Nielsen H. RNA: Methods and Protocols

Подождите немного. Документ загружается.

234 Nelson, Denisenko, and Bomsztyk

13. Thorne, A. W., Myers, F. A., Hebbes, T. R.

(2004) Native chromatin immunoprecipita-

tion. Methods Mol Biol 287, 21–44.

14. Solomon, M. J., Larsen, P. L., Varshavsky, A.

(1988) Mapping protein-DNA interactions

in vivo with formaldehyde: evidence that his-

tone H4 is retained on a highly transcribed

gene. Cell 53, 937–947.

15. Nelson, J. D., Denisenko, O., Sova, P., Bom-

sztyk, K. (2006) Fast chromatin immunopre-

cipitation assay. Nucleic Acids Res 34, e2.

16. Huebert, D. J., Kamal, M., O‘Donovan, A.,

Bernstein, B. E. (2006) Genome-wide analy-

sis of histone modiﬁcations by ChIP-on-chip.

Methods 40, 365–369.

17. Johnson, D. S., Mortazavi, A., Myers, R.

M., Wold, B. (2007) Genome-wide mapping

of in vivo protein-DNA interactions. Science

316, 1497–1502.

18. Chen, R., Weng, L., Sizto, N. C., Osorio, B.,

Hsu, C. J., Rodgers, R., Litman, D. J.

(1984) Ultrasound-accelerated immunoas-

say, as exempliﬁed by enzyme immunoas-

say of choriogonadotropin. Clin Chem 30,

1446–1451.

19. Nelson, J. D., Flanagin, S., Kawata, Y.,

Denisenko, O., Bomsztyk, K. (2008) Tran-

scription of laminin {gamma}1 chain gene in

rat mesangial cells: constitutive and inducible

RNA polymerase II recruitment and chro-

matin states. Am J Physiol Renal Physiol 294,

F525–F533.

20. Zager, R. A., Johnson, A. C., Naito, M.,

Bomsztyk, K. (2008) Maleate nephrotoxic-

ity: mechanisms of injury and correlates with

ischemic/hypoxic tubular cell death. Am J

Physiol Renal Physiol 294, F187–F197.

21. Denisenko, O., Bomsztyk, K. (2008)

Epistatic interaction between the K-

homology domain protein HEK2 and SIR1

at HMR and telomeres in yeast. J Mol Biol

375, 1178–1187.

22. Glover-Cutter, K., et al. (2008) RNA poly-

merase II pauses and associates with pre-

mRNA processing factors at both ends of

genes. Nat Struct Mol Biol 15(1), 71.

Part II

RNP Methods

Chapter 16

The Post-transcriptional Operon

Scott A. Tenenbaum, Jan Christiansen, and Henrik Nielsen

Abstract

A post-transcriptional operon is a set of monocistronic mRNAs encoding functionally related proteins

that are co-regulated by a group of RNA-binding proteins and/or small non-coding RNAs so that pro-

tein expression is coordinated at the post-transcriptional level. The post-transcriptional operon model

(PTO) is used to describe data from an assortment of methods (e.g. RIP-Chip, CLIP-Chip, miRNA

proﬁling, ribosome proﬁling) that globally address the functionality of mRNA. Several examples of post-

transcriptional operons have been documented in the literature and demonstrate the usefulness of the

model in identifying new participants in cellular pathways as well as in deepening our understanding of

cellular responses.

Key words: mRNA, RNA-binding proteins, miRNA, coordination, regulation.

