone H2B activation. Because DNA replication is 5 12 14 semiconservative, the pattern of CpG Figure 2. Summary of histone modifications. The amino acid sequences of the core methylation can be inherited, whereby histone tails are shown. Amino acid residues that are subject to epigenetic modifications maintenance DNA methyltransferases are shown. Of particular significance are mono-, di-, or trimethylation (M) at arginine or modify the replicated strand on the basis lysine residues; acetylation (A) at lysine residues; and phosphorylation (P) at serine of the pattern of CpG methylation in the residues. Some amino acids, such as lysine 9 of histone H3, can be subject to either template. DNA methylation as a gene-si- acetylation or methylation, but not both. lencing mechanism was further supported by identification of proteins with the amino-terminal tails, lysine residues tions and potential access for other tran- methyl-CpG binding domains. Such available for methylation include K4, K9, scription factors and RNA polymerase.26 methyl-CpG binding domains can act as K27, and K36 of histone H3 and K20 of adaptor proteins linking the methylated histone H4. For example, methylation of CpG islands to nucleosome and chroma- H3 K9 by the mammalian HMT HISTONE MODIFICATION IN tin remodeling factors such as histone Suv39h15 correlates with silent chroma- DEVELOPMENT deacetylases.8–10 tin,16 whereas actively expressed genes The complexity and types of modifi- are enriched in di- and trimethylated The realm of biomedical research en- cations that can affect the histone oc- H3K4.17 The conserved SET domain, compasses a multitude of disciplines and tamer have been studied in great detail first described in the yeast protein Set1,18 organisms, spanning everything from and form the basis of the histone code is the catalytic domain for methyltrans- prokaryotic biochemistry to human ge- hypothesis.11 More recently, the dy- ferases and is found in more than 50 po- netics and clinical practice. Thus, it is namic nature of many histone modifica- tential mammalian proteins. Among particularly satisfying when seemingly tions (Figure 2), whose mechanisms of these are the mammalian ALL-1,19 disparate fields converge to define new heritability are still unclear, has ALR,20 and MLL21 genes, the homo- paradigms that revolutionize how we prompted a more careful reexamination logues of the Drosophila trithorax group think about an old problem. That is pre- of the significance and impact of individ- of epigenetic regulators. cisely what happened when the fields of ual histone marks with respect to gene How does methylation at specific res- chromatin biochemistry met develop- function.12 From an epigenetic view- idues alter chromatin structure? One hy- mental genetics. point, the most important modifications pothesis is these modified histone tails Development and differentiation re- are methylation and acetylation of the interact with specific proteins. Indeed, quire the partitioning of the genome into histone tails, which extend out of the nu- the highly conserved chromodomain, active and inactive domains, or loci. This cleosome and are thus accessible to chro- found in some chromatin remodeling genomic organization must be stably in- matin remodeling and/or transcription factors, is able to recognize methylated herited through many rounds of cell di- regulatory proteins. Acetylation of his- lysine residues to promote heterochro- vision such that cellular memory is pre- tones H2A, 2B, 3, and 4 strongly correlates matin formation and suppression.22–24 served during development. Genes that with transcription activation in almost all Conversely, the WDR5 protein binds to control such epigenetic memory in a experimental systems. Furthermore, his- dimethyl-K4 of histone H3 and is re- complex metazoan were first discovered tone deacetylases mediate gene repres- sponsible for promoting trimethylation in the fruit fly, Drosophila, as enhancers sion and compaction into heterochro- as a mark for active gene expression.25 or suppressors of homeotic gene expres- matin.13,14 In contrast, methylation of Furthermore, histone H3 methylation at sion during early pattern formation.27,28 histones can correlate with gene activity K4 can provide docking sites for chroma- Genes that prevent inappropriate activa- or gene silencing, depending on which tin remodeling proteins to establish and tion or suppression of homeotic genes specific residues are modified. Within maintain open chromatin configura- fall into the polycomb group of epige-

