METHOD FOR THE PRODUCTION OF DOUBLE-STRANDED RNA

Abstract

The present invention relates to an isolated nucleotide acid sequence comprising a) the cDNAs of two strands of a target double-stranded RNA (dsRNA) separated by an autocatalytic intron flanked by exon fragments, and b) a plant viroid, wherein element (a) is inserted in the plant viroid sequence. The expression of this nucleotide sequence in a host cell, such as E. coli, along with a tRNA ligase, allows the production of high amounts of dsRNA specific for a target gene. This dsRNA can be used in the interfering RNA technology for silencing gene expression. Additionally, the present invention also relates to a method for producing said dsRNA.

Claims

1. An isolated nucleotide sequence comprising a) the cDNAs of two strands of a target dsRNA separated by an autocatalytic intron flanked by exons or exon fragments, and b) a plant viroid sequence, wherein element (a) is inserted in the plant viroid sequence.

2. The isolated nucleotide sequence according to claim 1, further comprising permuted autocatalytic intron-exon sequences between the plant viroid sequence and the cDNAs sequences.

3. The isolated nucleotide sequence according to claim 2, wherein the permuted autocatalytic intron-exon sequences derive from type I introns or the intron type I of the rRNA 26S from Tetrahymena thermophila.

4. The isolated nucleotide sequence according to claim 1, wherein the autocatalytic intron is an autocatalytic intron type I or an autocatalytic intron type II or an autocatalytic intron type III.

5. The isolated nucleotide sequence according to claim 4, wherein the autocatalytic intron type I is the autocatalytic intron type I of the rRNA 26S from Tetrahymena thermophila.

6. The isolated nucleotide sequence according to claim 1, wherein the plant viroid sequence is the Eggplant latent viroid (ELVd) sequence, the Avocado sunblotch viroid (ASBVd) sequence, the Peach latent mosaic viroid (PLMVd) sequence, or the Chrysanthemum chlorotic mottle viroid (CChMVd) sequence.

7. The isolated nucleotide sequence according to claim 1, wherein the target RNA is a gene from a plant pathogen or pest, or the β-actin gene from Ceratitis capitata.

8. The isolated nucleotide sequence according to claim 1, further comprising a nucleotide sequence encoding a tRNA ligase.

9. The isolated nucleotide sequence according to claim 8, wherein the tRNA ligase is the tRNA ligase from eggplant.

10. A vector comprising an isolated nucleotide sequence according to claim 1.

11. A host cell comprising an isolated nucleotide sequence according to claim 1 or a vector according to claim 10.

12. (canceled)

13. A method for producing dsRNA specific for a target gene comprising co-expressing in a host cell a) a nucleotide sequence according to claim 1 and b) a nucleotide sequence encoding a tRNA ligase, wherein the nucleotide sequences of a) and b) are in the same nucleic acid molecule or in different nucleic acid molecules.

14. The method according to claim 13, wherein the host cell is E. coli.

15. The method according to claim 13, wherein the tRNA ligase is the tRNA ligase from eggplant.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0107] FIG. 1 is a representation of the production of dsRNA in E. coli with the system derived from ELVd, using the autocatalytic intron of T. termophila to separate the inverted repeats.

[0108] FIG. 2 is a scheme of the intron-exon permutation strategy (IEP) and of the self-cleavage reaction of the introns of group I, in which the ligation of the exons and the circularization of the dsRNA is produced.

[0109] FIG. 3 is a representation of a circular double-stranded miniRNA with an extensive perfectly double-stranded region homologous to an endogenous gene of the pest or pathogen of interest, closed in both sides by the fragments of the exons.

[0110] FIG. 4 shows the sequence and predicted structure of minimum free energy of Eggplant latent viroid (ELVd; GenBank accession number AJ536613.1 (SEQ ID NO: 7)). The hammerhead ribozyme domain and self-cleavage site are indicated on grey background and by an arrowhead, respectively. The position U245-U246 where recombinant RNAs were inserted is indicated by an arrow.