1. Introduction

Gene expression is regulated at multiple levels. This chap-

ter explores a layer of regulation based on a model called

the post-transcriptional operon, a ribonucleopr o tein infrastruc-

ture in which multiple mRNAs are coordinately regulated by

RNA-binding proteins and/or small non-coding RNAs. Post-

transcriptional operons are analogous to classical prokaryotic,

polycistronic DNA operons in that they coordinate the expres-

sion of a subset of functionally related genes. However, they are

comprised of combinatorial RNA and protein complexes oper-

ating at the post-transcriptional level. The post-transcriptional

operon model or PTO was originally proposed by Jack Keene

and Scott Tenenbaum (1–3) and has more recently been reviewed

(4, 5). In this chapter, we initially provide an outline of many

of the ways in which coordination of gene expression can be

exerted. We then present the post-transcriptional operon model

H. Nielsen (ed.), RNA, Methods in Molecular Biology 703,

DOI 10.1007/978-1-59745-248-9_16, © Springer Science+Business Media, LLC 2011

237

238 Tenenbaum, Christiansen, and Nielsen

and describe key observations that emphasize the importance of

post-transcriptional regulation, especially in multicellular organ-

isms. Finally, we introduce many of the methods that are typically

being used to study this level of regulation and present several

examples of data in support of the PTO model.

2. Coordination

of Gene

Expression

Coordination o f gene expression is a fundamental feature of cellu-

lar life. It is a requirement for cells to grow proportionally, divide,

maintain cellular homeostasis, and respond adequately to external

stimuli. Coordination is obtained in multiple ways and at all levels

of gene expression. All known mechanisms involved in the pro-

cessing of genetic information require some regulation of at least

a fraction of the regulated genes, which will be affected to dif-

fering amounts, leading to coordination in gene expression. The

ﬁrst level that plays a role in the coordination of gene expression

is the physical linkage of some genes dictated by the architec-

ture of the genome. Here, the expression of some protein cod-

ing sequences is coordinated by being transcribed from a single

promoter into a polycistronic transcript. This is characteristic of

the prokaryotic operon organization of genes but is infrequent in

eukaryotes due to the cap-dependent translation of the mRNA.

However, in eukaryotes, the structure of the genome may con-

tribute to coordination by transcriptional interference between

transcription units. This is known from yeast in which cryptic

unstable transcripts regulate the expression of several genes ( 6).

Transcriptional interference may also be relevant in the many

sense/antisense pairs of transcripts that have been observed in

analyses of the human transcriptome (7). Other examples of coor-

dination mediated by the physical linkage of genes include the

globin gene clusters and certain transcription factor genes (Hox

genes) that include a homeobox DNA-binding sequence. The lat-

ter are found in a cluster in which the order of the genes is directly

correlated to the regions in the body plan they affect as well as the

timing of expression.

The chromatin structure of the genome provides coordina-

tion of gene expression at least at two levels. DNA found in the

heterochromatin fraction is largely transcriptionally inactive and

the formation and recession of facultative heterochromatin will

coordinately affect neighbouring genes. Within the euchromatin

fraction, the detailed chromatin structure governs the access of

transcription factors and thus the transcription of the genes. Here

too, the physical linkage of genes may inﬂuence their expression.

There are several examples of “gene expression neighbourhoods”

where genes are expressed in similar patterns due to their physical

linkage (8).

The Post-transcriptional Operon 239

Transcription initiation is one of the key regulatory steps in

gene expression and cell- and tissue-speciﬁc transcription factors

allow the coordinated expression of large sets of genes. Transcrip-

tion factors recognize short sequences and act in a combinato-

rial fashion. Furthermore, these factors are frequently modiﬁed

by post-translational modiﬁcations allowing modulation and ﬁne-

tuning of their regulatory function. More recent discoveries of

promoter-associated short RNA transcripts in both the sense and

antisense orientation at RNA pol II promoters may provide an

additional aspect to this step of regulation of gene expression

(9, 10).

The elongating polymerase is associated with numerous fac-

tors that participate in o r regulate co- or p ost-transcriptional

processing of the transcript. Splicing and polyadenylation are

highly regulated and coordinated events that are controlled by

some of these factors (11, 12). Thus, transcription-associated

processes provide another level at which coordination of expres-

sion can take place. The use of alter native upstream promoters

as well as alternative splicing and polyadenylation leads to for-

mation of mRNA isoforms. Widespread use of alternative tran-

scription start sites located >100 kb upstream was revealed by

the ENCODE pilot project that systematically analysed 1% of the

human genome (13). These transcripts result in the incorpora-

tion of exons derived from transcription units that were previously

annotated as independent as well as exons derived from what was

thought to be intergenic regions. Conventional alternative splic-

ing has been estimated to occur in primary transcripts of almost

all human genes with an average of approximately seven alterna-

tive splicing events per multiexon gene (14, 15). This is one of

the main contributions to diversity of gene expression in humans.

