The University of Chicago - Elisabet Pujadas PCBio Summer 2008 …ibd.uchicago.edu/undergrad_programs/pdf/elisabetPujadas.pdf · 2015-03-15 · Elisabet Pujadas PCBio Summer 2008

Elisabet Pujadas PCBio Summer 2008

High-throughput screening of enhancer elements and miRNA-target interactions

Differences in gene regulation are known to underlie phenotypic disparities between cell

types in multicellular organisms. In particular, transcriptional regulation has often been assumed to provide the best approximation to the regulatory state of a cell and has, indeed, yielded numerous insights into a multitude of biological systems. Gene regulatory networks have been described that focus mostly on interactions involving promoter sequences of a few core genes. Mammalian transcriptional regulation, however, relies very heavily on distal regulatory elements also known as enhancers. The lack of a TATA box, a transcriptional start site or other very unique features has rendered the identification and subsequent incorporation of enhancers into gene regulatory networks very slow, inefficient and rather serendipitous. We have developed a comprehensive approach that combines computational and experimental methods with a high-throughput robotic platform to allow for the identification and functional testing of putative enhancer elements governing cell specification. More specifically, we have chosen to study the specification of hematopoietic cell fates and the transcriptional program induced upon stimulation with Lipopolysaccharide (LPS), which is meant to represent the response of immune cells to a bacterial infection.

The regulatory regions of genes of interest are exhaustively searched using multiple-genome alignment tools, obtaining an average of 20 clustered evolutionary conserved regions (CECRs) of 500bp to 1kb in length for each gene analyzed. Each element is subcloned into a bacterial vector and electroporated into mammalian cells to be tested for function in a luciferase assay. We have developed a mutagenesis approach to try to determine the identity of the transcription factors responsible for the activity of given enhancers. We plan to complement these latter experiments with Chromatin Immunoprecipition (ChIP) and Electrophoretic Mobility Shift Assay (EMSA) to test for in vivo occupancy as well as in vitro binding, respectively.

Recently, I have started working on an experimental scheme that will take advantage of the robotic-based high-throughput approach already developed to functionally test miRNA-target interactions. I will discuss the strategies that we are currently considering and the design of the proof-of-principle experiments that will soon be carried out. By adding this new dimension to our efforts towards the construction of gene regulatory networks, we hope to gain deeper insight into the molecular mechanisms at work during cell fate specification and immune stimulation.

I have included two readings, both of them reviews. I suggest starting with “The evolution of gene regulation by transcription factors and miRNAs”. Read all figures, diagrams and boxes carefully. Also read main text until “Rates of evolution”. Familiarize yourselves with transcription factors and microRNAs and generally understand how their mechanisms of action are different. The second review, “microRNAs and the immune response” is to give you a sense of how much is known about the role of miRNAs in hematopoiesis and the immune response. The introduction and conclusion will be most helpful; there is no need to read the specifics known about each miRNA. Please, take a look at Fig. 2.

The emergence of complex, multicellular organisms was accompanied, and perhaps facilitated, by dramatic increases in the complexity of gene regulatory mecha-nisms1,2. At the level of transcriptional regulation, this can be clearly seen in the massive expansions of transcription-factor families and the pervasive com-binatorial control of genes by multiple transcription factors in higher organisms1,3 (BOX 1). At the level of post-transcriptional control, entirely new mechanisms of gene regulation arose, typified by a large and growing class of ~22-nucleotide-long non-coding RNAs, known as microRNAs (miRNAs), which function as repressors in all known animal and plant genomes4,5 (BOX 1). Although transcription factors and miRNAs are two of the best-studied gene regulatory mechanisms, there are many other layers of gene regulation, including: cell signal-ling; mRNA splicing, polyadenylation and localization; chromatin modifications; and mechanisms of protein localization, modification and degradation (FIG. 1).

But why have higher plants and animals evolved such complex, multilayered gene regulatory systems? Is combinatorial transcriptional regulation alone insuf-ficient to specify a developmental programme? What are the relative contributions of the various mechanisms of gene regulation to changes at the phenotypic level? Do all the different modes of gene regulation evolve in the same manner and at the same rate? Despite half a century of research on gene regulation, such questions have yet to be tackled seriously, because most of the effort in the field so far has been devoted to studying the evolution of transcriptional regulation. Although the

primacy of transcription as the necessary first step in gene expression is undeniable, this does not imply that transcriptional regulation has the largest effect on the final concentration of the active gene product, which is the most relevant quantity to the phenotype.

An important goal for future research is to eluci-date how complex gene regulatory networks evolve and how their evolution results in phenotypic change and speciation. However, a necessary first step towards this goal is to understand the basic principles that underlie the evolution of the individual regulators and their regulatory interactions with their target genes. Recent computational and experimental work has made it possible to begin to study the evolution of transcrip-tion factors, miRNAs and their binding sites, and to compare the rate and manner in which these two important regulatory mechanisms evolve. Here we mainly focus on animal evolution because most of the work on gene regulatory evolution has been carried out in animal systems, although we refer to plant evolution whenever possible.

Ultimately, a complete picture of the evolution of gene regulation will require a synthesis of information about all the diverse components of gene regulatory networks. As an initial step, we propose a simple model in which the evolution of transcriptional control of animal miRNA genes themselves is an important step in the successful acquisition of a novel miRNA. A corollary of this model is that there are miRNAs that are transcribed at low levels and in specific cell types, which might have little biological function in regulating target genes in trans.

*Center for Comparative Functional Genomics, Department of Biology, New York University, New York, New York 10003, USA.‡Max Delbrück Centrum für Molekulare Medizin, Robert-Rössle-Strasse 10, Berlin-Buch 13125, Germany.Correspondence to N.R. e-mail: [email protected]:10.1038/nrg1990

The evolution of gene regulation by transcription factors and microRNAsKevin Chen* and Nikolaus Rajewsky*‡

Abstract | Changes in the patterns of gene expression are widely believed to underlie many of the phenotypic differences within and between species. Although much emphasis has been placed on changes in transcriptional regulation, gene expression is regulated at many levels, all of which must ultimately be studied together to obtain a complete picture of the evolution of gene expression. Here we compare the evolution of transcriptional regulation and post-transcriptional regulation that is mediated by microRNAs, a large class of small, non-coding RNAs in plants and animals, focusing on the evolution of the individual regulators and their binding sites. As an initial step towards integrating these mechanisms into a unified framework, we propose a simple model that describes the transcriptional regulation of new microRNA genes.

NATURE REVIEWS | GENETICS VOLUME 8 | FEBRUARY 2007 | 93

REVIEWS

© 2007 Nature Publishing Group

Plants Animals

Drosha

Pri-miRNA

Helicase HelicaseMature miRNA

miRNA*

Mature miRNA

miRNA*

Pre-miRNA

Pri-miRNA

Nucleus

Cytoplasm

Pre-miRNA

Nucleus

Dicer

miRNA

miRNA*

Cytoplasm

DCL1

DCL1 [?]

miRNA

miRNA*

Transcription factors and miRNAs in developmentNo discussion of the evolution of gene regulation is com-plete without a consideration of the developmental roles of the regulators themselves, because these roles imply certain constraints on the evolvability of the regulatory relationships6,7. Furthermore, if we accept that develop-ment is the consequence of the unfolding of precise and robust spatio-temporal patterns of gene expression, then it is in the context of development that the evolution of gene regulation is most closely related to the evolution of organismal form3,7–9.

In complex multicellular organisms, transcription factors generally do not work in isolation, but instead, together with co-regulators, they form large networks of cooperating and interacting transcription factors3,8. It is widely believed that the rate of evolution of regula-tory relationships is not homogeneous over the entire network. For example, in one model that was proposed by Davidson and colleagues, animal developmental net-works can be decomposed into subnetworks, including: highly conserved ‘kernels’ that specify the spatial domain in which a particular body part will develop; ‘plug-in’

Box 1 | Transcription factors and microRNA genes — an overview

Transcription factors are proteins that either activate or repress transcription of other genes by binding to short cis-regulatory elements called transcription-factor binding sites that lie in the vicinity of the target genes. Transcription-factor binding sites, especially in developmental regulatory networks, are often organized into clusters called cis-regulatory modules (CRMs)3, which typically span a few hundred nucleotides and can contain dozens of binding sites for ~3–10 transcription factors. CRMs produce the initial spatio-temporal expression pattern of the target gene by ‘reading out’ the concentrations of multiple transcription factors that are present in a particular cell at a particular time. So, dependence on cellular context and combinatorial control are common themes in transcriptional regulation. Transcription factors are usually grouped into families on the basis of shared DNA-binding domains, which are an important determinant of transcription-factor binding specificity.

Mature microRNAs (miRNAs) are short (~22-nucleotide), non-coding ssRNAs that repress mRNAs post-transcriptionally by binding to partially complementary sites, called miRNA binding sites, in their target mRNAs. In animals, miRNA-mediated repression is often relatively weak, whereas transcription-factor-mediated repression can be much stronger. Mature miRNAs are cleaved from ~70-nucleotide hairpin structures, called precursor miRNAs (pre-mRNAs), by the enzyme dicer. Pre-miRNAs are in turn excised from a primary miRNA (pri-mRNA) transcript by the enzyme drosha. Pri-miRNAs are typically transcribed by RNA polymerase II104 and seem to possess promoter and enhancer elements that are similar to those of protein-coding genes (for example, REFS 105,106). They can be thousands of nucleotides long and contain multiple pre-miRNAs. In some cases, however, pre-miRNAs are contained in introns of protein-coding genes and are excised by the splicing machinery. In metazoans, pre-mRNAs are exported into the cytoplasm where they are processed into mature miRNAs, whereas in plants, miRNA maturation occurs within the nucleus. miRNAs are grouped into families on the basis of their target recognition motifs (BOX 3).

The predominant regulatory effect of miRNAs is to repress their target mRNAs; mechanisms for this include translational repression, mRNA cleavage, mRNA deadenylation or alteration of mRNA stability (reviewed in REF. 107,126). miRNA-mediated cleavage of mRNAs seems to be the exception in animals. By contrast, it is believed to be the dominant regulatory mode in plants. DCL1, dicer-like 1; miRNA*, the rapidly decaying strand that is complementary to the mature miRNA.

R E V I E W S

94 | FEBRUARY 2007 | VOLUME 8 www.nature.com/reviews/genetics


Transcriptional controlTranscription factors, chromatin state, combinatorial control, co-factors, alternative promoters, etc.

Post-transcriptional controlMicroRNAs, alternative splicing, alternative polyadenylation, RNA- binding proteins, etc.

mRNA

miRNAs

TF TF

components, such as signal transduction cassettes, which are re-used in multiple developmental contexts; termi-nal differentiation gene batteries that consist of all the genes that define a particular cell type; and ‘input–output switches’, such as Hox genes, which allow or disallow the action of particular developmental processes in specific spatio-temporal contexts3,7. The authors proposed that each of these types of subnetwork has its own set of evolutionary constraints, and that changes in different subnetworks result in qualitatively different types of change at the phenotypic level. For example, kernels are by definition highly conserved, and their consistency over long periods of evolutionary time might provide an explanation for the high degree of conservation of body plans within animal phyla. By contrast, differentiation gene sets, being the least pleiotropic of all the regulatory relationships, are easiest to alter, and such changes might result in species-level phenotypic differences (for details, see REFS 3,7).

Various studies have demonstrated that miRNAs have important roles in animal and plant development (see REFS 10–14 for a consideration of the developmental roles of miRNAs). Some well known examples include miRNAs with switch-like roles, such as lin-4 and let-7 in Caenorhabditis elegans developmental timing15,16 or the miRNAs that are involved in plant leaf or flower devel-opment (reviewed in REF. 13), and miRNAs that confer more general tissue or temporal identity, such as miR-1 in Drosophila melanogaster muscle development17 and miR-430 in the zebrafish maternal–zygotic transition18.

Much of the current evidence for an early develop-mental role for miRNAs is conflicting or difficult to interpret11. First, zebrafish embryos that lack maternal dicer, a protein that is required for miRNA biogenesis (BOX 2), progress through axis formation and regionaliza-tion19, a fact that strongly argues against a role for miRNAs in early zebrafish development. But dicer knockout mice20 and Arabidopsis thaliana that carry hypomorphic alleles of the dicer homolog DCL-1 die in early embryogenesis21. Second, although miRNAs have not been detected in early zebrafish and medaka embryos22,23, mature miRNAs have been detected in mice24 and D. melanogaster embryos25. Primary miRNA transcripts are spatially regulated in early D. melanogaster embryogenesis26, with the caveat that the processing of primary miRNAs into mature miRNAs (BOX 2) can be regulated24. Third, on the basis of miRNA knockdowns using 2′-O-methyl antisense oligoribonucleotides, it has been reported that miRNAs are involved in patterning the D. melanogaster embryo27, although a number of these results disagree with experimental data from genetic knockouts28.

The prevailing opinion (for example, REFS 3,29 and the references therein) seems to be that miRNAs as a class tend to function as lock-down mechanisms for already-differentiated states, or confer an additional layer of robustness or ‘noise’ reduction30,31 on the devel-opmental processes, rather than having fundamental roles in body-plan patterning. However, because the functions of only a few miRNAs have been dissected in detail, we believe that this point has not yet been proven unambiguously, at least not as a general principle.

trans-factor evolutionIndependent evolution of transcription factors and miRNAs in the plant and animal kingdoms. It is widely believed that the last common ancestor of plants and animals was unicellular, and therefore that animal and plant devel-opment evolved independently (for example, REF. 32). Both animal and plant development depends on ‘master’ transcription-factor regulators, perhaps most famously the MADS-box proteins in plants and homeodomain proteins in animals32. Both of these transcription-factor families, and most of the other transcription-factor families that are found in plants or animals, predate the divergence of the two kingdoms. However, at the sequence level, these ancient transcription-factor families are typically poorly conserved beyond the DNA-binding domain that defines the family. More importantly, their developmental functions seem to be different. For exam-ple, MADS-box proteins do not seem to have the same fundamental roles in animals as they do in plants; the opposite is true for homeodomain proteins. Furthermore, a substantial proportion of transcription-factor families are in fact kingdom-specific (for example, >45% in C. elegans and A. thaliana)33. So, despite the fact that many ancient transcription-factor families predate the divergence of plants and animals, the overall picture of plant and animal transcription-factor evolution involves the acquisition of novel transcription-factor families and the diversification of existing families.

