WO2017188707A1

WO2017188707A1 - Dicer substrate rna nanostructures with enhanced gene silencing effect and preparation method thereof

Info

Publication number: WO2017188707A1
Application number: PCT/KR2017/004389
Authority: WO
Inventors: Hyukjin Lee; Bo-Young Kim; Bo Ra Jang
Original assignee: Ewha Womans University
Current assignee: Ewha Womans University
Priority date: 2016-04-29
Filing date: 2017-04-29
Publication date: 2017-11-02
Anticipated expiration: 2018-10-29

Abstract

The present invention relates to a dicer substrate RNA nanostructure exhibiting enhanced gene silencing effects and a preparation method thereof. The RNA nanostructure of the present invention includes a plurality of the same or different RNAi sequences in a single RNA nanostructure, and thus can selectively, simultaneously and effectively regulates expressions of various genes. In addition, activation of RNAi sequences in the RNA nanostructure can be controlled in a programmed manner by controlling the opening or closing of the loops of the RNA nanostructure during production of the RNA nanostructure.

Description

DICER SUBSTRATE RNA NANOSTRUCTURES WITH ENHANCED GENE SILENCING EFFECT AND PREPARATION METHOD THEREOF

The present invention relates to dicer substrate RNA nanostructures with an enhanced gene silencing effect and preparation method thereof.

Nucleic acids, which have been recognized as genomes that store and transfer biological genetic information, have recently been recognized to have a potential as functional biopolymers based on the self-assembly property by complementary sequence of nucleic acid, and have been studied in various fields, including construction of nano-sized delicate structures using the same. As such, nucleic nanostructures to be prepared simply not only prepare nanostructures themselves but also provide a prospective field of study, which has great potentials that can be applicable to therapeutic fields, medical diagnostic fields, criminal investigation fields, environmental fields, and the like.

Specifically, since it has been known that a certain disease can be effectively treated by RNA interference (RNAi), there have been attempt to use RNA drugs to treat cancers and various gene-related diseases, and studies on various nucleic acid structures have been conducted to achieve effective gene silencing. The discovery of RNA interference has opened up an entirely new field of biology and medicine. The ability of RNAi to specifically silence target genes has yielded not only a new tool for basic research but also raised the concept of developing medicines based on RNAi.　 RNAi works through the targeting of mRNA via sequence-specific binding and results in degradation of target mRNA or its translational inhibition, leading to the loss of protein expression. This is pharmacologically achieved via the introduction of small 19-21 bp dsRNA molecules called small interfering RNA (siRNA). Since its discovery 10 years ago, siRNA has been widely investigated in vitro for its utility in treating various diseases, such as cancer, neurodegenerative and infectious diseases.

It was reported that 25-35 bp dsRNAs longer than small interfering RNAs (siRNAs) are recognized by the intracellular enzymatic protein dicer and exhibits an effective target gene-silencing effect by their conversion to 19-20 bp small interfering RNAs. Such relatively long dsRNAs have been used under the name “dicer substrate RNAs”. The roles of dicer protein in RNA interference are to cut microRNAs (miRNAs and shRNA) to suitable sizes and to help form RNA-induced silencing complex (RISC) to induce RNA interference. It was reported that, through this function of the dicer to form the RISC, dicer substrate RNAs exhibit a more effective gene-silencing effect compared to general siRNAs.

However, in disease treatment methods using such gene-silencing RNAi, there has long been a problem with limitation that cannot effectively deliver DNA drug in vivo due to the relatively large molecular weight (for example, 13-20 kDa) and polyanionic nature (for example, 40 negative phosphate charges), a RNA drug cannot be effectively delivered in vivo. Furthermore, there are also problems in that naked RNAs cannot freely cross the cell membrane, and in that unmodified naked RNAs are relatively unstable in blood and serum, as they are rapidly degraded by endonuclease or exonuclease, meaning that they have short half-lives in vivo. In an attempt to solve such problems, a method has been reported in which chemical modifications are introduced into RNA duplex structure so as to enhance biological activity without adversely affecting gene-silencing activity. That way, several chemical modifications to the backbone, base, or sugar of the RNA have been employed to enhance siRNA stability and activity.

[Non-Patent Documents]

Kuriyan and O'Donnell. J Mol Biol. 1993; 234: 915-925

The present invention is intended to provide a dicer substrate RNA nanostructure with an enhanced gene-silencing effect.

