Caitlin Smith has a B.A. in biology from Reed College, a Ph.D. in neuroscience from Yale University, and completed postdoctoral work at the Vollum Institute.
Long noncoding RNAs (lncRNAs) are mysterious molecules. They are almost-proteins—transcripts of about 200 or more nucleotides that appear not to encode proteins. Given their noncoding status, it is perhaps surprising that many lncRNAs are expressed in very specific anatomical or developmental patterns, suggesting that their regulation is of biological importance. At the cellular level, most lncRNAs, also called large intergenic noncoding RNAs (lincRNAs), are localized to the nucleus. In addition, most lncRNAs either overlap with genes that encode proteins, are transcribed antisense to genes that encode proteins or are expressed as intergenic or intronic regions.
But why spend energy tightly regulating the expression and localization of RNA molecules that don’t eventually end up as proteins? And what do they actually do? Recent advances in RNA sequencing are providing a look at these mysterious molecules, which can be perused in new databases such as the Human Body Map lincRNAs catalog developed by the Broad Institute, Harvard University and the Massachusetts Institute of Technology. Realizing the many lncRNAs out there whose functions we have yet to learn, it appears today’s knowledge is just the tip of the lncRNA iceberg.
Tools of the trade
One of the main challenges facing researchers in this field is that the tools are still evolving. An active area of tool development is creating microarrays with probes for different lncRNAs. For example, Agilent Technologies’ SurePrint G3 Gene Expression v2 microarray includes probes designed to the lincRNA and transcripts of uncertain coding potential (TUCP) catalogs developed by the Broad Institute, along with mRNA from the key public databases. According to Corinna Nunn, product manager for gene expression at Agilent Technologies, lncRNAs can have lower expression than coding RNAs, “so a microarray with wide dynamic range is important to ensure detection of these lower expressing genes.”
Affymetrix’s GeneChip® Gene 2.0 ST Arrays cover both coding mRNA and lincRNA on the same array and use transcripts from a combination of sources, including RefSeq, Ensemble and the Broad Institute’s Human Body Map lincRNAs and TUCP catalog. “Our ultra high-density arrays have unique probes that interrogate multiple loci on every exon of every transcript, ensuring that the most biologically relevant information can be captured,” says Tsetska Takova, director of strategic product marketing, gene expression arrays, reagents and instruments at Affymetrix.
Arraystar, which has been designing lncRNA microarrays since 2009, just announced its third-generation microarray for lncRNAs, the Arraystar Human LncRNA Microarray V3.0. This array profiles both mRNAs and lncRNAs. Ed Davis, technical support specialist at Arraystar, says that even though expression profiling is usually the first step in lncRNA research, developing expression microarrays for lncRNA remains challenging. One complication is that there are multiple databases of lncRNA. “To get around this problem,” Davis said, “we’ve created one comprehensive and reliable lncRNA database by pooling information from several public databases and landmark lncRNA publications.” Designing lncRNA probes is also difficult, because lncRNAs often overlap with nearby protein-coding genes. In addition, Davis says, “microarray probes are traditionally ‘gene-specific’; in other words, they’re targeted to the 3’ end of the transcript, and so cannot distinguish between multiple transcript isoforms resulting from alternative transcription start sites or differential exon splicing.” Davis adds that, instead, “Arraystar’s probes are ‘transcript-specific.’ By targeting the probes to exons and splice junctions, we can detect multiple transcript isoforms resulting from a single gene.”
Determining function and code
Even more mysterious than profiling lncRNAs is the task of elucidating their biological functions. “Unlike microRNAs,” says Davis, “lncRNAs function via diverse mechanisms, and it is very hard to deduce their functions according to the sequence.” In fact, “less than 1% of lncRNAs have been associated with a function,” says Takova. “However, increasingly, recent research is showing not only their role in tumorigenesis, but also in normal cell differentiation. With a deeper understanding of the roles of ncRNA in normal and diseases states, it is being hypothesized that ncRNAs may be used as diagnostic or predictive biomarkers.” [1]
How exactly is this accomplished? Davis says that in addition to expression profiling, investigating the protein-coding genes associated with particular lncRNAs has yielded valuable functional information. Chris Streck, market manager for RNA and epigenetics at Illumina, also believes our newfound ability to look at the entire transcriptome, instead of the coding RNA alone, is valuable. “Researchers are focusing on more integrative approaches, looking at multiple aspects, including RNA, DNA and regulatory functions,” he says. “This is driven home by the ENCODE [Encyclopedia of DNA Elements] project, including the role of noncoding RNA. We are seeing a high level of conversion from mRNA to whole transcriptome driven by activities like ENCODE, which, along with other disease-specific studies, has highlighted the importance of these noncoding RNAs.” The ENCODE project’s goal is to “identify all regions of transcription, transcription factor association, chromatin structure and histone modification in the human genome sequence,” according to the project’s website.
