Circular Concatemers of Ultra-Short DNA Segments Produce Regulatory RNAs

Abstract

In the ciliated protozoan Paramecium tetraurelia, Piwi-associated small RNAs are generated upon the elimination of tens of thousands of short transposon-derived DNA segments as part of development. These RNAs then target complementary DNA for elimination in a positive feedback process, contributing to germline defense and genome stability. In this work, we investigate the formation of these RNAs, which we show to be transcribed directly from the short (length mode 27 bp) excised DNA segments. Our data support a mechanism whereby the concatenation and circularization of excised DNA segments provides a template for RNA production. This process allows the generation of a double-stranded RNA for Dicer-like protein cleavage to give rise to a population of small regulatory RNAs that precisely match the excised DNA sequences. VIDEO ABSTRACT.

Keywords: DNA concatemers; DNA elimination; DNA repair; Dicer; Ligase IV; Paramecium; Piwi-interacting RNA; ciliates; circular DNA; small RNA; transcription; transposable elements.

Graphical abstract

Figure 1

Injection of DNA Resembling an Excised IES Is Sufficient to Trigger the iesRNA Pathway and Lead to Positive Feedback-Mediated DNA Excision (A) The current model for IES excision. An IES (red) with the TA repeats that mark its boundaries is represented in panel 1. The excisase PiggyMac cleaves at the ends leaving 4 nt 5′ overhangs centered on the TA (panel 2). The 5′ most nucleotide is resected (panel 2) and the broken DNA ends are annealed at the TA, then ligated. The missing nucleotide is filled in during the annealing-ligation process (green nucleotides) and a single TA dinucleotide is left in the genome (blue, panel 3). (B) A cryptic IES oligonucleotide corresponding to crypIESa with its 4 nt 5′ overhangs and 5′ phosphate. The gel represents the product of PCR performed with primers flanking the targeted region. PCR was performed after injected cells had been left to expand vegetatively for four divisions, and the PCR reactions were set up with ten cells each. Lane 1 is uninjected, and lane 2 is successfully injected as is seen from the presence of a lower band representing amplicon with a 27 bp deletion (arrowhead). Sequencing of the band demonstrates the precise excision of the crypIESa, with one TA removed and one retained. (C) Left: representative gels demonstrating excision of mating type IES (injection control) and crypIESb following co-injection of mating type IES-matching small RNAs (do not require Dcl5) and crypIESb DNA in control (lanes 1–4) versus Dcl5-silenced (lanes 5–8) cells. Arrowheads indicate bands corresponding to excision products. Right: co-injection of control RNA (mating type) and cryptic IES DNA was repeated to quantify the relative excision efficiencies in unsilenced (n = 13) and Dcl5-silenced (n = 19) cells. Charts are shown quantifying the relative efficiencies of excision. (D) Cryptic IES DNA with 3′ overhangs does not induce excision. DNA oligonucleotides for crypIESa (3′ overhangs) and crypIESb (control, 5′ overhangs) are shown. PCR products corresponding to excised cryptic IES are only found for crypIESb after coinjection of both oligonucleotides. This experiment was repeated with successful co-injection of crypIESa (3′ overhangs) and crypIESb (control, 5′ overhangs) in 12 cells, none of which exhibited excision of crypIESa. See also Figure S1.

Figure 2

Can iesRNAs Be Generated from Ligated Excised IESs? (A) Representation of the logistical problem of iesRNA transcription from excised IESs. A schematic excised dsDNA IES is shown in red, with its 4 nt 5′ overhangs. The corresponding iesRNA as deduced from small RNA mapping (Sandoval et al., 2014) is shown below in purple. The assembly of an RNA polymerase complex on a short (27 bp) DNA molecule is probably not possible. Even if it were, the RNA (purple) that corresponds to the 5′ end of the excised IES lacks a template for its 5′ nucleotide (indicated with a black arrow). (B) Two hundred iesRNAs that map only partially to IES ends, with overhangs of at least 12 bp, were identified, and the overhangs were mapped. These iesRNAs were chosen randomly by the software Geneious 8.1.8. Overhangs were sorted into groups according to where they map. The majority map at the ends of other IESs, which is a huge overrepresentation considering that the number of possible mapping sites for a random 12 nt sequence within an IES is much greater than at its ends. (C) The oligonucleotide corresponding to three ligated cryptic IESs is depicted, with the TAs that form the theoretical ligation points outlined. Gels of PCR products for each individual 27 bp cryptic IES (all on different chromosomes) are shown. Arrowheads indicate bands corresponding to excision products. Successful injection of the double-stranded oligonucleotide led to excision of all three cryptic IESs from the genome. This was successfully repeated in eight independent injected cells. All cells that exhibited excision of one cryptic IES following concatemer injection also exhibited excision of the other two.

