Angelo DePalma Ph.D. Writer GEN
Balance Differentiation and Proliferation to Make Stem Cell Culture Commercially Routine
One of the most surprising revelations from genomics, and since validated by the Human Genome Project, was that only 4% of the human genome codes for protein. At one time molecular biologists believed that noncoding genes amounted to little more than genomic noise. Junk DNA was the label used for any DNA in the genome that had no discernible function. It included large areas of the genome that today would still not be described as “genes” but may have a known function. We now understand that noncoding RNAs are essential to development and the everyday functioning of cells.
Over the last decade researchers have uncovered a virtual alphabet soup of small noncoding RNAs: snoRNA, microRNA, siRNA, exRNA, piRNA, gRNA, and other relatively low molecular-weight RNAs that function diversely within cells. In contrast to messenger RNA (mRNA) and transfer RNA (tRNA), short noncoding RNAs guide post-translational modification of proteins and DNA replication, and act as regulatory agents for a wide range of cellular processes.
Regulatory noncoding RNAs fall into two categories based on size. Transcripts shorter than 200 nucleotides, including those mentioned above, control developmental and physiologic pathways implicated in a large number of diseases.
Transcripts of at least 200 nucleotides, long noncoding RNAs (lncRNAs), comprise the majority of the human transcriptome. lncRNAs substantially outnumber small noncoding RNAs, for example at least 7,100 lncRNAs are known compared with slightly more than 3,000 microRNAs. Yet lncRNAs are relative newcomers. As of July 2014, only 127 have been characterized for biological function. Far from being mere genetic noise, lncRNAs are now recognized as functional in a variety of cellular processes,1 and are arguably the hottest area of RNA research today.
Analysis and Characterization
RNA-Seq and microarrays are the leading methods for detecting and characterizing lncRNA. RNA-Seq can be used as a shallow or deep-sequencing method for elucidating the transcriptome at a specific point in time. The method involves converting RNA to a library of cDNA fragments, followed by sequencing with read lengths ranging from 30 to about 400 base pairs. Number of reads obtained per sample depends on experimental design. These reads are mapped back to a reference to identify and count the transcripts present.
Microarray analysis is based on hybridization of cDNA or cRNA within high-density oligonucleotide microarrays. With the ability to address hundreds of thousands of unique transcripts in a single assay, microarray-based methods are high throughput for both sample processing and data analysis, relatively inexpensive, and use instrumentation that is readily accessible to most labs.
Both RNA-Seq and microarrays provide accurate relative measurements of lncRNAs but neither method gives absolute measurements1 because each introduces gene-specific biases, for example due to gene-length or sequence GC context. The biases are consistent, meaning the differences in signal/counts between samples can be determined but an absolute count is not accurate.
During the last few years several studies have pitted RNA-Seq and microarrays in head-to-head comparisons, with the aim of demonstrating superiority of one method over the other.
For several years, it was predicted that RNA-Seq would replace microarrays due to its greater flexibility. Investigators pointed out for example that RNA-Seq is sequence-naïve, whereas with microarray detection sequences must be known beforehand. Now, there is a better understanding of the impact of RNA-Seq read depth on experimental information content. This, combined with factors such as throughput, cost, sensitivity, and specificity, indicates that one technique or the other may provide more meaningful results depending on the context or the aim of the experiment.
In fact today’s consensus treats the two methods as complimentary. A recent study2 published by the Sequencing Quality Control Consortium concluded that: “…efficient transcript-specific measurements with good precision on microarrays for quantitative expression profiling could complement the power of RNA-Seq in the discovery and identification of new alternative transcripts. In other words, the novel transcripts found by RNA-Seq can lead to efficient measurements with good precision on microarrays, which can, in turn, aid in the confirmation and functional study of new transcript variants.”
Quick Win/Fast Fail
The strong association between lncRNA and disease, particularly cancer, raises the possibility for targeting these molecules, via knockdowns, through short-hairpin RNA (shRNA) or siRNA.