Similarly, it is generally accepted that miRNAs have evolved independently in the animal and plant king-doms, because there are no known homologous miRNAs between plants and animals (but see REF. 127 for some

Figure 1 | Gene regulation by transcription factors and microRNAs. One or more transcription factors activate transcription by binding to cis-regulatory sites, which are often, although not always, situated upstream of protein-coding genes. After transcription, one or more microRNAs bind to cis-regulatory sites, usually in the 3′ UTR of the mRNA, and repress protein translation. In addition to these two mechanisms, many more gene regulatory mechanisms work at either the transcriptional or post-transcriptional level, including: cell signalling; mRNA splicing, polyadenylation and localization; chromatin modifications; and mechanisms of protein localization, modification and degradation (not shown). TF, transcription factor.

R E V I E W S



Bilaterians Members of the animal kingdom that have bilateral symmetry — the property of having two similar sides, with definite upper and lower surfaces, and anterior and posterior ends.

Acoel flatworms A basal bilaterian clade that diverged from the rest of bilaterians before the split between protostomes and deuterostomes.

Synteny Collinearity in the order of genes or other DNA sequences in chromosomal regions of two species or in the same species.

Clade A group of organisms that includes a common ancestor and all of its descendants, representing a distinct branch on a phylogenetic tree.

Cnidarians Radially symmetrical animals that have sac-like bodies with only one opening. They include jellyfish, corals, hydra and anemones.

evidence to the contrary) and miRNA biogenesis and the mechanism of miRNA-mediated gene repression are sig-nificantly different between plants and animals (BOX 1). Further evidence for this comes from the phylogenetic distribution of miRNAs, specifically the apparent lack of miRNAs in sponges and fungi, as discussed below. To a first approximation then, miRNA genes and most transcription-factor families have evolved independently in the animal and plant kingdoms.

Deep conservation of transcription factors and miRNAs. It is well known that many transcription factors are highly conserved over large evolutionary distances, and some have similar developmental roles in diverse spe-cies. Hox genes, which regulate development along the anterior–posterior axis in most animals, are the textbook example of this phenomenon. Other examples in verte-brates and many invertebrates include the paired box 6 (Pax6) genes, which direct eye and anterior nervous system development, and Csx/Nkx2-5/Tinman genes, which direct visceral mesoderm and heart development (reviewed in REF. 8). Although such broad similarities in function are intriguing, it is worth noting that they do not necessarily imply complete functional redundancy between distant homologues, or even evolutionarily conserved developmental roles34.

Likewise, many miRNAs seem to be extremely well conserved. The best known example is let-7, which is phylogenetically conserved in all bilaterians that have been tested so far, with the single exception of acoel flatworms35,36. Furthermore, the exact sequence of the mature form of let-7, its temporal expression pattern and in some cases its syntenic position in the genome are conserved35,36. Some other well known examples are muscle-specific miR-1, which is conserved in nematodes,

mammals and flies (REF. 17 and the references therein) and miR-7, the mature form of which is perfectly con-served between mammals and flies, and which lies in an intron of the same host gene in both clades.

Several groups have used bioinformatic sequence comparisons or northern blots to study the conserva-tion of miRNAs across many animal species37–39. They found that 18–30 miRNA families seem to be conserved in all bilaterians that have been studied, depending on the stringency of the bioinformatic methods used. Three of these miRNA families were also found in cnidarians, but none was found in sponges38,39. As previ-ously discussed, this supports the idea of independent evolution of miRNAs in plants and animals. One of the cnidarian miRNAs, miR-10, has been detected by northern blots, and is particularly interesting because it is present in a Hox cluster and regulates Hox genes in Drosophila and vertebrates38,40. Overall, the level of sequence conservation of many miRNAs is generally high, although sequence conservation need not imply functional conservation23.

Further evidence for deep conservation of animal miRNA genes at a genome-wide level can be found in cross-clade comparisons of highly conserved motifs (HCMs) in 3′ UTRs. Xie et al. showed that HCMs in vertebrates are highly enriched in miRNA-binding sites41. We and others have extended these results to nematodes and flies42,43, and have shown that HCMs are highly conserved between vertebrates, nematodes and flies (Supplementary information S1 (figure)). Because many HCMs are expected to represent miRNA recogni-tion motifs, these analyses provide indirect evidence for deep conservation of miRNA genes.

The sampling of miRNAs in plants is less extensive than in animals, with most of the plant miRNAs that are known so far having been discovered in A. thaliana (for example, REF. 44). Nevertheless, several examples of deeply conserved plant miRNAs are known from sequencing of miRNAs in moss45–47. Some of these examples are particularly impressive because their regu-latory relationships with their targets also seem to be conserved, a point we return to below.

Lineage-specific expansions of transcription factors and miRNAs. Lineage-specific expansions of tran-scription-factor families are common and are widely believed to have an important role in both plant and animal diversification and complexity1,3,9. Interestingly, there is evidence that the expansion of transcription-factor families is greater in plants than in animals48–50. In principle, this could reflect a fundamental difference between plant and animal biology. However, an alter-native explanation is that A. thaliana has undergone a recent whole-genome duplication, whereas the putative whole-genome duplications in animal lineages are more ancient: immediately after a large-scale duplication event, a transcription-factor loss might be deleterious if it changes the relative concentrations of the set of transcription factors that are expressed in a cell51, lead-ing to increased rates of transcription-factor retention in plants compared with animals.

Box 2 | MicroRNA gene discovery: bioinformatics and experimental methods

The two general approaches to microRNA (miRNA) gene discovery are bioinformatics and experimental methods (reviewed in REFS 108,109). Generally, bioinformatics methods use RNA-folding algorithms (for example, REFS 110,111) to search for approximate hairpin structures in non-coding and non-repetitive regions of the genome, and filter them using patterns of evolutionary conservation. Known examples of miRNA precursors are used as training examples for machine learning algorithms to discriminate between true predictions and false positives (reviewed in REFS 108,109). Predictions are generally verified by northern blots, PCR or microarray analysis. Naturally, bioinformatic predictions have false-positive rates and can miss species-specific miRNAs, as many, but not all112, of the current methods use evolutionary conservation as an indicator of biological function.

The traditional experimental approach to miRNA discovery is cloning and sequencing (for example, REF. 113). This approach successfully detects species-specific miRNAs, but tends to miss miRNAs that are expressed at low levels, in a small number of cells or only under particular cellular conditions. More recently, high-throughput sequencing methods, especially 454 sequencing114, have become popular for surveying small RNA populations (for example, REF. 115).

Currently, 328 miRNAs have been annotated in the human genome and 199 in Arabidopsis thaliana, the metazoan and plant species in which their small RNA complements have been well surveyed116. Several groups have shown that metazoan miRNAs probably regulate thousands of genes in mammals41,117,118, flies62,119, and nematodes120. For example, in humans, known miRNAs make up >1% of the gene repertoire and are thought to regulate >30% of all protein-coding genes.

Both computational and sequencing approaches indicate that there are likely to be many more miRNAs, many of which are lineage-specific (for example, REFS 86,92).

R E V I E W S



ChIP-chip analysis A method that combines chromatin immunoprecipitation with microarray technology to identify in vivo targets of a transcription factor.

Several detailed studies in plants52 and animals53,54 have shown that miRNA families can expand by the same processes of tandem, segmental and whole-genome dupli-cation as protein-coding genes. Plant miRNA families tend to be larger and their members more similar to each other than animal miRNA families (reviewed in REF. 55). This indicates that the expansion of plant families is more recent, and that the main effect of having multiple paralo-gous copies of the same miRNA in plants is to increase dosage. By contrast, family members of animal miRNAs might tend to have synergistic but functionally distinct roles, as shown for the let-7 family in C. elegans56,57.

cis-element evolutionLarge-scale rewiring of miRNA-mediated regulatory relationships over large evolutionary distances in animals, but deep conservation in plants. Although miRNAs them-selves seem to be highly conserved, there are only a few miRNA-target regulatory relationships that are known to be conserved over large evolutionary distances in ani-mals (for example, between vertebrates and Drosophila species). There are two cases, namely let-7:lin-41 and let-7:let-60–RAS, in which the target relationship has been experimentally dissected in C. elegans, and for which there is evidence for conservation in mammals that was obtained from computational target predictions and reciprocal expression patterns of the miRNA and mRNA in cell lines58,59. Another miRNA-target relationship, lin-4:lin-28, has been established experimentally in C. elegans, and there is computational and gene-expression evidence that it is conserved in a number of other animal species, including mammals60,61. However, even in these cases, cau-tion in interpretation is warranted, as the existence of the same regulatory relationship in highly diverged species does not always imply that it is evolutionarily conserved as opposed to independently evolved. For example, the well known Pax6 transcription factor is involved in mammalian and Drosophila eye development, and ectopic expression of the mouse homologue in Drosophila can induce the development of ectopic Drosophila eyes, but it is not clear whether even this regulatory relationship represents conservation or convergent evolution3.

These isolated examples aside, the conservation of com-putationally predicted miRNA targets between vertebrates, Drosophila and nematodes seems to be close to what is expected by chance28,42,43,62. These results seem to be robust, even when accounting for the false-positive and false-negative rates of the target prediction algorithms and possible errors in assignments of homologous genes and miRNAs42. Together with the high level of conservation that is observed for the miRNA genes themselves, this indi-cates that miRNA regulatory networks have undergone extensive rewiring during animal evolution.

In plants, a number of miRNA regulatory relation-ships are known to be conserved between A. thaliana and moss46,47, which diverged over 400 million years ago. These relationships are supported by strong experimental evidence, with verification by 5′-RACE (rapid amplifi-cation of cloned ends) of the cleavage products of the target mRNA (BOX 2). Because many of these regulatory relationships are involved in crucial biological processes

(for example, auxin signalling) this indicates that at least some miRNAs have held central positions within plant developmental regulatory networks for a long time.

Because transcription-factor binding sites are diffi-cult to predict computationally, it is harder to decipher the global dynamics of transcription-factor binding site turnover over large evolutionary distances using com-putational methods. However, the recent emergence of large-scale experimentally defined transcription-factor binding-site data (reviewed in REF. 63), particularly from ChIP-chip analysis, might make this problem trac-table in the near future. Towards this end, an interesting study that compared the targets of the RNA-binding protein Pumilio in D. melanogaster and yeast 64 found that, although the binding affinity of the regulator had remained virtually the same in these two species, its targets had diverged almost completely. Therefore, the high conservation of trans-acting factors and low overall conservation of cis-regulatory sites might be common to many regulatory mechanisms.

High turnover of binding sites even over short evolu-tionary distances. A number of groups have studied the turnover of experimentally verified transcription-factor binding sites between humans and mice65, and between various Drosophila species66–69. The general conclusion is that sequence conservation for known binding sites is sur-prisingly low (for example, ~50% for D. melanogaster and Drosophila pseudoobscura67), although this picture is com-plicated by the fact that selection is more likely to work at the level of an entire cis-regulatory module than on an individual binding site (BOX 1). In particular, cis-regulatory modules often contain redundant binding sites and, as elegantly demonstrated by Ludwig et al.70,71, there can be compensatory mutations that maintain the function of the enhancer despite the loss of individual binding sites.

Similar studies have not been carried out for miRNAs due to the paucity of experimentally verified miRNA binding sites. However, several recent microarray-based studies have indicated that the rate of binding-site con-servation is also around 50% (between humans and mice or between zebrafish, Tetraodon and Fugu)18,72,73. Furthermore, computational miRNA target predictions indicate that many lineage-specific miRNA binding sites exist in Drosophila and vertebrates (N.R., unpub-lished observations). Finally, we have used human SNP data and population-genetics techniques to show that 30–50% of non-conserved miRNA binding sites in the human genome might be functional when the mRNA and miRNA are expressed in the same tissue74. It should be noted that terms such as ‘non-conserved binding sites’ generally refer to cases in which a binding site cannot be aligned to its homologous sequence. Because these align-ments often suffer from technical problems (for example, almost all alignment algorithms are unable to deal with genomic rearrangements on various scales), ‘non-conserved’ sites might actually turn out to be conserved in a strict evolutionary sense. Nevertheless, these studies together indicate that both miRNA and transcription-factor binding sites are gained and lost quickly over short evolutionary distances.

R E V I E W S



Taking these observations to the species level, it has been shown that human promoter regions harbour a surprisingly high amount of variation that significantly affects expression levels, presumably by interfering with transcriptional control (reviewed in REF. 75). Examples of functional polymorphism have also been identified recently in miRNA binding sites in humans76 and sheep77, and sets of SNPs in predicted miRNA binding sites in the human and mouse genomes have been collected by several groups74,77. By contrast, resequencing of miRNA genes in humans showed almost no polymorphism in the sequences of mature miRNA genes78, consistent with the higher levels of constraint on trans-acting regulators compared with cis-regulatory sites.

Rates of evolutionIt has been suggested that repressors should evolve faster than activators, as there are many ways to repress a gene but relatively few ways to activate it79. As tran-scription factors can function as activators or repressors, but all known miRNAs work as repressors, one might expect miRNA binding sites to evolve faster than transcription-factor binding sites.

However, a more fundamental difference between transcription factors and miRNAs is that transcription-factor binding sites are typically ‘fuzzy’ (that is, the same transcription factor can bind to many similar DNA

sequences, possibly with different binding affinities), whereas many miRNA binding sites exhibit almost exact Watson–Crick complementarity, either to the first 6–8 bases from the 5′ end of the mature miRNA in animals, or to the entire mature miRNA in plants (BOX 3). Therefore, under neutral evolution, one would expect that it is more difficult to destroy a functional transcription-factor binding site than to create a new one, whereas the converse would be expected for miRNA binding sites.

Plant and animal miRNA binding sites might also evolve at different rates. Plant miRNA binding sites are typically found in coding regions, and if we assume that non-synonymous sites are highly constrained and syn-onymous sites are neutrally evolving, then approximately one-third of the ~22 bases in a plant miRNA binding site can accommodate a substitution without highly deleterious consequences for the organism. Therefore, the sizes of the mutational target for plant and animal miRNA binding sites are comparable, implying that the probabilities of losing a plant or animal miRNA bind-ing site are similar under simple neutral evolutionary models. On the other hand, the length of a plant miRNA binding site means that in theory it would be virtually impossible for a plant gene to gain a new miRNA bind-ing site by point mutation, whereas the same is not true for animals.