2062 Journal of the American Society of Nephrology J Am Soc Nephrol 19: 2060–2067, 2008 www.jasn.org SPECIAL ARTICLE netic regulators, whereas the trithorax lian embryonic stem cells, many impor- eral inhibition of gene suppression group function to maintain homeotic tant regulatory genes seem to possess mechanisms is critical for maintaining gene activity. Although many individual neither fully active nor fully inactive his- pluripotency. genes were identified in screens for mod- tone modifications; rather, they contain ifiers of homeotic gene activity, the bio- some elements of both. This so-called chemical function of the respective pro- bivalent epigenetic mark is thought to KIDNEY DEVELOPMENT teins and their effects on chromatin were provide plasticity in the stem cell, be- unknown. Meanwhile, investigators us- cause these genes are poised to assume During the past 40 yr, progress in under- ing such diverse model systems as tetera- either active or inactive chromatin states standing the genetic basis of kidney de- hymena and yeast discovered histone depending on which direction the cells velopment has been remarkable.43,44 methylation as an essential component differentiate.35,36 Thus, as stem cells dif- What began as a model system for epi- of gene silencing and heterochromatin ferentiate along lineage pathways, more thelial-mesenchymal inductive interac- formation. Histone H3K9 methylation and more genes assume epigenetic marks tions, early kidney development has now correlated with the inactive chromatin in specific for a particular cell type. Such a become a paradigm for organogenesis, yeast15,29 and was later found as an early model would require interactions of epi- epithelial cell differentiation, branching epigenetic mark on the mammalian X genetic imprinting enzymes with DNA morphogenesis, and complex patterning chromosome.30 The importance of his- binding proteins that recognize specific events. The adult kidney is composed of tone methylation as a developmental genes at precise times and respond to po- many specialized epithelial, endothelial, epigenetic mark became clear when a sitional information in the embryo. Un- and stromal cell types; however, the polycomb protein complex was shown to fortunately, few such factors have been functional components of the kidney, have H3K9 methyltransferase activity described. Thus, linking changes in cell the renal epithelial cells, share a common and contain not only the SET domain en- type–specific epigenetic patterning to the lineage that is specified quite early in de- zymes but also numerous co-factors essential regulatory proteins that deter- velopment. This lineage is first apparent whose activity had been described in mine cell lineages, as defined genetically shortly after gastrulation, in a region of yeast. Enhancer of Zeste and Extra Sex through mutant analysis, remains to be mesoderm called intermediate, because Combs are polycomb group proteins fully realized. it lies between the axial, or somitic, me- that form a complex capable of methyl- Genetic studies of flies first pointed soderm, and the lateral plate mesoderm ating histone H3K9 and H3K27 to si- to the stability and heritability of epi- along the mediolateral axis (Figure 3). lence expression.31,32 Theses studies genetic modifications during develop- What defines the intermediate meso- linked the genetics of polycomb group ment. The principle of genome stabil- derm and how does it arise? This question genes to biochemical activity on silent ity in differentiated cells was further is fundamental to our understanding of re- chromatin. Similarly, the trithorax underscored by the difficulty in clon- nal epithelial cell lineage determination, group proteins TRX and Ash1 contain ing mammalian embryos by somatic because the fate decisions made in the me- SET domains and histone methyltrans- cell nuclear transfer; however, recent soderm as it becomes more specialized ferase activity able to mark H3K4, a mark