Identification and comparison of m6A modifications in glioblastoma non-coding RNAs with MeRIP-seq and Nanopore dRNA-seq.

EPIGENETICS 2023, VOL. 18, NO. 1, 2163365 https://doi.org/10.1080/15592294.2022.2163365 RESEARCH PAPER Identification and comparison of m6A modifications in glioblastoma non-coding RNAs with MeRIP-seq and Nanopore dRNA-seq Raulas Krusnauskasa, Rytis Stakaitisb, Giedrius Steponaitisb, Kristian Almstrupc,d, and Paulina Vaitkienea a Laboratory of Molecular Neurobiology, Neuroscience Institute, Lithuanian University of Health Sciences, Eiveniu str. 4, LT50161, Kaunas, Lithuania; bLaboratory of Molecular Neurooncology, Neuroscience Institute, Lithuanian University of Health Sciences, Eiveniu str. 4, LT50161, Kaunas, Lithuania; cDepartment of Growth and Reproduction, Rigshospitalet, University of Copenhagen, GR-5064, Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen, Denmark; dInternational Center for Research and Research Training in Endocrine Disruption of Male Reproduction and Child Health (Edmarc), Rigshospitalet, University of Copenhagen, GR-5064, Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen, Denmark ABSTRACT The most prominent RNA modification – N6-methyladenosine (m6A) – affects gene regulation and cancer progression. The extent and effect of m6A on long non-coding RNAs (lncRNAs) is, however, still not clear. The most established method for m6A detection is methylated RNA immunoprecipitation and sequencing (MeRIP-seq). However, Oxford Nanopore Technologies recently developed direct RNA-seq (dRNA-seq) method, allowing m6A identification at higher resolution and in its native form. We performed whole transcriptome sequencing of the glioblastoma cell line U87-MG with both MeRIP-seq and dRNA-seq. For MeRIP-seq, m6A peaks were identified using nf-core/chipseq, and for dRNA-seq – EpiNano pipeline. MeRIPseq analysis revealed 5086 lncRNAs transcripts, while dRNA-seq identified 336 lncRNAs tran scripts from which 556 and 198 were found to be m6A modified, respectively. While 24 lncRNAs with m6A overlapped between two methods. Gliovis database analysis revealed that the expression of the major part of identified overlapping lncRNAs was associated with glioma grade or patient survival prognosis. We found that the frequency of m6A occurrence in lncRNAs varied more than 9-fold throughout the provided list of 24 modified lncRNAs. The highest m6A frequency was detected in MIR1915HG, THAP9-AS1, MALAT1, NORAD1, and NEAT1 (49–88nt), while MIR99AHG, SNHG3, LOXL1-AS1, ILF3-DT showed the lowest m6A frequency (445–261nt). Taken together, (1) we provide a high accuracy list of 24 m6A modified lncRNAs of U87-MG cells; (2) we conclude that MeRIP-seq is more suitable for an initial m6A screening study, due to its higher lncRNA coverage, whereas dRNA-seq is most useful when more indepth analysis of m6A quantity and precise location is of interest. ARTICLE HISTORY Received 01 July 2022 Revised 15 December 2022 Accepted 23 December 2022 KEYWORDS lncRNA; m6A; epitranscriptome; glioblastoma; MeRIP-seq; nanopore dRNAseq Abbreviations: (dRNA-seq) direct RNA-seq, (GBM) glioblastoma, (LGG) low-grade glioma, (lncRNAs) long non-coding RNAs, (m6A) N6-methyladenosine, (MeRIP-seq) methylated RNA immunoprecipitation and sequencing, (ncRNA) non-coding RNA, (ONT) Oxford Nanopore Technologi; Lietuvos Mokslo Taryba Background First chemical RNA modification was discovered by Waldo E. Cohn and Elliot Volkin in 1951 [1]. Almost 70 years later, this knowledge expanded to more than 150 identified RNA modifications [2]. Interestingly, the field of epi-transcriptomics recently regained its interest due to characterization of the regulatory machinery of N6-methyladenosine (m6A) modifica tion [3,4] and a breakthrough of its efficient recogni tion methods [5,6]. N6-methyladenosine is a reversibly methylated adenosine, which is co-/posttranscriptionally installed by ‘writers’ [7], interpreted by ‘readers’ [8], and removed by ‘erasers’ [4,9]. Thorough mapping and identification of specific RNA molecules containing m6A revealed modifica tion importance and role in gene expression [10], development [11], and disease progression [12]. Recent development of novel approaches for m6A identification revealed limitations of the firstgeneration methods and highlighted directions for further improvements, primarily related to increased registration sensitivity and bioinformatics [13]. Furthermore, majority of the studies solely focuses on the methylation of coding RNAs, while ignoring CONTACT Raulas Krusnauskas Laboratory of Molecular Neurobiology, Neuroscience Institute, Lithuanian University of Health Sciences, Eiveniu str. 4, LT50161, Kaunas, Lithuania Supplemental data for this article can be accessed online at https://doi.org/10.1080/15592294.2022.2163365 © 2023 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 2 R. KRUSNAUSKAS ET AL. hugely important and m6A modified non-coding RNA (ncRNA) molecules [14]. Human genome contains over 16,000 genes encoding long non-coding RNAs (lncRNAs) (Gencode v30) which leads to production of more than 30,000 transcripts [15]. Long non-coding RNAs (lncRNAs) interact with DNA, RNA, and proteins [16–19], and by doing so affect gene regulation, proliferation, infiltration, and metastasis of tumour cells. Methylated RNA immunoprecipitation and sequencing (MeRIP-seq) and m6A iCLIP (miCLIP) experiments revealed that XIST lncRNAs contain 78 m6A marks [20]. Later, Patil et al. showed that m6A is crucial for XIST functionality which was supported by interactions with key components of m6A machinery [21]. Another lncRNA, metastasisassociated lung adenocarcinoma transcript 1 (MALAT1) also contains high level of m6A modifi cations [22]. Furthermore, MALAT1 m6As not only affect its cellular location and binding properties but also change its structure [22]. Currently, m6A is primarily recognized by MeRIP-seq method based on anti-m6A antibodies [5,6]. These antibodies specifically bind to methy lated adenosines and enable identification of the approximate location of modifications. More recently, identification of RNA modifications was introduced by Oxford Nanopore Technologies (ONT) direct RNA-seq (dRNA-seq) method [13]. dRNA-seq enables precise detection of RNA mod ifications including m6A in its native form, by pulling molecules through the membrane contain ing embedded nanopore particles. ONT sequen cing device registers nucleotides by measuring disruptions in the electric current intensity as an RNA molecule passes through the pore. Even though dRNA-seq is an established method which can register RNA modifications at a single nucleotide resolution, its efficiency and accuracy require further improvements [13]. The aim of the study was (1) to define a list of m6A locations in lncR (...truncated)