Box 3 | MicroRNA target prediction: bioinformatics and experimental methods

Most microRNA (miRNA) targets have been identified using bioinformatics methods (reviewed in REFS 88,109,121). In fact, miRNAs are one of the few classes of trans-acting regulatory factors for which computational approaches can, with reasonable confidence, successfully predict a large number of cis-regulatory binding sites. This is primarily owing to the fact that miRNAs recognize their targets at least partly on the basis of simple sequence complementarity between the miRNA and its binding site. In other words, pure knowledge of the sequence of a miRNA is sufficient to predict many targets. This is typically not yet possible for transcription factors, for which large training data sets or other experimental information are needed to accurately identify targets computationally122,123.

In plants, miRNA binding sites are usually contained in coding regions and have extensive complementarity to the mature miRNA. Therefore, many plant miRNA binding sites have been successfully discovered using relatively straightforward bioinformatics screens89,124. Because the usual mechanism of repression in plants is mRNA cleavage, predicted binding sites can be verified with confidence using 5′-RACE (rapid amplification of cloned ends) to identify the cleavage products. In addition, smaller off-target effects are also possible89.

Animal miRNA binding sites usually lie in 3′ UTRs of target mRNAs and exhibit imperfect complementarity to the mature miRNA (reviewed in REFS 88,109,121). Many target prediction methods are based on a model in which miRNA–mRNA binding is nucleated by an exact Watson–Crick complementary match to the first 6–8 bases from the 5′ end of the mature miRNA. Estimates of false-positive rates on the basis of comparative genomics (see REF. 88 and references therein), population genetics74 and experimental assays95 all indicate that the accuracy of these algorithms is high. For example, the estimated accuracy for targets that are conserved in humans, chimpanzees, mice, rats and dogs is 50–85% (REFS 74,88). However, other classes of miRNA sites have also been predicted, such as imperfect miRNA sites without an exact Watson–Crick match to the first 6–8 bases of the miRNA (reviewed in REFS 88,109,121).

Various assays have been developed to verify predicted animal miRNA targets, for example, by expressing the miRNA in a localized domain in vivo while simultaneously expressing and monitoring the target mRNA in a broader domain (for example, REF. 119). However, the number of target sites that have been validated in vivo under endogenous conditions and by mutagenesis of the predicted sites is extremely small. An entirely different approach from computational methods is to assay the expression levels of mRNAs directly, using the observation that miRNAs can not only downregulate protein levels, but also downregulate mRNA levels of their targets (for example, REF. 125). Therefore, a direct approach to target discovery is to knockout or overexpress the miRNA and use microarrays to identify the genes that show expression changes (for example, REFS 18,72,73), and to correlate these changes with 3′ UTR sequence motifs, for example, by a linear regression model73. Because of the additional information, evolutionary conservation filters on the target-site predictions can be relaxed. Therefore, this method is expected to have a higher sensitivity than approaches that are purely sequence-based, although it suffers from the possibility of indirect effects.

Several databases and web resources have been developed to store and maintain predicted target sites and sites with experimental support (reviewed in REF. 88,121).

R E V I E W S



Quantitative models of transcription-factor and miRNA binding-site evolution. One promising way of making the qualitative arguments of the previous section more quantitative involves developing models of binding-site evolution on the basis of point substitutions. Note that this approach does not accommodate the acquisition of new binding sites by large-scale rearrangements such as transpositions (for example, REF. 80), or small-scale rearrangements such as tandem duplications81.

Stone and Wray 82 calculated the expected time for the appearance of a new transcription-factor bind-ing site by neutral point substitution and concluded that this time was short (for example, 55,000 years for two 6-bp sites to evolve within a 200-bp region in D. melanogaster 82), assuming that all members of the population evolve independently. The problem was revisited under the more realistic assumption that the members of the population are related by descent83–85. Although the methods and models differed, the three studies arrived at similar conclusions. First, neutral mutation is too slow to efficiently evolve new binding sites by point substitutions. Second, positive selection on partial binding sites can effectively speed up the rate of evolution, making it feasible to evolve new binding sites. Third, the time taken to create new sites scales linearly with the length of the regulatory region (for example, the promoter or enhancer regions for a transcription-factor binding site, or the 3′ UTR for a miRNA binding site) but exponentially with the length of the binding site. Fourth, the base composition of the region is important, par-ticularly the presence of ‘pre-sites’ that are a single point mutation away from being a functional binding site.

As a specific example, consider a 1-kb region of non-coding DNA with equal base composition in the human genome. Assume that the binding sites for a miRNA and a transcription factor are each 8 bp long, but a miRNA requires exactly eight matches, whereas a transcription factor requires any seven matches out of eight. Durrett and Schmidt calculated that, given neutral point substi-tutions, it would take ~650 million years for the miRNA binding site to appear in the absence of a pre-site, and ~375,000 years in the presence of a pre-site. By contrast, the transcription-factor binding site would take ~60,000 years to appear85.

The rate of acquisition of miRNA genes versus transcrip-tion factors. As more genome sequences are completed, it emerges that few novel transcription-factor families have arisen since the divergence between animals and plants3,79 (although the number of transcription factors in each family can be different in individual genomes). The situation seems to be different for miRNAs. As a result of a combination of bioinformatics and sequencing efforts, it is now apparent that the process of miRNA creation is both active and ongoing (for example, REFS 37,38,86,87) (BOX 2). For example, it seems that the human genome alone might contain more than 1,000 miRNAs86, of which many have been proposed to be primate-specific or even human-specific87. It should be noted that this global picture of transcription-factor-gene versus miRNA-gene acquisition ignores important issues such

as combinatorial transcriptional control, co-factors and mutations outside of the DNA-binding domain that can affect transcription-factor binding specificity, as well as difficulties in computational identification of homolo-gous miRNAs in different genomes. Nevertheless, even taking these points into account, the difference in the rates of creation of transcription-factor families and miRNA families remains striking, and it seems reason-able to propose that the speed of creation of new miRNA families has been faster in animal evolution than that of new transcription-factor families.

Do animals need so many miRNAs, and if so, why? Also, given that an animal miRNA can apparently target easily hundreds of genes (BOX 3) (reviewed in REF. 88), how can new miRNAs be acquired with such apparent ease without seriously disrupting the existing regulatory network of the organism? Even in plants in which the specificity of miRNA–mRNA binding is higher than in animals, significant off-target effects can occur and so similar issues can arise89. To address these questions, we next consider models of how new transcription factors and miRNAs evolve.

Creating new trans-factorsTranscription factors typically contain multiple functional domains, which mediate binding to DNA, interactions with other proteins and the subcellular localization of the transcription factor. A transcription factor with a new binding specificity can be created by duplication of an existing transcription factor followed by mutations, often, although not always, in the DNA-binding domain. Transcription factors can also evolve by the acquisition or loss of one of its other functional domains. For example, the loss of a transcriptional activation domain could turn an activator into a repressor, whereas the acquisition of a new protein–protein interaction domain that facilitates heterodimerization with a novel binding partner could significantly alter the targets of the transcription factor. It is often assumed that mutations in the coding sequence of the transcription factor itself are likely to be highly deleterious because they potentially affect the expres-sion of many downstream target genes. However, recent examples of transcription factors that are important in Drosophila development that have significantly diverged in sequence and function indicate that this assumption might be worth revisiting (reviewed in REF. 34).

A duplication-mutation model accounts also for the evolution of at least some miRNA genes. For example, human miR-10a and miR-100 are homologues but differ by a single nucleotide insertion-deletion in the predicted target recognition region of the respective mature miRNAs. If this were the predominant mode of miRNA-family creation, one might expect that miRNA target recognition motifs would not be randomly dis-tributed in sequence space, but would show a propensity to cluster in groups of similar sequences. Although this problem has not been fully studied, preliminary bioin-formatic analyses indicate that the distribution of these sequences seems to be close to random (K.C. and N.R., unpublished observations). If true, this would indicate that a duplication-mutation process accounts for a

R E V I E W S



Deleterious targets

DNADNA

a

Neutral or advantageous targets

miRNAs

Deleterious targets

bNeutral or advantageous targets

miRNAs

Deep sequencing Sequencing to high coverage, where coverage (or depth) corresponds to the average number of times that a nucleotide is sequenced.

relatively small fraction of miRNAs, and that instead most new miRNA families arise de novo.

One model of de novo miRNA acquisition, the inverted duplication model, suggests that new miRNAs evolve by inverted duplication of a stretch of coding sequence followed by subsequent erosion of the sequence into an imperfect hairpin structure90. In some cases, part of the coding sequence itself might be duplicated before the creation of the miRNA, so the miRNA shows homol-ogy to two regions of the target gene128. This model is attractive for plant miRNAs because it accounts for the long stretches of sequence similarity that are required between the miRNA and its target.

The inverted duplication model seems less appropriate for animal miRNAs, because the length of complemen-tary sequence in miRNA binding sites is much smaller in animals than in plants, and no examples of newly formed miRNAs arising in this manner are known in animals. A second model, the random creation model, proposes that new miRNAs simply arise randomly from existing hairpin structures in the genome54,91. Hairpin structures are gen-erally abundant in eukaryotic genomes — for instance, Bentwich et al.92 identified 11 million hairpins in the human genome in a bioinformatic screen. Therefore, the problem of creating a new miRNA in animals might be less involved with creating a new hairpin structure, but rather with appropriately transcribing an existing hairpin struc-ture in the genome and providing the prerequisite signals for biogenesis of the new miRNA (for example, signals for processing by the RNase III Drosha) (BOX 1).

A model of transcriptional control of new miRNAs. Because the minimal binding site of an animal miRNA is short, a new miRNA should be able to target many mRNAs simply by chance, and many of these interactions are likely to be selectively deleterious, as is the case for all types of mutations. Indeed, the existence of many genes for which the presence of a miRNA binding site would be deleterious (‘anti-targets’) was proposed93 and later

demonstrated by several groups in various species using microarray data and computational studies72,73,89,94,95.

These observations raise the question of how a new miRNA could ever be acquired without seriously impair-ing the fitness of the organism. We propose that one way this could happen is if the miRNA were initially tran-scribed only weakly, and in a specific tissue or at a specific developmental stage. Current studies indicate that mul-tiple sites for the same miRNA in the same target mRNA are needed to generate a strong regulatory effect of the miRNA on this target (REF. 88 and references therein). If this is true, natural selection can eliminate slightly delete-rious miRNA sites over time — recall that it is easier for an mRNA to lose a miRNA site than to gain one — while also maintaining or creating beneficial binding sites for the new miRNA. Once this elimination process is com-plete, the expression level of the miRNA can be increased and its tissue-specificity can be relaxed (FIG. 2).

Although this model is clearly a simplified picture of a complex process, it makes the testable prediction that more-recently evolved miRNAs should be expressed weakly and in specific spatio-temporal domains. Indeed, this prediction is generally supported by the available miRNA expression data (for example, REFS 87,96), which indicate that more-recently acquired human miRNAs are more likely to be weakly expressed than ancient con-served miRNAs. A recent study by Berezikov et al.87, in which miRNA expression in the human and chimpanzee brain was determined using deep sequencing (BOX 2), is also consistent with this prediction.

The model puts the enormous number of putative small RNA transcripts that are being revealed by deep sequencing efforts in perspective: it suggests that a number of them are randomly transcribed hairpins that might not have a significant biological role as trans-acting regulators, although it does not preclude the possibility that they could acquire such functionality in the future. The model is consistent with the hypothesis proposed by Sempere et al.38 and Prochnik et al.39 that the acquisi-tion of new miRNAs contributed to the acquisition of novel tissue types and organs in animal development. It is also consistent with an intriguing idea put forward by Davidson7 that it is relatively easy to evolve a new miRNA gene that targets a specific sequence motif, but the same is not true of transcription factors, because the sequence specificity of a protein is a complex function of its amino-acid sequence. So, if a transcription factor were to acquire a new domain of expression (for example, in a new tissue), it would be expected to regulate genes that it already regulates in its original domain of expression, probably leading to deleterious effects. However, assum-ing that a miRNA with an arbitrary target sequence can evolve easily, such complications can be avoided, and consequently it might be easier to insert a miRNA into a developmental network than a transcription factor.

The expression and conservation of miRNAs are not as well understood in plants compared with animals. The correlation between the age of a miRNA and its expression level is expected to be weaker in plants than in animals for two reasons. First, a newly arising plant miRNA is expected to have relatively few targets compared with

Figure 2 | A model of the acquisition of a new microRNA. a | Initially, a new microRNA (miRNA) is expressed at low levels and in a specific spatio-temporal domain. It has many targets that appear at random in the genome, many of which are selectively slightly deleterious and a few of which are selectively neutral or advantageous. b | At a later point in time, natural selection has purged many deleterious targets from the genome while preserving many of the neutral or advantageous targets, and the expression of the miRNA can increase without being highly deleterious to the organism.

R E V I E W S



an animal miRNA, and if these are either highly benefi-cial or highly deleterious, then selection can drive up or down the expression level of the miRNA more quickly. Second, a single plant miRNA molecule can target and cleave many mRNAs, whereas it seems that animal miRNAs must be bound to their targets to confer repression, a process that is reversible under specific conditions97,98. Nonetheless, it is likely that some plant miRNAs that are weakly expressed or tissue specific have little or no biological function as trans-acting regulators128.

ConclusionWith our rapidly advancing knowledge of the different mechanisms of gene regulation in higher eukaryotes, we can begin to consider the evolutionary implications of these different mechanisms within a unified framework. In the past, much work focused on a synthesis between transcriptional regulation and cell signalling mecha-nisms3,99; here we have concentrated on the evolution of transcription factors and miRNAs. Ultimately, all other mechanisms of gene regulation should be brought into the discussion in order to form a holistic picture of the evolution of gene regulation.

Many open questions and directions for future research remain. This Review gives a local view of the evolution of individual regulators and binding sites as a necessary first step to understanding the evolution of gene regulation as a whole. In the future, it will be necessary to move towards the broader view of the evolution of developmental regu-latory networks, and from there, towards the even bigger picture of changes in organismal form. The model of Davidson and colleagues that is discussed above is one promising way of thinking about the evolution of network structure and body plans on a global scale. Our current knowledge of how transcription factors, miRNAs, signal-ling pathways and other regulators are wired together into developmental networks is much too rudimentary to make any sensible statements about the effect of different regulatory mechanisms on global network evolution.

However, as our knowledge of the developmental roles of these regulatory mechanisms increases, it should be possible to extend the model to account for these differ-ent components. For example, if it indeed turns out that miRNAs tend to work at the periphery of developmental networks to confer additional layers of robustness, and not as the primary agents of developmental patterning, then this might lead us to postulate a certain amount of evolutionary pliability for miRNA-mediated regulation. It will also be interesting to investigate whether different eukaryotic lineages, particularly plants and animals, use different regulatory mechanisms in similar ways or not.