advances in histone biochemistry and may be irreversible. For example, by em- that correlates with gene expres- stem cell technology suggest that, given bryonic day 10.5 in the mouse, the poste- sion.19,21,33 Epistasis experiments (one the right circumstances, the genome rior aspect of the intermediate mesoderm gene inhibiting another) in the fly using may ultimately be more amenable to consists of an aggregate of cells called meta- mutants in TRX and PcG suggested that epigenetic reprogramming than previ- nephric mesenchyme. This mesenchyme is the trithorax group genes function not as ously thought. Several key break- awaiting inductive WNT signals from the classical activators but rather as suppres- throughs have altered prevailing ureteric bud,45 which will promote aggre- sors of silencing by the polycomb dogma. First, the identification of his- gation and mesenchyme-to-epithelial con- genes.34 Although many issues still need tone demethylases clearly suggests version; however, these inductive signals, to addressed, one interpretation of such that, like acetylation, specific nucleo- normally provided by the ureteric bud, can experiments is that repression is a default some methylation marks can be come from many other tissues, such as spi- state that will ultimately win out if active erased.37–39 Second, the ability to renal cord, WNT1-expressing cells, or even epigenetic marks are not maintained. program somatic cells into embryonic LiCl, an activator of the canonical WNT If active or repressed chromatin re- stem cells by introducing a limited set signaling pathway. The idea that meta- quires positive or negative histone meth- of genes that were known to maintain nephric mesenchyme is already pro- ylation marks, then what is the status of pluripotency dramatically shifts both grammed to make only renal epithelial chromatin in undifferentiated, pluripo- the scientific and ethical boundaries in precursors prompted Saxen44 to propose tent cells? This question is being ad- the stem cell field.40,41 In fact, histone the concept of a permissive inductive sig- dressed systematically in embryonic demethylation of the repressive H3K9 nal, rather than instructive signal that re- stem cells. So far, the results have been epigenetic mark is a key event in stem programs the fate of the mesenchyme. The surprising. In undifferentiated mamma- cell self-renewal,42 suggesting that gen- metanephric mesenchyme may be epige-

J Am Soc Nephrol 19: 2060–2067, 2008 Epigenetics, Development, and the Kidney 2063 SPECIAL ARTICLE www.jasn.org

Cap chyme generates the definitive kidney. ABCmesenchyme Such A-P patterning is likely to involve the Mesonephric AS tubules homeotic or Hox genes, which are known Nephric Neural duct to specify segmented identity in the fly and tube A-P position in the mouse. Indeed, loss of BMP Somite Ureteric Metanephric the Hox11 group of genes results in com- Intermediate bud mesenchyme plete renal agenesis as a result of the sup- Lateral mesoderm plate pression of more posterior markers such as Renal GDNF and Six2.53 The intersection among vesicle genes specifying the mediolateral axis and Figure 3. The kidney is specified from intermediate mesoderm. (A) A cross-section the A-P axis may provide a unique molec- through a mammalian embryo shortly after gastrulation. The mesoderm becomes spec- ular address to position the metanephric ified into paraxial, intermediate, and lateral plate. The paraxial mesoderm makes somites, kidney. This idea is supported by recent whereas the intermediate mesoderm will make the nephric duct and metanephric mes- data on the regulation of Six2 by a complex enchyme. Bone morphogenetic protein (BMP) signals derived from the lateral plate are of proteins that includes Hox11, Pax2, and thought to promote expression of intermediate mesodermal markers, such as Pax2/8 and Eya1 in the metanephric mesenchyme.54 Osr1. Axial signals (AS) may counteract the BMP effect to suppress intermediate markers Given that the metanephric mesen- in the paraxial mesoderm. (B) A longitudinal view of the intermediate mesoderm shortly chyme is already programmed to gener- before kidney induction, anterior is top. The nephric duct is a bilateral epithelial tube from which mesonephric tubules are induced