[0111] FIG. 5 is a 1% agarose gel stained with ethidium bromide in which the electrophoretic migration of the empty plasmid pLELVd (lane 2) is compared to those of five independent clones of pLELVd-β-actin (lanes 3 to 7). Lane 1 corresponds to a ladder of DNA markers whose sizes in base pairs (bp) are indicated on the left.

[0112] FIG. 6 is a 5% polyacrylamide gel, containing 8 M urea, stained with ethidium bromide in which total Escherichia coli RNA preparations were analyzed. Lanes 1 to 4 correspond to E. coli clones transformed with empty pLELVd, lane 5 corresponds to a ladder of RNA markers whose size in nucleotides (nt) is indicated on the left, lanes 6 to 7 correspond to E. coli clones transformed with pLELVd-β-actin. White, gray and black arrows point to ELVd RNA, ELVd-dsRNA-β-actin, and the spliced intron RNA, respectively.

[0113] FIG. 7 are two 5% polyacrylamide gels, the second containing 8 M urea, stained with ethidium bromide. A total RNA preparation from an Escherichia coli clone transformed with pLELVd-β-actin was separated in the first gel (left) under non-denaturing conditions. After staining the gel, the band pointed by the arrow was cut from the gel and applied onto a second gel (right) that was run under denaturing conditions. The arrow on the right points to the spliced circular dsRNA, which is also schematized in the figure. Both gels contain RNA markers on the left.

[0114] FIG. 8. RNA from E. coli clones co-transformed with plasmids to produce dsRNA including the ELVd scaffold or not and the tRNA ligase or not, as indicated, was purified and analyzed by PAGE and ethidium bromide staining. Recombinant dsRNA accumulation was quantified (in fluorescence arbitrary units) using an image analyzer. Normalized average fluorescence is plotted. Error bars represent standard deviation (n=5).

[0115] FIG. 9. RNA from E. coli clones co-transformed with plasmids to express the tRNA ligase and different versions of the construct to express a 100-bp dsRNA with different exon sizes was purified and analyzed by denaturing PAGE. The gel was stained with ethidium bromide. Lane 1, RNA marker with sizes (in nt) on the left; lanes 2 to 5, RNAs from constructs with 10, 8, 4 and 2 exons, respectively. Black and gray arrows point the positions of the recombinant dsRNAs and the spliced introns, respectively.

EXAMPLES

[0116] Due to the instability of DNA inverted repeats in E. coli cells when plasmids to produce dsRNA to induce RNAi in pests and pathogens were built, bacterial colonies bearing the expected constructs were not obtained. In order to alleviate the well-known problems with DNA inverted repeats in E. coli, the cDNAs corresponding to the two strands of the target dsRNAs were separated with a 433-bp cDNA (SEQ ID NO: 6) corresponding to T. termophila 26S rRNA Group-I intron (GenBank accession number V01416.1 [SEQ ID NO: 23]) plus 10 bp of both native flanking exons. In this way, while the intron cDNA will separate the inverted repeat at the DNA level, the intron would self-splice from the transcript allowing formation of the dsRNA.