Such global trends in network evolution have a natu-ral counterpart in local subnetwork motifs that exist within the overall network. One particularly interesting example of such a motif is a feedback loop that involves multiple transcription factors and miRNAs. Examples of this motif have been identified in recent work on neuron cell-fate determination100 and vulval develop-ment101 in C. elegans, and granulocytic differentiation in humans102. Whether multicomponent feedback loops and other complex subnetwork motifs are common fea-tures of developmental networks, and whether they have any significance for organismal traits and evolution, are intriguing questions for future research. Within the Davidson model, signalling cassettes function as plug-in components that are re-used repeatedly in regulatory networks. Likewise, certain subnetwork motifs could also potentially form re-usable plug-in components.

Since the work of Mary-Claire King and Allan Wilson three decades ago103, scientists have asked whether changes in gene regulation or protein sequence have made bigger contributions to phenotypic differences between species. Today, we are well positioned to broaden the question to ask about the relative contributions of the evolution of different mechanisms of gene regulation to the evolution of phenotypic diversity in animals and plants. The long journey towards a comprehensive understanding of the evolution of gene regulation is only beginning.

1. Levine, M. & Tjian, R. Transcription regulation and animal diversity. Nature 424, 147–151 (2003).

2. Moore, M. J. From birth to death: the complex lives of eukaryotic mRNAs. Science 309, 1514–1518 (2005).

3. Davidson, E. H. The Regulatory Genome: Gene Regulatory Networks in Development and Evolution (Academic, New York, 2006).This book synthesizes several decades of work on gene regulation, animal development and evolution, with an emphasis on transcriptional regulation.

4. Ambros, V. The functions of animal microRNAs. Nature 431, 350–355 (2004).

5. Bartel, D. P. MicroRNAs: genomics, biogenesis, mechanism and function. Cell 116, 281–297 (2004).

6. Wittkopp, P. J. Genomic sources of regulatory variation in cis and in trans. Cell. Mol. Life Sci. 62, 1779–1783 (2005).

7. Davidson, E. H. & Erwin, D. H. Gene regulatory networks and the evolution of animal body plans. Science 311, 796–800 (2006).

8. Carroll, S., Grenier, J. & Weatherbee, S. From DNA to Diversity: Molecular Genetics and the Evolution of Animal Design (Blackwell Scientific, Malden, 2005).

9. Doebley, J. & Lukens, L. Transcriptional regulators and the evolution of plant form. Plant Cell 10, 1075–1082 (1998).

10. Kloosterman, W. P. & Plasterk, R. H. The diverse functions of microRNAs in animal development and disease. Dev. Cell 11, 441–450 (2006).

11. Kosik, K. S. The neuronal microRNA system. Nature Rev. Neurosci. 7, 911–920 (2006).

12. Carrington, J. C. & Ambros, V. Role of microRNAs in plant and animal development. Science 301, 336–338 (2003).

13. Jones-Rhoades, M. W., Bartel, D. P. & Bartel, B. MicroRNAs and their regulatory roles in plants. Annu. Rev. Plant Biol. 57, 19–53 (2006).

14. Alvarez-Garcia, I. & Miska, E. A. MicroRNA functions in animal development and human disease. Development 132, 4653–4662 (2005).

15. Lee, R. C., Feinbaum, R. L. & Ambros, V. The C. elegans heterochronic gene lin-4 encodes small RNAs with antisense complementarity to lin-14. Cell 75, 843–854 (1993).

16. Reinhart, B. J. et al. The 21-nucleotide let-7 RNA regulates developmental timing in Caenorhabditis elegans. Nature 403, 901–906 (2000).

17. Sokol, N. S. & Ambros, V. Mesodermally expressed Drosophila microRNA-1 is regulated by Twist and is required in muscles during larval growth. Genes Dev. 19, 2343–2354 (2005).

18. Giraldez, A. J. et al. Zebrafish MiR-430 promotes deadenylation and clearance of maternal mRNAs. Science 312, 75–79 (2006).

19. Giraldez, A. J., Cinalli, R. M., Glasner, M. E., Thomson, J. M. & Baskerville, S. MicroRNAs regulate brain morphogenesis in zebrafish. Science 308, 833–838 (2005).

20. Bernstein, E. et al. Dicer is essential for mouse development. Nature Genet. 35, 215–217 (2003).

21. Schauer, S. E., Jacobsen, S. E., Meinke, D. W. & Ray, A. DICER-LIKE1: blind men and elephants in Arabidopsis development. Trends Plant Sci. 7, 487–491 (2002).

22. Wienholds, E. et al. MicroRNA expression in zebrafish embryonic development. Science 309, 310–311 (2005).

23. Ason, B. et al. Differences in vertebrate microRNA expression. Proc. Natl Acad. Sci. USA 103, 14385–14389 (2006).

24. Thomson, J. M. et al. Extensive post-transcriptional regulation of microRNAs and its implications for cancer. Genes Dev. 20, 2202–2207 (2006).

25. Aravin, A. A. et al. The small RNA profile during Drosophila melanogaster development. Dev. Cell 5, 337–350 (2003).

26. Aboobaker, A. A., Tomancak, P., Patel, N., Rubin, G. M. & Lai, E. C. Drosophila microRNAs exhibit diverse spatial expression patterns during embryonic development. Proc. Natl Acad. Sci. USA 102, 18017–18022 (2005).

27. Leaman, D. et al. Antisense-mediated depletion reveals essential and specific functions of microRNAs in Drosophila development. Cell 121, 1097–1108 (2005).

28. Brennecke, J., Stark, A. & Cohen, S. M. Not miR-ly muscular: microRNAs and muscle development. Genes Dev. 19, 2261–2264 (2005).

R E V I E W S



29. Hornstein, E. & Shomron, S. Canalization of development by microRNAs. Nature Genet. 38, S20–S24 (2006).

30. Li, Y., Wang, Y., Lee, J. A. & Gao, F. B. MicroRNA-9a ensures the precise specification of sensory organ precursors in Drosophila. Genes Dev. 20, 2793–2805 (2006).

31. Cohen S. M., Brennecke, J. & Stark A. Denoising feedback loops by thresholding — a new role for microRNAs. Genes Dev. 20, 2769–2772 (2006).

32. Meyerowitz, E. M. Plants compared to animals: the broadest comparative study of development. Science 295, 1482–1485 (2002).

33. Riechmann, J. L. et al. Arabidopsis transcription factors: genome-wide comparative analysis among eukaryotes. Science 290, 2105–2110 (2000).A detailed study of the evolution of transcription- factor families in plants and animals, which showed that they have diversified greatly during eukaryotic evolution.

34. Hsia, C. C. & McGinnis, W. Evolution of transcription factor function. Curr. Opin. Genet. Dev. 13, 199–206 (2003).

35. Pasquinelli, A. E. et al. Conservation of the sequence and temporal expression of let-7 heterochronic regulatory RNA. Nature 408, 86–89 (2000).This paper demonstrated that let-7, one of the two founding miRNAs, is highly conserved in animals.

36. Pasquinelli, A. E. et al. Expression of the 22 nucleotide let-7 heterochronic RNA throughout the metazoa: a role in life history evolution? Evol. Dev. 5, 372–378 (2003).

37. Hertel, J. et al. The expansion of the metazoan miRNA repertoire. BMC Genomics 7, 25 (2006).

38. Sempere, L. F., Cole, C. N., McPeek, M. A. & Peterson, K. J. The phylogenetic distribution of metazoan microRNAs: insights into evolutionary complexity and constraint. J. Exp. Zool. B. Mol. Dev. Evol. 306, 575–588 (2006).The authors studied the evolution of miRNAs in bilaterians, cnidarians and sponges, and proposed that the acquisition of new miRNAs has had an important role in the development of new animal organs.

39. Prochnik, S. E., Rokhsar, D. S. & Aboobaker, A. A. Evidence for a microRNA expansion in the bilaterian ancestor. Dev. Genes Evol. 14 Nov 2006 (doi: 10.1007/s00427-006-0116-1).

40. Lemons, D. & McGinnis, W. Genomic evolution of Hox gene clusters. Science 313, 1918–1922 (2006).

41. Xie, X. et al. Systematic discovery of regulatory motifs in human promoters and 3′ UTRs by comparison of several mammals. Nature 434, 338–345 (2005).

42. Chen, K. & Rajewsky, N. Deep conservation of miRNA-target relationships and 3′ UTR motifs in vertebrates, flies and nematodes. Cold Spring Harb. Symp. Quant. Biol. 12 Dec 2006 (doi:10.1101/sqb.2006.71.039).

43. Chan, C. S., Elemento, O. & Tavazoie, S. Revealing posttranscriptional regulatory elements through network-level conservation. PLoS Comput. Biol. 1, e69 (2006).

44. Lu, C. et al. MicroRNAs and other small RNAs enriched in the Arabidopsis RNA-dependent RNA polymerase-2 mutant. Genome Res. 16, 1276–1288 (2006).

45. Arazi, T. et al. Cloning and characterization of micro-RNAs from moss. Plant J. 43, 837–848 (2005).

46. Floyd, S. K. & Bowman, J. L. Gene regulation: ancient microRNA target sequences in plants. Nature 428, 485–486 (2004).

47. Axtell, M. J. & Bartel, D. P. Antiquity of microRNAs and their targets in land plants. Plant Cell 17, 1658–1673 (2005).These authors showed that many miRNAs and miRNA-target relationships are well conserved in plants.

48. Shiu, S. H., Shih, M. C. & Li, W. H. Transcription factor families have much higher expansion rates in plants than in animals. Plant Physiol. 139, 18–26 (2005).

49. Blanc, G. & Wolfe, K. H. Functional divergence of duplicated genes formed by polyploidy during Arabidopsis evolution. Plant Cell 16, 1679–1691 (2004).

50. Seoighe, C. & Gehring, C. Genome duplication led to highly selective expansion of the Arabidopsis thaliana promoterome. Trends Genet. 20, 461–464 (2004).

51. Davis, J. C. & Petrov, D. A. Do disparate mechanisms of duplication add similar genes to the genome? Trends Genet. 21, 548–551 (2005).

52. Maher, C., Stein, L. & Ware, D. Evolution of Arabidopsis microRNA familes through duplication events. Genome Res. 16, 510–519 (2006).

53. Tanzer, A., Amemiya, C. T., Kim, C. B. & Stadler, P. F. Evolution of microRNAs located within Hox gene clusters. J. Exp. Zool. B. Mol. Dev. Evol. 304, 75–85 (2005).

54. Tanzer, A. & Stadler, P. F. Molecular evolution of a microRNA cluster. J. Mol. Biol. 339, 327–335 (2004).

55. Li, A. & Mao, L. Evolution of plant microRNA gene families. Cell Res. 28 Nov 2006 (doi:10.1038/sj.cr.7310113).

56. Abbott, A. L. et al. The let-7 microRNA family members mir-48, mir-84, and mir-241 function together to regulate developmental timing in Caenorhabditis elegans. Dev. Cell 9, 403–414 (2005).

57. Hayes, G. D., Frand, A. R. & Ruvkun, G. The mir-84 and let-7 paralogous microRNA genes of Caenorhabditis elegans direct the cessation of molting via the conserved nuclear hormone receptors NHR-23 and NHR-25. Development 133, 4631–4641 (2006).

58. Johnson, R. M. et al. RAS is regulated by the let-7 microRNA family. Cell 120, 635–647 (2005).

59. Schulman, B. R. M., Esquela-Kerscher, A. & Slack, F. J. Reciprocal expression of Lin-41 and the microRNAs Let-7 and Mir-125 during mouse embryogenesis. Dev. Dyn. 234, 1046–1054 (2005).

60. Moss, E. G. & Tang, L. Conservation of the heterochronic regulator Lin-28, its developmental expression and microRNA complementary sites. Dev. Biol. 258, 432–442 (2003).

61. Yang, D.-H. & Moss, E. G. Temporally regulated expression of Lin-28 in diverse tissues of the developing mouse. Gene Expr. Patterns 3, 719–726 (2003).

62. Grun, D., Wang, Y., Langenberger, D., Gunsalus, K. C. & Rajewsky, N. microRNA target predictions across seven Drosophila species and comparison to mammalian targets. PLoS Comput. Biol. 1, e13 (2005).

63. Elnitski, L., Jin, V. X., Farnham, P. J. & Jones, S. J. M. Locating mammalian transcription factor binding sites: A survey of computational and experimental techniques Genome Res. 16, 1455–1464 (2006).

64. Gerber, A. P., Luschnig, S., Krasnow, M. A., Brown, P. O. & Herschlag, D. Genome-wide identification of mRNAs associated with the translational regulator PUMILIO in Drosophila melanogaster. Proc. Natl Acad. Sci. USA 103, 4487–4492 (2006).

65. Dermitzakis, E. T. & Clark, A. G. Evolution of transcription factor binding sites in mammalian gene regulatory regions: conservation and turnover. Mol. Biol. Evol. 19, 1114–1121 (2002).

66. Dermitzakis, E. T., Bergman, C. M. & Clark, A. G. Tracing the evolutionary history of Drosophila regulatory regions with models that identify transcription factor binding sites. Mol. Biol. Evol. 20, 703–714 (2003).

67. Emberly, E., Rajewsky, N. & Siggia, E. D. Conservation of regulatory elements between two regions of Drosophila. BMC Bioinformatics 4, 57 (2003).

68. Richards, S. et al. Comparative genome sequencing of Drosophila pseudoobscura: chromosomal, gene and cis-element conservation. Genome Res. 15, 1–18 (2005).

69. Moses, A. M. et al. Large-scale turnover of functional transcription factor binding sites in Drosophila. PLoS Comput. Biol. 2, e130 (2006).

70. Ludwig, M. Z., Bergman, C. M., Patel, N. H. & Kreitman, M. Evidence for stabilizing selection in a eukaryotic enhancer element. Nature 403, 564–567 (2000).An elegant demonstration of the complex evolution of the well studied D. melanogaster even-skipped stripe 2 enhancer.

71. Ludwig, M. Z. et al. Functional evolution of a cis-regulatory module. PLoS Biol. 3, e93 (2005).

72. Krutzfeldt, J. et al. Silencing of microRNAs in vivo with ‘antagomirs’. Nature 438, 685–689 (2005).

73. Sood, P., Krek, A., Zavolan, M., Macino, G. & Rajewsky, N. Cell-type-specific signatures of microRNAs on target mRNA expression. Proc. Natl Acad. Sci. USA 103, 2746–2751 (2006).

