February 15, 2016 (Vol. 36, No. 4)
Richard A. A. Stein M.D., Ph.D.
MicroRNAs Are More Plentiful and More Subtle In Action Than Was Once Suspected
One of the unexpected findings of the Human Genome Project was that over 98% of the human genome does not encode for proteins. Once dismissed as “junk” genomic material, non-protein-coding DNA is now appraised more highly. Or to be more precise, at least some portions of non-protein-coding DNA are thought to serve important biological functions.
For example, some stretches of DNA give rise to a noncoding but still functional kind of RNA called microRNA. MicroRNAs have increasingly emerged in recent years as key regulators of biological processes and pathways.
During the years since their discovery, a key question in the biology of microRNAs has focused on the number of microRNAs encoded in the genome. Between 1993 and 2015, approximately 1,900 human genome loci were discovered to produce microRNAs and were added to miRBbase, the public database that catalogues and annotates microRNA molecules.
The cataloguing of microRNAs work has been pursued with extra urgency since 2004, the year the connection between microRNAs and human disease was first demonstrated. “When this connection was made, it launched a whole new field,” says Isidore Rigoutsos, Ph.D., professor of pathology, anatomy, and cell biology and director of the Computational Medicine Center at Thomas Jefferson University.
Another Set of MicroRNAs Emerges
“We wanted to know how many microRNA-producing loci really exist in humans,” recalls Dr. Rigoutsos. In a study published in 2015, Dr. Rigoutsos and colleagues analyzed datasets from 1,323 individuals that represented 13 different tissues and identified an additional 3,356 such genomic loci that produce (at least) 3,707 novel microRNAs.
“We basically tripled the number of locations in the human genome that are now known to encode microRNAs,” asserts Dr. Rigoutsos. Considering that each microRNA regulates up to hundreds of different mRNAs, and that each mRNA is regulated by tens of microRNAs, this finding adds a new layer of complexity to the regulatory dynamics of the human transcriptome.
The newly unveiled microRNAs and previously characterized microRNAs have distinct expression patterns. While 50–60% of the microRNAs previously deposited into the miRBase are expressed in multiple tissues, only about 10% of the newly discovered microRNAs are shared across multiple tissue types. Also, most of the newly found microRNAs show tissue-specific expression.
Using Argonaute CLIP-seq data, Dr. Rigoutsos and colleagues showed that similar percentages of the two sets of microRNAs were in complex with Argonaute proteins. “This shows that these novel microRNAs participate in RNA interference just as frequently as the miRBase microRNAs,” contends Dr. Rigoutsos.
In a comparative analysis between the human microRNA datasets and the chimpanzee, gorilla, orangutan, macaque, mouse, fruit fly, and mouse genomes, Dr. Rigoutsos and colleagues discovered that almost 95% of the newly unveiled microRNAs were primate-specific, and over 56% of them were found only in humans.
“We are seeing many human microRNAs that do not exist in the mouse,” states Dr. Rigoutsos. “This means that the mouse models engineered to capture human disease cannot recapitulate the interactions mediated by these microRNAs.”
Interest in IsomiRs Grows
In the years since the biology of microRNAs started receiving increasing attention, the conventional view has been that one microRNA locus generates one microRNA. However, once deep sequencing became widely available, microRNA variants that showed differences at their 5′- or 3′-termini have been described.
“It was initially presumed that these variants were likely the result of the enzyme Dicer not being sufficiently accurate when processing microRNA precursors,” notes Dr. Rigoutsos. Subsequent research revealed that microRNAs are more dynamic than previously thought, with each precursor being able to generate multiple mature microRNA species known as isomiRs.
To gain insight into the biology of isomiRs, Dr. Rigoutsos and colleagues analyzed genomic datasets from 452 individuals participating in the 1000 Genomes Project. The datasets comprised five different populations and two races. In addition, each population was represented by an even number of men and women.
This collection allowed the abundance of microRNA isoforms to be examined with respect to population, gender, and race. “We found that isomiRs have expression profiles that are population-, race-, and gender-dependent,” informs Dr. Rigoutsos.
All the transcriptome data that this analysis was based on came from immortalized B cells. “These are cells that normally are not associated with gender differences, but molecularly we found, in these cells, differences between men and women of the same population and race,” explains Dr. Rigoutsos.
Expanding these observations to disease states, Dr. Rigoutsos and colleagues collected isomiR profiles from tissue affected by breast cancer, and compared them with isomiR profiles from control breast tissue. The investigators found that the isomiR profiles also depend on tissue state (healthy vs. diseased), on disease subtype, and on the patient’s race.
For example, their analysis identified several miR-183-5p isoforms that were upregulated in white triple-negative breast cancer patients compared to control breast samples, but not in black/African-American triple-negative breast cancer patients. In an in vitro phase of this study, three isoforms of this microRNA species were overexpressed in human breast cancer cell lines.
“We found very little overlap in the gene sets that were affected by each of these isoforms,” emphasizes Dr. Rigoutsos. Despite being generated simultaneously by the same locus, each of the three isoforms affected distinct groups of genes, thus exerting different effects on the transcriptome.
“As the relative abundance of these isoforms changes ever so slightly from patient to patient, it will affect the corresponding gene groups slightly differently,” concludes Dr. Rigoutsos. “In the process, it creates a new molecular background in each patient.”
MicroRNAs Point to Therapeutic Strategies against Colorectal Cancer
“We are using microRNAs as modulators to overcome chemotherapy resistance in colorectal cancer,” says Jingfang Ju, Ph.D., associate professor of pathology and co-director of translational research at Stony Brook University School of Medicine. Resistance to chemotherapy is one of the major challenges in the clinical management of malignancies, including colorectal cancer. Chemotherapy is usually unable to eliminate cancer stem cells, which may become even more resistant over time, and several microRNAs have been implicated in this process.
