MODIFIED FROM ORIGINAL PHOTO. © 2009, THE ANN ARBOR NEWS. ALL RIGHTS RESERVED. REPRINTED WITH PERMISSION. fter getting the tion by as much as 90%.1 Interestingly, the other genes could also switch on gene data back from the findings confirmed work by Kevin Morris, mkZgl\kbimbhg'Ebd^
. went against the
N nucleic acids, the RNAs targeted to the oth groups had trouble pub- O same promoter also silenced gene expres- grain of everything lishing their papers. Corey
MISSI lbhgZmma^e^o^eh_mkZgl\kbimbhg'ÊPa^g Zg] CZghpldbÍl phkd pZl
PER [Ram] saw the silencing, she thought we knew about rejected by Science before it was published she had done something wrong,” says in Nature Chemical Biology bg CZgnZkr WITH CZghpldb'ÊLa^]b]gÍmpZgmmhlahpf^ma^ how small RNAs 2007,3pabe^EbZg]IeZ\^[Zmme^]_hkmph ED data because she thought it was supposed years and faced four rejections before INT regulate gene to be a negative result.” they finally published their paper in the The data just didn’t make sense: Sin- expression.” Proceedings of the National Academy of
ED. REPR 2
V gle-stranded peptide nucleic acids bind Sciences in November 2006. So, when ER
S directly to unwound DNA at the transcrip- —John Rossi . A well characterized process known as RNA itively that the researchers had observed saw a phenomenon that was similar to WS interference, or RNAi. So how could they `^g^Z\mboZmbhg%Ê[nmbmieZgm^]ma^l^^]l ours,” adds Place. both be causing the same effect? With in our minds that maybe activation could However, neither group offered a mablhg^Ög]bg`%ÊZeemahl^mabg`lmaZmrhn be occurring,” says Corey. To investigate the plausible mechanism for how RNA acti- ARBOR NE thought you could predict just flew out the trend further, the researchers switched to a oZmbhgfb`amh\\nk'ÊB]b]gÍmmabgdmaZm NN pbg]hp%ËCZghpldblZrl' cell line with much lower background activ- ma^ ^ob]^g\^ maZm TEb Zg] IeZ\^V aZ] E A H ÊBmmhhdfhgmal[^_hk^p^\hgobg\^] ity levels of their gene of interest, the human really supported the mechanism they were ourselves that the results were real,” says progesterone receptor, and the data became k^ihkmbg`%ËlZrlChagKhllb%Zfhe^\neZk 2009, T Corey. But eventually they did. They tested glaringly obvious: The same RNAs that did geneticist at City of Hope Comprehensive © several different double-stranded RNAs— not inhibit transcription could now trigger Cancer Center in Duarte, Calif., who was O. T O each 21 nucleotides long, just like standard, as much as 10- to 20-fold increases in gene Zk^ob^p^k_hkma^iZi^k'ÊBmp^gmZ`Zbglm H P small interfering RNAs used in RNAi. All expression. The activation effect was real. the grain of everything we knew about of these RNAs perfectly matched regions Before they could publish the findings, how small RNAs regulate gene expression. IGINAL of the DNA promoter but had little to no however, another lab published results The data were clear, but were they looking OR 2 M overlap with the gene’s mRNA, to ensure showing essentially the same effect. A at something indirect? We wanted more the RNAs were acting on transcription, m^Zfe^][rEhg`& ED FRO not translation. Ultimately, in September Place at the University of California, San Zg]ma^rTEbZg]IeZ\^Vg^o^k]b]ma^f'Ë I F I +)). MOD introduced RNAs could inhibit transcrip- targeting the DNA promoters of three ÊBaZo^ZaZk]mbf^mh^qieZbgma^Ög]bg`l% Ω May 2009 THE SCIENTIST 35 RNA and what mechanism can explain it,” says Argonaute proteins—known members the same direction as mRNA transcription Thomas Tuschl, an RNA expert at New of the RNAi pathway—help facilitiate Zlp^eeZlbgma^hiihlbm^ÉhkÊZgmbl^gl^ËÉ York’s Rockefeller University, in an email. RNA-RNA interactions. Thus, the new- orientation. These ubiquitous noncoding ÊBml^^fllhb]bhlrg\kZmb\Zg]li^\bZeZm found connection between transcrip- RNAs probably served some purpose, the moment,” adds Timothy Nilsen, an tional regulation and Argonaute proteins perhaps even in RNA activation, the sci- RNAi researcher at Case Western Reserve indicated that RNA activation’s target entists reasoned. Ngbo^klbmrbg 36 THE SCIENTIST May 2009 coding mRNA An active model: non-coding RNA Three views of