74. Chen, K. & Rajewsky, N. Natural selection on human miRNA binding sites inferred from SNP data. Nature Genet. 38, 1452–1456 (2006).This paper examined polymorphism in human miRNA binding sites and used population genetic techniques to estimate levels of selective constraint on these binding sites.

75. Rockman, M. V. & Wray, G. A. Abundant raw material for cis-reguatory evolution in humans. Mol. Biol. Evol. 19, 1991–2004 (2002).A systematic survey of the literature on naturally occurring polymorphisms in human promoter regions that significantly affect gene expression levels.

76. Abelson, J. F. et al. Sequence variants in SLITRK1 are associated with Tourette’s syndrome. Science 310, 317–320 (2005).

77. Clop, A. et al. A mutation creating a potential illegitimate microRNA target site in the myostatin gene affects muscularity in sheep. Nature Genet. 38, 813–818 (2006).The clearest example to date of a naturally occurring SNP in a miRNA binding site that affects a gross phenotypic trait.

78. Iwai, N. & Naraba, H. Polymorphisms in human pre-miRNAs. Biochem. Biophys. Res. Commun. 331, 1439–1444 (2005).

79. Wray, G. A. et al. The evolution of transcriptional regulation in eukaryotes. Mol. Biol. Evol. 20, 1377–1419 (2003).A comprehensive review of the evolution of transcription factors and their binding sites.

80. Balhoff, J. P. & Wray, G. A. Evolutionary analysis of the well characterized endo16 promoter reveals substantial variation within functional sites. Proc. Natl Acad. Sci. USA 102, 8591–8596 (2005).

81. Sinha, S. & Siggia, E. D. Sequence turnover and tandem repeats in cis-regulatory modules in Drosophila. Mol. Biol. Evol. 22, 874–885 (2005).

82. Stone, J. & Wray, G. Rapid evolution of cis-regulatory sequences via local point mutations. Mol. Biol. Evol. 18, 1764–1770 (2001).

83. Berg, J., Willmann, S. & Laessig, M. Adaptive evolution of transcription factor binding sites. BMC Evol. Biol. 4, 42 (2004).

84. MacArthur, S. & Brookfield, J. F. Y. Expected rates and modes of evolution of enhancer sequences. Mol. Biol. Evol. 21, 1064–1073 (2004).

85. Durrett, R. & Schmidt, D. Waiting for regulatory sequences to appear. Ann. Appl. Probab. (in the press).

86. Berezikov, E. et al. Many novel mammalian microRNA candidates identified by extensive cloning and RAKE analysis. Genome Res. 16, 1289–1298 (2006).

87. Berezikov, E. et al. Diversity of microRNAs in human and chimpanzee brain. Nature Genet. 38, 1375–1377 (2006).These authors used 454 sequencing to show that many miRNAs that are expressed in the human brain are not conserved in the chimpanzee brain.

88. Rajewsky, N. microRNA target predictions in animals. Nature Genet. 38, S8–S13 (2006).

89. Schwab, R. et al. Specific effects on microRNAs on the plant transcriptome. Dev. Cell 8, 517–527 (2005).

90. Allen, E. et al. Evolution of microRNA genes by inverted duplication of target gene sequences in Arabidopsis thaliana. Nature Genet. 36, 1282–1290 (2004).

91. Svoboda, P. & Di Cara, A. Hairpin RNA: a secondary structure of primary importance. Cell. Mol. Life Sci. 63, 901–918 (2006).

92. Bentwich, I. et al. Identification of hundreds of conserved and nonconserved human microRNAs. Nature Genet. 37, 766–770 (2005).

93. Bartel, D. P. & Chen, C. Z. Micromanagers of gene expression: the potentially widespread influence of metazoan microRNAs. Nature Rev. Genet. 5, 396–400 (2004).

94. Farh, K. K. et al. The widespread impact of mammalian microRNAs on mRNA repression and evolution. Science 310, 1817–1821 (2005).

95. Stark, A., Brennecke, J., Bushati, N., Russell, R. B. & Cohen, S. M. Animal microRNAs confer robustness to gene expression and have a significant impact on 3′ UTR evolution. Cell 123, 1133–1146 (2005).

96. Baskerville, S. & Bartel, D. P. Microarray profiling of microRNAs reveals frequent coexpression with neighboring miRNAs and host genes. RNA 11, 241–247 (2005).

97. Bhattacharyya, S. N., Habermacher, R., Martine, U., Closs, E. I. & Filipowicz, W. Relief of microRNA-mediated translational repression in human cells subjected to stress. Cell 125, 1111–1124 (2006).

98. Ashraf, S. I., McLoon, A. L., Sclarsic, S. M. & Kunes, S. Synaptic protein synthesis associated with memory is regulated by the RISC pathway in Drosophila. Cell 124, 191–205 (2006).

R E V I E W S



99. Davidson, E. H. Genomic Regulatory Systems. Development and Evolution (Academic, San Diego, 2001).

100. Johnston, R. J. J., Chang, S., Etchberger, J. F., Ortiz, C. O. & Hobert, O. MicroRNAs acting in a double-negative feedback loop to control a neuronal cell fate decision. Proc. Natl Acad. Sci. 102, 12449–12454 (2005).

101. Yoo, A. S. & Greenwald, I. LIN-12/Notch activation leads to microRNA-mediated down-regulation of Vav in C. elegans. Science 310, 1330–1333 (2005).

102. Fazi, F. et al. A minicircuitry comprised of microRNA-223 and transcription factors NFI-A and C/EBPα regulates human granulopoiesis. Cell 123, 819–831 (2005).

103. King, M. C. & Wilson, A. C. Evolution at two levels in humans and chimpanzees. Science 188, 107–116 (1975).

104. Lee, Y. et al. MicroRNA genes are transcribed by RNA polymerase II. EMBO J. 23, 4051–4060 (2004).

105. Johnson, S. M., Lin, S. Y. & Slack, F. J. The time of appearance of the C. elegans let-7 microRNA is transcriptionally controlled utilizing a temporal regulatory element in its promoter. Dev. Biol. 259, 364–379 (2003).

106. Biemar, F. et al. Spatial regulation of microRNA gene expression in the Drosophila embryo. Proc. Natl Acad. Sci. 102, 15907–15911 (2005).

107. Engels, B. M. & Hutvagner, G. Principles and effects of microRNA-mediated post-transcriptional gene regulation. Oncogene 25, 6163–6169 (2006).

108. Berezikov, E., Cuppen, E. & Plasterk, R. H. Approaches to microRNA discovery. Nature Genet. 38, S2–S7 (2006).

109. Bentwich, I. Prediction and validation of microRNAs and their targets. FEBS Lett. 579, 5904–5910 (2005).

110. Zuker, M. Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res. 31, 3406–3415 (2003).

111. Hofacker, I. L. Vienna RNA secondary structure server. Nucleic Acids Res. 31, 3429–3431 (2003).

112. Pfeffer, S. et al. Identification of virus-encoded microRNAs. Science 304, 734–736 (2004).

113. Lagos-Quintana, M., Rauhut, R., Lendeckel, W. & Tuschl, T. Identification of novel genes coding for small expressed RNAs. Science 294, 853–858 (2001).

114. Margulies, M. et al. Genome sequencing in microfabricated high-density picolitre reactors. Nature 437, 376–380 (2005).

115. Lu, C. et al. Elucidation of the small RNA component of the transcriptome. Science 309, 1567–1569 (2005).

116. Griffiths-Jones, S., Bateman, A., Marshall, M., Khanna, A. & Eddy, S. R. Rfam: an RNA family database. Nucleic Acids Res. 31, 439–441 (2003).

117. Krek, A. et al. Combinatorial microRNA target predictions. Nature Genet. 37, 495–500 (2005).

118. Lewis, B. P., Shih, I. H., Jones-Rhoades, M. W., Bartel, D. P. & Burge, C. B. Prediction of mammalian microRNA targets. Cell 115, 787–798 (2003).

119. Brennecke, J., Stark, A., Russell, R. B. & Cohen, S. M. Principles of microRNA-target recognition. PLoS Biol. 3, e85 (2005).

120. Lall, S. et al. A genome-wide map of conserved microRNA targets in C. elegans. Curr. Biol. 16, 460–471 (2006).

121. Sethupathy, P., Megraw, M. & Hatzigeorgiou, A. G. A guide through present computational approaches for the identification of mammalian microRNA targets. Nature Methods 3, 881–886 (2006).

122. Tompa, M. et al. Assessing computational tools for the discovery of transcription factor binding sites. Nature Biotechnol. 23, 137–144 (2005).

123. Siggia, E. D. Computational methods for transcriptional regulation. Curr. Opin. Genet. Dev. 15, 214–221 (2005).

124. Rhoades, M. W. et al. Prediction of plant microRNA targets. Cell 110, 513–520 (2002).

125. Lim, L. P. et al. Microarray analysis shows that some microRNAs downregulate large numbers of target mRNAs. Nature 433, 769–773 (2005).

126. Pillai, R. S., Bhattacharyya, S. N. & Filipowicz, W. Repression of protein synthesis by miRNAs: how many mechanisms? Trends Cell Biol. 2 Jan 2007 (doi:10.1016/j.tcb.2006.12.007).

127. Arteaga-Vazquez, M., Caballero-Perez, J. & Vielle-Calzada, J. P. A family of microRNAs present in plants and animals. Plant Cell 22 Dec 2006 (doi:10.1105/tpc.106.044420).

128. Rajagopalan, R., Vaucheret, H., Trejo, J., Bartel, D. P. A diverse and evolutionarily fluid set of microRNAs in Arabidopsis thaliana. Genes Dev. 20, 3407–3425.

AcknowledgementsWe regret that due to space constraints, we are unable to cite the work of many colleagues who have made key contribu-tions to the field. We thank K. Birnbaum for insightful discus-sions about gene regulation in plants, and V. Ambros, M. Hammell, S. Small and N. Sokol for helpful comments on a preliminary version of the manuscript.

Competing interests statementThe authors declare no competing financial interests.

DATABASESThe following terms in this article are linked online to:Entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=genelet-7 | miR-1 | miR-7 | miR-10 | miR-100 | Pax6 | Nkx2-5 |UniProtKB: http://ca.expasy.org/sprotdicer | Pumilio

FURTHER INFORMATIONThe Rajewsky laboratory web site: http://www.mdc-berlin.de/rajewskyNYU Department of Biology and Center for Comparative Functional Genomics: http://www.nyu.edu/fas/dept/biology

SUPPLEMENTARY INFORMATIONSee online article: S1 (figure) Access to this links box is available online

R E V I E W S



microRNAs and the immune responseMark A. Lindsay1,2

1 Respiratory Research Group, Wythenshawe Hospital, School of Translational Sciences, University of Manchester M23 9LT, UK2 Biopharmaceutics Research Group, Airways Disease, National Heart and Lung Institute, Imperial College, London SW3 6LY, UK

Review

microRNA (miRNA)-mediated RNA interference hasbeen identified as a novel mechanism that regulatesprotein expression at the translational level. Recentpublications have provided compelling evidence that arange of miRNAs are involved in the regulation of immu-nity, including the development and differentiation of Band T cells, proliferation of monocytes and neutrophils,antibody switching and the release of inflammatorymediators. In this review, we examine what is presentlyknown of the function and mechanism of action of thesemiRNAs in the regulation of the innate and acquiredimmune response.

IntroductionThe posttranscriptional regulation of gene expression viamicroRNA (miRNA)-mediated RNA interference (RNAi)was first identified in Caenorhabditis elegans in 1998 [1](Box 1 summarizes the pathway of miRNAmediated inter-ference). Since this initial observation, >700 miRNAs(miRNA registry at www.sanger.ac.uk/Software/Rfam/mirna/) have been identified in mammalian cells and havebeen shown to be involved in a range of physiologicalresponses, including development, differentiation andhomeostasis [2,3]. Interestingly, it is now clear that miR-NAs regulate many aspects of the immune response. Thishas emerged from studies that revealed selective expres-sion of miR-181a in the thymus and miR-223 in the bonemarrow and indicated their involvement in the differen-tiation of pluripotent hematopoietic stem cells (HSCs) intothe various blood cells lineages including B and T cells [4–7]. Subsequent reports have identified functions for indi-vidual miRNAs such as miR-150, miR-181a and miR-17�92 during T- and B-cell differentiation [4,8], whereasmiR-17�92 and miR-223 are implicated in myeloid pro-duction [9–12]. Experimentation has also revealed roles formiRNAs during the activation of the innate and acquiredimmune response. Thus, miR-146, miR-155 and miR-223are thought to regulate the acute inflammatory responseafter the recognition of pathogens by the Toll-like receptors(TLR) [13–16], whereas miR-155 and miR-181a are impli-cated in B- and T-cell responses [8,13–15]. Although notreviewed in this article (for review, see Refs. [17,18]),miRNAs have also been shown to have a direct antiviralaction by targeting crucial proteins and pathways used byhepatitis C virus [19,20], primate foamy virus type 1 [21]and vesicular stomatitis virus [22] (Table 1).

In the following sections, we will review the role ofmiRNAs during the development and differentiation ofimmune cells and in the regulation of the innate and

Corresponding author: Lindsay, M.A. ([email protected]).

