Promoter-Bound Full-Length Intronic Circular RNAs-RNA Polymerase II Complexes Regulate Gene Expression in the Human Parasite Entamoeba histolytica

Abstract

Ubiquitous eukaryotic non-coding circular RNAs are involved in numerous co- and post-transcriptional regulatory mechanisms. Recently, we reported full-length intronic circular RNAs (flicRNAs) in Entamoeba histolytica, with 3'ss-5'ss ligation points and 5'ss GU-rich elements essential for their biogenesis and their suggested role in transcription regulation. Here, we explored how flicRNAs impact gene expression regulation. Using CLIP assays, followed by qRT-PCR, we identified that the RabX13 control flicRNA and virulence-associated flicRNAs were bound to the HA-tagged RNA Pol II C-terminus domain in E. histolytica transformants. The U2 snRNA was also present in such complexes, indicating that they belonged to transcription initiation/elongation complexes. Correspondingly, inhibition of the second step of splicing using boric acid reduced flicRNA formation and modified the expression of their parental genes and non-related genes. flicRNAs were also recovered from chromatin immunoprecipitation eluates, indicating that the flicRNA-Pol II complex was formed in the promoter of their cognate genes. Finally, two flicRNAs were found to be cytosolic, whose functions remain to be uncovered. Here, we provide novel evidence of the role of flicRNAs in gene expression regulation in cis, apparently in a widespread fashion, as an element bound to the RNA polymerase II transcription initiation complex, in E. histolytica.

Keywords: Pol II-CTD; circRNA; flicRNA; gene promoter; splicing; transcription regulation.

Figure 1

flicRNAs associated with RNA Pol II-CTD domain. (A) PCR amplifications were used to verify the establishment of single and double amoeba transfectants. Appropriate primers were used to amplify the 2240 bp HA-Pol II-CTD-specific (CTD) and 287 bp HA-ΔGU-specific fragments in single and HA-Pol II-CTD/ΔGU (CTD/ΔGU) double transfectants. DNA fragments were resolved in 1 or 2% agarose gels. HA-V transfectants were used as negative controls, and actin was amplified as a loading control. Minus RT controls were also carried out. M, molecular weight marker. (B) Two nuclear localization signals (KLKSENKLEIRRKNGIK and LRKKNFKSIEERLSSKQGRL, residues 178–194 and 314–333, respectively) import the Pol II-CTD domain into the nucleus of amoeba transformants. The merged immunofluorescence confocal microscopy (bottom right panel; 100× magnification) representative image shows signals for anti-HA antibodies and FIT-C conjugated secondary antibodies (green channel, top right panel; 60% positive cells of 3 fields, 100 cells each); DAPI-stained nuclear DNA (blue channel, top left panel); phase contrast (bottom left panel); scale bar = 20 μm; 75% at least three areas of 100. (C) CLIP validation: densitometric quantitation of flicX13 amplified using circular RT-PCR, from CLIP RNA samples of HA IP, nuclear input, Pol II-CTD IP (CTD IP), and Pol II-CTD/ΔGU (CTD/ΔGU IP) immunoprecipitations. Amplified products were resolved in 2.5% agarose gels (bottom). (D) Different flicRNAs were amplified as in (C). To monitor the cotranscriptional nature of CLIPs, the spliceosomal U2 snRNA (U2) was amplified, and U6 snRNA (U6) was used as the corresponding negative control. The Hsp70 transcript was amplified to test the specificity of the HA antibodies. The cytoplasmic circular RNA circ670e2 was amplified to monitor cellular fractionation, and actin was amplified as the circ670e2 loading control. The plot shows the average abundance of amplicons of two independent experiments. (E) Divergent RT-PCR was used to amplify ribosomal flicRNAs RpS14 and RpL12 from cytosolic RNA fractions (red asterisks) and RNA purified from Pol II-CTD CLIPs (CTD IP). Products were resolved in 2.4% agarose gels. m, molecular weight marker. Statistically significant results (ANOVA, Turkey test a posteriori. p ≤ 0.01, *).

