to develop class II HDAC inhibitors as therapeutics for treating metabolic disorders.

Given that pharmacological HDAC inhibitors are already in development as potential anticancer agents, this strategy seems particularly promising. Importantly, the two new studies also raise a number of complex questions that are critical to understanding the regulation of metabolic homeostasis. First, the studies highlight the importance of acetylation in controlling Foxo function. Given Foxos involvement in numerous physiological processes, this regulatory mechanism is expected to inuence not only metabolism but also cell growth, proliferation, apoptosis, and longevity. However, the relative contributions of Sirtuins and class IIa HDACs for controlling Foxo activity still remain unknown. It will also be important to understand why gluconeogenic gene expression is regulated by two distinct pathways: HDAC inhibition

by AMPK/SIK1/2 and CRTC2 inhibition by SIK1/2. One speculation is that the perceived redundancy allows intricate, context-dependent regulation of gluconeogenesis or of Foxo function more generally. Finally, the physiologic consequences of this regulatory mechanism might reach beyond glucose and lipid homeostasis and may include Foxomediated control of life span, cell survival, and growth. The combined power of Drosophila and mouse genetics will clearly provide key insights into these intriguing questions.

Biddinger, S.B., and Kahn, C.R. (2006). Annu. Rev. Physiol. 68, 123158. Guarente, L. (2006). Nature 444, 868874. Haigis, M.C., and Sinclair, D.A. (2010). Annu. Rev. Pathol. 5, 253295. Mihaylova, M.M., Vasquez, D.S., Ravnskjaer, K., Denechaud, P.-D., Yu, R.T., Alvarez, J.G., Downes, M., Evans, R.M., Montminy, M., and Shaw, R.J. (2011). Cell 145, this issue, 607621. Purushotham, A., Schug, T.T., Xu, Q., Surapureddi, S., Guo, X., and Li, X. (2009). Cell Metab. 9, 327338. Rodgers, J.T., Lerin, C., Gerhart-Hines, Z., and Puigserver, P. (2008). FEBS Lett. 582, 4653.

REFERENCES Altarejos, J.Y., and Montminy, M. (2011). Nat. Rev. Mol. Cell Biol. 12, 141151. Berdeaux, R., Goebel, N., Banaszynski, L., Takemori, H., Wandless, T., Shelton, G.D., and Montminy, M. (2007). Nat. Med. 13, 597603.

Viollet, B., Guigas, B., Leclerc, J., Hebrard, S., Lantier, L., Mounier, R., Andreelli, F., and Foretz, M. (2009). Acta Physiol. (Oxf.) 196, 8198. Wang, B., Moya, N., Niessen, S., Hoover, H., Mihaylova, M.M., Shaw, R.J., Yates, J.R., III, Fischer, W.H., Thomas, J.B., and Montminy, M. (2011). Cell 145, this issue, 596606.

A Hormone Sends Instant Messages to the Genome

Andreas Prokesch1,2 and Mitchell A. Lazar1,*
1Division of Endocrinology, Diabetes, and Metabolism, Department of Medicine, The Institute for Diabetes, Obesity, and Metabolism, University of Pennsylvania School of Medicine, Philadelphia, PA 19104, USA 2Institute for Genomics and Bioinformatics, Graz University of Technology, A-8010 Graz, Austria *Correspondence: lazar@mail.med.upenn.edu DOI 10.1016/j.cell.2011.04.016