1471-4906/$ – see front matter � 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.it.2008.04

acquired immune response (for additional reviews, seeRefs. [23–26]). Although expression profiling has indicateda role for a host of miRNAs (Table 1), we shall focus onthose for which we have functional and mechanistic infor-mation that is predominantly derived from studies usingtransgenic or knockout mouse models. With regard to theevaluation of mechanism, it must be emphasized that thedifficulty in predicting the mRNA targets of individualmiRNAs means that this process is often problematic.Thus, miRNAs are believed to either block mRNA trans-lation or reduce mRNA stability after imperfect binding ofthe guide strand to miRNA-recognition elements (MREs)within the 30 untranslated region (UTR) of target genes.Originally, it was believed that the specificity of thisresponse was mediated by the ‘seed’ region, which islocalized at residues 2–8 at the 50 end of the miRNA guidestrand (see Box 1). However, it now seems that this was anoversimplification and that miRNA targeting is influencedby additional factors such as the presence and cooperationbetween multiple MREs [27,28], the spacing betweenMREs [28,29], proximity to the stop codon [28], positionwithin the 30 UTR [28], AU composition [28] and targetmRNA secondary structure [30]. This problem is com-pounded by the usual problems associated with the deter-mination of the role of a gene, in this casemiRNAs, throughoverexpression or inhibition. This is particularly the casewhen undertaking overexpression studies in either cells ortransgenic mice, because nonphysiological levels of miR-NAs expression are likely to lead to downregulation ofincorrect targets.

miRNA-16: constitutive modulation of inflammationExpression profiling has shown that miRNA-16 is found athigh levels in most cells and tissues including thoseinvolved in inflammation such as monocytes, neutrophils,B cells and both CD4+ and CD8+ T cells [31]. This ubiqui-tous expression suggests that miR-16 might have a genericfunction, possibly in preventing cell cycle progression [32].Interestingly, a publication by Jing et al. [33] has shownthat miR-16 is required for the rapid degradation ofproteins that contain AU-rich elements (AREs) in their30 untranslated region, which includes most inflammatorymediators [33]. This process requires the miRNA proces-sing componentsDicer, Ago/eiF2C familymembers and theARE binding protein, tristetraprolin and involves bindingbetween theUAAAUAUU sequence inmiR-16 andAU-richsequences in cytokines and chemokines such as tumornecrosis factor a (TNFa), interleukin-8 (IL-8) and IL-6[33]. Because there are no reports that miR-16 expressionis changed in response to immune modulators, in additionto a potential role in cell cycle progression, one could

.004 343

http://www.sanger.ac.uk/Software/Rfam/mirna/

http://www.sanger.ac.uk/Software/Rfam/mirna/

mailto:[email protected]

http://dx.doi.org/10.1016/j.it.2008.04.004

Box 1. Pathway of microRNA-mediated RNA interference

microRNAs (miRNAs) are short double-stranded RNA molecules of

�19–23 nucleotides in length. They are produced from full-length,

RNA polymerase II transcripts called pri-miRNA after cleavage by

two RNase III enzymes called Drosha [58,59] and Dicer [60], in

association with a range of accessory proteins. Initial cleavage by

Drosha within the nuclear compartment produces a hairpin RNA of

�65 nucleotides known as pre-miRNAs. These pre-miRNAs are

transported into the cytoplasm by exportin 5 [61] and further

processed by Dicer [60] to produce the mature miRNAs, which are

double-stranded RNA molecules with a length of 19–23 nucleotides.

The actions of miRNAs are mediated by the miRNA-induced

silencing complex (miRISC), which is composed of multiple proteins

including a member of the double-stranded RNA binding protein

argonaute family (Ago) [62,63]. Using one strand of the miRNA,

called the guide strand (the red strand in Figure I), the miRISC is

believed to either block mRNA translation, reduce mRNA stability or

induce mRNA cleavage, after imperfect binding to miRNA recogni-

tion elements (MREs) within the 30 and 50 untranslated region (UTR)

of target mRNA genes [64–66]. Significantly, this redundancy within

the miRNA:mRNA binding region means that individual miRNAs are

able to target multiple mRNA targets and has led to speculation that

miRNAs might have a similar role to transcription factors [67,68].

Figure I. Pathway of microRNA-mediated RNA interference.

Review Trends in Immunology Vol.29 No.7

speculate that the highmiR-16 levels in inflammatory cellsmight also restrict the production of inflammatorymediators under nonstimulated conditions and/or thatmiR-16 binding might cooperate with other miRNA toregulate immunity.

miR-17�92 cluster: a ubiquitious regulator of B-cell,T-cell and monocyte developmentThemiR-17�92 cluster is located on chromosome 13 and iscomposed of six miRNAs (miR-17, miR-18a, miR-19a, miR-20a, miR-19b-1 and miR-92–1). Interestingly, a process ofduplication and deletion in early vertebrate evolutionhas resulted in the production of two additional paralogclusters, miR-106a�363 (miR-106a, miR-18b, miR-20b,miR-19b-2, miR-92–2 and miR-363) and miR-106b�25(miR-106b, miR-93 and miR-25), which are located onchromosome X and chromosome 6, respectively.

Previous studies indicated that miR-17�92 might beinvolved in the regulation of B cells after reports showingthat this cluster is located within a region of DNA that isamplified in B-cell lymphomas [34] and that overexpres-sion of both miR-17�92 and the transcription factor c-myc

344

produced by B-cell lymphomas in mice [35]. Subsequentstudies using miR-17�92 and Dicer knockout mice haveindicated that this cluster is required for the pro-B to pre-B-cell transition during B-cell development [36,37](Figure 1). Indeed, expression of miR-17�92 was shownto peak in pre-B-cells, where it inhibited cell death throughthe downregulation of the proapoptotic protein Bim [36](Figure 2). Significantly, miR-17�92 knockout mice werenonviable at birth and demonstrated severely hypoplasticlungs and a ventricular septal defect in the heart [36]. Forthis reason, it was necessary to perform these studies usingliver-derived B cells that were grafted into lethally irra-diatedmice. This contrasted with transgenic mice in whichmiR-17–92 was selectively induced in both T and Blymphocytes that were viable but that developed lympho-proliferative disease and autoimmunity, ultimately culmi-nating in premature death [38]. As might be expected ifmiR-17�92 positively regulated B-cell development, theseanimals were found to have increased numbers of activatedB cells. However, the major expansion was in the numberof activated CD4+ T cells and to a lesser extent, CD8+ Tcells, which implied that miR-17�92 also regulated T-celldevelopment [38]. Importantly, as with the knockout stu-dies, the increase in B- and T-cell population was shown toresult from enhanced proliferation and survival after thedownregulation of Bim and the tumour enhancer phospha-tase and tensin homology (PTEN) [38]. This thereby facili-tated the pro- to pre-transition of both B and T cells. Insummary, the authors of this report speculated that thephenotypic changes after transgenic expression of miR-17�92 resulted in breakdown in peripheral T-cell toler-ance, expansion of the CD4+ T cells that activates B cells,leading to the development of autoimmunity.

In humans, investigations using cord blood CD34+ hae-matopoietic progenitor cells that had been induced todifferentiate into monocytes after exposure to macro-phage-colony stimulating factor (M-CSF) indicated thatmiR-17-5p, miR-20a and miR-106a expressions areinvolved in monocyte differentiation and maturation [12](Figure 1). Examination of the mechanism showed thatreduced miR-17–5p, miR-20a and miR-106a expressionresulted in increased levels of their target protein, acutemyeloid leukaemia-1 (AML-1). AML-1 is a DNA-bindingsubunit of the transcription factor core-binding factor(CBF), which has been shown to induce monocyte differ-entiation and maturation by increasing the expression ofthe receptor for M-CSF, as well as other cytokines such asIL-3 and granulocyte macrophage-colony stimulating fac-tor (GM-CSF) [12] (Figure 2). Significantly, AML1 was alsofound to bind within the promoter region of miR-17–5p,miR-20a and miR-106a and inhibit their transcription.Thus, a reduction in miR-17–5p, miR-20a and miR-106alevels establishes a negative feedback loop involvingincreased AML1 and M-CSF receptor expression and thesubsequent downregulation in miR-17–5p, miR-20a andmiR-106a expression [12] (Figure 2).

miRNA-146a: negative regulator of the innateimmune responseThe miR-146 family is composed of two members, miR-146a andmiR-146b, that are located on chromosomes 5 and

Table 1. Summary of miRNA involvement in the immune response

miRNA Function Transcriptional

regulation

Targets Refs

miR-15a Decreased expression in chronic lymphocytic leukaemia Bcl-2 [43,77–79]

miR-16 Binds to UA rich elements in the 30 UTR and induces TNFa mRNA

degradation

TNFa [33]

miR-21 Increased expression in B-cell lymphoma and chronic lymphocytic

leukaemia

[43,80]

miR-17–5p In combination with miR-20a and miR-106a inhibits monocyte proliferation,

differentiation and maturation

AML-1 AML-1 [12]

miR-17�92

cluster

Regulates pro- to pre- transition during B- and T-cell development Bim, PTEN [36–38]

miR-20a In combination with miR-17–5p and miR-106a inhibits monocyte

proliferation, differentiation and maturation

AML-1 AML-1 [12]

miR-24 Inhibits replication of vesicular stomatitis virus [22]

miR-29a Down-regulated in B-cell chronic lymphocytic leukemia Tcl-1 [81]

miR-32 Inhibits replication of primate foamy virus type 1 [21]

miR-93 Inhibits replication of vesicular stomatitis virus [22]

miR-106a In combination with miR-17–5p and miR-20a inhibits monocyte proliferation,

differentiation and maturation

AML-1 AML-1 [12]

miR-122 Required for hepatitis C proliferation in liver [19]

miR-125b Expression downregulated by LPS and oscillations in expression after

exposure to TNFa

NFkB TNFa [15]

miR-146a Expression induced in macrophages and alveolar/bronchial epithelial

following activation of TLR-2, -4 and -5 or exposure to TNFa and IL-1b.

NFkB IRAK1, TRAF6 [14,16]

miR-146b LPS induced expression induced in macrophages IRAK1, TRAF6 [14]

miR-150 Increased expression leads to suppression of B-cell formation by blocking in

pro- to pre-B cell transition. Decreased expression in chronic lymphocytic

leukaemia (CLL)

[43,82]

miR-155 Increased expression in Hodgkin and non-Hodgkin lymphomas and chronic

lymphocytic leukaemia (CLL)

[39–43]

miR-155 Required for normal production of isotype-switched, high-affinity IgG1

antibodies in B-cells, determines Th1 and Th2 differentiation and positive

regulator of antigen induced responses in T-cells

AP-1 PU.1, c-Maf [44–46,48]

miR-155 Increased expression following activation of the innate immune response.

Inhibits inflammatory mediator release and stimulates granulocyte and

monocyte proliferation

AP-1 [13–15,47]

miR-181a Positive regulator of B-cell development and CD4+ T-cell selection, activation

and sensitivity.

SHP-2,

PTPN22,

DUSP5, DUSP6

[4,8,49]

miR-196 Induced by IFNb and inhibits replication of hepatitis C virus [20]

miR-223 Negative regulator of neutrophil proliferation and activation PU.1, C/EBPa,

NFI-A,

Mef2c, IGFR [9,11,50]





This table describes all the miRNAs that have been implicated in the immune response. It describes their function, transcription factors that are known to regulate their

expression and potential targets, including viral targets.

Abbreviations: AML, acute myeloid leukaemia; AP-1, activator protein; Bcl, B-cell lymphoma; Tcl, T-cell lymphoma; C/EBP, CCAAT-enhancer binding protein; DUSP, dual

specificity phosphatase; IGFR, insulin-like growth factor receptor; IRAK, IL-1 receptor activated kinase; JNK, c-jun N-terminal kinase; Maf, musculoaponeurotic fibrosarcoma:

Mef, myeloid ELF-1 like factor; miRNA, microRNA; NF, nuclear factor; PTEN, phosphatase and tensin homolog; PTP, protein tyrosine phosphatase; SHP, Src homology 2

domain–containing protein-tyrosine phosphatase; TLR, Toll-like receptor; TRAF, TNF receptor-associated factor.


10, respectively. Evidence that miR-146a and miR-146bmight be involved in the innate immune response was firstprovided by Taganov et al. [14], who showed increasedexpression in the human monocytic THP-1 cell line inresponse to lipopolysaccharide (LPS) [14] (Figure 2). Sub-sequent studies have shown that this is a general responsein myeloid cells activated through TLR-2, -4 or -5 bybacterial and fungal components or after exposure to theproinflammatory cytokines, TNF-a or IL-1b [13–15]. Bycontrast, miR-146a expression is not increased after acti-vation of the TLR-3, -7 or -9 in response to viral andbacterial nucleic acids (Figure 2) [14]. Interestingly, thisresponse does not seem to be restricted to inflammatorycells because we also observed increased miRNA-146abut not miRNA-146b expression in lung epithelial cells[16] and airway smooth muscle cells (H. Larner-Svensson

et al., personal communication) after IL-1b stimulation.Inspection of the upstream promoter region of miRNA-146a identified several potential transcription factor bind-ing sites including interferon regulatory factor (IRF3),IRF7 and CCAAT enhancer-binding protein-b (C/EBPb),although it was shown that NF-kB is responsible for theLPS-induced response [14]. At the present time, the tran-scriptional regulators of miR-146b are unknown.

To determine the biological role of miRNA-146a and b,examination of the public databases predicted severalpotential targets for miR-146a and miR-146b includingthe IL-1 receptor associated kinase 1 (IRAK1) and TNFreceptor-associated factor 6 (TRAF6). Significantly, IRAK1and TRAF6 are known to be part of the common signallingpathway derived from TLR-2,-4 and -5 and the IL-1b re-ceptor,which led to speculation that increasedmiRNA-146a

345

Figure 1. Overview of the role of microRNAs (miRNAs) during the differentiation and maturation of immune cells. miRNAs have been shown to regulate multiple steps in

the development of immune cells including lymphocytes (B and T cells) and myeloid cell (monocytes and neutrophils). CD, cluster designation; CLP, common lymphocyte

progenitor; CMP, common myeloid progenitor; GMP, granulocyte monocytic progenitor; HSC, hematopoietic stem cell; MDP, myeloid dendritic progenitor.


or miR-146b expression might act in a negative feedbackpathway. Indeed, this contention was supported by studiesshowing downregulation in the activity of a luciferase repor-ter fused to the 30UTRof IRAK1orTRAF6and co-expressedwith a miR-146a or miR-146b containing plasmid [14].Increased miR-146a but not miR-146b expression, wasrecently confirmed in alveolar epithelial cells and shownto negatively regulate IL-1b–induced IL-8 and RANTES(regulated on activation, normal T-cell expressed andsecreted) [16]. Interestingly, increasedmiRNA-146aexpres-sion was only seen at high IL-1b concentrations, whichindicated that negative feedback is only activated duringsevere inflammation and that this might be crucial in pre-venting potentially dangerous inflammation from spirallingout of control. However, examination of the mechanismshowed that this was not caused by downregulation ofIRAK1 or TRAF6 but instead occurred at the translational

346

level, through an as yet unidentified mechanism (Figure 2)[16].

miRNA-155: a regulator of T- and B-cell maturationand the innate immune responseA single miR-155 transcript has been identified that isprocessed from within the second exon of the non–protein-encoding gene, bic. Interestingly, the observation thatmiR-155 is overexpressed in human B-cell lymphomasand studies showing that transgenic expression of miR-155 in mouse B cells produced malignancies have led tospeculation that this miRNA is involved in B-cell differ-entiation and proliferation [39–43].

