Fig 1.
Back-spliced CRM1 RNA in the maize centromere.
(A) Procedure for anti-CENH3 RIP and subsequent high-throughput sequencing and cDNA library screening. (B) BLAST results of the back-spliced CRM1 reads from the anti-CENH3 RIP-seq, input-seq data, and the 354-nt RNA from the anti-CENH3 RIP cDNA library. The arrow shows the back-splicing site. (C) The location of the 607-bp combined sequence in CRM1, and the back-spliced form of the 354-nt RNA. The red line represents the 607-bp sequence. (D) Distribution of the 354-bp sequence on CRM1. The first track of each panel represents the centromeric region indicated by CENH3 enrichment, and the peak height represents the RPM value (0–1). The other 4 tracks represent the distributions of the 354-bp, 269-bp, 85-bp, and CRM1 sequences along a specific region of cen5. The lower panel shows a detailed version of the information displayed in the upper panel. The arrows inside the rectangular bars represent the directions of the sequences. (E) The public raw genome sequencing data (including Pacbio [65×] and Illumina [100×] reads) and 4 anti-CENH3 ChIP-seq datasets from B73 (including 1 generated in this study and 3 from public resources), together with 1 input-seq dataset, were mapped to the assumed 354-bp DNA. Only one read from anti-CENH3 ChIP-seq dataset was matched to the region containing the back-spliced junction site (purple line). All the other reads show no covering the back-spliced junction site. The data underlying this figure can be found in the GEO with accession numbers GSE124242, SRR3018834, SRR2000635, SRR2000640, SRR2000648, SRX1472849, and SRX1452310 and on Github (https://github.com/sxx-ying/maize-centromere-circRNA). CB, chromatin binding; CENH3, centromeric H3 variant; ChIP-seq, chromatin immunoprecipitation following high-throughput sequencing; Chr, chromosome; CRM, centromeric retrotransposon; Gag, gag protein; GEO, Gene Expression Omnibus; input-seq, input sequencing; nt, nucleotides; PR, protease; RIP, RNA immunoprecipitation; RIP-seq, RIP sequencing; RNH, RNase H; RPM, reads per million; RT, reverse transcriptase.
Table 1.
The number of merged fragments from anti-CENH3 RIP mapped to the CRM elements.
Fig 2.
Detection of the full-length circular CRM1 RNAs.
(A) The 354-nt RNA was stable after an RNase R treatment. Cenh3 mRNA was used as a linear RNA control. The black arrows in the right panel show the positions of primers used. (B) Divergent primers F4+R2 were used to detect the existence of the 354-nt circular RNA (left panel). The right panel shows the form of the 354-nt circular RNA. (C) Divergent primers F2+R2 targeting the 269-nt sequence confirmed the existence of the 607-nt circular RNA (left panel). The right panel shows the form of the 607-nt circular RNA. (D) Divergent primers F2+R2 also confirmed the existence of 277- to 296-nt circular RNAs (left panel). The right panel shows the form of the 277- to 296-nt circular RNA. In (B–D), the upper models show the position of primers on the 354-nt sequences, while the lower models show the amplified sequences. The right schematic diagrams show the circular RNA with corresponding sizes. (E) Divergent PCR showed that the 607-nt and 277- to 296-nt RNAs were stable after RNase R treatment. The right panel shows the compositions of the amplified sequences mentioned in Figs 2C, 2D and S2E. (F) A northern blot was performed using B73 RNA purified with biotinylated antisense or sense oligos, then detected using digoxin-labeled 25-bp antisense or sense probes. The probe was located in the 269-nt region. The RNA was run in 3% denaturing formaldehyde agarose gel. (G) AFM image of the circular CRM1 RNAs. The RNAs were purified by biotinylated sense oligo, followed by RNAse R treatment. The white arrows indicate the circular RNA. The scale bar is 800 nm. In (A–E), the yellow, red, green, and purple arrows represent the directions of the sequences. The short arrows under or above the sequences represent the positions of the primers. The data underlying this figure can be found in S1 Raw Images. AFM, atomic force microscopy; CRM, centromeric retrotransposon; nt, nucleotides; RT-PCR, reverse transcription PCR.
Fig 3.
Circular CRM1 RNAs induce chromatin loops in the centromere.