Bridging a gap between transcriptomics and the study of cis-acting elements (cistromics), Hah et al. (2011) apply a next-generation sequencing technique to gain an unprecedented view of the changes in RNA synthesis that occur following estrogen receptor activation in human breast cancer cells.
The regulation of eukaryotic gene expression is remarkably complex, with the transcriptome in any given cell affected by the epigenome, which is acted upon by transcription factors binding to cis-acting elements (the cistrome) and RNA polymerases that generate messenger RNAs (mRNAs) (Figure 1). In this issue, Hah et al. (2011) bridge transcriptomics and cistromics to provide surprising insights into the genome-wide distribution and dynamics of transcription in response to an external stimulusin this case, the exposure of human breast cancer cells to the hormone estrogen. Techniques based on next-generation sequencing, most notably chromatin immunoprecipitation sequencing (ChIPseq) and RNA-seq are greatly contributing to our understanding of transcriptional regulation. ChIP-seq enables researchers to localize transcription factor binding or specic nucleosome structures dening cistromes in a given cell type or tissue under different conditions (Lupien and Brown, 2009). RNA-seq makes it possible to assay many species of transcripts, including mRNAs, noncoding RNAs, and small RNAs, while providing exact information about alternative splice variants

Cell 145, May 13, 2011 2011 Elsevier Inc. 499

in the surveyed biological at later time points are rst sample (Ozsolak and Milos, loaded with Pol2, and only 2011). Although combining later does the GRO-seq signal these approaches is a powerextend into the gene body. ful means of gaining biologThese insights into the kiical insights, one limitation is netics of transcription cannot that transcriptomes reprebe gleaned from measuring sent steady-state mRNA steady-state mRNA levels levels, which are determined that take hours to accumulate. by the rate of mRNA synMoreover, owing to the large thesis divided by the rate of number of nongenic tranmRNA degradation. RNA scripts detected, the GROdegradation is highly reguseq method was able to lated, leading to posttranassign three to four times scriptional alteration of gene more of the ERa cistrome to expression that is largely transcribed regions than was independent of the cistromes possible using transcriptome that govern mRNA synthesis. data (Welboren et al., 2009; Thus, to appreciate the true Carroll et al., 2006). impact of the cistromes, there GRO-seq also reveals new is a need to measure the aspects of transcription that synthesis of nascent RNA cannot be learned by ChIPmolecules and thereby the seq for Pol2. Some 16% of the Figure 1. The Missing Link between the Cistrome and the Trannumerator of the mRNA syntranscripts stem from divergent scriptome thesis/mRNA degradation transcription, a phenomenon Estradiol-bound dimers of the estrogen receptor bind either at distal equation (Figure 1). that has recently been shown enhancers or at promoters of their target genes, thereby dening the estrogen Lis and colleagues recently to be widespread in the gereceptor (ER) cistrome. This binding leads to initiation of transcription by RNA polymerase 2 (Pol2) and eventually to synthesis of nascent RNA chains. The invented a technique to nome and may promote a steady-state of the cytoplasmic RNA poolthe transcriptomeis determined address this gap. Called nucleosome-poor environment by the rate of RNA synthesis divided by the rate of RNA degradation. GRO-seq GRO-Seq (genome-wide nuat promoters where transcrip(genome-wide nuclear run-on followed by next-generation sequencing) connects the cistrome and the transcriptome by mapping the quantity and clear run-on followed by tional rates need to be high genome-wide distribution of transcriptionally engaged RNA polymerases. PIC, next-generation sequencing) (Seila et al., 2009). Consistent preinitiation complex; TSS, transcription start site. (Core et al., 2008), it is based with this, E2 regulates hundreds of divergent transcripts on classical nuclear run-on experiments in which ongoing transcrip- breast cancer cells treated with 17b- that correlate with accumulation of the tion is studied in isolated nuclei under estradiol (E2), a natural ligand of the estro- protein-coding mRNA. In addition, Hah conditions that allow labeling of nascent, gen receptor, for 10, 40, and 160 min. et al. uncover antisense transcription and actively transcribed RNA molecules, Using new bioinformatic algorithms for transcripts from Pol1 and Pol3some which can now be puried and deeply transcript calling and quantitation, the regulated by E2that cannot be identied sequenced in a genome-wide manner. authors report numerous insights into by Pol2 ChIP-seq because it does not More recently, this lab employed GRO- estrogen receptor signaling and tran- deliver strand-specic information. seq to compare the distribution of tran- scriptional regulation in general, some of The authors also noted short pieces of scriptionally engaged RNA polymerase 2 which relate directly to the difference RNA transcribed from enhancers, as was (Pol 2) in embryonic stem cells and differ- between measuring the rate of transcrip- recently observed in neurons using RNAentiated mouse embryonic broblasts tion versus the level of mature transcript. seq and Pol 2 ChIP-seq (Kim et al., 2010). (Min et al., 2011). In another study, GROFor example, a surprisingly high per- The observation of RNAs emanating from seq was used to demonstrate that the centage (26%) of transcripts active at enhancers using different methods in very male-specic lethal (MSL) complex in one or more time point during the experi- different cell systems suggests their Drosophila enhances transcription by ment were regulated by E2 treatment, general existence. Although their function facilitating the progression of Pol2 across whereas 10% respond after only remains obscure, their locations are active X-linked genes, thus contributing to 10 min. In addition, direct comparison of consistent with the nding of ER and other new understanding of sex-specic gene transcription rate with steady-state tran- transcription factors at enhancers far from dosage differences on X chromosomes scriptomes (from microarrays) shows transcription start sites (the dark matter (Larschan et al., 2011). Hah et al. now that the synthesis of protein-coding tran- of the genome) that has been one of the utilize GRO-seq to provide exciting new scripts is transiently responsive to E2 major surprises of cistromics. insights into hormone action. in contrast to the gradual increase in It is clear that transcriptome analyses, The authors measure the instantaneous mRNA levels. Further, the promoters of especially RNA-seq, will continue to prorate of RNA synthesis in MCF-7 human most genes that are maximally induced vide valuable biological information. Logic
500 Cell 145, May 13, 2011 2011 Elsevier Inc.