The present invention is also intended to provide a method for preparing the RNA nanostructure.

The present invention is also intended to show that the use of the RNA nanostructure can selectively, simultaneously and effectively regulate multiple target genes that are expressed in cells.

The present invention provides an RNA nanostructure having radially extending arms which are K in number, wherein K is an integer ranging from 3 to 90.

Each of the arms of the RNA nanostructure according to the present invention consists of a stem and a loop,

wherein

1) the stem of each arm consists of a core and an RNAi sequence,

i) the core is 2-20 nucleotides (nt) in length, and

ii) the RNAi sequence is double-stranded sequence which consists of sense and antisense sequences(strands) for any gene, and is 19 or 20 nt in length,

wherein the antisense sequence contains a nucleotide in which a U (uracil)-A (adenine) base pair is present at 5'-end or the second nucleotide from the 5'-end, and the sense sequence contains a nucleotide in which a G (guanine)-C (cytosine) base pair is present at 5'-end,

the antisense sequence contains a nucleotide in which 4 or more U (uracil)-A (adenine) base pairs are present within the 7th nucleotides from the 5'-end (seed region),

2)　the RNAi sequences included in the stem of each arm are the same or different, and

3) the loop of each arm consists of 2-30 nucleotides and is open or closed, respectively, and where the loop is open, it has a single-stranded RNA (overhang) of 2-10 nt in length at the 3'-end of the RNAi sequence.

In one embodiment of the present invention, the 5'-end of the antisense sequence of the RNAi sequence included in the stem of each arm in the RNA nanostructure of the present invention may towards the loop (“forward structure”), or the 5'-end of the sense sequence may towards the loop (“reverse structure”).

In one embodiment of the present invention, where the loop of each arm in the RNA nanostructure of the present invention is open, the nanostructure may have a single-stranded RNA (overhang) of 2 nt in length at 3'-end.

In one embodiment of the present invention, the core of each arm in the RNA nanostructure of the present invention may have a G (guanine)-C (cytosine) contents (G-C contents) of 50% or more.

In one embodiment of the present invention, K in the RNA nanostructure of the present invention may be 3 or 4, but is not limited thereto.

The present invention also provides a method for preparing an RNA nanostructure having radially extending arms which are K in number, the method comprising the steps of:

(A) preparing a circular DNA template strand having radially extending arms which are K in number, wherein K is an integer ranging from 3 to 90, and each of the arms consists of a stem and a loop,

wherein

1) the stem of each arm consists of a core and an RNAi sequence,

i) the core is 2-20 nucleotides (nt) in length, and

ii) the RNAi sequence is double-stranded sequence which consists of sense and antisense sequences for any gene, and is 19 or 20 nt in length,

the antisense strand contains a nucleotide in which 4 or more U (uracil)-A (adenine) base pairs are present within the 7th nucleotides from the 5'-end (seed region),

3) the loop of each arm consists of 2-30 nucleotides, and at least one of the loops of each arm consists of 8-30 nt in length;

(B) treating the circular DNA template strand, prepared in step (A), with RNA polymerase, and resulting in an amplified RNA product by RCT (rolling circle transcription);

(C) treating the amplified RNA product with DNase I to remove the circular DNA template strand;

(D) treating the amplified RNA product, from which the circular DNA template strand was removed, with at least one DNA helper, and treating the amplified RNA product with RNase H to cleave a portion forming a RNA/DNA double strand with the DNA helper; and

(E) obtaining the RNA nanostructure by self-assembly of the cleaved amplified RNA product.

In one embodiment of the present invention, the step (A) may comprises the steps of:

a) preparing DNA template strands which is K in number, in which sense and antisense sequences for any gene are present at both ends of the DNA template strand, respectively, while a portion for forming the loop of the circular DNA template strand is interposed between the sense strand and the antisense strand, and a portion for forming the core of the circular DNA template strand is present at the same end of each of the sense strand and the antisense strand;

b) hybridizing the DNA template strands which are K in number to allow complementary sequences to bind to each other, thereby preparing a DNA template strand having nicks which are K in number; and

c) treating the DNA template strand having nicks which are K in number, with DNA ligase, thereby preparing the circular DNA template strand.

However, the method for preparing the circular DNA template strand is not limited thereto.

In one embodiment of the present invention, the DNA helper in step (D) may be a DNA helper that is 2-30 nt in length and is 2'-O-methylated at a specific part of nucleotide. The specific part of nucleotide may be 3'-end, 5'-end or any middle part of the nucleotide, including at least 4 of un-modified parts.