A central role for miR-155 in the regulation of T- and B-cell responses during the acquired immune response hasemerged from studies in knockout mice, because thesewere shown to be immunodeficient and failed to develop

Figure 2. Role of microRNAs (miRNAs) in the regulation of the immune response. The illustration summaries what is know of the mechanism by which individual miRNAs

(red boxes) interact with transcription factors (green circles) and other proteins (yellow boxes) to regulate cell responses (blue circles) during immune cell development and

the innate (monocytes, macrophages and neutrophils) and acquired (B and T cells) immune response. Stimulation (") or inhibition (>) should be determined by following

how the changes in the individual miRNA expression (red squares) impacts on the indicated biological response (blue circles). AML, acute myeloid leukaemia; AP-1,

activator protein; BCR, B-cell receptor; C/EBP, CCAAT-enhancer binding protein; DUSP, dual specificity phosphatase; IGFR, insulin-like growth factor receptor; IRAK, IL-1

receptor activated kinase; JNK, c-jun N-terminal kinase; M-CSF, macrophage colony stimulating factor; Maf, musculoaponeurotic fibrosarcoma: Mef, myeloid ELF-1 like

factor; NF, nuclear factor; PTEN, phosphatase and tensin homolog; PTP, protein tyrosine phosphatase; RA, retinoic acid; SHP, Src homology 2 domain–containing protein-

tyrosine phosphatase; TCR, T-cell receptor; TLR, Toll-like receptor; TRAF, TNF receptor-associated factor.


a protective response to bacteria after immunisation[44,45] (Figure 1). Interestingly, miR-155 knockout micealso undergo age-related lung airway remodelling, which ischaracterised by increased collagen deposition, smoothmuscle mass and inflammatory cell infiltrate within thebroncholaveolar lavage [44]. These phenotypic obser-vations seem to be in part mediated through involvementof miR-155 in B-cell production of isotype-switched, high-affinity IgG1 antibodies and during the development of B-cell memory [45,46]. The reduction in IgG1 antibodies wasnot the result of impaired somatic hypermutation and classswitch recombination. Instead, the observed reduction inthe size of the germinal centers suggested that reducedantibody production and B-cell memory was the result of afailure to select high affinity plasma B cells [45,46]. Exam-ination of themechanism showed that miR-155 attenuatedthe expression of the transcription factor PU.1, which wasshown to downregulate IgG1 levels [46] (Figure 2).

As with B cells, it seems that miR-155 is involved in T-cell differentiation [44,45]. Thus, naı̈ve T cells derived from

miR-155 knockout mice were shown to have an increasedpropensity to differentiate into Th2 rather than Th1 cells,with the concomitant production of Th2 cytokines such asIL-4, IL-5 and IL-10 [44,45]. It has been speculated thatthis bias results from miR-155 targeting of c-Maf (muscu-loaponeurotic fibrosarcoma), a transcription factor that isknown to be a potent transactivator of the IL-4 promoter, akey cytokine in the development of Th2 cells [44]. Withregard to the acute immune response, the T lymphocyteshad an impaired response and showed attenuated IL-2 andinterferon g (IFNg) release in response to antigens [44,45](Figure 2).

A link between miR-155 and the innate immuneresponse was suggested from studies showing increasedexpression after LPS (via TLR-4) and lipoprotein (via TLR-2) stimulation in monocytes or macrophages and in thesplenocytes of mice that had been inoculated with Salmo-nella enteritidis–derived LPS [13–15] (Figure 2). In con-trast to miR-146a, increased miR-155 expression was alsoseen after activation of the innate response by viral- and

347


bacterial-derived nucleotides including poly(I:C) (viaTLR-3) andCpG (viaTLR-9), although increasedexpressionin response to IFNb and IFNg seemed to be secondary to theautocrine release of TNFa [13]. Tili et al. [15] showed thatmiR-155 expression oscillated after TNFa stimulation, withan initial drop at 30 min followed by an increase at 60 min[15]. This led them to speculate that miR-155 might exertboth positive and negative actions on the expression oftarget proteins that included theNF-kB signalling proteins,IKKb and IKKe, as well as FADD (Fas-associated deathdomain) and Ripk1 (receptor interacting serine-threoninekinase 1) [15,42]. In contrast to miR-146a, miRNA-155seems to positively regulate the release of inflammatorymediators during the innate immune response because itwas shown to increase TNF-a production in human embryo-nic kidney cells (HEK-293) cells by relieving the 30 UTR–mediated posttranscriptional inhibition. This observationwas supported by studies in Em-miR-155 transgenic micethat overexpress miR-155 in B cells, which demonstrate anelevated level of serumTNFaand increased susceptibility toseptic shock [15]. Interestingly, a recent report has shownthat LPS induced strong but transient miR-155 expressionin mouse bone marrow cells and indicated that this is likelyto drive granulocyte/monocyte expansion [47]. The possibleinvolvement ofmiR-155 in thedevelopment of acutemyeloidleukaemia (AML) was revealed from studies of the effect oflong-term miR-155 overexpression. In these studies, viral-mediated transfection of miR-155 into hematopoietic stemcells (HSCs) and engraftment into lethally irradiated miceproduced some of the pathological features characteristic ofmyeloid neoplasia [47]. This led these authors to speculatethat the well-established link between inflammation andcancer might involve chronic upregulation of miR-155,which could predispose these individuals to the develop-ment of myeloproliferative disorders.

The proinflammatory transcription factors, AP-1 andNF-kB, have both been reported to regulate miR-155expression [13,45,48]. Thus, in macrophages, the actionof TLR-3 and TNFa is mediated via AP-1 [13], whereas theresponse to LPS is via NF-kB [19]. Similarly, BCR cross-linking in a human B-cell line was shown to induce miR-155 expression via activation of the extracellular-regulatedkinase (ERK) and c-jun N-terminal kinase (JNK) pathwayand the subsequent recruitment of FosB and JunB to AP-1binding site [48] (Figure 2).

miRNA-181a: regulation of B-cell differentiation andCD4+ T-cell selection, activation and sensitivityEarly studies by Chen et al. [4] demonstrated that miR-181a was selectively expressed in thymus-derived B cellsand identified a positive role for miR-181a in B-cell differ-entiation after showing that ecotopic expression in mousehematopoietic precursor cells led to an increased fraction ofB cells [4]. Significantly, these overexpression studies alsoshowed a reduction in CD8+ killer T cells, which suggesteda potential role in the regulation in T-cell differentiation(Figure 1). However, despite subsequent studies thatshowed dramatic changes in miR-181a expression duringthe various stages of T-cell differentiation, these obser-vations were often contradictory, and the role of miR-181aduring T-cell development is still inconclusive [8,49].

348

In contrast to T-cell differentiation, miR-181a has beenshown to positively regulate the T-cell receptor (TCR)-mediated response to antigen in CD4+ T cells. Thus, bymeasuring changes in the intracellular Ca2+ transient andIL-2 release, these investigators showed that miR-181aoverexpression amplifyied the strength and sensitivity ofTCR-mediated activation [8] (Figure 2). In addition, thesestudies showed a qualitative change that permitted T cellsto respond to antagonists as agonists and indicated thatmiR-181a, at least in part, regulated the positive andnegative selection of T cells during thymic development[8]. Examination of the mechanism demonstrated that thisdid not result from changes in the expression of surfacereceptors but instead involved the coordinated downregu-lation of multiple phosphatases. These included the tyro-sine phosphatases, Src homology 2 domain–containingprotein-tyrosine phosphatase (SHP)-2 and protein tyrosinephosphatase (PTP)-N22 and the ERK-specific, dual speci-ficity phosphatases (DUSP)-5 and -6, which are all knownto negatively regulate the TCR signalling pathway. Specifi-cally, it has been speculated that changes in miR-181amight act like a rheostat by regulating protein phosphoryl-ation levels. In this model, increased miR-181a andreduced phosphatase levels would lead to increased basalphosphorylation of protein kinases such as Lck and ERK.Under these condition, T cells would exhibit increased TCRsignalling and a reduced T-cell activation threshold [8](Figure 2).

miRNA-223: a regulator of neutrophil proliferationand activationExpression profiling has shown that miR-223 expression isassociated with myeloid cells in the bone marrow and thatincreased expression of miR-223 is involved in the differ-entiation of myeloid precursors into granulocytes such asneutrophils [4,9,11] (Figure 1). Interestingly, recent obser-vations inmiR-223 knockout mice, which were shown to beviable and healthy, revealed increased numbers of gra-nulocyte progenitors in the bone marrow and hypermatureneutrophils in the circulation. This implied that miR-223was involved in the negative regulation of maturation butnot differentiation of granulocytes [11]. Mechanistic stu-dies showed that the action of miR-223 is probablymediated through downregulation of either myeloidELF-1-like factor (Mef)-2c, a transcription factor that pro-motes myeloid progenitor proliferation, or the insulin-likegrowth factor receptor (IGFR) [11] (Figure 2). Paradoxi-cally, an earlier report by Fazi et al. [9] reported that miR-223 was a positive regulator of granulocyte differentiation.However, these studies were performed in a retinoic acid(RA)-treated acute promyelocytic leukemic cell line and therole of miR-223 might be dependent on the timing ofmiRNA overexpression or miRNA inhibition [9]. Func-tional studies in circulating neutrophils obtained fromthe knockout animals indicated that miR-223 was notinvolved in extravasagation, migration or phagocytosisof the bacterium Escherichia coli but that reduced miR-223 resulted in enhanced oxidative burst and killing of thefungi, Candida albicans (Figure 2). In addition, theseanimals spontaneously developed inflammatory lung path-ology and exhibited exaggerated tissue destruction after

Box 2. Modulating microRNAs as a novel therapeutic

approach

To determine their biological function, it is possible to both inhibit

and overexpress individual microRNAs (miRNAs). Inhibition is

performed by using antisense sequences (called miRNA antago-

mirs) targeted to the miRNA guide strand that block the interaction

with the miRNA recognition elements within the 30 untranslated

region (UTR) of the target mRNA genes [4,69]. To increase their

binding affinity and stability in biological fluids, these antagomirs

often contain 2’-O-methyl, phosphorothioate or locked nucleic acid

substitutions [12,54]. Significantly, two studies have successfully

used chemically stabilized antagomirs to target miRNA-122a in the

liver of mice after intravenous administration and shown that this

regulates cholesterol metabolism [70,71]. To overexpress miRNAs,

investigators have transfected cells with chemically synthesized

miRNAs (called miRNA mimics) or expressed pre-miRNA from

plasmids and viral vectors [72]. Although it might be possible to

extend the use of virally expressed miRNAs as a therapeutic

approach, there have been no reports on the use of miRNA mimics

in animals. Thus, although it has been possible to develop stabilized

siRNA as potential therapeutic agents [73–76], miRNAs are known to

contain noncomplementary bulge regions that makes them highly

susceptible to RNase degradation in biological fluids. In addition to

stability, the major hurdle to the development of miRNA antagomirs

and miRNA mimics is the delivery of these large, negatively charged

oligonucleotides into relevant cells and tissues [73]. Until this can be

addressed, the application of miRNA-based therapeutics is likely to

be restricted to those tissues that are amenable to topical

administration including liver (after intravenous injection), lung,

eye, vagina and skin.


endotoxin challenge. Of relevance to this observation, wehave recently demonstrated a rapid and selective increasein miR-223 expression in lung and airway epithelia afterexposure of mice to aerosolized endotoxin [10]. Thus, ifmiR-223 also acts as a negative modulator of the inflam-matory response in these cell types, this might contributeto the lung pathologies observed in miR-223 knockoutmice.

Multiple transcription factors have been implicated inregulating miR-223 expression during granulocyte differ-entiation and maturation. In the report showing positiveregulation of RA-induced granulocyte differentiation bymiR-223, Fazi et al. [9] showed that miR-223 expressionwas increased by C/EBP-a and attenuated by nuclearfactor (NF)I-A. This investigation also showed that miR-223 targeted the downregulation of NFI-A, which wouldthereby established a negative feedback loop that pro-moted granulopoiesis [9]. A role of C/EBPa in regulatingmiR-223 expression was confirmed by Fukao et al. [50],who also demonstrated increasedmiR-223 expression afteractivation of another myeloid specific transcription factor,PU.1 (Figure 2).

ConclusionAlthough investigators have only recently identified arole for miRNA-mediated RNA interference in theimmune response, several interesting trends haveemerged. The first indication that miRNAs, includingmiR-155, miR-181a and miR-223, might regulate thedevelopment of immune cells was their differentialexpression in the thymus and bone marrow and reportsthat changes in expression were associated with the de-velopment of various leukaemias [4,39–43,51]. Signifi-cantly, subsequent studies have confirmed that miR-155 [44,45,47], miR-181a [4,8,49] and miR-223[9,11,50], as well as those miRNAs derived from themiR-17�92 cluster [36–38], are indeed involved in thedevelopment and differentiation of lymphoid and myeloidcells. In addition, mechanistic studies have suggestedthat this developmental process is crucially dependentupon interactions between miRNAs and transcriptionfactors (Figure 2), which has also been observed duringdevelopment of other tissues such as skeletal muscle andheart [52–55]. Given that aberrant expression of miRNAshas been associated with development of leukaemias,howmight this be linked to development of immune cells?Interestingly, increased expression of miR-155, miR-181aand miR-17�92 has been shown to promote the survivaland proliferation of both lymphoid and myeloid cells.Following this observation, Xiao et al. [38] speculatedthat somatic mutation and amplification of these miRNAsmight result in a reduction in the activation threshold,increased proliferation and enhanced survival. In turn,this could lead to expansion of this cellular populationand a higher probability of accumulating subsequentmutations that are required to produce the malignanttransformation [38].

At the present time, miRNAs seem to regulate theresponses associated with acquired immunity, such asantibody production [45,46] and inflammatory mediatorrelease [44,45] through their impact on the development

and differentiation of immune cells. Thus, there are noreports that activation of TCR and BCR signalling influ-ences T- and B-cell responses through changes in miRNAexpression. This contrasts with activation of the innateimmune response that has been shown to induce rapidchanges in the levels of miR-146a [14,16] and miR-155expression [13–15,47]. These miRNAs have been reportedto regulate acute inflammatory responses such as therelease of IL-8 and RANTES either through the targetingof proteins involved in the signalling pathway (IRAK1,TRAF6, IKKb, IKKe) or by modulating the translation ofthe inflammatory mediators themselves (i.e. TNFa, IL-8and RANTES). At the present time, the mechanism bywhich miR-146a and miR-155 influence translation isunknown, although, like miR-16 [33], these might regulatemRNA stability via AU-rich elements that are found in the30UTR ofmany inflammatorymediators [56]. Of relevance,a recent in silico survey of the interactions between miR-NAs and immune genes unexpectedly found preferentialtargeting of proteins involved in the regulation of AU-richelements and miRNA metabolism [57]. Furthermore, thisreport indicated that major targets of miRNAs are tran-scription factors, cofactors and chromatin modifiers ratherthan receptors, their ligands or inflammatory mediators[57].