suggests that steady-state protein levels, which largely determine cellular functions, will correlate better with steady-state mRNA levels than with the rate of mRNA transcription, although this should be formally proven. But for those interested in the direct effects of transcription factors or of the environment (including hormonal signals) upon transcription, the ability to measure the rates of RNA synthesis in a genome-wide manner the RNA synthesomeopens new doors to discovery.
ACKNOWLEDGMENTS A.P. was supported by a GEN-AU mobility grant from the Austrian Ministry of Science and

Research. Cistromic work in the Lazar lab is funded by NIH DK45586, DK43806, DK49780, and NURSA (DK062434). REFERENCES Carroll, J.S., Meyer, C.A., Song, J., Li, W., Geistlinger, T.R., Eeckhoute, J., Brodsky, A.S., Keeton, E.K., Fertuck, K.C., Hall, G.F., et al. (2006). Nat. Genet. 38, 12891297. Core, L.J., Waterfall, J.J., and Lis, J.T. (2008). Science 322, 18451848. Hah, N., Danko, C.G., Core, L., Waterfall, J.J., Siepel, A., Lis, J.T., and Kraus, W.L. (2011). Cell 145, this issue, 622634. Kim, T.K., Hemberg, M., Gray, J.M., Costa, A.M., Bear, D.M., Wu, J., Harmin, D.A., Laptewicz, M., Barbara-Haley, K., Kuersten, S., et al. (2010). Nature 465, 182187.

Larschan, E., Bishop, E.P., Kharchenko, P.V., Core, L.J., Lis, J.T., Park, P.J., and Kuroda, M.I. (2011). Nature 471, 115118. Lupien, M., and Brown, M. (2009). Endocr. Relat. Cancer 16, 381389. Min, I.M., Waterfall, J.J., Core, L.J., Munroe, R.J., Schimenti, J., and Lis, J.T. (2011). Genes Dev. 25, 742754. Ozsolak, F., and Milos, P.M. (2011). Nat. Rev. Genet. 12, 8798. Seila, A.C., Core, L.J., Lis, J.T., and Sharp, P.A. (2009). Cell Cycle 8, 25572564. Welboren, W.J., van Driel, M.A., Janssen-Megens, E.M., van Heeringen, S.J., Sweep, F.C., Span, P.N., and Stunnenberg, H.G. (2009). EMBO J. 28, 14181428.

Cell 145, May 13, 2011 2011 Elsevier Inc. 501