Figure 2

Inverse correlation of flicX13 and RabX13 gene expression. (A) 5 mM boric acid (BA) reduces the production of flicRNAs flicX13 (lanes 1–2), flic170 (lanes 3–4), and flic670i1 (asterisks in lanes 5–6), assessed by circular RT-PCR using specific primers and total RNA isolated from BA treated (+) and untreated (−) amoebas. DNA fragments were visualized on 2.5% agarose gels. Densitometric analyses of the results are shown above. (B) RT-qPCR was used to monitor the effect of BA, for 2 and 5 h, on the production of flicX13 and the expression of mRabX13 in Pol II-CTD transfectants (CTD). As controls, we used untreated DbrΔC and ΔGU trophozoites. They elicit flicRNA accumulation/slight mRNA reduction and flicRNA reduction/mRNA accumulation, respectively. Relative expression of RNA variants is presented compared to one in the HA control. Logarithmic values are the average of three independent experiments; statistically significant results (ANOVA and Turkey test a posteriori. p ≤ 0.01, *) are shown.

Figure 3

RNA Pol II binds to gene promoters regardless of their flicRNA-forming capacity. (A) E. histolytica trophozoites were transiently transfected with the empty vector (pHA, lanes 1) or with the HA-tagged major subunit of the RNA polymerase II (HA-Pol II, lanes 2) and selected with low concentrations of G418 (4 µg/mL). Overexpression of HA-Pol II was monitored by Western blot using anti-HA antibodies compared to anti-actin (EHI_107290). Using these amoeba transfectants, ChIP assays detect RNA Pol II binding sites on selected flicRNA-forming promoters (B), as well as on promoters that do not form flicRNAs (C). PCR amplified DNA fragments were resolved in 2.5% agarose gels. Input fractions of sonicated DNAs are shown. Tris/SDS elution (E), heat reversed-crosslinked (ΔE), and DNase Q treated (ΔEQ) samples were analyzed. In each step of the ChIP assay, the actin coding region was used as a control. The position of the primers relative to their promoters and the first codon are shown. Based on their sequence motifs in large promoters, two DNA fragments were analyzed.

Figure 4

flicRNAs are found on the promoters of their parental genes and alter the expression of flicRNA-forming and no flicRNA-forming genes. (A) RNA was extracted from ΔE fractions of HA-Pol II ChIP aliquots and used to perform circular RT-PCR to detect immunoprecipitated flicRNAs. (B) After 2 h or 5 h boric acid-induced flicRNA production reduction, the expression of the non-flicRNA-forming genes, EhNudC, EhSMC, and EhRNF (2 h sample only), and the flicRNA-forming genes, EHI_192510, EHI_169670, and EHI_014170, was analyzed by RT-qPCR, compared to untreated control amoebas, set as 1 in each case. The logarithmic plot shows that all expression profiles are statistically significant (ANOVA and Turkey test a posteriori. p ≤ 0.01).

Figure 5

Entamoeba histolytica mRNA and circular RNA biogenesis, transport, and functions. The model described in the main text depicts our current understanding of the protein complexes involved in each process. Components of the spliceosome previously described in our group [31]. Putative methylases and nuclear export proteins that have been previously identified by proteomic analysis of pre-messenger ribonucleoparticles are highlighted in red. Activation/silencing (green upwards/red downwards arrows, respectively) of virulence-related and nuclear gene transcripts (black asterisk) are indicated, as well as the flicRNAs (red asterisk) described here. Green and blue arrowheads indicate adenine N6-methylation and cytosine-5-methylation, respectively. The question mark denotes unproven methylation events. A green bubble with a question mark represents additional factor(s) facilitating flicRNA-spliceosome-Pol II-CTD interaction involved in transcriptional regulation.