Finding Primary Targets of Transcriptional Regulators

10 downloads 6178 Views 185KB Size Report
... online as a Cell Cycle E-publication: http://www.landesbioscience.com/journals/cc/abstract.php?id=1521 .... Genes Dev 2002; 16:235-44. 4. Buck MJ, Lieb JD.
[Cell Cycle 4:3, 356-357; March 2005]; ©2005 Landes Bioscience

Finding Primary Targets of Transcriptional Regulators Extra Views

Received 12/27/04; Accepted 01/05/04

OT D

IEN

I thank Dr. Keith Yamamoto for his support and comments on writing this article, and Drs. Marc Van Gilst and Eric Bolton for critiques of the manuscript. Research in the Yamamoto laboratory is supported by grants from NIH. I am supported by a postdoctoral fellowship from the American Heart Association.

Transcriptional regulators control cell differentiation, development, the cell cycle and other physiological responses by modulating specific gene expression networks, which are a complex mixture of “primary” targets, defined as those genes at which a particular transcriptional regulator directly acts, and secondary effects, defined as expression changes that are consequent to the primary regulatory events. Thus, at primary targets, the regulator occupies genomic response elements either by direct DNA binding or, at “tethering” response elements, through protein:protein interactions with other bound factors.1 Because it is the primary target genes that set in motion the processes eventually seen as global physiological effects, an understanding of complex gene networks requires efficient methods to distinguish primary from secondary targets. Techniques such as DNA microarrays fail to make this distinction. There are several methods to identify primary target genes (Table 1). If the consensus DNA-binding site of a transcriptional regulator is known, a bioinformatic/computational approach that accommodates the sequence degeneracy known among binding sites at different natural response elements could be applied to identify putative response elements in the genome, and consequently, predict potential primary target genes. Candidates identified by this approach must be interrogated by chromatin immunoprecipitation (ChIP), to determine whether these response elements are indeed occupied in vivo. The candidate target genes must also be shown to be regulated at the transcriptional level, using functional assays such as loss of function (via RNAi) or overexpression of the regulator. Bioinformatic approaches are limited in their ability to predict more complex response elements. Notably, some transcriptional regulators bind to a wide variety of DNA sequences, many of which differ substantially from characterized consensus elements, thus complicating bioinformatic strategies. Furthermore, at response elements where the regulator of interest is recruited via the “tethering”mechanism, primary target genes will go undetected using bioinformatic approaches. In the past several years, the development of another method, ChIP on chip, has provided a comprehensive strategy to identify primary target genes.2-4 In this approach, a ChIP assay is first used to crosslink transcriptional regulators to their response elements, and then the precipitated DNA fragments are subjected to microarray analysis to identify nearby genes. Quantitative real-time PCR (qPCR) and RNAi, as discussed above, are then used to assess whether or not these candidate genes are transcriptionally regulated in vivo. Unlike the bioinformatics approach, ChIP on chip requires no prior knowledge of DNA binding sequences. However, it is a complicated and expensive procedure that requires availability of DNA microarrays that include upstream and intronic regulatory regions; especially for complex metazoans, arrays of this sort are not commonly available. At present, the microarrays commonly employed for most ChIP on chip experiments include

ON

ACKNOWLEDGMENTS

.D

chromatin immunoprecipitation (ChIP), transcription, transcriptional regulators, primary target gene, response element

CE

KEY WORDS

IST

RIB

Previously published online as a Cell Cycle E-publication: http://www.landesbioscience.com/journals/cc/abstract.php?id=1521

Transcriptional regulators directly regulate the transcription of primary target genes by occupying genomic response elements proximal to the target promoters. In turn, regulation of primary target genes triggers subsequent physiological events by acting on distinct biological pathways and modulating the expression of secondary target genes. Thus, distinguishing primary and secondary target genes is critical for our understanding of mechanisms underlying the biological effects of a regulator. In this article, four distinct strategies for identification of primary target genes are compared and discussed: bioinformatic approaches, ChIP on chip, and two novel strategies, ChIP scanning and in vitro genomic selection. Overall, these approaches complement each other, and depending on available information and resources, the appropriate methods can be chosen to identify primary target genes for a particular regulator.

.

Correspondence to: Jen-Chywan Wang; Department of Cellular and Molecular Pharmacology; University of California, San Francisco; S574 Genentech Hall, 600 16th Street; San Francisco, California 94107-2280 USA; Tel.: 415.476.2251; Email: [email protected]

ABSTRACT

UT E

Jen-Chywan Wang

SC

NOTE ADDED IN PROOF

©

20

05

LA

ND

ES

BIO

After this article was accepted, an alternative and comprehensive approach to identify primary target genes was published (Impey et. al., Defining the CREB regulon: a genomewide analysis of transcription factor regulatory regions. Cell, 2004 vol 119, p1041-1054). As this approach is based on ChIP, its advantages and limitations are similar to those of ChIP scanning.

