NEW YORK (GenomeWeb) – Small RNA cloning and sequencing strategies can affect microRNA expression signatures, confounding efforts to study the roles of the small RNAs in biology, according to new research.
As such, data should be compiled using a combination of techniques in addition to high-throughput sequencing approaches, according to a team led by investigators from the University of Tours in France.
To come to their conclusions, the scientists were studying Marek's disease virus 1 (MDV-1), an avian herpesvirus that naturally infects chickens. It also induces T cell lymphomas and therefore is used as a natural model for virus-induced lymphagenesis.
Previously, a number of MDV-1 miRNAs had been identified from small RNA libraries established from infected chicken embryonic fibroblasts using deep sequencing technology. MDV-1 sRNAs have also been identified in various samples including spleen tumors using various sequencing approaches.
"One observation of these studies was the striking discrepancy of miRNA expression patterns that existed between different biological samples or even between identical biological samples analyzed with different technologies," the investigators wrote in the Journal of Biotechnology this month. Even if some of the differences are associated with the phase of the virus cycle that was analyzed in the different studies, "the viral miRNA relative frequencies seemed to be influenced by the small RNA sequencing technology used in the different laboratories."
This, they added, suggests that miRNA cloning procedures or sequencing could influence miRNA representation in libraries — a finding borne out by their own experiments in MDV-1.
Specifically, the scientists compared the influence of classical Sanger and high-throughput Illumina sequencing on miRNA cloning frequencies generated from the same biological samples obtained in vivo — peripheral blood leukocytes from infected and non-infected animals — and in vitro — two MDV-transformed cell cultures derived from MDV-induced lymphomas.
They found qualitative and quantitative variations were found in the data depending on the strategy used. The highest variation was seen with the mature miRNA miR-M7-5p, which is derived from the latency-associated transcript, they wrote.
This miRNA's cloning frequency was 50 percent of the viral miRNA counts when a small-scale sequencing approach was used. However, its frequency was 100 times less abundant when determined through deep sequencing.
"Northern blot analysis showed a better correlation with the miRNA frequencies found by the small-scale sequencing approach," according to the Journal of Biotechnology paper.
"Altogether, our data show that cloning and sequencing strategies introduce several biases in the establishment of viral and cellular miRNA repertoires," the team concluded. "These biases must be considered since sequencing bias influences the choice of the miRNAs that will be studied in depth to explore functional aspects of the miRNA expression."
In the end, next-generation sequencing data alone are "limited" for assessing the absolute copy number of transcripts, the group added. "Thus, the quantification of small RNA should be addressed by compiling data obtained by using different techniques" such as microarrays, RT-PCR, and Northern blot analyses to bolster high-throughput sequencing data.