“We reasoned that we could provide new modulatory approaches to target this small cell population and allow chemotherapy, radiotherapy, or immunotherapy to eliminate resistant populations or at least prolong long-term survival,” explains Dr. Ju.
In a retrospective study in which colorectal patient samples were used, Dr. Ju and colleagues revealed that hsa-miR-140-5p expression progressively decreases from normal tissues to primary colorectal cancer tissue, and that it shows a further decrease in liver and lymph node metastases. The experimental overexpression of hsa-miR-140-5p inhibited colorectal cancer stem cell growth by disrupting autophagy, and in a mouse model of disease it abolished tumor formation and metastasis.
In addition to hsa-miR-140-5p, Dr. Ju and colleagues recently identified hsa-miR-129 and found that it, too, has therapeutic potential. Specifically, they showed that hsa-miR-129 enhanced the sensitivity of colorectal cancer cells to 5-fluorouracil, pointing toward its ability to function as a tumor suppressor.
One of the mechanisms implicated in this process was the ability of miR-192 to inhibit protein translation of several important targets. These include Bcl-2 (B-cell lymphoma 2), a key anti-apoptotic protein; E2F3, a major cell cycle regulator; and thymidylate synthase, an enzyme that is inhibited by 5-fluorouracil.
The NIH recently awarded a $3 million grant to establish the Long Island Bioscience Hub (LIBH), which is part of the NIH’s Research Evaluation and Commercialization Hub (REACH) program and represents a partnership between the Center for Biotechnology, Stony Brook University, Cold Spring Harbor Laboratory, and Brookhaven National Laboratory. One of the technology development grants, as part of the first funding cycle of this initiative, will support a feasibility investigation of hsa-miR-129-based therapeutics in colon cancer, an effort led by Dr. Ju. “We are further exploring this novel mechanism,” states Dr. Ju. “We anticipate conducting pharmacokinetic studies and moving to a clinical trial in the future.”
MicroRNA Insights Gleaned from Host-Virus Interactions
“We observed that when a poxvirus is artificially engineered to encode a microRNA, the microRNA is destroyed along with all the microRNAs from the host cell,” says Benjamin R. tenOever, Ph.D., professor of microbiology at the Icahn School of Medicine, Mount Sinai Hospital. Previously, Dr. tenOever’s group reported that a single vaccinia virus-encoded gene product, VP55, is sufficient to achieve this effect. The group also found that the protein adds nontemplate adenosines to the 3′-end of microRNAs associated with the RNA-induced silencing complex.
“This tool provides a powerful means to address anything relevant to microRNA biology,” asserts Dr. tenOever.
In a recent study, Dr. tenOever and colleagues used a codon-optimized version of VP55 produced from an adenovirus-based vector to study the impact microRNA deletion would have on our normal response to virus infection. “We found that after administration of the vector and rapid ablation of microRNA expression, there is very little that happens over the first one to two days,” informs Dr. tenOever. During the first 24–48 hours after VP55 delivery, the elimination of cellular microRNAs impacted less than 0.35% of the over 11,000 genes expressed in the cell. After 9 days, however, almost 20% of the genes showed significant changes in expression.
“MicroRNAs are very powerful and influential in controlling the biology of the cell but they do so over the long term,” declares Dr. tenOever. These findings are in agreement with knowledge that has accumulated over the years about microRNA biology, which established that microRNAs play a central role in determining how cells differentiate during development.
“While microRNAs can act on hundreds of mRNAs, their action requires several days of fine-tuning to have long-term consequences,” adds Dr. tenOever. This finding suggests miRNAs are unable to significantly contribute to the acute response to virus infection.
The one exception to this observation was that, even though very few genes were affected in the first 48 hours after VP55 delivery, several genes encoding chemokines were impacted. These included chemokines responsible for recruiting antigen-presenting cells, neutrophils, and other immune cells.
An in vivo analysis of mouse lung tissue 48 hours after vector administration confirmed that several cytokines were specifically upregulated, resulting in immune cell infiltration following the degradation of all microRNAs. These results indicate that the acute viral infection is largely independent of microRNAs, and that microRNAs are primarily involved in the adaptive response to infection and other longer term processes.
MicroRNA Biomarkers Reveal Molecular Pathways of Kidney Damage
“Our approach involves looking at microRNAs from the perspective of biomarkers as a readout for kidney damage,” says Vishal S. Vaidya, Ph.D., associate professor of medicine and environmental health at Brigham and Women’s Hospital, Harvard Medical School, and Harvard T.H. Chan School of Public Health. “At the same time, we are exploring their utility as therapeutics.”
A large number of medications and occupational toxins cause kidney damage, but many tests to assess kidney function and damage are not sufficiently sensitive or specific, opening the need for novel diagnostic strategies. MicroRNAs, which are differentially expressed between healthy and diseased states, are promising as early biomarkers for impaired renal function.
“MicroRNAs can also provide information about which pathways are active and which targets can be druggable,” points out Dr. Vaidya.
In a study that used microRNAs and proteins to provide a combined biomarker signature, Dr. Vaidya and colleagues examined two patient cohorts, one presenting with acetaminophen-induced kidney injury and the other one with cisplatin-induced kidney damage. “Protein biomarkers provide sensitivity, and microRNAs offer mechanistic insight,” explains Dr. Vaidya.
This approach helped visualize metabolic pathways that are altered in the kidney during toxic injury. “The biggest challenge, from a therapeutic perspective, is that microRNAs regulate many mRNAs and, therefore, impact many proteins,” concludes Dr. Vaidya.