Non-coding RNAs: Lightning strikes twice

R436

Dispatch

Non-coding RNAs: Lightning strikes twice Eric G. Moss

A second case has been found of a nematode gene involved in developmental timing that encodes a short, non-coding RNA. Both RNAs are expressed at specific times and appear to repress target genes by interacting with their 3′ untranslated regions. A coincidence? Or does this pathway attract small RNA regulators? Address: Cell and Developmental Biology, Fox Chase Cancer Center, Philadelphia, Pennsylvania 19111, USA. Current Biology 2000, 10:R436–R439 0960-9822/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved.

Genetic analysis often yields surprises when investigators clone a gene known previously only by its mutant phenotype. That was the case in 1993, when the gene lin-4 of the nematode Caenorhabditis elegans was found to encode an exceptionally small, non-coding RNA [1]. The lin-4 gene was considered an oddity, and no-one knew whether another like it would be found. But Reinhart et al. [2] have now reported that the gene let-7 also encodes a small noncoding RNA. These two RNAs are of completely different sequence, but the fact that they are both found in the same developmental control pathway underscores the question of whether there are more such RNAs to be found in other pathways or other species. The pathway

Both lin-4 and let-7 are ‘heterochronic’ genes, a set of genes defined by their special mutant phenotypes [3]. Heterochronic phenotypes fall into two classes: precocious and retarded. Precocious mutants skip developmental events that are characteristic of one of the animal’s four larval stages, causing subsequent development to occur one stage too early. Retarded mutants reiterate stagespecific events, so that subsequent development is either delayed or never occurs. Collectively, the heterochronic genes regulate timing in all tissues of the animal except the gonad. Their signature characteristic is that they explicitly regulate the temporal component of cell fates, just as other kinds of developmental control genes regulate the spatial component. The heterochronic gene pathway contains at least 12 known genes. The group started with four, lin-4, lin-14, lin-28 and lin-29, which were described by Ambros and Horvitz [4]. Victor Ambros has characterized this pathway in great detail, and his original genetic analysis [5], carried out independently of molecular information, accurately predicted the precise molecular relationships among these four heterochronic genes for a variety of developmental events. The

lin-14 gene is required for first stage events to occur on time, and lin-28 for second stage events. The lin-4 gene represses both lin-14 and lin-28 when their jobs are done to allow later events to occur. And lin-29 is needed for the animal to exit larval development and differentiate into an adult. The let-7 gene is among the more recently characterized members of the set with heterochronic mutant phenotypes. Others include lin-41, lin-42 and daf-12 (reviewed in [3]). The genetic and molecular relationships among these heterochronic genes and the four original ones have not been fully worked out, and one suspects that they are complex. Furthermore, it is likely that more heterochronic genes remain to be found, because the pathway has a few holes in it. Nevertheless, the picture emerging out of the molecular characterization of the heterochronic pathway is becoming clearer. The regulation

The lin-4 loss-of-function mutants show a retarded phenotype, in which first stage events are repeated in subsequent stages (Figure 1). This is the opposite of the lin-14 loss-of-function mutants, which have a precocious phenotype, skipping first stage events. The precocious lin-14 mutation is epistatic to the retarded lin-4 mutation, which is the basis of the conclusion that lin-4 normally represses lin-14. There are two rare gain-of-function alleles of lin-14 that show a retarded phenotype, just like that of lin-4, as if they were freed of lin-4 repression. The expression pattern of LIN-14 protein is consistent with the genetic expectations: the level of the Lin-14 protein is maximal early in larval development, then decreases rapidly [6]. Interestingly, the abundance of the lin-14 mRNA does not change over time, implying that the repression of lin-14 expression occurs after transcription [7,8]. The lin-4 RNA is itself temporally regulated: it is off from the start of larval development to near the end of the first larval stage, and fully on by the time lin-14 is repressed (Figure 1) [9]. The two rare gain-of-function alleles of lin-14 were of particular interest because they could suggest how lin-4 represses lin-14. When these mutations were sequenced, they were found to be rearrangements that disrupt much of the 3′ untranslated region (3′ UTR) of the lin-14 RNA [10]. The complementarity

As soon as lin-4 was found to encode a small RNA, the immediate question was how does it repress lin-14 expression through the 3′ UTR of the latter’s RNA? Both the Ambros and Ruvkun groups searched the

