A comprehensive joint analysis of the long and short RNA transcriptomes of human erythrocytes

Doss et al. BMC Genomics (2015) 16:952 DOI 10.1186/s12864-015-2156-2 RESEARCH ARTICLE Open Access A comprehensive joint analysis of the long and short RNA transcriptomes of human erythrocytes Jennifer F. Doss1,2, David L. Corcoran2, Dereje D. Jima2,3, Marilyn J. Telen4, Sandeep S. Dave2,3 and Jen-Tsan Chi1,2* Abstract Background: Human erythrocytes are terminally differentiated, anucleate cells long thought to lack RNAs. However, previous studies have shown the persistence of many small-sized RNAs in erythrocytes. To comprehensively define the erythrocyte transcriptome, we used high-throughput sequencing to identify both short (18–24 nt) and long (>200 nt) RNAs in mature erythrocytes. Results: Analysis of the short RNA transcriptome with miRDeep identified 287 known and 72 putative novel microRNAs. Unexpectedly, we also uncover an extensive repertoire of long erythrocyte RNAs that encode many proteins critical for erythrocyte differentiation and function. Additionally, the erythrocyte long RNA transcriptome is significantly enriched in the erythroid progenitor transcriptome. Joint analysis of both short and long RNAs identified several loci with co-expression of both microRNAs and long RNAs spanning microRNA precursor regions. Within the miR-144/451 locus previously implicated in erythroid development, we observed unique co-expression of several primate-specific noncoding RNAs, including a lncRNA, and miR-4732-5p/-3p. We show that miR-4732-3p targets both SMAD2 and SMAD4, two critical components of the TGF-β pathway implicated in erythropoiesis. Furthermore, miR-4732-3p represses SMAD2/4-dependent TGF-β signaling, thereby promoting cell proliferation during erythroid differentiation. Conclusions: Our study presents the most extensive profiling of erythrocyte RNAs to date, and describes primate-specific interactions between the key modulator miR-4732-3p and TGF-β signaling during human erythropoiesis. Keywords: Erythrocyte, microRNA, Long noncoding RNA, TGF-β Background Human erythrocytes provide gas transport throughout the body and comprise the majority of cells in whole blood. Human diseases related to red blood cells (RBCs) or erythrocytes, such as anemia or malaria, affect hundreds of millions of people worldwide and present huge health concerns. Although we have gained a significant understanding of how these diseases occur and many treatment options are available, we still cannot explain many aspects of these erythrocyte diseases. Since the * Correspondence: 1 Department of Molecular Genetics and Microbiology, Duke University, Durham, NC 27710, USA 2 Center for Genomic and Computational Biology, Duke University, Durham, NC 27708, USA Full list of author information is available at the end of the article precise regulation of both coding and noncoding RNAs is essential for erythrocyte development, these erythrocyte diseases are often accompanied by significant transcriptome changes. During erythropoiesis, microRNAs actively regulate proliferation and/or differentiation of erythroid cells during physiological and pathological adaptions [1]. Dysregulation of various other long and short RNAs also leads to several disease states such as ineffective erythropoiesis and anemias. Circulating erythrocytes can be easily obtained by blood drawing. Accordingly, a detailed analysis of the erythrocyte transcriptome may provide an accessible window into the developmental history and pathophysiology of erythrocytes. However, such an analysis has been deemed impossible since circulating erythrocytes were thought to lack any genetic materials. © 2015 Doss et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Doss et al. BMC Genomics (2015) 16:952 During terminal maturation of erythrocytes, the nucleus is extruded from progenitors, leading to anucleate cells with no further RNA production. Therefore, mature erythrocytes were once thought to lack RNAs and have significantly lower signals from RNA-binding dyes such as methylene blue. However, many studies have shown erythrocytes do contain diverse and abundant small RNA species [2], including noncoding RNAs Y1 and Y4 [3] as well as microRNAs [4, 5]. We have previously shown that higher levels of miR-144 and miR-451 reflect the hemolytic phenotype and malaria resistance of sickle erythrocytes, respectively [6, 7]. Additionally, both miR-451 and miR-144 reside in a locus that is regulated by GATA-1, are highly induced during erythroid differentiation, and are critical to erythropoiesis [8]. Therefore, extensive profiling of the RBC transcriptome is critical for an understanding of erythrocyte biology. The RNA composition of erythrocytes may also change during long-term storage for future blood transfusion [9]. However, previous transcriptomic analyses were limited to known erythrocyte microRNAs using microarrays [4], or known microRNAs from mixed reticulocyte (immature red blood cell) and erythrocyte populations using sequencing [10]. In addition, it is not clear whether erythrocytes also contain long (large-sized) RNAs that may provide valuable insights into their development and adaptations. With the recent advances in high-throughput sequencing technologies, it is possible to perform RNASeq to identify both known and unknown transcripts. Here, we employed high-throughput sequencing to characterize both short (small, 18–24 nt) and long (large, > 200 nt) RNAs in human erythrocytes. For long RNA profiling, we prepared RNA-Seq libraries using a protocol that allows for identification of both polyadenylated and non-polyadenylated RNAs. A total of 6843 transcripts were expressed in all three analyzed erythrocyte samples. While this number is far less than that of typical nucleated cells, these analyses established a surprisingly diverse RBC transcriptome. In parallel, short RNA sequencing libraries were prepared and the miRDeep pipeline was utilized to identify both known and putative microRNAs. From these analyses, we identified in mature erythrocytes an abundant, diverse set of microRNAs that include both known and putative microRNAs. The joint analysis of transcriptomes identified several loci with expression of both long and short RNAs, suggesting their coordinated regulation of expression or processing. Furthermore, we performed a functional investigation of the uncharacterized, primate-specific miR-4732-3p within the miR-144/451 locus. MiR-4732-3p was predicted to target both SMAD2 and SMAD4 (...truncated)