more rostrally and the ureteric bud grows ate renal epithelia, whether this restric- out at the posterior aspect. Adjacent to the posterior nephric duct is the metanephric tion in fate is in part determined by mesenchyme, an aggregate of cells already programmed to generate renal epithelia. (C) epigenetic changes at specific loci that act As the ureteric bud invades the mesenchyme and undergoes branching morphogenesis, to limit the developmental potential of inductive WNT signals from the bud induce the mesenchyme to aggregate around the these renal stem cells remains to be de- tips, the cap mesenchyme, and become polarized to form a primitive epithelial renal termined. Recently, the Pax2 protein was vesicle. Each renal vesicle will generate a single nephron and reconnect to the branching linked to a histone methyltransferase ureteric bud, which generates the collecting duct system. complex containing the trithorax homo- logues ALR/Mll2 and Mll3 through in- netically programmed to respond to WNT together with Lim1 can also make pro- teractions with the adaptor protein, signals in a limited way, such that only a nephric tissue,49 although Osr1 has a PTIP.55 Originally identified because of certain type of epithelia and stromal deriv- similar activity all by itself.50 The activa- its interaction with Pax proteins,56 PTIP atives can arise. Thus, cell fate decisions for tion of Pax2/8, Lim1, and Osr1 genes contains multiple BRCT domains, at making renal epithelia have already begun must be position dependent and require least one of which is a phospho-serine– and can likely be traced back to specializa- local environmental cues. James and binding domain.57 Several laboratories tion of the intermediate mesoderm after Schultheiss51,52 suggested that low con- have independently shown that PTIP is gastrulation. centrations of bone morphogenetic pro- part of a histone H3K4 methyltransferase Genetic mutations in the mouse and teins diffusing from the lateral plate me- complex.58,59 Because Pax2 can be serine/ tissue manipulation in the chick embryo soderm promote Osr1 and Pax2 threonine phosphorylated in response to have helped to define some critical com- expression, hence intermediate meso- Wnt signals,60,61 the interaction with ponents of intermediate mesoderm derm formation, whereas axial signals PTIP promotes H3K4 methylation at identity and, by inference, metanephric may inhibit bone morphogenetic pro- kidney-specific loci in response to induc- mesenchyme specification. The DNA- teins to suppress the intermediate phe- tive signals. Although PTIP is not kidney binding proteins Pax2, Lim1, and Osr1 notype. Thus, the Pax2/8 expression do- specific, it may serve as an adaptor pro- (odd skipped related) all are required main could be initiated at the boundary tein linking Pax2 and other tissue-spe- and in some cases sufficient to specify in- between axial and lateral plate mesoderm cific DNA-binding proteins to the epi- termediate mesoderm from surrounding by the actions of two opposing gradients. genetic machinery as important cell lateral plate and paraxial mesoderm. The The intermediate mesoderm is initially fate decisions are being made (Figure epistatic interactions among these three defined by its position along the mediolat- 4). Consistent with this idea, PTIP mu- genes are still unclear. In the mouse, eral axis; however, as development pro- tants are postgastrulation lethal and Pax2 is still expressed in Lim146 or Osr147 ceeds, it is clear that anterior-posterior show a global reduction in H3K4 meth- mutants, although Lim1 is not expressed (A-P) patterning is necessary for the deriv- ylation.55,62 These data suggest that in Pax2/8 double mutants.48 At the time atives of the intermediate mesoderm to as- Pax2/8 function to provide locus and of mesodermal specialization, ectopic sume their appropriate fates. For example, tissue specificity for epigenetic cues to expression of Pax2 can expand the do- more anterior intermediate mesoderm restrict the developmental potential of main of intermediate mesoderm in the generates the mesonephric tubules, whereas the intermediate mesoderm to the re- chick embryo.48 In the frog embryo, Pax2 the more posterior metanephric mesen- nal lineage.