Dispatch

R437

Figure 1 The time of expression of the lin-4 and let-7 RNAs, and the phenotypes caused by mutations in them. Each RNA appears at a specific stage during the animal’s larval development. The lin-4 RNA is produced by the end of the first larval stage, and continues thereafter. The let-7 RNA is present by the fourth larval stage. When lin-4 is mutant, the developmental events of the first stage are reiterated in subsequent stages (shown by the repeated red rectangles); the molting cycle that punctuates the stages continues indefinitely. When let-7 is mutant, the hypodermal cell fates are reiterated once after the fourth stage (shown by the repeated blue rectangle). The colored blocks represent the developmental events that are normally specific to each larval stage.

Expression Hatch

lin-4 RNA

let-7 RNA

Phenotype Wild type

lin-4(–)

let-7(–)

L1 L2 L3 L4 Adult

Current Biology

3′ UTR of lin-14 for sequences that might be complementary to the lin-4 RNA — and they found seven of them [1,7]. The model apparent to all involved was clear: the specificity for the action of lin-4 on lin-14 comes from complementarity between the small RNA and sequences in the mRNA of its target (Figure 2). Indeed, the 3′ UTR of lin-14 can confer on a lacZ reporter a time-specific repression that is dependent on both lin-4 and the complementary sequences [11]. Subsequently, another gene predicted by genetic analysis to be repressed by lin-4, lin-28, was also found to have a single complementary sequence in its 3′ UTR (Figure 2) [12]. Repression of lin-28 also fails to occur in a lin-4 mutant, and deleting just 15-nucleotides complementary to the lin-4 RNA from the lin-28 3′ UTR causes the persistence of lin-28 expression and a retarded phenotype. This showed that lin-4 can act on two entirely different genes whose only apparent similarity is having sequences complementary to the little RNA in their 3′ UTRs. So what gene or genes does let-7 regulate? The let-7 gene appears to be important near the end of development (Figure 1) [2]. The let-7 RNA is not fully expressed until the fourth and final larval stage. When let-7 is mutant, the animals have a retarded phenotype: at the fourth molt, certain hypodermal cells repeat their larval cell division pattern instead of differentiating as they should. So the targets of the let-7 RNA are expected to be genes that act near the end of larval development and to be regulated when let-7 is expressed. Unlike the lin-4/lin-14 situation, no gain-of-function mutations have been identified that specifically affect late developmental stages of C. elegans. In an attempt to get around this problem, Reinhart et al. [2] used genetic epistasis analysis to try to identify candidates for let-7

targets. Predicting that a precocious mutation in a gene downstream of let-7 would be epistatic to a retarded let-7 mutation, they combined the let-7 mutation with mutations in each of the genes lin-14, lin-28, lin-41 and lin-42. In each case, they found that there is a mutual partial suppression of both the precocious and retarded phenotypes. These ambiguous results are consistent with let-7 either regulating, or being regulated by, any of the other genes. Because let-7 is expressed well after lin-14 and lin-28 are repressed, however, it should not be necessary to propose that let-7 regulates them. But as lin-41 and lin-42 have phenotypes apparently restricted to late stages, Reinhart et al. [2] propose that these are regulatory targets of let-7. The analogy to lin-4 predicts that let-7 should have some complementarity to sequences in the 3′ UTRs of the RNAs encoded by its target genes. Indeed, Reinhart et al. [2] found stretches of complementarity to let-7 in lin-41 and lin-42 RNAs. The complementarity is discontinuous and requires drawing some bulges and loops, but that was done as well with lin-4 and its target RNAs (Figure 2). Complementary sites to let-7 can also be found in the 3′ UTRs of lin-14, lin-28 and daf-12, though the significance of these is not apparent. Clearly, what was needed was a demonstration that let-7 can act through the 3′ UTR of a gene in a way similar to lin-4. To address whether let-7 might have an effect on the expression of lin-41, Reinhart et al. [2] attached the 3′ UTR of lin-41 to a lacZ reporter. They found that expression of a reporter with the whole lin-41 3′ UTR was repressed in adult animals in the presence of let-7, but not in a let-7 mutant. A reporter lacking a region containing the let-7-complementary sequences was not repressed, even in the presence of let-7. Slack et al. [13] found that expression of a reporter in which the sequence encoding green fluorescent protein (GFP) was fused to

