A tour through the transcriptional landscape of platelets

Abstract

The RNA code found within a platelet and alterations of that code continue to shed light onto the mechanistic underpinnings of platelet function and dysfunction. It is now known that features of messenger RNA (mRNA) in platelets mirror those of nucleated cells. This review serves as a tour guide for readers interested in developing a greater understanding of platelet mRNA. The tour provides an in-depth and interactive examination of platelet mRNA, especially in the context of next-generation RNA sequencing. At the end of the expedition, the reader will have a better grasp of the topography of platelet mRNA and how it impacts platelet function in health and disease.

Figure 1

A tour through the transcriptional landscape of a platelet. This tour through the transcriptional landscape in platelets is divided into 8 segments, which are listed in this figure.

Figure 2

General features of an mRNA transcript as visualized by RNA-seq. This figure shows RNA-seq read coverage plots and individual reads for platelet GP9 mRNA as it appears (ie, with introns removed) in the platelet (A) or as viewed along genomic coordinates in a genome browser (B). The blue annotations along the bottom are the consensus RefSeq annotations of the gene structure. Individual exons are represented by the blue bars, and introns (only in panel B) by the connecting blue lines. The thicker regions of the bars distinguish the coding region from the 5′ and 3′ untranslated regions (UTRs). Arrows indicate the strand orientation (forward [+] or reverse [−]) of the annotation. For GP9, there are 3 exons and 2 introns. In GP9, the 5′ UTR ends just beyond the start of the third exon (at the ATG start codon of the coding region), and the 3′ UTR starts near the end of the third exon (after the TAA stop codon). Above the annotation are the individual reads that align to the transcript. For simplicity, not all reads are shown here. For visualization purposes, the individual reads are colored according to the exon to which they align. Above the reads is the read coverage map. This is generated by plotting the sums of mapped reads that overlap (cover) each nucleotide position along the genomic coordinates (x-axis). The y-axis represents the number of reads. The examples are as illustrated in Integrated Genome Viewer. See it yourself: To see GP9 in platelets via the UCSC Genome Browser, follow the Web link > access instructions for site 1 (see Table 1). For individual reads, follow the Web link > access instructions for site 2 (see Table 1). Search for GP9 in the browser.

Figure 3

Platelet transcripts are capped, tailed, and spliced. (A) The techniques we have used to capture and concentrate poly-A mRNA or 5′-capped mRNA from platelets. The captured RNAs are concentrated by immunoprecipitation techniques. For poly-A pull-down, the bead-conjugated oligo-dT sequence binds to the poly-A tail. For the 5′-cap pull-down, a bead-conjugated hyperaffinity mutant enzyme binds to the 7mG cap at the 5′ end of the mRNA. (B) Bar graph of the RNA-seq expression estimates (RPKM) of a histone transcript HIST1H4B and the transcript coding for the platelet thrombin receptor protease-activated receptor 1 (F2R) following mRNA isolation by a 5′-cap or poly-A tail pull-down. (C) RNA-seq coverage graph of FBJ murine osteosarcoma viral oncogene homolog B (FOSB), which appears mostly unspliced in platelets. Note that reads align to both exons and introns. Compare this with GP9 in Figure 2, which has very few reads mapping to introns. RPKM, read (or fragment) per kilobase normalized to a million.

Figure 4

Platelet transcripts have variable 5′ or 3′ UTRs. (A) Shown are coverage graphs of the 5′ end of GFI1B in platelets. The top panel represents coverage of all reads mapping to the 5′ end of GFI1B. The red box is drawn around the region that maps beyond the known transcript annotation (blue). The bottom panels show only those reads that contain the sequence tag ligated onto the 5′ end at time of sample preparation (for details, see Osman et al). The 5′-tagged reads are only the extreme 5′-terminal reads and therefore mark the transcription start site. (B) Coverage graph of the 3′ end of purinergic receptor P2Y, G-protein coupled 1 (P2RY1) in platelets. The red box is drawn around the region that maps beyond the known 3′ UTR annotation. miRNA target analysis of this additional sequence identified several miRNA binding sites. A representative subset of the target sites is marked underneath the graph. See it yourself: To see GFI1B and P2RY1 in platelets via the UCSC Genome Browser, follow the Web link > access instructions for site 1 (see Table 1). For individual reads, follow the Web link > access instructions for site 2 (see Table 1). Search for GFI1B or P2RY1 in the browser. Note that the 5′-terminal reads in the bottom panels of A were processed from the data set published by Osman et al and are therefore not yet available in a browser-friendly format.