2064 Journal of the American Society of Nephrology J Am Soc Nephrol 19: 2060–2067, 2008 www.jasn.org SPECIAL ARTICLE

A MII2 histone and DNA methyltransferases in Pax2 Ptip complex aging cells is an area of research that will need to be explored further if the stability of the epigenome in aging cells is to be Pax DNA fully understood. binding site B MII2 CONCLUSIONS Ptip complex Pax2* The study of epigenetics is proving to be fertile ground. How the genome is packaged and replicated in different cell types C MII2 is fundamental to maintaining cell lin- complex Ptip NURF eage restriction and differentiated phe- BPTF Pax2 notypes in developing and adult mam- (CH3) 3 Pol II mals. The complex machinery that methylates DNA, modifies histones, and alters chromatin structure is being char- acterized at the biochemical level. A ma- Figure 4. A model of Pax mediates gene activation in the metanephric mesenchyme. jor challenge ahead is to link this ma- (A) The Pax2 protein recognizes a DNA-binding site and becomes phosphorylated (*), chinery to locus- and tissue-specific perhaps in response to WNT signals. (B) The nuclear adaptor protein PTIP localizes to the Pax2 protein and recruits the Mll2 histone methyltransferase complex to promote tri- factors that drive development and dis- methylation at H3K4. (C) Trimethylation of H3K4 promotes nucleosome remodeling ease in mammalian organ systems, in- through interactions with BPTF, a subunit of the NURF complex. Pax2 and the Mll2 cluding the kidney. complex are no longer required at the binding site. Remodeling attracts the RNA polymerase II complex, and gene activation commences. ACKNOWLEDGMENTS EPIGENETICS AND RENAL tions that aberrant histone methylation DISEASE is oncogenic, including translocations of This work was funded in part by National In- the H3K4 methyltransferase ALL/MLL1 stitutes of Health grants DK073722 and Although the study of epigenetic phe- in acute lymphocytic leukemia65 and the DK054740 to G.R.D. nomena in disease has been widely inves- overexpression of the H3K27 methyl- I am especially grateful to Sanj Patel for tigated, I believe there are two specific areas transferase EZH2 in prostate cancer.66 sharing preliminary data and many stimulat- in which chromatin modifications play a Furthermore, overexpression of the his- ing discussions. particularly important role. In aging and tone demethylases LSD1, the Jumonji cancer, the stability of the differentiated domain–containing protein 2, and Ju- state must be determined, at least in part, monji interacting protein PLU-1 corre- DISCLOSURES by the inherent memory of the genome lates with prostate, esophageal, and None. within the mature cell. The hypermethyl- breast cancer.67 Thus, the balance be- ation of promoter regions in cancer cells tween histone methylation and demeth- can lead to the inactivation of tumor sup- ylation is likely to be critical for main- REFERENCES pressor genes.63 For counteracting this ef- taining the differentiated state in many fect, inhibitors of DNA methylation have adult tissues. 1. Hore TA, Rapkins RW, Graves JA: Construc- been developed for treating leukemias and Chronic renal disease is in part a function and evolution of imprinted loci in mam- other hematologic malignancies; however, tion of age. In the glomerulus, aging mals. Trends Genet 23: 440–448, 2007 a longstanding issue in this field has always podocytes exhibit altered morphology 2. Solter D: Differential imprinting and expression of maternal and paternal genomes. been whether DNA methylation is the and patterns of gene expression.68,69 In Annu Rev Genet 22: 127–146, 1988 cause of gene inactivation or is merely mammals, reduced levels of DNA meth- 3. Campbell KH, McWhir J, Ritchie WA, Wilmut the result of inactivation. At least in one ylation correlates with age,70 suggesting I: Sheep cloned by nuclear transfer from a case, chromatin inactivation as mea- that repression of gene expression may cultured cell line. Nature 380: 64–66, 1996 sured by changes in histone methylation be lost over time. Loss of epigenetic re- 4. Valley CM, Willard HF: Genomic and epi- patterns seems to precede DNA methyl- pression has also been observed in aging genomic approaches to the study of X chromosome inactivation. Curr Opin Genet Dev 71 ation at an important tumor suppressor mice at X-linked and imprinted loci. 16: 240–245, 2006 locus.64 There are many other indica- The potential maintenance functions of 5. Ng HH, Bird A: DNA methylation and chro-