Therapeutics based on lncRNA are years or decades away. Much more immediate is the potential of lncRNA as biomarkers for disease states (diagnosis), predicting the course of a disease (prognosis), or the efficacy of treatment (companion diagnostic), for example during clinical trials.
lncRNA activity has been associated with a wide diversity of disorders including diabetes, pain, neurologic diseases, muscular dystrophy, coronary artery disease, autoimmune disease, and rare diseases such as Prader-Willi syndrome, Down’s syndrome, fragile X syndrome, and others. Thus the potential for lncRNA-based diagnostic and prognostic tests, both in therapeutic and clinical development settings, is enormous.
In part due to their dysregulation in cancer, some scientists believe lncRNA to be the “master drivers of carcinogenesis.”3 For example lncRNA PVT1, which is upregulated in non-small cell lung cancer compared with adjacent normal tissue, significantly correlates with histologic grade and metastasis,4 and is an independent marker for survival in lung cancer patients. In vitro experiments suggest that knockdown of PVT1 inhibits cell proliferation, migration, and invasion.
lncRNA disease specificity suggests a significant role for these molecules as biomarkers for a drug’s effectiveness during clinical studies, particularly as part of “quick win, fast fail” strategies.
Development-stage drugs fail for a variety of technical (as opposed to business) reasons, including poor oral bioavailability or pharmacokinetics, toxicity, or lack of efficacy. All quick win, fast fail approaches involve establishing a set of criteria defining proof of concept (POC) that minimizes risk for molecules promoted beyond Phase I.5 Thus more molecules fail early and fewer enter mid- to late-stage clinical testing, but those that progress have a higher likelihood of approval. Avoidance of expensive late-stage trials spares resources that may be applied to discovery and preclinical development.
The availability of robust biomarkers that compliment clinical endpoints is essential for determining POC, particularly during Phase I. Prognostic biomarkers that antecede clinical endpoints are especially useful, particularly for oncology drugs that require multiple doses or prolonged administration to take effect. Even in situations where concrete clinical endpoints are available (e.g., blood pressure or lipid lowering), biomarkers related to safety and efficacy will be essential to arriving at go/no-go decisions.
In a preclinical setting, lncRNA may serve the normal functions of biomarkers, namely to determine mode of action, proof of concept, toxicity, and potency. In vivo, they fulfill the roles of exploratory biomarkers and indicators of efficacy, safety/toxicity, enrollment suitability, and as companion diagnostics.6 Similar utility is available in Phases II and III. Of particular interest to quick win, fast fail is indication of efficacy. Post-marketing, lncRNA biomarkers will add to the diagnostic toolbox required for personalized medicine.7 While not yet reality, utilization of lncRNA testing as companion diagnostics will one day become routine during all clinical testing phases as well as post-marketing.
Quick Win/Fast Fail
The strong association between lncRNA and disease, particularly cancer, raises the possibility for targeting these molecules, via knockdowns, through short-hairpin RNA (shRNA) or siRNA.
Therapeutics based on lncRNA are years or decades away. Much more immediate is the potential of lncRNA as biomarkers for disease states (diagnosis), predicting the course of a disease (prognosis), or the efficacy of treatment (companion diagnostic), for example during clinical trials.
lncRNA activity has been associated with a wide diversity of disorders including diabetes, pain, neurologic diseases, muscular dystrophy, coronary artery disease, autoimmune disease, and rare diseases such as Prader-Willi syndrome, Down’s syndrome, fragile X syndrome, and others. Thus the potential for lncRNA-based diagnostic and prognostic tests, both in therapeutic and clinical development settings, is enormous.
In part due to their dysregulation in cancer, some scientists believe lncRNA to be the “master drivers of carcinogenesis.”3 For example lncRNA PVT1, which is upregulated in non-small cell lung cancer compared with adjacent normal tissue, significantly correlates with histologic grade and metastasis,4 and is an independent marker for survival in lung cancer patients. In vitro experiments suggest that knockdown of PVT1 inhibits cell proliferation, migration, and invasion.
lncRNA disease specificity suggests a significant role for these molecules as biomarkers for a drug’s effectiveness during clinical studies, particularly as part of “quick win, fast fail” strategies.