RNA activation introduced RNA genomic DNA Argonaute promoter COREY & JANOWSKI The introduced RNA binds an antisense RNA transcript and recruits Argonaute to the target antisense transcript gene promoter to activate transcription. promoter MORRIS The introduced RNA targets a long antisense RNA transcript and together with Argonaute cuts up the RNA leading to an increase in gene transcription. antisense transcript sense transcript LI & PLACE The introduced RNA binds tiny non-coding promoter sense RNA transcripts in the promoter region, recruits Argonaute and activates gene expression. r^Zk[^_hk^ May 2009 THE SCIENTIST 37 RNA of interest with the green-fluorescence protein gene. This allowed them to quan- A natural turn-on tify levels of transcription and pinpoint the target site for RNA activation. After mutating the promoter DNA and looking iny snippets of genetic code called microRNAs (miRNAs) regularly whir at changes in the intensity of the green around the cell, fine-tuning gene expression. More than 400 miRNAs glow, they showed that their activating T have been identified in the human genome, and each one can regulate RNAs were homing in on the promoter hundreds of different genes, says David Bartel, an RNA biologist at the White- region itself, similar to what Corey and head Institute in Cambridge, Mass. Like most noncoding RNAs, these small, CZghpldbikhihl^%Zg]ghmZ]hpglmk^Zf cellular supervisors were long thought only to repress genes by interfering with target, as Morris argues. However, Place messenger RNA and reducing protein production. But recent research shows that believes that his activating RNAs are miRNAs are much more versatile. actually targeting tiny, noncoding RNA Last year, UCSF’s Robert Place and Long-Cheng Li followed up on their transcripts in the promoter region that initial observation of RNA activation by scanning the gene promoter region code in the same direction as mRNA, not for natural miRNA target sites. They found one matching the miRNA-373 and Zgmbl^gl^KG:'ÊHnkZ\mboZmbg`KG:lZk^ showed that this endogenous RNA could significantly upregulate transcription not targeting antisense transcripts and in the same way as synthetic RNA (Proc Natl Acad Sci, 105:1608–13, 2008). p^Ík^`^mmbg`Z\mboZmbhg%ËIeZ\^lZrl'ÊMaZm “MicroRNAs had only been considered to suppress gene activity,” says Place. offsets their [Corey and Morris’s] models This study was “proof of principle that microRNAs can have the opposite right away.” effect and can lead to gene activation.” Manl%EbZg]IeZ\^Zk^phkdbg`pbma Other researchers are also seeing the same effects. For example, Raghu a third, as yet unpublished explanation Vemuganti, a neuroscientist at the University of Wisconsin–Madison, profiled for RNA activation. And they’re making miRNAs in rat brains and discovered that several appeared to activate transcrip- ma^fhlmh_bm'ÊP^Ío^]^lb`g^]Zik^mmr tion (J Cereb Blood Flow Metab, 29:675–87, 2009). “Bioinformatics-wise, I can standard set of rules for acquiring acti- already say confidently that this is happening in the brain,” says Vemuganti. vating, double-stranded RNA that we can “Experimentally, we still need to prove that.” now apply to any promoter that we want,” These weren’t the first examples of naturally occurring RNAs that could get Place says. To date, his team has activated transcription revved up, however. In 2004, Fred Gage, a neuroscientist at the around 15 to 20 genes, often with several Salk Institute in La Jolla, Calif., found a small, double-stranded RNA in rat neural different RNAs that all work for the same stem cells that interacted with proteins to activate genes important for neuron gene, he adds. function (Cell, 116:779–93, 2004). “There’s lots and lots of RNA that exists in the Importantly, however, all three models nucleus,” says Gage. “Now we think of it as junk or background noise, but it may of transcriptional activation share one not be. There’s enormous room here for mischief given the fact that there’s this key feature in common. They all zero in plethora of RNA that’s there.” on the same central target of transcrip- What’s more, endogenous miRNAs