J Am Soc Nephrol 19: 2060–2067, 2008 Epigenetics, Development, and the Kidney 2065 SPECIAL ARTICLE www.jasn.org

matin modification. Curr Opin Genet Dev 9: 20. Goo YH, Sohn YC, Kim DH, Kim SW, Kang yltransferase activity of a Drosophila Poly- 158–163, 1999 MJ, Jung DJ, Kwak E, Barlev NA, Berger SL, comb group repressor complex. Cell 111: 6. Chen T, Li E: Establishment and mainte- Chow VT, Roeder RG, Azorsa DO, Meltzer 197–208, 2002 nance of DNA methylation patterns in mam- PS, Suh PG, Song EJ, Lee KJ, Lee YC, Lee 33. Beisel C, Imhof A, Greene J, Kremmer E, mals. Curr Top Microbiol Immunol 301: JW: Activating signal cointegrator 2 belongs Sauer F: Histone methylation by the Dro- 179–201, 2006 to a novel steady-state complex that con- sophila epigenetic transcriptional regulator 7. Klose RJ, Bird AP: Genomic DNA methyl- tains a subset of trithorax group proteins. Ash1. Nature 419: 857–862, 2002 ation: The mark and its mediators. Trends Mol Cell Biol 23: 140–149, 2003 34. Klymenko T, Muller J: The histone methyl- Biochem Sci 31: 89–97, 2006 21. Milne TA, Briggs SD, Brock HW, Martin ME, transferases Trithorax and Ash1 prevent 8. Jones PL, Veenstra GJ, Wade PA, Vermaak Gibbs D, Allis CD, Hess JL: MLL targets SET transcriptional silencing by Polycomb group D, Kass SU, Landsberger N, Strouboulis J, domain methyltransferase activity to Hox proteins. EMBO Rep 5: 373–377, 2004 Wolffe AP: Methylated DNA and MeCP2 re- gene promoters. Mol Cell 10: 1107–1117, 35. Bernstein BE, Mikkelsen TS, Xie X, Kamal M, cruit histone deacetylase to repress tran- 2002 Huebert DJ, Cuff J, Fry B, Meissner A, scription. Nat Genet 19: 187–191, 1998 22. Min J, Zhang Y, Xu RM: Structural basis for Wernig M, Plath K, Jaenisch R, Wagschal A, 9. Wade PA, Gegonne A, Jones PL, Ballestar E, specific binding of Polycomb chromodo- Feil R, Schreiber SL, Lander ES: A bivalent Aubry F, Wolffe AP: Mi-2 complex couples main to histone H3 methylated at Lys 27. chromatin structure marks key developmen- DNA methylation to chromatin remodelling Genes Dev 17: 1823–1828, 2003 tal genes in embryonic stem cells. Cell 125: and histone deacetylation. Nat Genet 23: 23. Jacobs SA, Khorasanizadeh S: Structure of 315–326, 2006 62–66, 1999 HP1 chromodomain bound to a lysine 36. O’Neill LP, VerMilyea MD, Turner BM: Epi- 10. Nan X, Ng HH, Johnson CA, Laherty CD, 9-methylated histone H3 tail. Science 295: genetic characterization of the early embryo Turner BM, Eisenman RN, Bird A: Transcrip- 2080–2083, 2002 with a chromatin immunoprecipitation pro- tional repression by the methyl-CpG-bind- 24. Nielsen PR, Nietlispach D, Mott HR, Cal- tocol applicable to small cell populations. ing protein MeCP2 involves a histone laghan J, Bannister A, Kouzarides T, Murzin Nat Genet 38: 835–841, 2006 deacetylase complex. Nature 393: 386–389, AG, Murzina NV, Laue ED: Structure of the 37. Metzger E, Wissmann M, Yin N, Muller JM, 1998 HP1 chromodomain bound to histone H3 Schneider R, Peters AH, Gunther T, Buettner 11. Jenuwein T, Allis CD: Translating the histone methylated at lysine 9. Nature 416: 103– R, Schule R: LSD1 demethylates repressive code. Science, 293: 1074–1080, 2001 107, 2002 histone marks to promote androgen-recep- 12. Berger SL: The complex language of chro- 25. Wysocka J, Swigut T, Milne TA, Dou Y, tor-dependent transcription. Nature 437: matin regulation during transcription. Na- Zhang X, Burlingame AL, Roeder RG, Brivan- 436–439, 2005 ture 447: 407–412, 2007 lou AH, Allis CD: WDR5 associates with his- 38. Shi Y, Lan F, Matson C, Mulligan P, Whet- 13. Rusche LN, Kirchmaier AL, Rine J: The es- tone H3 methylated at K4 and is essential for stine JR, Cole PA, Casero RA, Shi Y: Histone tablishment, inheritance, and function of si- H3 K4 methylation and vertebrate develop- demethylation mediated by the nuclear lenced chromatin in Saccharomyces cerevi- ment. Cell 121: 859–872, 2005 amine oxidase homolog LSD1. Cell 119: siae. Annu Rev Biochem 72: 481–516, 2003 26. Wysocka J, Swigut T, Xiao H, Milne TA, 941–953, 2004 14. Kurdistani SK, Grunstein M: Histone acetyla- Kwon SY, Landry J, Kauer