Overall, there is now compelling evidence that miRNAsare involved in the regulation of the immune response, andmodulation of their activity might ultimately provide anovel therapeutic approach in the treatment of inflamma-tory disease and certain leukaemias (Box 2).

AcknowledgementsM.A.L. is supported by the Wellcome Trust University Award (076111).

349


References1 Fire, A. et al. (1998) Potent and specific genetic interference by double-

stranded RNA in Caenorhabditis elegans. Nature 391, 806–8112 Williams, A.E. (2008) Functional aspects of animal microRNAs. Cell.

Mol. Life Sci. 65, 545–5623 Stefani, G. and Slack, F.J. (2008) Small non-coding RNAs in animal

development. Nat. Rev. Mol. Cell Biol. 9, 219–2304 Chen, C.Z. et al. (2004) MicroRNAs modulate hematopoietic lineage

differentiation. Science 303, 83–865 Cobb, B.S. et al. (2005) T cell lineage choice and differentiation in the

absence of the RNase III enzyme Dicer. J. Exp. Med. 201, 1367–13736 Cobb, B.S. et al. (2006) A role for Dicer in immune regulation. J. Exp.

Med. 203, 2519–25277 Muljo, S.A. et al. (2005) Aberrant T cell differentiation in the absence of

Dicer. J. Exp. Med. 202, 261–2698 Li, Q.J. et al. (2007) miR-181a is an intrinsic modulator of T cell

sensitivity and selection. Cell 129, 147–1619 Fazi, F. et al. (2005) A minicircuitry comprised of microRNA-223 and

transcription factors NFI-A and C/EBPalpha regulates humangranulopoiesis. Cell 123, 819–831

10 Moschos, S.A. et al. (2007) Expression profiling in vivo demonstratesrapid changes in lung microRNA levels following lipopolysaccharide-induced inflammation but not in the anti-inflammatory action ofglucocorticoids. BMC Genomics 8, 240

11 Johnnidis, J.B. et al. (2008) Regulation of progenitor cell proliferationand granulocyte function by microRNA-223. Nature 451, 1125–1129

12 Fontana, L. et al. (2007) MicroRNAs 17-5p-20a-106a controlmonocytopoiesis through AML1 targeting and M-CSF receptorupregulation. Nat. Cell Biol. 9, 775–787

13 O’Connell, R.M. et al. (2007) MicroRNA-155 is induced during themacrophage inflammatory response. Proc. Natl. Acad. Sci. U. S. A.104, 1604–1609

14 Taganov, K.D. et al. (2006) NF-kappaB-dependent induction ofmicroRNA miR-146, an inhibitor targeted to signaling proteins ofinnate immune responses. Proc. Natl. Acad. Sci. U. S. A. 103,12481–12486

15 Tili, E. et al. (2007) Modulation of miR-155 and miR-125b Levelsfollowing Lipopolysaccharide/TNF-{alpha} Stimulation and TheirPossible Roles in Regulating the Response to Endotoxin Shock. J.Immunol. 179, 5082–5089

16 Perry, M.M. et al. (2008) Rapid changes in microRNA-146aexpression negatively regulate the interleukin-1b inducedinflammatory response in human lung alveolar epithelial cells. J.Immunol. 180, 5689–5698

17 Cullen, B.R. (2006) Viruses and microRNAs. Nat. Genet. 38 (Suppl 1),S25–S30

18 Sarnow, P. et al. (2006) MicroRNAs: expression, avoidance andsubversion by vertebrate viruses. Nat. Rev. Microbiol. 4, 651–659

19 Jopling, C.L. et al. (2005) Modulation of hepatitis C virus RNAabundance by a liver-specific MicroRNA. Science 309, 1577–1581

20 Pedersen, I.M. et al. (2007) Interferon modulation of cellularmicroRNAs as an antiviral mechanism. Nature 449, 919–922

21 Lecellier, C.H. et al. (2005) A cellular microRNA mediates antiviraldefense in human cells. Science 308, 557–560

22 Otsuka, M. et al. (2007) Hypersusceptibility to vesicular stomatitisvirus infection in Dicer1-deficient mice is due to impaired miR24 andmiR93 expression. Immunity 27, 123–134

23 Taganov, K.D. et al. (2007) MicroRNAs and immunity: tiny players in abig field. Immunity 26, 133–137

24 Lodish, H.F. et al. (2008) Micromanagement of the immune system bymicroRNAs. Nat. Rev. Immunol. 8, 120–130

25 Calin, G.A. and Croce, C.M. (2006) MicroRNA signatures in humancancers. Nat. Rev. Cancer 6, 857–866

26 Croce, C.M. and Calin, G.A. (2005) miRNAs, cancer, and stem celldivision. Cell 122, 6–7

27 Doench, J.G. et al. (2003) siRNAs can function as miRNAs. Genes Dev.17, 438–442

28 Farh, K.K. et al. (2005) The widespread impact of mammalianMicroRNAs on mRNA repression and evolution. Science 310, 1817–1821

29 Saetrom, P. et al. (2007) Distance constraints between microRNAtarget sites dictate efficacy and cooperativity. Nucleic Acids Res. 35,2333–2342

350

30 Long, D. et al. (2007) Potent effect of target structure on microRNAfunction. Nat. Struct. Mol. Biol. 14, 287–294

31 Landgraf, P. et al. (2007) A mammalian microRNA expression atlasbased on small RNA library sequencing. Cell 129, 1401–1414

32 Linsley, P.S. et al. (2007) Transcripts targeted by the microRNA-16family cooperatively regulate cell cycle progression.Mol. Cell. Biol. 27,2240–2252

33 Jing, Q. et al. (2005) Involvement of microRNA in AU-rich element-mediated mRNA instability. Cell 120, 623–634

34 Ota, A. et al. (2004) Identification and characterization of a novel gene,C13orf25, as a target for 13q31-q32 amplification in malignantlymphoma. Cancer Res. 64, 3087–3095

35 He, L. et al. (2005) A microRNA polycistron as a potential humanoncogene. Nature 435, 828–833

36 Ventura, A. et al. (2008) Targeted deletion reveals essential andoverlapping functions of the miR-17 through 92 family of miRNAclusters. Cell 132, 875–886

37 Koralov, S.B. et al. (2008) Dicer ablation affects antibody diversity andcell survival in the B lymphocyte lineage. Cell 132, 860–874

38 Xiao, C. et al. (2008) Lymphoproliferative disease and autoimmunity inmice with increased miR-17-92 expression in lymphocytes. Nat.Immunol. 9, 405–414

39 Kluiver, J. et al. (2005) BIC and miR-155 are highly expressed inHodgkin, primary mediastinal and diffuse large B cell lymphomas. J.Pathol. 207, 243–249

40 Eis, P.S. et al. (2005) Accumulation of miR-155 and BIC RNA in humanB cell lymphomas. Proc. Natl. Acad. Sci. U. S. A. 102, 3627–3632

41 Metzler, M. et al. (2004) High expression of precursor microRNA-155/BIC RNA in children with Burkitt lymphoma. Genes ChromosomesCancer 39, 167–169

42 Costinean, S. et al. (2006) Pre-B cell proliferation and lymphoblasticleukemia/high-grade lymphoma in E{micro}-miR155 transgenic mice.Proc. Natl. Acad. Sci. U. S. A. 103, 7024–7029

43 Fulci, V. et al. (2007) Quantitative technologies establish a novelmicroRNA profile of chronic lymphocytic leukemia. Blood 109, 4944–4951

44 Rodriguez, A. et al. (2007) Requirement of bic/microRNA-155 fornormal immune function. Science 316, 608–611

45 Thai, T.H. et al. (2007) Regulation of the germinal center response bymicroRNA-155. Science 316, 604–608

46 Vigorito, E. et al. (2007) microRNA-155 regulates the generation ofimmunoglobulin class-switched plasma cells. Immunity 27, 847–859

47 O’Connell, R.M. et al. (2008) Sustained expression of microRNA-155 inhematopoietic stem cells causes a myeloproliferative disorder. J. Exp.Med. 205, 585–594

48 Yin, Q. et al. (2008) B-cell receptor activation induces BIC/miR-155expression through a conserved AP-1 element. J. Biol. Chem. 283,2654–2662

49 Neilson, J.R. et al. (2007) Dynamic regulation of miRNA expression inordered stages of cellular development. Genes Dev. 21, 578–589

50 Fukao, T. et al. (2007) An evolutionarily conserved mechanism formicroRNA-223 expression revealed by microRNA gene profiling. Cell129, 617–631

51 Chen, C.Z. and Lodish, H.F. (2005) MicroRNAs as regulators ofmammalian hematopoiesis. Semin. Immunol. 17, 155–165

52 Naguibneva, I. et al. (2006) The microRNA miR-181 targets thehomeobox protein Hox-A11 during mammalian myoblastdifferentiation. Nat. Cell Biol. 8, 278–284

53 Zhao, Y. et al. (2005) Serum response factor regulates a muscle-specificmicroRNA that targets Hand2 during cardiogenesis. Nature 436, 214–220

54 Care, A. et al. (2007) MicroRNA-133 controls cardiac hypertrophy.Nat.Med. 13, 613–618

55 Rao, P.K. et al. (2006) Myogenic factors that regulate expression ofmuscle-specific microRNAs. Proc. Natl. Acad. Sci. U. S. A. 103, 8721–8726

56 Barreau, C. et al. (2006) AU-rich elements and associated factors: arethere unifying principles? Nucleic Acids Res. 33, 7138–7150

57 Asirvatham, A.J. et al. (2008) MicroRNA targets in immune genes andthe Dicer/Argonaute and ARE machinery components. Mol. Immunol.45, 1995–2006

58 Lee, Y. et al. (2003) The nuclear RNase III Drosha initiates microRNAprocessing. Nature 425, 415–419


59 Denli, A.M. et al. (2004) Processing of primary microRNAs by theMicroprocessor complex. Nature 432, 231–235

60 Zhang, H. et al. (2004) Single processing centermodels for humanDicerand bacterial RNase III. Cell 118, 57–68

61 Yi, R. et al. (2003) Exportin-5 mediates the nuclear export of pre-microRNAs and short hairpin RNAs. Genes Dev. 17, 3011–3016

62 Liu, J. et al. (2004) Argonaute2 is the catalytic engine of mammalianRNAi. Science 305, 1437–1441

63 Meister, G. et al. (2004) Human Argonaute2 mediates RNA cleavagetargeted by miRNAs and siRNAs. Mol. Cell 15, 185–197

64 Lewis, B.P. et al. (2003) Prediction of mammalian microRNA targets.Cell 115, 787–798

65 Doench, J.G. and Sharp, P.A. (2004) Specificity of microRNA targetselection in translational repression. Genes Dev. 18, 504–511

66 Didiano, D. andHobert, O. (2006) Perfect seed pairing is not a generallyreliable predictor formiRNA-target interactions.Nat. Struct.Mol. Biol.13, 849–851

67 Lewis, B.P. et al. (2005) Conserved seed pairing, often flanked byadenosines, indicates that thousands of human genes are microRNAtargets. Cell 120, 15–20

68 Krek, A. et al. (2005) Combinatorial microRNA target predictions.Nat.Genet. 37, 495–500

69 Meister, G. et al. (2004) Sequence-specific inhibition of microRNA- andsiRNA-induced RNA silencing. RNA 10, 544–550

70 Esau, C. et al. (2006) miR-122 regulation of lipid metabolism revealedby in vivo antisense targeting. Cell Metab. 3, 87–98

71 Krutzfeldt, J. et al. (2005) Silencing of microRNAs in vivo with‘antagomirs’. Nature 438, 685–689

72 Zeng, Y. et al. (2002) Both natural and designedmicro RNAs can inhibitthe expression of cognate mRNAs when expressed in human cells.Mol.Cell 9, 1327–1333

73 de Fougerolles, A. et al. (2007) Interfering with disease: a progressreport on siRNA-based therapeutics.Nat. Rev. Drug Discov. 6, 443–453

74 John, M. et al. (2007) Effective RNAi-mediated gene silencing withoutinterruption of the endogenous microRNA pathway. Nature 449, 745–747

75 Soutschek, J. et al. (2004) Therapeutic silencing of an endogenous geneby systemic administration of modified siRNAs. Nature 432, 173–178

76 Zimmermann, T.S. et al. (2006) RNAi-mediated gene silencing in non-human primates. Nature 441, 111–114

77 Calin, G.A. et al. (2002) Frequent deletions and down-regulation ofmicro- RNA genes miR15 and miR16 at 13q14 in chronic lymphocyticleukemia. Proc. Natl. Acad. Sci. U. S. A. 99, 15524–15529

78 Cimmino, A. et al. (2005) miR-15 and miR-16 induce apoptosis bytargeting BCL2. Proc. Natl. Acad. Sci. U. S. A. 102, 13944–13949

79 Raveche, E.S. et al. (2007) Abnormal microRNA-16 locus with syntenyto human 13q14 linked to CLL in NZB mice. Blood 109, 5079–5086

80 Lawrie, C.H. et al. (2007) MicroRNA expression distinguishes betweengerminal center B cell-like and activated B cell-like subtypes of diffuselarge B cell lymphoma. Int. J. Cancer 121, 1156–1161

81 Pekarsky, Y. et al. (2006) Tcl1 expression in chronic lymphocyticleukemia is regulated by miR-29 and miR-181. Cancer Res. 66,11590–11593

82 Zhou, B. et al. (2007) miR-150, a microRNA expressed in mature B andT cells, blocks early B cell development when expressed prematurely.Proc. Natl. Acad. Sci. U. S. A. 104, 7080–7085

351

Documents

The University of Chicago - Elisabet Pujadas PCBio Summer 2008 …ibd.uchicago.edu/undergrad_programs/pdf/elisabetPujadas.pdf · 2015-03-15 · Elisabet Pujadas PCBio Summer 2008