The repertoire and features of human platelet microRNAs

Abstract

Playing a central role in the maintenance of hemostasis as well as in thrombotic disorders, platelets contain a relatively diverse messenger RNA (mRNA) transcriptome as well as functional mRNA-regulatory microRNAs, suggesting that platelet mRNAs may be regulated by microRNAs. Here, we elucidated the complete repertoire and features of human platelet microRNAs by high-throughput sequencing. More than 492 different mature microRNAs were detected in human platelets, whereas the list of known human microRNAs was expanded further by the discovery of 40 novel microRNA sequences. As in nucleated cells, platelet microRNAs bear signs of post-transcriptional modifications, mainly terminal adenylation and uridylation. In vitro enzymatic assays demonstrated the ability of human platelets to uridylate microRNAs, which correlated with the presence of the uridyltransferase enzyme TUT4. We also detected numerous microRNA isoforms (isomiRs) resulting from imprecise Drosha and/or Dicer processing, in some cases more frequently than the reference microRNA sequence, including 5' shifted isomiRs with redirected mRNA targeting abilities. This study unveils the existence of a relatively diverse and complex microRNA repertoire in human platelets, and represents a mandatory step towards elucidating the intraplatelet and extraplatelet role, function and importance of platelet microRNAs.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1. The small RNA profile of human platelets.

(A) Size distribution of usable reads detected in each of the platelet RNA pool 1 and 2. Most of the RNA sequences are 21 to 23 nucleotides (nt) in length, representing 88% and 83% of the sequences obtained from pool 1 and pool 2, respectively. (B) Classification of the small RNA species isolated and sequenced from human platelet RNA pool 1 and 2. The proportion of sequences (%) matching to each RNA category is shown. MicroRNAs represent approximately 80% of the small RNA species between 18 and 30 nt in length in human platelets.

Figure 2. The annotated microRNA profile of human platelets.

(A) The HTS profiling of platelet microRNAs, which is shown in order of decreasing number of reads. (B) Distribution of the 15 most abundant mature microRNA families detected in human platelets. The let-7 microRNA family represented 48% of a total of 492 different mature microRNAs, whereas the 15 most abundant microRNAs accounted for more than 90% of all platelet microRNAs.

Figure 3. Post-transcriptional modifications of platelet microRNAs.

(A) The relative frequency of nucleotide mutation ordered by position on the mature microRNA sequences. The majority of the nucleotide modifications were observed at the 3′ end of microRNAs, whereas the sequence of their seed region was relatively well preserved. (B) Analysis of post-transcriptional modifications affecting two highly expressed platelet microRNAs, miR-223-3p and let-7f-5p. Whereas miR-223-3p showed preferential uridylation, let-7f-5p exhibited predominant adenylation. These modifications represented 29% and 13% of the expressed forms of miR-223-3p and let-7f-5p, respectively. (C) Prevalence of the modifications observed in microRNAs that displayed more than 10% modified sequences in abundance. MicroRNAs for which adenylated/uridylated sequences were at least 10 times more abundant than other modifications were classified as predominant adenylated/uridylated microRNAs. Similar level (e.g. within a 10-fold range) of adenylated and uridylated forms were detected for the majority (56%) of microRNAs. Exclusive or predominant adenylation was observed for 43% of the 5p-microRNAs as compared to 12% of the 3p-microRNAs. On the contrary, exclusive or predominant uridylation was more frequent in 3p-microRNAs (27%), as compared to 5p-microRNA (7%).

Figure 4. Platelets possess microRNA terminal nucleotidyltransferase activity.

(A–B) Uridylation (A) and adenylation (B) assays were performed by incubating synthetic miR-223 or let-7a mature microRNAs or microRNA duplexes with protein extracts, prepared from Meg-01 cells or platelets isolated from 3 healthy volunteers, in the presence of α-32P-labeled UTP or ATP respectively, at 30°C for 90 min. Uridylated or adenylated forms of microRNAs were detected by denaturing PAGE and autoradiography. (C) Western blot detection of TUT4 uridyltransferase (left panel) and GLD2 adenyltransferase (right panel) enzymes in Meg-01 and platelet extracts using anti-TUT4 and anti-GLD2 antibodies, respectively.

Figure 5. Detection of multiple microRNA isoforms in human platelets.

The expression level of the main isoforms of miR-140-3p are shown. The expected mature microRNA sequence, as retrieved from miRBase, is highlighted in bold and does not correspond to the most frequently encountered isoform. As shown on the left, most of the miR-140-3p isoforms may result from a combination of imprecise processing by Drosha and/or Dicer, including the most abundant, whose cleavage sites on the pre-microRNA species are shifted towards the 3′ end by a single nucleotide. Of notice 2 major miR-140-3p isomiRs population coexist in platelets depending on 5′clivage by Dicer, either at the canonical position (blue bars) or harboring a 1 nt cleavage shift (red bars).

Figure 6. Redirection of platelet miR-140-3p targeting upon a 1-nt shift of the 5′ cleavage site.

(A) mRNA target prediction for the reference mature miR-140-3p sequence and the miR-140-3p isomiR harboring a 1-nt 5′ shift using TargetScan. The reference isomiR potentially regulates 240 different mRNA targets, whereas the shifted isomiR potentially regulates 367 mRNA targets. Notably, only 50 of these mRNAs are predicted to be regulated by both forms of miR-140-3p. (B) Experimental validation of the target specificity of the 1-nt 5′ shifted isomiR for CAP1 mRNA. HEK 293 cells were co-transfected with miR-140 microRNA duplexes, containing either the reference microRNA sequence (Reference) or the most abundant 1-nt 5′ shifted isomiR (Shifted), and a reporter gene construct in which the adenylate cyclase-associated protein 1 (CAP1) 3′UTR was inserted downstream of the Rluc reporter gene. Base pairing complementarity between miR-140-3p microRNAs and their binding sites is shown in the upper panel. Base pairing involving the microRNA seed region is highlighted in color. The blue X denotes the loss of base pairing of nt 2 of the miR-140-3p reference isoform, which may explain its lower efficiency in regulating CAP1 mRNA 3′UTR expression. Rluc and Fluc activities were measured, and the values were normalized to those obtained with a non-relevant RNA duplex (n = 3 to 4 experiments, in duplicate) (lower panel). ** p