M, Tackett AJ, 39. Tsukada Y, Fang J, Erdjument-Bromage H, tion and deacetylation in yeast. Nat Rev Mol Chait BT, Badenhorst P, Wu C, Allis CD: A Warren ME, Borchers CH, Tempst P, Zhang Cell Biol 4: 276–284, 2003 PHD finger of NURF couples histone H3 ly- Y: Histone demethylation by a family of 15. Rea S, Eisenhaber F, O’Carroll D, Strahl BD, sine 4 trimethylation with chromatin remod- JmjC domain-containing proteins. Nature Sun ZW, Schmid M, Opravil S, Mechtler K, elling. Nature 442: 86–90, 2006 439: 811–816, 2006 Ponting CP, Allis CD, Jenuwein T: Regula- 27. Ringrose L, Paro R: Polycomb/Trithorax re- 40. Okita K, Ichisaka T, Yamanaka S: Generation tion of chromatin structure by site-specific sponse elements and epigenetic memory of of germline-competent induced pluripotent histone H3 methyltransferases. Nature 406: cell identity. Development 134: 223–232, stem cells. Nature 448: 313–317, 2007 593–599, 2000 2007 41. Wernig M, Meissner A, Foreman R, Bram- 16. Litt MD, Simpson M, Gaszner M, Allis CD, 28. Schwartz YB, Pirrotta V: Polycomb silencing brink T, Ku M, Hochedlinger K, Bernstein Felsenfeld G: Correlation between histone mechanisms and the management of BE, Jaenisch R: In vitro reprogramming of lysine methylation and developmental genomic programmes. Nat Rev Genet 8: fibroblasts into a pluripotent ES-cell-like changes at the chicken beta-globin locus. 9–22, 2007 state. Nature 448: 318–324, 2007 Science 293: 2453–2455, 2001 29. Nakayama J, Rice JC, Strahl BD, Allis CD, 42. Loh YH, Zhang W, Chen X, George J, Ng 17. Santos-Rosa H, Schneider R, Bannister AJ, Grewal SI: Role of histone H3 lysine 9 meth- HH: Jmjd1a and Jmjd2c histone H3 Lys 9 Sherriff J, Bernstein BE, Emre NC, Schreiber ylation in epigenetic control of heterochro- demethylases regulate self-renewal in em- SL, Mellor J, Kouzarides T: Active genes are matin assembly. Science 292: 110–113, bryonic stem cells. Genes Dev 21: 2545– tri-methylated at K4 of histone H3. Nature 2001 2557, 2007 419: 407–411, 2002 30. Heard E, Rougeulle C, Arnaud D, Avner P, 43. Dressler GR: The cellular basis of kidney de- 18. Briggs SD, Bryk M, Strahl BD, Cheung WL, Allis CD, Spector DL: Methylation of histone velopment. Annu Rev Cell Dev Biol 22: 509– Davie JK, Dent SY, Winston F, Allis CD: His- H3 at Lys-9 is an early mark on the X chro- 529, 2006 tone H3 lysine 4 methylation is mediated by mosome during X inactivation. Cell 107: 44. Saxen L: Organogenesis of the kidney. In: Set1 and required for cell growth and rDNA 727–738, 2001 Developmental and Cell Biology Series, silencing in Saccharomyces cerevisiae. 31. Czermin B, Melfi R, McCabe D, Seitz V, Im- 19th Ed., edited by Barlow PW, Green PB, Genes Dev 15: 3286–3295, 2001 hof A, Pirrotta V: Drosophila enhancer of White CC, Cambridge, UK, Cambridge Uni- 19. Nakamura T, Mori T, Tada S, Krajewski W, Zeste/ESC complexes have a histone H3 versity Press, 1987, pp 1–173 Rozovskaia T, Wassell R, Dubois G, Mazo A, methyltransferase activity that marks chro- 45. Carroll TJ, Park JS, Hayashi S, Majumdar A, Croce CM, Canaani E: ALL-1 is a histone mosomal Polycomb sites. Cell 111: 185– McMahon AP: Wnt9b plays a central role in methyltransferase that assembles a super- 196, 2002 the regulation of mesenchymal to epithelial complex of proteins involved in transcrip- 32. Muller J, Hart CM, Francis NJ, Vargas ML, transitions underlying organogenesis of the tional regulation. Mol Cell 10: 1119–1128, Sengupta A, Wild B, Miller EL, O’Connor mammalian urogenital system. Dev Cell 9: 2002 MB, Kingston RE, Simon JA: Histone meth- 283–292, 2005

2066 Journal of the American Society of Nephrology J Am Soc Nephrol 19: 2060–2067, 2008 www.jasn.org SPECIAL ARTICLE