(A) Anti-S9.6 RIP-qPCR was used to confirm the R-loop formation by 354-, 607-, and 277- to 296-nt circular CRM1 RNAs. Zm00001d007960 RNA was used as a negative control, and rRNA was used as a positive control. Chromatin-binding RNA was used for RIP. Actin was used as an internal reference gene. (B) Regions chosen for detecting the ssDNA sites are marked as 85–1, 253–1, 269–1, and 269–2. (C) ssDNA sites in CRM1 were checked using an S1 nuclease treatment of the nuclear DNA. DNA with no S1 nuclease treatment was used as a control template. The 607-left sequence was used as an internal reference gene. (D and E) Potential chromatin loops were induced by circular RNA inside a single CRM1 element (D) and between two CRM1 elements (E). Red, green, and yellow lines represent the 85-, 269-, and 253-bp regions, respectively. Black lines represent sequences on the left side of the 85-bp sequence and the right side of the 269-bp sequence. The blue ovals represent circular CRM1 RNAs. ①, ‘①’, ②, and ③ represent the broken ends on the two sides of the 253-bp sequence, the left side of the 85-bp sequence, and the right side of the 269-bp sequence. (F) 3C-PCR confirms the potential ligations of chromatin loops after DpnII digestion. The left panel shows the PCR results in the undigested, unligated samples and 3C samples under potential ligation forms. The right panel shows the sequences from the bands on the left, including the expected sequences, the first and the second part of the expected sequences, and the amplified sequences. (G and H) 3C-qPCR shows chromatin interactions inside a single CRM1 element (G) and between two CRM1 elements (H). The interaction frequencies between two DpnII-digested fragments were normalized to the 3C control template from the digested and ligated centromeric BAC clone and an internal reference gene, SAM. In (A), (C), (G), and (H), the columns and error bars represent the relative value and standard error of the means (n = 3). In (A) and (C), the P values were determined using a Student t test: *P < 0.05, **P < 0.01. The data underlying this figure can be found in S1 Data and S1 Raw Images. 3C, chromatin conformation capture; BAC, bacterial artificial chromosome; CRM, centromeric retrotransposon; IgG, Immunoglobulin G; nt, nucleotides; qPCR, quantitative PCR; RIP, RNA immunoprecipitation; ssDNA, single-strand DNA.
Fig 4.
Decreased chromatin loops and CENH3 localizations in the CRM1 regions of the RNAi plants.
(A and B) RT-qPCR analysis of the levels of the 354-, 607-, and 277- to 296-nt circular RNAs in the T1 generation of RNAi plants 5 and 18, with HiII as the control. (C) Level of linear RNAs (RNA-85, RNA-269, and RNA-85+269) in the T1 generation of RNAi plants, with HiII as the control. (D) Seedlings of the T1 RNAi plants. (E) Anti-S9.6 RIP quantification of R-loops in the circular CRM1 RNAs in the T1 RNAi plants. (F) Anti-S9.6 RIP quantification of R-loops in the linear RNAs in the T1 RNAi plants. (G and H) 3C-qPCR analysis of the chromatin interactions inside a single CRM1 element (G) and between two CRM1 elements (H) in the T1 generation of RNAi line 5. Data were normalized to the cross-link frequencies of the 3C control template composed of DpnII-digested and ligated centromeric BAC and the internal reference SAM. The primers used were the same as in Fig 3. (I) CENH3 signals in the T1 RNAi plants. Blue indicates DAPI. Red indicates the CENH3 signals. Bar = 10 μm. In (A–C,) and (E–H), the columns and error bars represent the relative value and standard error of the means (n = 3). In (A–C), (E), and (F), Actin was used as an internal reference gene, the P values were determined using a Student t test: *P < 0.05, **P < 0.01. The data underlying this figure can be found in S1 Data. 3C, chromatin conformation capture; BAC, bacterial artificial chromosome; CENH3, centromeric H3 variant; CRM1, centromeric retrotransposon; IgG, Immunoglobulin G; RIP, RNA immunoprecipitation; RNAi, RNA interference; RT-qPCR, reverse transcription quantitative PCR.
Fig 5.
Conserved back-splicing process of retrotransposons in numerous crops.
(A) The procedure of protoplast transformation using in vitro–transcribed CRM1 RNA. The digestion sites are marked with red triangles. (B and D) Three 354-nt–like back-spliced RNAs from transformed oat (B) and soybean (D) protoplasts. The red frames mark the labeled digestion sites. The first 3 tracks show the sequences from transformed protoplasts, and the last tracks show the 354-nt RNA sequence. (C and E) The detailed information of the 3 back-spliced sequences from oat (C) and soybean (E) protoplasts. The left panels show the sequence position on the 1,671-nt sequence. The right panel shows the final back-spliced sequences. The dotted lines mark the intermediate region. The green bars show the downstream parts of the 1,671-nt sequence, and the red bars indicate the upstream parts. (F) A 323-nt back-spliced RNA from wheat retrotransposons, consisting of a 288-nt (green) and a 135-nt (red) sequence. (G) Detailed information for the 323-nt and 221-nt back-spliced sequences. The left panel shows the position of the sequence on the retrotransposon, while the right panel shows the final back-spliced sequences. The dotted lines indicate the intermediate region, green bars represent the downstream regions, and red bars indicate the upstream regions. CRM, centromeric retrotransposon; nt, nucleotides.
Fig 6.
The model for roles of circular CRM1 RNAs in centromere structure and function.
Circular CRM1 RNAs can bind to the chromatin through R-loops to induce chromatin loop formation. The chromatin binding of circular RNAs can repress the formation of R-loops by related linear RNAs. Higher numbers of R-loops and lower numbers of chromatin loops lead to lower levels of CENH3 localization in the centromere. CENH3, centromeric H3 variant; CRM, centromeric retrotransposon.