can also stimulate gene expression mbhgZek^`neZmbhg3ghg\h]bg`KG:'EZlm at the post-transcriptional level, after the mRNA has been produced. Yale December, a quartet of papers appeared University molecular biologists Joan Steitz and Shobha Vasudevan discovered in Science showing that more than 90% that miRNAs can switch from turning protein production on and off depending of the genome is transcribed—forward, on the activity of the cell as a whole: miRNAs repressed translation in prolif- backward, and all around—and that the erating or cycling cells, but activated translation in quiescent cells, they found promoters of active genes are not, in effect, (Science, 318:1931–34, 2007). Êjnb^m'ËMaZmf^Zglma^k^ÍlZehmh_KG:mh The same basic silencing components—including Argonaute and RNA go around, possibly with many different binding proteins—were involved in both translational activation and repression, hidden regulatory effects. suggesting, yet again, that gene activation and gene silencing might simply be ÊBm phne] [^ lahkmlb`am^] mh mabgd two sides of the same coin. “Maybe the associations are different, but the basic that there’s only one pathway to anything Argonautes and the basic microRNAs are the same in activation and repression,” we’re seeing; redundancy is built into the says Vasudevan. “It looks like microRNA can do a lot of other things and the system,” says Morris. On this point, all mechanism is still mysterious.” ma^k^l^Zk\a^kl\ZgZ`k^^'ÊBmehhdlebd^ there are many forms by which you can induce RNA activation,” says Place. In the meantime, even without an agreed- upon explanation, the phenomenon is k^Ze%lZrl 38 THE SCIENTIST May 2009 work for opening up more questions than munity have now warmed up to the idea of Gingeras, head of functional genomics at it answers, but another way of looking at KG:Z\mboZmbhg'Ê:m_Z\^oZen^%ma^k^ÍlZo^kr Oct4 and Nanog, with an eye to reprogram- Northwestern University in Evanston, References ming adult cells to an embryonic-like state. Bee'%Z`k^^l'ÊBgm^kflh_paZmlfZeeKG:l *';':'CZghpldb^mZe'%ÊBgab[bmbg``^g^^qik^llbhgZm Morris is now working together with are actually doing, it’s not all that fun- transcription start sites in chromosomal DNA with the Scripps tech-transfer office to com- damentally new—it’s still RNAi,” he says. antigene RNAs,” Nat Chem Biol, 1:216–22, 2005. mercialize his approaches, while Corey Corey, too, now concedes the point. +'E'<'Eb^mZe'%ÊLfZee]lKG:lbg]n\^mkZgl\kbimbhgZe activation in human cells,” Proc Natl Acad Sci, and Place’s teams both licensed their ÊPa^gp^ÖklmlmZkm^]]hbg`mabli^hie^ 103:17337–42, 2006. technologies to Alnylam Pharmaceuticals, didn’t know about these noncoding RNA ,';':'CZghpldb^mZe'%Ê:\mboZmbg``^g^^qik^llbhgbg a Cambridge, Mass.–based company spe- transcripts, and now they do,” he says. mammalian cells with promoter-targeted duplex \bZebsbg`bgKG:b&[Zl^]ma^kZi^nmb\l'ÊBmÍl ÊLhbmÍlcnlmZghma^k[kZg\ah_ma^KG:b RNAs,” Nat Chem Biol, 3:166–73, 2007. a very logical extension for us, but obvi- pathway. The RNAi pathway has many -'C'AZg^mZe'%ÊIkhfhm^k&Zllh\bZm^]KG:blk^jnbk^] for RNA-directed transcriptional gene silencing in ously it could represent a whole new plat- fingers reaching into many pies.” human cells,” Proc Natl Acad Sci, 104:12422–27, _hkf%ËlZrl:egreZfÍl\ab^_^q^\nmbo^Chag Lmbee%ghm^o^krhg^bl\hgobg\^]'ÊMa^ 2007. FZkZ`Zghk^'ÊBmÍllmbee^oheobg`bgm^kflh_ whole effect of RNA on transcription in .'C'<'L\apZkms^mZe'%Ê:gmbl^gl^mkZgl\kbimlZk^ how reliable and predictive the observa- mammals is viewed with a bit of skepticism targets for activating small RNAs,” Nat Struct tions are, but there certainly are a number by the whole field,” says Nilsen. Tuschl is Molec Biol, 15:842–48, 2008. /'D'O'Fhkkbl^mZe'%Ê;b]bk^\mbhgZemkZgl\kbimbhg of gene targets we’ve been able to activate.” Zghma^khg^h_ma^ahe]hnml3ÊBmg^^]lfhk^ directs both transcriptional gene activation With a few working models to toss thorough biochemistry, and more genes to and suppression in human cells,” PLoS Genet, around, some researchers in the RNA com- be targeted before I get sold on it.” Thomas 4:e1000258, 2008. May 2009 THE SCIENTIST 39