356

Cell Cycle

2005; Vol. 4 Issue 3

Finding Primary Targets of Transcriptional Regulators

Table 1

Comparison of strategies for identification of primary target genes Bioinformatic/

ChIP on chip

ChIP scanning

In vitro genomic selection

computational search Requires prior knowledge of DNA binding sequences of transcriptional regulators

yes

no

no

no

Requires functional Analysis

yes

yes

no perform functional screen first

yes

Requires antibody or purified protein

no but Ab is required to perform ChIP to confirm identified targets

Ab

Ab

protein

Requires genomic array

no

yes

no

no

Identifies primary targets regulated via tethering response elements

no

yes

yes

no

Resolution of response element identification

10-20 bp transcriptional regulator binding site

usually 1-2 kb

500 bp

200-300 bp

only 1–2 kb upstream of the promoter. Thus, primary targets that host response elements located outside of this segment will elude identification. ChIP scanning is a related but distinct strategy to identify primary target genes.1,5-7 In this approach, primary and secondary target genes for a particular transcriptional regulator are first identified by a conventional microarray analysis. ChIP is then employed to identify those genes directly occupied by the transcriptional regulator of interest. Thus, the microarray expression survey provides a functional assay that constrains the number of genes that are then assessed by ChIP; importantly, genomic microarrays are not required. Rather, qPCR is employed for rapid, quantitative and economical ChIP screens. We commonly design primer sets for 500 bp segments representing ~3 kb upstream of a candidate gene; in a 96-well plate, with 6 primer sets per gene plus a control, 15 genes can be screened.1 Indeed, this approach should be readily adaptable to a 384-well format, or even higher throughput using custom chips. In addition to its efficiency for identifying primary targets, another advantage of ChIP scanning is that the in vivo transcriptional regulator binding sites are readily localized to a small region, thus facilitating identification of core binding sequences for factors of interest. Notably, ChIP scanning can be designed to survey entire genomic regions including exons and introns, which commonly contain potential response elements. Both ChIP scanning and ChIP on chip can identify target genes and locate response elements for regulators bound at tethering response elements, as well as those occupied by cofactors of interest.7 As with ChIP on chip, high quality antibodies are essential in ChIP scanning experiments. Moreover, variations in regulator conformation or complex geometry can render epitopes undetectable in particular response element contexts. Two or more antibodies that detect distinct epitopes may be useful in such cases. In addition, if antibodies are not available, one can express tagged-transcriptional regulators, and then perform ChIP experiments with the corresponding tag-specific antibodies. This strategy has been widely applied in studies of transcriptional regulatory mechanisms in yeast. “In vitro genomic selection” is a novel approach that does not rely on antibodies to identify primary target genes. This strategy was exploited to identify C. elegans genomic DNA fragments that bind www.landesbioscience.com

directly to the purified DNA binding domain of a transcriptional regulator, DAF-12.8 The genes closely linked to these DNA fragments were then shown by qPCR to be regulated in vivo by DAF-12. In theory, this approach should be readily applicable to DNA binding transcriptional regulators in mammals and other complex metazoans. Similar to ChIP on chip and ChIP scanning, in vitro genomic selection does not require prior knowledge of DNA binding sequences of transcriptional regulators. Furthermore, comparative analyses of isolated fragments could lead to the identification of consensus DNA binding sites for a regulator, as in the case of DAF-12.8 The average fragment size isolated from this approach is 200–300 bp; such segments can be readily manipulated for further identification of response elements. Notably, however, this approach, unlike ChIP scanning and ChIP on chip, detects only that subset of primary target genes in which the regulator of interest binds directly to DNA. In summary, two new approaches, ChIP scanning and in vitro genomic selection, provide new alternatives for identifying the primary target genes of transcriptional regulators. These strategies complement other methods, such as bioinformatic/computational search and ChIP on chip, for target gene identification (Table 1). Thus, one can choose one or multiple approaches to identify primary targets, depending on available information and resources, such as antibodies or purified proteins. Overall, the revelation of primary targets will not only facilitate understanding of the pathways that govern biological functions of specific transcriptional regulators, but will also advance analyses of the modes by which regulators control gene transcription, and of the evolution and function of transcriptional regulatory networks. References 1. Wang JC, Derynck MK, Nonaka DF, Khodabakhsh DB, Haqq C, Yamamoto KR. Chromatin immunoprecipitation (ChIP) scanning identifies primary glucocorticoid receptor target genes. Proc Natl Acad Sci USA 2004; 101:15603-8. 2. Ren B, Cam H, Takahashi Y, Volkert T, Terragni J, Young RA, Dynlacht BD. E2F integrates cell cycle progression with DNA repair, replication, and G2/M checkpoints. Genes Dev 2002; 16:245-56. 3. Weinmann AS, Yan PS, Oberley MJ, Huang TH, Farnham PJ. Isolating human transcription factor targets by coupling chromatin immunoprecipitation and CpG island microarray analysis. Genes Dev 2002; 16:235-44. 4. Buck MJ, Lieb JD. ChIP-chip: Considerations for the design, analysis, and application of genome-wide chromatin immunoprecipitation experiments. Genomics 2004; 83:349-60.

Cell Cycle

357

Finding Primary Targets of Transcriptional Regulators

5. Zeller KI, Haggerty TJ, Barrett JF, Guo Q, Wonsey DR, Dang CV. Characterization of nucleophosmin (B23) as a Myc target by scanning chromatin immunoprecipitation. J Biol Chem 2001; 276:48285-91. 6. Haggerty TJ, Zeller KI, Osthus RC, Wonsey DR, Dang CV. A strategy for identifying transcription factor binding sites reveals two classes of genomic c-Myc target sites. Proc Natl Acad Sci USA 2003; 100:5313-8. 7. Kirmizis A, Bartley SM, Kuzmichev A, Margueron R, Reinberg D, Green R, Farnham PJ. Silencing of human polycomb target genes is associated with methylation of histone H3 Lys 27. Genes Dev 2004; 18:1592-605. 8. Shostak Y, Van Gilst MR, Antebi A, Yamamoto KR. Identification of C. elegans DAF-12-binding sites, response elements, and target genes. Genes Dev 2004; 18:2529-44.

358

Cell Cycle

2005; Vol. 4 Issue 3