Development-stage drugs fail for a variety of technical (as opposed to business) reasons, including poor oral bioavailability or pharmacokinetics, toxicity, or lack of efficacy. All quick win, fast fail approaches involve establishing a set of criteria defining proof of concept (POC) that minimizes risk for molecules promoted beyond Phase I.5 Thus more molecules fail early and fewer enter mid- to late-stage clinical testing, but those that progress have a higher likelihood of approval. Avoidance of expensive late-stage trials spares resources that may be applied to discovery and preclinical development.
The availability of robust biomarkers that compliment clinical endpoints is essential for determining POC, particularly during Phase I. Prognostic biomarkers that antecede clinical endpoints are especially useful, particularly for oncology drugs that require multiple doses or prolonged administration to take effect. Even in situations where concrete clinical endpoints are available (e.g., blood pressure or lipid lowering), biomarkers related to safety and efficacy will be essential to arriving at go/no-go decisions.
In a preclinical setting, lncRNA may serve the normal functions of biomarkers, namely to determine mode of action, proof of concept, toxicity, and potency. In vivo, they fulfill the roles of exploratory biomarkers and indicators of efficacy, safety/toxicity, enrollment suitability, and as companion diagnostics.6 Similar utility is available in Phases II and III. Of particular interest to quick win, fast fail is indication of efficacy. Post-marketing, lncRNA biomarkers will add to the diagnostic toolbox required for personalized medicine.7 While not yet reality, utilization of lncRNA testing as companion diagnostics will one day become routine during all clinical testing phases as well as post-marketing.
First Success: Prostate Cancer
Prostate cancer, the sixth-leading cause of death in men worldwide, lacks sufficiently reliable, robust diagnostics for detecting and patient monitoring. Prostate-specific antigen (PSA) testing, while specific for prostate tissue is not cancer specific, thereby resulting in significant over-diagnosis, while biopsy and imaging approaches lead to false negatives. There is also great need for markers distinguishing between indolent and aggressive forms of this malignancy.
PCA3, a lncRNA, is overexpressed in prostate cancer cells compared with normal prostate tissue by a factor of 66, and not expressed at all in nonprostate cells. The U.S. Food and Drug Administration approved a commercial PCA3 test, PROGENSA® PCA3 Assay (Gen-Probe) in 2012. The test is indicated in men with elevated PSA and negative biopsies.
Since the approval of PROGENSA many new ncRNAs associated with prostate cancer have been discovered.8 These include six lncRNAs that specifically overexpress in prostate malignancies, and seven microRNAs that are either up- or downregulated in prostate and other cancers. While the PCA3 test is diagnostic, attempts to use it as a prognostic or predictive tool have failed. Panels of noncoding microRNAs appear to be better suited for this purpose, as well as for detecting metastases, but so far none have reached clinical testing.
Conclusion
The interest generated by the relatively small number of lncRNAs that have been characterized suggests that future studies will reveal even more relevant associations between lncRNA activity and human health and disease. With fewer than 2% of known lncRNA investigated, and given the success thus far, there is every reason for optimism. While it is still too early to rely on lncRNA as primary or independent diagnostics, scientists already use these molecules to enhance the diagnostic, prognostic, and disease-monitoring capabilities of existing biomarkers.
For more information on lncRNA go to http://www.nxtbook.com/nxtbooks/gen/affymetrix/.
References
1 Front Genet. 2014; 5: 379.
2 Nature Biotechnology 32, 903–914 (2014)
3 Cancer Discov. 2011 Oct;1(5):391-407.
4 Int J Clin Exp Pathol. 2014 Sep 15;7(10):6929-35.
5 Nature Reviews Drug Discovery 9, 203-214 (March 2010)
6 Public Workshop — Clinical Flow Cytometry in Hematologic Malignancies, Silver Springs, MD. February 25-26, 2013.
7 Cellular Oncology, August 2014
8 Biomed Res Int. 2014; 2014: 591703. Accessed at http://goo.gl/rdO74J.