46. Tsang TE, Shawlot W, Kinder SJ, Kobayashi BRCT-domain containing protein PTIP links Kehlen A, Garbe JC, Stampfer MR, Dam- A, Kwan KM, Schughart K, Kania A, Jessell PAX2 to a histone H3, lysine 4 methyltrans- mann R: Chromatin inactivation precedes TM, Behringer RR, Tam PP: Lim1 activity is ferase complex. Dev Cell 13: 580–592, 2007 de novo DNA methylation during the pro- required for intermediate mesoderm differ- 56. Lechner MS, Levitan I, Dressler GR: PTIP, a gressive epigenetic silencing of the entiation in the mouse embryo. Dev Biol novel BRCT domain-containing protein in- RASSF1A promoter. Mol Cell Biol 25: 223: 77–90, 2000 teracts with Pax2 and is associated with ac- 3923–3933, 2005 47. James RG, Kamei CN, Wang Q, Jiang R, tive chromatin. Nucleic Acids Res 28: 2741– 65. Krivtsov AV, Armstrong SA: MLL transloca- Schultheiss TM: Odd-skipped related 1 is 2751, 2000 tions, histone modifications and leukaemia required for development of the metaneph- 57. Manke IA, Lowery DM, Nguyen A, Yaffe MB: stem-cell development. Nat Rev Cancer 7: ric kidney and regulates formation and dif- BRCT repeats as phosphopeptide-binding 823–833, 2007 ferentiation of kidney precursor cells. Devel- modules involved in protein targeting. Sci- 66. Varambally S, Dhanasekaran SM, Zhou M, opment 133: 2995–3004, 2006 ence 302: 636–639, 2003 Barrette TR, Kumar-Sinha C, Sanda MG, 48. Bouchard M, Souabni A, Mandler M, Neu- 58. Issaeva I, Zonis Y, Rozovskaia T, Orlovsky K, Ghosh D, Pienta KJ, Sewalt RG, Otte AP, buser A, Busslinger M: Nephric lineage Croce CM, Nakamura T, Mazo A, Eisenbach Rubin MA, Chinnaiyan AM: The polycomb specification by Pax2 and Pax8. Genes Dev L, Canaani E: Knockdown of ALR (MLL2) re- group protein EZH2 is involved in progres- 16: 2958–2970, 2002 veals ALR target genes and leads to alter- sion of prostate cancer. Nature 419: 624– 49. Carroll TJ, Vize PD: Synergism between ations in cell adhesion and growth. Mol Cell 629, 2002 Pax-8 and lim-1 in embryonic kidney devel- Biol 27: 1889–1903, 2007 67. Shi Y: Histone lysine demethylases: Emerg- opment. Dev Biol 214: 46–59, 1999 59. Cho YW, Hong T, Hong S, Guo H, Yu H, Kim ing roles in development, physiology and 50. Tena JJ, Neto A, de la Calle-Mustienes E, D, Guszczynski T, Dressler GR, Copeland disease. Nat Rev Genet 8: 829–833, 2007 Bras-Pereira C, Casares F, Gomez-Skarmeta TD, Kalkum M, Ge K: PTIP associates with 68. Wiggins JE, Goyal M, Sanden SK, Wharram JL: Odd-skipped genes encode repressors MLL3- and MLL4-containing histone H3 ly- BL, Shedden KA, Misek DE, Kuick RD, Wig- that control kidney development. Dev Biol sine 4 methyltransferase complex. J Biol gins RC: Podocyte hypertrophy, “adapta- 301: 518–531, 2007 Chem 282: 20395–20406, 2007 tion,” and “decompensation” associated 51. James RG, Schultheiss TM: Bmp signaling 60. Cai Y, Brophy PD, Levitan I, Stifani S, with glomerular enlargement and glomeru- promotes intermediate mesoderm gene ex- Dressler GR: Groucho suppresses Pax2 losclerosis in the aging rat: Prevention by pression in a dose-dependent, cell-autono- transactivation by inhibition of JNK-medi- calorie restriction. J Am Soc Nephrol 16: mous and translation-dependent manner. ated phosphorylation. EMBO J 22: 5522– 2953–2966, 2005 Dev Biol 288: 113–125, 2005 5529, 2003 69. Wiggins JE, Goyal M, Wharram BL, Wiggins 52. James RG, Schultheiss TM: Patterning of the 61. Cai Y, Lechner MS, Nihalani D, Prindle MJ, RC: Antioxidant ceruloplasmin is expressed avian intermediate mesoderm by lateral plate Holzman LB, Dressler GR: Phosphorylation by glomerular parietal epithelial cells and and axial tissues. Dev Biol 253: 109–124, 2003 of Pax2 by the c-Jun N-terminal kinase and secreted into urine in association with glo- 53. Wellik DM, Hawkes PJ, Capecchi MR: Hox11 enhanced Pax2-dependent transcription ac- merular aging and high-calorie diet. JAm paralogous genes are essential for meta- tivation. J Biol Chem 277: 1217–1222, 2002 Soc Nephrol 17: 1382–1387, 2006 nephric kidney induction. Genes Dev 16: 62. Cho EA, Prindle MJ, Dressler GR: BRCT do- 70. Fraga MF, Esteller M: Epigenetics and ag- 1423–1432, 2002 main-containing protein PTIP is essential for ing: the targets and the marks. Trends 54. Gong KQ, Yallowitz AR, Sun H, Dressler GR, progression through mitosis. Mol Cell Biol Genet 23: 413–418, 2007 Wellik DM: A Hox-Eya-Pax complex regu- 23: 1666–1673, 2003 71. Bennett-Baker PE, Wilkowski J, Burke DT: lates early kidney developmental gene ex- 63. Jones PA, Laird PW: Cancer epigenetics Age-associated activation of epigenetically pression. Mol Cell Biol 27: 7661–7668, 2007 comes of age. Nat Genet 21: 163–167, 1999 repressed genes in the mouse. Genetics 55. Patel SR, Kim D, Levitan I, Dressler GR: The 64. Strunnikova M, Schagdarsurengin U, 165: 2055–2062, 2003

J Am Soc Nephrol 19: 2060–2067, 2008 Epigenetics, Development, and the Kidney 2067