[0117] For this purpose, we amplified by PCR using the Phusion high-fidelity DNA polymerase (Thermo Scientific) two different DNAs corresponding to a 100-bp fragment of C. capitata beta-actin mRNA (GenBank accession number EU665679.1 [SEQ ID NO: 24]) in both orientations (SEQ ID NO: 1 and SEQ ID NO: 2). We also amplified by PCR a DNA corresponding to T. termophila 26S rRNA group-I intron with fragments of flanking exons (SEQ ID NO: 6), from a plasmid obtained by gene synthesis. PCR products were separated by electrophoresis in a 1% agarose gel and the gel stained with ethidium bromide. DNAs of the expected size were eluted from the gel using silica-gel spin columns. The three DNAs were assembled between the positions T245 and T246 of the ELVd-AJ536613.1 cDNA with duplicated hammerhead ribozyme domains (SEQ ID NO: 11) in plasmid pLELVd (Daròs, J. A., Aragonés, V., and Cordero, T. (2018) Sci. Rep., 8, 1904), by means of the Gibson assembly reaction using the NEBuilder HiFi DNA assembly master mix (New England Biolabs). Competent E. coli cells were finally electroporated with the products of the Gibson assembly reactions and transformed bacteria selected in Luria-Bertani (LB) plates containing 50 μg/ml ampicillin. We grew five independent E. coli clones in liquid LB cultures and purified the corresponding plasmids. An analysis by electrophoresis in a 1% agarose gel that was stained with ethidium bromide showed all of them being the same size and exhibiting a migration delay consistent with the inserted DNAs (FIG. 5, compare lane 2 with lanes 3 to 7). The expected sequence SEQ ID NO: 12) was experimentally confirmed in one of these plasmids. This plasmid was used to co-electroporate the RNase III-deficient HT115(DE3) strain of E. coli that lacks RNase III (Timmons, L., Court, D. L., and Fire, A. (2001) Gene, 263, 103-112), along with plasmid pLtRnISm (Daròs et al., 2018, cited ad supra) from which eggplant tRNA ligase (SEQ ID NO: 13 and SEQ ID NO: 14) is constitutively expressed. As controls, pLtRnISm was also co-electroporated with the plasmid to express the empty ELVd (pLELVd) (Daròs et al., 2018, cited ad supra). Nine independent colonies (four for the empty ELVd control) were picked from LB plates containing 50 μg/ml ampicillin and 34 μg/ml chloramphenicol and grown for 24 hours in liquid Terrific Broth (TB) medium. Total RNA was extracted from the cells using a 1:1 mix of phenol and chloroform (pH 8.0) (Daròs et al., 2018), and analyzed by polyacrylamide gel electrophoresis (PAGE) in denaturing conditions (8 M urea) (Daròs et al., 2018, cited ad supra). Two prominent bands above the 600-nt and slightly below the 400-nt RNA markers were observed in the RNA extracts from bacteria transformed with the plasmid to express the dsRNA (FIG. 6, lanes 6 to 14, gray and black arrows respectively). Remarkably, these bands exhibited a fluorescence intensity comparable to those from the endogenous E. coli rRNAs (FIG. 6, upper part of the gel), indicating a large accumulation in E. coli. RNA extracts from the empty ELVd controls exhibited a single prominent band above the 400-nt marker that, according to our previous analysis (Fadda, Z., Daròs, J. A., Fagoaga, C., Flores, R., and Duran-Vila, N. (2003) J. Virol., 77), corresponds to the 333-nt-long circular ELVd RNA (FIG. 6, lanes 1 to 4, white arrow). Circular RNAs migrate slower than the linear counterparts of the same size in denaturing conditions. The identity of the different RNA species indicated in the FIG. 2 was confirmed by Northern-blot hybridization analysis using .sup.32P-labelled RNA probes complementary to the ELVd and the T. termophila intron RNAs.

[0118] Taken together, these results indicate that, while the T. thermophila cDNA serves to stabilize the inverted repeats in the E. coli expression plasmids, the intron very efficiently self-splices from the primary transcript in bacterial cells facilitating the accumulation of large amounts of a circular RNA product that consists of an ELVd scaffold from which the dsRNA bulges. This is schematized in FIG. 1. The dsRNAs produced in E. coli with this method showed the same insecticide activities as the counterparts of the same sequence obtained using the standard methodologies, such as two-strand transcription from two opposite promoters.