Ewan Gibb, a postdoctoral researcher in the lab of Wan Lam at the BC Cancer Agency, Integrative Oncology, is currently studying the unfolding roles of lncRNA in disease—at least, based on the little we currently know about lncRNAs. “Despite the fact that many lncRNAs have been genomically annotated, functional evidence is missing for the vast majority,” says Gibb. “While I am confident there are certainly thousands of functional lncRNAs, it is still important to remember that pervasive transcription does not necessarily support pervasive function. Studies which go beyond transcriptional profiling to elucidate the lncRNA function are crucial to overcoming this hurdle.” As an example of work in this vein, he cites a group from the department of biology and biotechnology at Sapienza University of Rome, who found that a lncRNA controls muscle differentiation by functioning as a competing endogenous RNA. [2] An additional challenge, says Gibb, is figuring out the structure-function relationship of lncRNA. “Current work suggests lncRNAs have domain-based structures, much like protein-coding genes,” he says, “but precisely how these structures are organized, or how they equate to lncRNA function is an open question.”
Roles in cancer and other diseases
One function is becoming quite clear— lncRNAs play a powerful role in disease. How do lncRNAs exert their potent influence over disease processes? Indeed, as Anne Bergstrom Lucas, senior research scientist at Agilent Technologies, says, when researchers at the Broad Institute knocked down particular lncRNAs associated with p53, it affected the expression of hundreds of genes that p53 normally repressed. [3] “The idea that lncRNAs play a direct role in tumor-suppressor or oncogenic pathways paves the way for identifying new novel cancer therapeutics,” she says. The latest contributions to the ENCODE project, says Bergstrom Lucas, also “demonstrate that nearly 80% of the noncoding genome is playing a role in gene regulation.”
Gibb is researching how to identify and characterize the functions of lncRNAs that are aberrantly expressed in human cancer. “Typically, identification of lncRNAs is achieved through computational analysis of deep-sequencing datasets, while elucidating the function of these unique genes relies on classical biochemistry methods,” says Gibb. “A deeper understanding of lncRNA function in human cancer will not only expand the number of potential target cancer genes, but may also facilitate development of novel anti-cancer therapies, such as gene regulation mediated by antisense RNAs or targeting lncRNA-protein interactions.”
Many examples point to lncRNAs not only as regulators, but also as biomarkers for diseases—such as the lncRNA known as HOTAIR. [4] “HOTAIR is increased in expression in primary breast tumors and metastases, and its expression level in primary tumors is a powerful predictor of eventual metastasis and death,” says Arraystar’s Davis. Moving closer to clinical applications, Davis cites another lncRNA called prostate cancer antigen 3 (PCA3), which is highly overexpressed in prostate cancer. It also happens to be found in urine, which makes for easy testing. “A commercial kit, called the Progensa PCA3 test, which is the first urine-based molecular test to help determine a need for repeat prostate biopsies, has been approved for clinical application by the FDA recently,” he says. Arraystar’s LncRNA Microarray also is playing a role in investigating lncRNA in hepatocellular carcinoma, in which a lncRNA called lncRNA-HEIH is highly expressed. [5]
The disease-regulating importance of lncRNAs is not limited to cancer. They also play important roles in heritable conditions, notes Gibb, in which lncRNA deregulation has been associated with brachydactyly and HELLP syndrome. Davis also points to an example in Alzheimer’s research, in which a lncRNA was shown to stabilize the mRNA for a crucial enzyme in the Alzheimer’s disease pathway. Can lncRNAs help us learn something about disease regulation that their coding counterparts have not yet shown us? “In my opinion, the most exciting advance is that increasing evidence suggest[s] lncRNAs are closely associated with major human diseases, and may have better performance in disease diagnosis and prognosis compared with protein-coding RNAs,” says Davis. Hopefully, learning more about the mysterious world of noncoding RNAs will help arm us with disease-fighting weapons for years to come.
References
[1] Mitra, et al., “A central role for long non-coding RNA in cancer,” Frontiers in Genetics, 3:17, 2012.
[2] Cesana, et al., “A long noncoding RNA controls muscle differentiation by functioning as a competing endogenous RNA,” Cell, 147:358-369, 2011.
[3] Huarte, et al., “A large intergenic noncoding RNA induced by p53 mediates global gene repression in the p53 response,” Cell, 142:409-419, 2010.
[4] Gupta, et al., “Long non-coding RNA HOTAIR reprograms chromatin state to promote cancer metastasis,” Nature, 464:1071-1076, 2010.
[5] Yang, et al., “Long noncoding RNA high expression in hepatocellular carcinoma facilitates tumor growth through enhancer of zeste homolog 2 in humans,” Hepatology, 54:1679-1689, 2011.
The image at the top of the page is from ArrayStar.