R438

Current Biology Vol 10 No 12

Figure 2 5'

5'

AA U 3' A C A A G A U U U lin-14 3′ UTR G G U C G U A C CA G U U CA AGC A C U G CC G C A lin-4 U A U RNA 5' 3'

U 3' A U G G U C lin-28 3′ UTR G A U C G U A C CA G U U CA AGC A C U G C G C G lin-4 C A U U RNA U 5'

genes. RNA interference, or RNAi, is a gene-silencing phenomenon mediated by double-stranded RNA. Zamore and colleagues [14] have shown that the double-stranded RNA is cleaved into RNAs of 21–23 nucleotides, and these short RNAs lead to specific degradation of mRNA of the complementary sequence. The short RNAs that mediate RNAi and the heterochronic RNAs act by entirely different mechanisms, but both get their specificity from relatively short stretches of complementarity. 3'

The future 5'

5' U 3' U G U A A U U lin-41 3′ UTR A A U C G A U A U C GU G C U G C A G U U A A A U C let-7 G C RNA G U A C G A U 5'

3'

A 3' U U G U A A U U lin-14 3′ UTR A G U C G A A U A A U U C U G C G G U AU A A U C G C let-7 G U RNA A C G A U 5'

3'

Current Biology

Models for base pairing between the lin-4 and let-7 RNAs and their proposed target RNAs. The sequences of the small RNAs have been confirmed, but their target sites are defined mostly by complementarity to the small RNA. There are several ways to draw the possible basepairing, and none of the structures shown has been verified experimentally. Additional complementary sites for lin-4 in lin-14 and lin-41, and for let-7 in lin-14, lin-28, lin-41, lin-42 and daf-12 have been proposed [1,2,7].

the entire lin-41 gene was also repressed late in development. The level of lin-41 RNA over time has not yet been determined, but it seems that, like lin-4, let-7 acts by way of complementarity to sequences in the 3′ UTR of its targets. The mechanism

Assuming both lin-4 and let-7 act by base-pairing of their short RNA products with the 3′ UTRs of their target RNAs, the important question is: how do they affect gene expression? It is quite clear that lin-14 is regulated by the lin-4 RNA at a step after transcription, but exactly which step has been mysterious. Recently, Olsen and Ambros [8] made the fascinating finding that, even when lin-4 is acting, lin-14 mRNA is present in abundance associated with polyribosomes. It looks as if the mRNA is engaged in translation, but no protein accumulates. Tantalizingly similar observations have been made for the regulation of the nanos gene during Drosophila embryogenesis (I. Clark and E. Gavis, personal communication). The lin-4 and let-7 genes are the only ones known to encode small RNAs that specifically regulate other genes. But recently, RNAs of similar size have been found to have a profound effect on gene expression — although in these cases they are not the products of naturally occurring

Are there genes encoding other tiny RNAs out there? The C. elegans lin-4 and let-7 genes were identified by virtue of their mutant phenotypes — lightning struck twice where the phenotypes were clear, even though they present small targets for mutagenesis. In genomic sequences, genes that act via RNAs, rather than proteins, are almost invisible, because they do not have open reading frames. But sophisticated computational methods, as well as comparative genome sequencing efforts, should greatly improve our ability to see them in the expanse of noncoding sequences [15,16]. As for the two heterochronic RNAs, the challenges for the future include finding out whether they act in a similar way. What is really the basis of the specificity in the interaction between each small RNA and its targets? What are the proteins involved? The parallel study of these two RNAs makes the analysis of their action substantially more powerful. Once we make progress on these, we can expect to have a much easier time when the next tiny RNA gene is discovered. Acknowledgements I am grateful to Eric Lambie and Kathy Seggerson for valuable comments on the manuscript and to Frank Slack for communicating unpublished results.