On top of this, alternative polyadenylation is increasingly being

observed, in particular during development. In general, mouse

genes tend to be expressed with longer 3



UTRs as embryonic

development progresses by a mechanism that involves suppression

of short type polyA sites (16). Conversely, shor t type alternative

polyA sites are favoured at late stages of differentiation following

stimulation of resting T cells (17).

Once formed, the fate of an mRNA is governed by the pro-

teins with which it is associated (18). Proteins associate with

the mRNA during transcription and are exchanged during the

life of the mRNA. A prominent example of this is the exchange

of nuclear versions of cap- and polyA tail-binding proteins with

the corresponding cytoplasmic versions upon export to the cyto-

plasm. Following export many, if not all, mRNAs are localized

within the cytoplasm. mRNAs that encode membrane-associated

proteins or proteins that are secreted ar e speciﬁcally localized to

the cytoplasmic side of the endoplasmic reticulum where they are

translated. Fractionation studies have shown that 44% of mRNAs

in human cells belong to this class (19). Around 100 examples

240 Tenenbaum, Christiansen, and Nielsen

of individual mRNA species that are speciﬁcally located within

the cytoplasm are known in diverse organisms (20)andmore

than 400 mRNAs have been identiﬁed in mammalian neuronal

dendrites (21).

The overall level of translation is coordinately regulated, at

least in part, by the availability of initiation factor eIF2 and by the

trimeric eIF4F discriminating between capped and non-capped

mRNAs. The non-capped mRNAs are characterized by the pres-

ence of internal ribosomal entry sites (IRES) or other 5



UTR

elements (22) that mediate alternative translation initiation in sit-

uations of cellular stress or during mitosis when the cap-binding

eIF-4E subunit is absent. Besides regulation of the overall level

of translation, regulation of individual mRNAs and groups of

mRNAs take place by association of protein factors and microR-

NAs. Some protein factors operate in an all or none fashion but

others are combinatorial. MicroRNAs regulate more than 30%

of human genes at the translational level. Individual microRNAs

have many targets and conversely, each mRNA exhibits target

sequences for several microRNAs. Thereby a combinatorial type

of coordinated regulation that serves to ﬁne-tune gene expression

is established. Recent advances in ribosome proﬁling using high-

throughput sequencing (23) are likely to give a more detailed

picture of coordinated regulation than previous methods. Decay

of the mRNA is the ultimate alternative to translation and is not

only an important step in the regulation of individual mRNAs

but also one that displays many examples of coordinated reg-

ulation, in particular of mRNA decay initiated by Argonaute-2

induced endonucleolytic cleavage in the 3



UTR or AU-rich ele-

ment dependent de-adenylation.

Finally, at the protein level there are several crucial mecha-

nisms for coordinated regulation of sets of proteins with related

functions. Proteins may be co-localized because they share a

common sorting signal. They may be regulated in a concerted

fashion because they are substrates of the same kinases and/or

phosphatases or other modiﬁcation enzymes that inﬂuence their

biological properties. Eventually, they may be degraded simulta-

neously by being targeted by the same ubiquitin ligase that marks

the proteins for degradation by the proteasome.