Figure 3

Detection of RNA Corresponding to Concatenated IESs (A) Schematic representation of concatenated IESs forming a DNA circle. The corresponding RNA transcript is a repetitive sequence of the concatenated IESs. Outward-directed primers Fwd and Rev on IES A amplify one repeat of the RNA transcript. Individual IESs are shown in different colors. (B) Following reverse transcription (RT) of total RNA during late development, cDNA was amplified using primer pairs as depicted in (A). Upper: representative gels showing bands (arrows) amplified. Lower: representative result of sequencing. The amplicon contains four different concatenated IESs (1–4) in addition to the primer-containing IES, resulting from transcribed circular DNA. Annotated IESs overlap with their terminal TA dinucleotide, resulting from the concatenation event. (C) Sequence of amplicon from Fwd and Rev primers as in (B), including one additional circle repeat. Primer binding site- containing IESs are annotated in letters A and B, and internal IES are numbered (1–8). Additional concatenated IES transcripts and annotation of IESs are shown in Figure S2 and Table S2.

Figure 4

Sequencing of Small RNAs in Control versus Ligase IV-Silenced Cells (A) RNA reads by length, normalized to 23 nt reads for each sample. Green bars are RNAs that map to macronuclear sequences not including IESs (mature MAC sequence). Red bars are RNAs that map to IES sequences. iesRNAs are IES matching RNAs (red columns) of around 26–31 nt, evident in the control but not in the Ligase IV silencing experiment. Ligase IV silencing leads to an ∼100-fold reduction in iesRNAs (see text). (B) Mapping of 27–31 nt RNAs (iesRNAs) to long (>150 bp) versus short (<31 bp) IESs in control and Ligase IV silencing. Reads are normalized to the number that map to short IESs, for the purposes of clarity as the number of iesRNAs in the Ligase IV silencing is very small. Proportionally more iesRNAs map to long IESs in the Ligase IV-silenced cells than in the control, which is consistent with a model whereby RNA polymerase assembly is a prerequisite of iesRNA production and unligated <31 bp IESs are too short to allow for polymerase assembly. (C) The ratio of iesRNAs that map to middle portions (31^st–150^th bp) versus ends (1^st–30^th bp) of IESs increases drastically in the Ligase IV-silenced cells. This is consistent with a model whereby IES ligation allows transcription through IEs ends. See also Figure S3.

Figure 5

A Model for Small RNA Production from Short DNA Fragments in Paramecium (1) Developing macronuclear genomic DNA is targeted for excision by short RNAs. (2) Excised DNA fragments of different lengths. Blue: >200 bp, other colors: <200 bp. (3) Circularization of longer excised DNA fragments (blue), concatenation and circularization of short fragments (multicolored) by Ligase IV. (4) Bidirectional transcription to form dsRNA. (5) Cleavage of long dsRNA by Dcl5 to produce iesRNAs. (6) iesRNAs target their complementary sequences for excision in a positive feedback loop.

Figure S1

Related to Figure 1 (A) silencing of Dcl5 was confirmed by PCR across an IES that is dependent on iesRNAs, and thus Dcl5, for complete excision. The upper band (arrowhead) corresponds to the PCR product when the IES is retained. Retention is seen in the Dcl5 silenced cells, demonstrating that Dcl5 silencing was successful. (B) co-injection of three cryptic IESs leads to excision of all three in the same cells. Lanes 1 correspond to successfully injected cells and lanes 2 correspond to uninjected cells. (C) Left: crypIESa modified with 3 overhangs and inosine replacing the 3′ most A on the upper strand. Right: gel from PCR to check excision. In two out of three injected cells, inosine rescues the excision.

Figure S2

RNA Transcripts of Concatenated IESs, Related to Figure 3 (A) RNA transcript of a repetitive IES sequence, presumably deriving from a circularised IES (> 200nt in length). Primer sites on the 221bp long IES are annotated, which amplify each IES end followed by the sequence of the opposite IES end. (B) Sequencing results of transcripts from concatenated IES. Primer binding site containing IESs are annotated in letters (A-E) and internal IESs in numbers (1-27). Gaps in sequence annotation result from lack of corresponding sequences in the database. Sequencing the Paramecium MIC genome is expected to reveal additional IESs that are not documented yet, but are present in these concatemers. iesRNAs exclusively map IESs and also map the gap sequences, suggesting these are non-annotated IESs. Point substitutions in the IES sequence are marked in blue and likely derive from sequence amplification with GoTaq® polymerase that lacks proofreading activity. In several cases, only partial IES sequences are detected, suggesting additional cleavage of these IESs occurs either while excision from the genome or after concatenation and circle formation. One example is IES C in A, where nucleotides 5-227 of the total 287bp long IES are forming the repeat.

Figure S3

Related to Figure 4 Survival test following Ligase IVa and IVb silencing. Cells (n = 30) were scored for growth and survival two days after re-feeding following autogamy. In the Ligase IV silenced cells, almost all cells died, confirming that silencing had taken place.