References 1. Lee RC, Feinbaum RL, Ambros V: The C. elegans heterochronic gene lin-4 encodes small RNAs with antisense complementarity to lin-14. Cell 1993, 75:843-854. 2. Reinhart BJ, Slack FJ, Basson M, Pasquinelli AE, Bettinger JC, Rougvie AE, Horvitz HR, Ruvkun G: The 21-nucleotide let-7 RNA regulates developmental timing in Caenorhabditis elegans. Nature 2000, 403:901-906. 3. Slack F, Ruvkun G: Temporal pattern formation by heterochronic genes. Annu Rev Genet 1997, 31:611-634. 4. Ambros V, Horvitz HR: Heterochronic mutants of the nematode Caenorhabditis elegans. Science 1984, 226:409-416. 5. Ambros V: A hierarchy of regulatory genes controls a larva-to-adult developmental switch in C. elegans. Cell 1989, 57:49-57. 6. Ruvkun G, Giusto J: The Caenorhabditis elegans heterochronic gene lin-14 encodes a nuclear protein that forms a temporal developmental switch. Nature 1989, 338:313-319. 7. Wightman B, Ha I, Ruvkun G: Posttranscriptional regulation of the heterochronic gene lin-14 by lin-4 mediates temporal pattern formation in C. elegans. Cell 1993, 75:855-862. 8. Olsen PH, Ambros V: The lin-4 regulatory RNA controls developmental timing in Caenorhabditis elegans by blocking LIN-14 protein synthesis after the initiation of translation. Dev Biol 1999, 216:671-680. 9. Feinbaum R, Ambros V: The timing of lin-4 RNA accumulation controls the timing of postembryonic developmental events in Caenorhabditis elegans. Dev Biol 1999, 210:87-95.

Dispatch

10. Wightman B, Burglin TR, Gatto J, Arasu P, Ruvkun G: Negative regulatory sequences in the lin-14 3′-untranslated region are necessary to generate a temporal switch during Caenorhabditis elegans development. Genes Dev 1991, 5:1813-1824. 11. Ha I, Wightman B, Ruvkun G: A bulged lin-4/lin-14 RNA duplex is sufficient for Caenorhabditis elegans lin-14 temporal gradient formation. Genes Dev 1996, 10:3041-3050. 12. Moss EG, Lee RC, Ambros V: The cold shock domain protein LIN-28 controls developmental timing in C. elegans and is regulated by the lin-4 RNA. Cell 1997, 88:637-646. 13. Slack FJ, Basson M, Liu Z, Ambros V, Horvitz HR, Ruvkun G: The lin-41 RBCC gene acts in the C. elegans heterochronic pathway between the let-7 regulatory RNA and the LIN-29 transcription factor. Mol Cell 2000, 5:659-669. 14. Zamore PD, Tuschl T, Sharp PA, Bartel DP: RNAi: double-stranded RNA directs the ATP-dependent cleavage of mRNA at 21 to 23 nucleotide intervals. Cell 2000, 101:25-33. 15. Eddy SR: Noncoding RNA genes. Curr Opin Genet Dev 1999, 9:695-699. 16. Shabalina SA, Kondrashov AS: Pattern of selective constraint in C. elegans and C. briggsae genomes. Genet Res 1999, 74:23-30.

R439

If you found this dispatch interesting, you might also want to read the April 2000 issue of

Current Opinion in Genetics & Development which included the following reviews, edited by Adrian Bird and Tony Kouzarides, on Chromosomes and expression mechanisms: DNA double strand break repair in mammalian cells Peter Karran Replication and recombination intersect Kenneth J Marians Mismatch repair defects in cancer Josef Jiricny and Minna Nyström-Lahti Sloppier copier DNA polymerases involved in genome repair Myron F Goodman and Brigette Tippin Telomere transitions in yeast: the end of the chromosome as we know it Julia Promisel Cooper Think global, act local – how to regulate S phase from individual replication origins Philippe Pasero and Etienne Schwob Promoter targeting and chromatin remodeling by the SWI/SNF complex Craig L Peterson and Jerry L Workman mRNA stability in eukaryotes Philip Mitchell and David Tollervey Locus control regions and epigenetic chromatin modifiers Richard Festenstein and Dimitris Kioussis The HP1 protein family: getting a grip on chromatin Joel C Eissenberg and Sarah CR Elgin Making noise about silence: repression of repeated genes in animals James A Birchler, Manika Pal Bhadra and Utpal Bhadra DNA methylation, a key regulator of plant development and other processes E Jean Finnegan, W James Peacock and Elizabeth S Dennis Dynamics of DNA methylation pattern Chih-Lin Hsieh The uniqueness of the imprinting mechanism Frank Sleutels, Denise P Barlow and Robert Lyle The full text of Current Opinion in Genetics & Development is in the BioMedNet library at http://BioMedNet.com/cbiology/jgen