[0119] Next, we designed a novel strategy to separate the active dsRNA from the undesired ELVd scaffold. For this purpose, we took advantage of the permuted intron-exon (PIE) reaction (Price, J. V, Engberg, J., and Cech, T. R. (1987) J. Mol. Biol., 196, 49-60; Puttaraju, M. and Been, M. (1992) Nucleic Acids Res., 20, 5357-5364). We built a new plasmid in which a cDNA corresponding to the T. termophila 26S rRNA group-I intron with 10 nt of the flanking exons (SEQ ID NO: 6) was split and permuted, and the 3′ fragment (SEQ ID NO: 15) inserted between the 5′ moiety of ELVd and the plus strand of the dsRNA, and the 5′ fragment (SEQ ID NO: 16) inserted between the minus strand of the dsRNA and the 3′ moiety of the ELVd (see FIG. 2). We built this plasmid by amplifying the DNAs by PCR with the Phusion high-fidelity DNA polymerase and assembling the fragment by the Gibson reaction. The sequence of the plasmid was confirmed by sequencing. The selected plasmid, along pLtRnISm to co-express the eggplant tRNA ligase, was used to electroporate competent E. coli cells. Transformed bacteria were selected in LB plates containing ampicillin and chloramphenicol. Next, we grew liquid cultures in TB, containing the same antibiotics, and extracted total RNA from bacteria using phenol:chloroform. The RNA preparation was analyzed by PAGE under non-denaturing and denaturing (8 M urea) conditions (FIG. 7). The alteration of electrophoretic migration from non-denaturing to denaturing conditions confirmed the production in E. coli of a circular RNA molecule consisting of a 100-bp dsRNA locked at both ends by the exon fragments (FIG. 3), as a result of the self-cleavage of the regular intron I and the permuted intron I (FIG. 2).

[0120] In order to analyze the importance of the viroid sequences during production of recombinant double-stranded RNA (dsRNA), the viroid sequence from a previously described construct (SEQ ID NO: 17) expressing a Ceratitis capitata beta-acting 100-bp dsRNA using the viroid-type I intron system and including sequences for the viroid scaffold (SEQ ID NO: 25) was removed using standard molecular biology techniques, generating a new version of this plasmid (SEQ ID NO: 26).

[0121] E. coli was transformed with these two plasmids and those expressing the eggplant tRNA ligase or an empty version of it. Five colonies were picked from each transformation, selected in ampicillin and chloramphenicol, and liquid cultures were grown for 24 hours. The production of bacterial dsRNA RNA was extracted and analyzed by polyacrylamide gel electrophoresis (PAGE) followed by ethidium bromide staining of the gel. Bands corresponding to the dsRNA product were quantified using an image analyzer. Results demonstrated that the viroid scaffold significantly has a positive effect on recombinant dsRNA accumulation, since average accumulation in E. coli co-transformed with the plasmid that expresses de viroid sequences and the tRNA ligase exhibited two-fold more dsRNA than those co-transformed with the plasmid with no viroid sequences and the empty version of the tRNA ligase (FIG. 8).

[0122] Additionally, to gain insight on exon fragment requirements in the viroid-based system to produce recombinant dsRNA in E. coli, starting from the construct that expresses a 100-bp dsRNA corresponding to Ceratitis capitata beta-actin mRNA, intron processing and recombinant dsRNA accumulation with successive exon deletions was analyzed. E. coli was co-transformed with the plasmid to express the tRNA ligase and those to produce the recombinant dsRNA with 10, 8, 6 and 2-nt-long exons (SEQ ID NO: 27, 28, 29 and 30). Transformed clones were selected in plates with ampicillin and chloramphenicol and liquid cultures grown for 24 h. RNA was extracted from bacteria and analyzed by PAGE and ethidium bromide staining. Results demonstrated that the exon size, from 10 to 2 nucleótidos, has no effect on intron self-cleavage efficiency and recombinant dsRNA production (FIG. 9).

[0123] In conclusion, we have worked out a new method to produce dsRNA that can be used in RNAi-based biotechnology applications. The new method is based on the activity of two versions of a self-splicing group-I intron, the second in a permuted form. The first intron stabilizes the DNA inverted repeats that are required to transcribe a dsRNA in E. coli, while is efficiently processed from the final transcript. The second (permuted) intron cleaves the dsRNA from the viroid scaffold producing a circular molecule consisting of the dsRNA locked at both ends by the exon fragments (FIG. 3).