Figure 5

Platelets express mRNA splice variants. Shown are coverage graphs of PEAR1 (A), tissue factor pathway inhibitor (TFPI) (B), and GFI1B (C) in platelets. (A) Note that there are 4 different Ensembl annotations listed underneath PEAR1. The red box is drawn around the second exon, which distinguishes the second annotation from the other 3 annotations. (B) The annotations for 2 major isoforms of TFPI, α (iso A) and β (iso B), are distinguished by their 3′ ends. Note that platelets are rich in isoform α but not isoform β. Exon 2 (exon 2 is very small, so it is difficult to see the annotation here), which can also be removed by alternative splicing, is also present in the platelet. (C) The arrow points to a putative novel exon of GFI1b. Annotations for an exon at this region are not present in any of the Ensembl, UCSC, or RefSeq databases. The red box outlines examples of paired reads that map from the unannotated novel exon to the known second exon suggesting that it is part of the GFI1B transcript. See it yourself: To see PEAR1, TFPI, and GFI1B in platelets via the UCSC Genome Browser, follow the Web link > access instructions for site 1 (see Table 1). For individual reads, follow the Web link > access instructions for site 2 (see Table 1). Search for PEAR1, TFPI, or GFI1B in the browser.

Figure 6

Antisense transcripts in platelets. Shown are positive and negative strand coverage graphs of the 5′ end of CD109 and of AK124950, an antisense transcript in platelets. Positive strand reads are on the top panel in blue, and negative strand reads are in the bottom panel in red. Both the expression of CD109 and the expression of an antisense transcript (AK124950, surrounded by the red box) transcribed from the opposite strand of DNA at the 5′ end of CD109 can be inferred when the read orientation is taken into account. See it yourself: To visualize CD109 and its antisense transcript in platelets via the UCSC Genome Browser, follow the Web link > access instructions for site 3 or 4 (see Table 1). Search for CD109 or AK124950 in the browser. Note that only stranded data (only site 3 or 4 of Table 1) can distinguish antisense from sense transcripts.

Figure 7

Single nucleotide resolution of glycoprotein Ib alpha (GP1BA) in platelets. Individual reads from platelets from 2 different individuals are shown. The arrow points to the location of the SNP rs6065, which associates with platelet counts in GWASs. Looking at the individual nucleotide sequences compared with the reference sequence indicates that the reads from the bottom individual match the reference cyotosine (reference matches are grayed out), whereas approximately half of the reads from the top individual contain the C to T polymorphism. See it yourself: To visualize rs6065 in GP1BA expressed in platelets via the UCSC Genome Browser, follow the Web link > access instructions for site 2 (see Table 1). Search for rs6065 and zoom in to the center of the region, or alternatively, type chr17:4 836 371-4 836 392 in the search box.

Figure 8

RNA-seq facilitates discovery in platelets. The table summarizes the various transcript features, and their respective functions, that are identifiable through RNA-seq in platelets. Each feature in platelets can be analyzed to provide mechanistic insights into platelet function (1), as biomarkers or modulators of disease (2), or for their role in megakaryocyte development and function (3). As an example, the vessel in (1) depicts platelets that express 2 different splice variants: a predominant 3-exon variant (purple-green-red) and a minor 2-exon variant (purple-red). In vessel (2), which is atherosclerotic, the predominant isoform in platelets is shifted to the 2-exon isoform that may serve as a biomarker for atherosclerosis and may additionally contribute to disease pathogenesis. As depicted in (3), a combination of genetics and environmental signals reaching the bone marrow megakaryocyte dictate the expression of the RNAs, and the features of RNA, that are captured as the platelets are formed.