Long non-coding RNAs (lncRNAs) have diverse regulatory roles in gene expression that implicate them in numerous biological processes and diseases. Expression profiling of lncRNAs has therefore become essential to unravel how genes are expressed and regulated. Since lncRNA expression during development or pathogenesis is often more specific than for mRNAs, lncRNAs may serve as biomarkers.
Microarrays are the preferred analysis platform for lncRNA expression profiling. Since lncRNAs are expressed at much lower abundance relative to mRNA, their true expression levels may be masked using RNA sequencing: Due to the difficulty of transcript reconstruction from short RNA sequencing reads, less than 35% of the RNA transcript isoforms are correctly reconstructed by state-of-the-art RNA-seq.
By using splice junction-specific or unique exon-specific probes, Arraystar LncRNA microarrays detect RNAs at the transcript level unambiguously and reliably. Microarray RNA targets are generated by T7 promoter-driven linear amplification, which better preserves the fidelity of native RNA abundance while avoiding distortions introduced by exponential PCR amplifications during RNA sequencing.
Inherent challenges exist in transcript-profiling lncRNA using RNA sequencing. These include short NGS sequencing reads, low splice junction lncRNA sequence read abundance, and even lower transcript and isoform levels, as well as the computational complexities of dealing with alternative splicing.
With commonly used RNA-seq sequencing aligners HISAT2, TopHat2, and Mapsplice2, the sensitivity of detecting both the mRNA and lncRNA isoforms assisted with the reference transcriptome is less than 20%. Finding new transcript isoforms requires de novo transcript assembly from the sequencing reads, but even here de novo assemblers perform poorly for transcript isoforms, and detect only a small percent of lncRNAs.
To compound this issue, different algorithms and software packages generate highly variable exon models for transcript isoforms, and widely varying profiling values, even from the same dataset.
The challenges for transcript-specific profiling are well known, as reported in Nature Biotechnology: “Expression profiling of alternative transcripts requires knowledge of all the alternative transcript forms of a gene, and involves combining information across the transcripts,” and “for genome-scale RNA-seq, this is particularly difficult because of the sampling noise from low read counts for many transcripts.” Clearly, the need exists for a new approach to long-read NGS for transcript-specific lncRNA studies.
Writing in PLoS One, Liu and coworkers reached the same conclusion, that “300 million fragments are required for the detection of a specific human alternative splicing event with 80% power.”
Based on the analysis of GENCODE transcripts, the expression level of lncRNAs is about one-tenth that of mRNAs. Many lncRNAs exist at only a few copies per cell. Despite their low abundance, lncRNAs operate effectively because of how they regulate the target genes. For example, some lncRNAs are tethered locally to the neighboring genes to profoundly change their expression.
Arraystar LncRNA global gene expression microarrays profile both lncRNAs and entire sets of known mRNAs on the same array. Each RNA transcript, its variants and isoforms, are reliably detected by a splice junction-specific probe or a unique exon sequence. Arraystar LncRNA experimental protocols have been optimized for sensitive, robust, and accurate data generation, including analyses for microarray raw data processing, data quality control, gene expression clustering and heat map visualization, differentially expressed lncRNAs and mRNAs, lncRNA subcategories, regulatory relationships between lncRNAs and mRNAs, gene ontology, and pathway analysis.
Figure 1: An example of transcript isoform-specific detection. The same BCL2L1 gene can produce functionally different transcript isoforms: anti-apoptotic/oncogenic BCL-XL, pro-apoptoic/tumor suppressive BCL-XS, and intronic antisense lncRNA ENST00000412972. Arraystar Human LncRNA Microarray uses array probes unique to the transcript-specific exons (red bars) or slice junction (blue bars) to unambiguously, reliably, and accurately detect these individual transcript isoforms.
Arraystar LncRNA microarray probes provide unambiguous isoform detection by using transcript isoforms annotated from large datasets at the species level. Additionally, transcript models include those based on full-length, experimentally validated cDNA sequences, while splice junction-specific probes and unique exon probes ensure reliable quantification (Figure 1).
The performance of Arraystar LncRNA microarrays is explained by their reliance on well-annotated, experimentally validated lncRNAs instead of “anonymous,” partially degraded RNA fragments mistaken for lncRNAs. These full-length lncRNAs have been characterized by at least one Pubmed-indexed research paper, thus represent a gold standard for lncRNAs. Full annotation is increasingly viewed as the shortest path for understanding the biological functions of lncRNAs (Figure 2).
Figure 2. Arraystar LncRNA Microarrays come available with rich annotation and analyses dedicated for noncoding regulatory lncRNAs that are very different from the traditionally profiled coding mRNAs (additional advanced analysis packages may be required). Expression data: Include both but separate lncRNA and mRNA data sets; Epigenetic landscape: promoter driven p-lncRNAs and enhancer driven e-lncRNAs; Genomic context: genomic locations and regulatory potentials with the closest coding genes; True 5’-end: for experimentally supported transcript full length status; Nuclear/Cytosol: subcellular localization to imply functional mechanism; lincRNA catalogs: long intergenic lncRNAs that can act on coding genes at long distance; Antisense: able to target and regulate transcript isoforms of its own host gene; ceRNA: competing endogenous lncRNAs that can sequester miRNAs down-regulating the coding mRNAs; Conservation: from ultra-conserved lncRNAs across species, ancient lncRNAs having fundamental cellular functions, to less conserved emerging lncRNAs in primates and human; Disease-associated: lncRNAs known to have disease association; Biological process: lncRNAs with known functions in biology; Tissue-specific: specific expression for functional or biomarker indications.
Arraystar microarrays have supported lncRNAs research for years. In a paper published in Nature Cell Biology this year, Arraystar LncRNA microarrays helped researchers profile the total cellular and lipid-bound RNAs to identify the key lncRNA link to pathologically significant triple-negative breast cancers.
In another study, published in Cell, investigators used an Arraystar LncRNA microarray to identify differential expression of BRAR4-lncRNA activated in advanced breast cancer.
A third study, in Cancer Discovery, identified GClnc1 as a differentially expressed lncRNA that cis-activates gastric cancer using the Arraystar LncRNA microarray.
As the vast majority of RNA transcriptome are non-coding, lncRNAs present a unique view into the regulatory apparatus of cells. Unfortunately NGS, which has served genomics so well, lacks the coverage, sensitivity, and quantification accuracy to profile and characterize lncRNAs adequately enough to elucidate their biological roles associated with disease and health.
Arraystar LncRNA arrays overcome the shortcomings of NGS by employing array probes that simultaneously and sensitively capture and measure lncRNAs often functionally at low abundance. They are the method of choice for researchers to study lncRNAs at the transcript level unambiguously and reliably.
Webinar: Arraystar LncRNA Microarray profiling
Review: Benefits of Arraystar LncRNA arrays