Clearly, coordination of gene expression operates at sev-

eral levels that are intimately related. This interrelatedness fol-

lows from simple observations, including that every transcription

regulator must be translated and every translation factor must

be transcribed. It is also clear that biological processes are too

complex to be studied exhaustively by focusing at single or indi-

vidual levels of regulation. For example, UV radiation of eukary-

otic cells activate pathways that include signal transduction, DNA

synthesis, transcriptional regulation of both initiation and elonga-

tion rate, alternative splicing, translational repression, and protein

turnover. Some of these pathways are independent, while others

The Post-transcriptional Operon 241

are interdependent, and pr esently there are no means of studying

the response in its entirety. Thus, the key question is whether new

concepts, such as the PTO model provides a simplifying principle

that can be applied to understand some of the observations made

by experiments and to provide testable predictions.

3. The Post-

transcriptional

Operon Model

of Co-regulated

Gene Expression

The post-transcriptional operon model is based on the existence

of distinct mRNP (messenger ribonucleoprotein) particles as a

central regulatory entity of the cell. These are relatively stable in

their protein content but are dynamic and can change upon cellu-

lar perturbation. A post-transcriptional operon has been described

as a set of monocistronic mRNAs that encode functionally related

proteins (analogous to the classical polycistronic operon of bacte-

ria) that are regulated by a group of RNA-binding proteins and

small non-coding RNAs so that protein expression is coordinated

at the post-transcriptional level. Thus, an R NA operon comprises

many different mRNA species, any of which can be distributed

between multiple post-transcriptional operons. The term RNA

regulon is sometimes used as a further extension of the parallel to

regulation at the DNA level in bacteria. A DNA regulon is a set

of operons that are co-regulated by a protein. One example is the

catabolite activator protein (CAP; also known as cAMP receptor

protein, CRP) that co-regulates the transcription of several poly-

cistronic mRNAs. By analogy, an RNA regulon represents sets of

monocistronic mRNAs in eukaryotes that are co-regulated by a

protein factor or a family of RNA-binding proteins.

The post-transcriptional operon model is applied to events

that take place both during transcription (e.g. splicing and

polyadenylation) and after the formation of the mature mRNA.

These events are clearly separable and the inclusion of both may

seem confusing. However, this is justiﬁed by the facts that co-

and post-transcriptional events in many cases are coordinated and

that several RNA-binding proteins, that are involved in processing

(e.g. splicing factors such as U2AF, SR, and PTB), are also found

to be pr eferentially associated with different species of mature

mRNAs (4).

A term that is often used in conjunction with the post-

transcriptional operon model is the ribonome. This term refers to

the total cellular complement of regulated RNAs and their reg-

ulatory factors functioning dynamically in time and space within

RNP complexes. The ribonome is highly dynamic and responsive

to cellular perturbations and is distinct from the transcriptome,

which is comprised of the steady-state RNA complement of the

expressed genes from the genome.

242 Tenenbaum, Christiansen, and Nielsen

Several methods have been speciﬁcally developed to

facilitate the study of post-transcriptional gene expression includ-

ing RIP-Chip (RNA-binding protein Immunoprecipitation-

microchip proﬁling, also called ribonomics), in which an antibody

speciﬁc to the targeted RBP is used to immunoprecipitate

endogenously formed mRNP complexes followed by quantitative

analysis by microarrays (24, 25). In variations of the technique,

SAGE (serial analysis of gene expression) or high-throughput

sequencing (also known as “deep sequencing” or RNA-seq) is

used to characterize the mRNAs. The related CLIP-Chip (cross-

linking and immunoprecipitation) uses in vivo cross-linking prior

to the immunoprecipitation step in order to faithfully repre-

sent interactions taking place within cells. A related experimental

approach uses tagged mRNAs to purify mRNPs followed by char-

acterization of the mRNA as well as the protein complement of

the isolated material. Given that microRNAs function in a manner

equivalent to RNA-binding proteins in the post-transcriptional

operon hypothesis, microRNA proﬁling and target characteriza-

tion are frequently used with the assortment of methods that

are relevant in this context. Finally, ribosome proﬁling and other

methods that globally address functionality of mRNAs can be

used together with RIP-Chip and related methods to obtain a

more detailed description of the post-transcriptional operon in

question.

3.1. Basic

Observations in

Support of the

Importance of

Post-transcriptional

Regulation and the

Post-transcriptional

Operon Model

There are a number of basic observations that indirectly sup-

port the idea of the post-transcriptional operon. First, the tran-

scriptome and the proteome are poorly correlated. This has been

shown in several studies of the expression of the yeast genome,

and is probably even more pronounced in the expression of the

mammalian genome. In a study of two mouse haematopoietic

cell lines representing distinct stages of myeloid differentiation, as

well as experimentally perturbed livers, it was observed that the

differential expression of mRNA (up or down) can only account

for 40% at most of the variation of protein expression (26). These

studies are suppor ted by numerous observations of the expres-

sion of individual genes that all point to an important role of

post-transcriptional regulation. Second, pre-mRNAs are generally

processed into several isoforms by alternative splicing and other

mechanisms. Differential expression of these must necessarily be

based on post-transcriptional events. In many cases isoforms are

produced by alternative polyadenylation to encode identical pro-

teins expressed from mRNAs that differ in their UTRs resulting

in differential regulation by miRNAs and RNA-binding proteins

targeting the 3



-UTR (16, 17). Third, RNA-binding proteins are

very diverse and function in a modular and cooperative fashion

that promotes target binding and facilitates a diversity of biolog-

ical functions (27). The number of RNA interacting proteins in

The Post-transcriptional Operon 243

eukaryotes has been suggested to be approximately 2,500, almost

twice the number of DNA-binding proteins (2, 28). Fourth, ther e

are many examples of RNA-binding proteins that regulate their

own expression as well as the expression of other RNA-binding

proteins (4, 5). This is a property that would be expected if

post-transcriptional operons constitute a major regulatory layer of

gene expression. Taken together, post-transcriptional regulation

is mandatory for proper gene expression. Regulation of mRNAs

by RNA-binding proteins and small non-coding RNAs is the most

obvious way to obtain this, but other models may also be relevant.

3.2. Examples of

Post-transcriptional

Operons

Recent reviews discuss numerous examples (4, 5, 20)ofdatain

support of post-transcriptional operons based predominantly on

RIP-Chip and CLIP-Chip experiments and a list comprising more

than 30 examples of RIP-Chip analysis of speciﬁc RNA-binding

proteins can be found in (20). These examples include RNA

operons encoding mRNAs involved in mRNA splicing,

polyadenylation, export, translation, and decay. Moreover, addi-

tional examples include studies based on tagged RNA-binding

proteins. The current view is that many RNA operons operate

predominantly at the level of translation and mRNA stability

and several examples focusing on these functions will be detailed

below.

A clear cut example of a post-transcriptional operon derives

from a study of Pumilio RNA-binding proteins (PUFs) in yeast

(28). These proteins repress translation of their target mRNAs

by binding to elements located in the 3



UTR. Five different

members were each tagged and the mRNAs associated with indi-

vidual PUFs were characterized. Distinct groups of 40–220 dif-

ferent mRNAs with striking similarities in terms of functions and

subcellular localization of the proteins they encode were found

associated with each PUF member. Among them, Puf3p bound

to a subset of 154 mRNAs, 135 of which could be assigned

to mitochondrial proteins. The involvement of Puf3p in mito-

chondrial biogenesis has subsequently been the focus of a more

detailed study that found even more targets for Puf3p and cor-

roborated the physical localization of mRNAs in the vicinity of

mitochondria ( 29). Thus, it can be concluded that Puf3p deﬁnes

a post-transcriptional operon involved in synthesis and motility of

mitochondria.

Several examples of post-transcriptional operons are related to

mRNA turnover. Half-lives of mRNAs in humans vary tremen-

dously and decay of mRNA is an important regulatory step.

An intuitive example is that of the non-polyadenylated mRNAs

encoding histones. Most histone mRNAs are required for syn-

thesis of the major histones (H1, H2A, H2B, H3, and H4) dur-

ing the S-phase of the cell cycle. The mRNAs are regulated at

several steps, including decay, by the attachment of a stem-loop