In one embodiment of the present invention, the RNA polymerase in step (B) may be T7-RNA polymerase. However, the RNA polymerase is not limited thereto, and may be any RNA polymerase.

In one embodiment of the present invention, in step (B), the emplified RNA product can be obtained by treating the circular DNA template strand with a primer together with the RNA polymerase and using RCT, wherein the primer may be a RNA sequence of 12-30 nt in length complementary to the loop of DNA template strand of 8-30 nt in length.

In one embodiment of the present invention, one or more of the loops of the circular DNA template strand in step (A) may comprise an RNA polymerase promoter which is recognized by the RNA polymerase.

In one embodiment of the present invention, K may be 3 or 4. However, K is not limited thereto.

The present invention also provides an RNA nanostructure prepared by the method.

The present invention also provides a method for inhibiting the expression of a target gene by the use of the RNA nanostructure.

The present invention provides a method for treating disease, comprising administering the RNA nanostructure to a subject to inhibit the expression of a target gene in the subject. Examples of diseases that may be treated by the present invention include, but are not limited to, various cancers including thyroid cancer, gastric cancer, colorectal cancer, lung cancer, liver cancer, breast cancer, prostate cancer, gallbladder cancer, bile duct cancer, pancreatic cancer, and blood cancer such as non-Hodgkin lymphoma and the like; autoimmune diseases including rheumatoid arthritis, psoriasis, Crohn's　disease and the like; amyloidosis including hereditary ATTR amyloidosis; hemophilia and other bleeding disorders; complement-mediated diseases including infectious disease; hepatic porphyria; alpha-1 antitrypsin deficiency; primary hyperoxaluria type 1; beta-thalassemia; iron overload disorder; hereditary angiodema, and the like. In addition, the present invention may be used in various genetic medicine programs.

The present invention relates to RNA nanostructure with an enhanced gene-silencing effect as a dicer substrate, and preparation method thereof. The RNA nanostructure of the present invention comprises a plurality of the same or different RNAi sequences in a single RNA nanostructure, and thus can selectively, simultaneously and effectively regulates the expression of multiple genes. In addition, in preparation of the RNA nanostructure of the present invention, activation of the RNAi sequences can be controlled in a programmed manner by controlling the opening or closing of the loop of the RNA nanostructure.

FIG. 1 is a schematic view showing Y-shaped nanostructures having three arms, among RNA nanostructures produced by the method of the present invention. Illustratively, the RNAi sequence included in each arm is labled with GFP, RFP or BFP, which is a fluorescent protein, respectively.

FIG. 2 is a schematic view showing an RNA nanostructure produced by the method of the present invention.

FIG. 3 is a schematic view showing a case in which an RNAi sequence included in the stem of the arm of an RNA nanostructure was cleaved in the structure. The RNAi sequence is a 19-bp double-stranded sequence containing a sense sequence and an antisense sequence, and has a 2-nt overhang at the 3'-end of each of the sense sequence and the antisense sequence.

FIG. 4 is a schematic view showing a forward structure (left) and a reverse structure (right).

FIG. 5 is a schematic view showing the design of an RNAi sequence included in the stem of the arm of an RNA nanostructure. The RNA nanostructure of the present invention is designed as shown in the top of FIG. 5.

FIG. 6 shows a reverse structure which an RNAi sequence is 19+2 nt.

FIG. 7 shows a forward structure which an RNAi sequence is 19+2 nt.

FIG. 8 shows a reverse structure which an RNAi sequence is 20+2 nt.

FIG. 9 shows a forward structure which an RNAi sequence is 20+2 nt.

FIG. 10 shows a mechanism by which the RNA nanostructure of the present invention induces gene silencing when it is administered to a subject.

FIG. 11 shows a method of preparing an RNA nanostructure by preparing a circular DNA template strand, producing an RNA nanostructure by RCT using the circular DNA template strand, allowing DNA helper to bind complementarily to one or more of the loops of the RNA nanostructure, and opening the loop by RNase H.

FIG. 12 shows a DNA template strand in which nicks are present in the loops (left), and a DNA template strand in which nicks are present in the stems (right).

FIG. 13 more specifically shows the structure of a DNA template strand in which nicks are present in the stems.

FIG. 14 is a schematic view showing a mechanism by which RNA polymerase binds in a method for preparing an RNA nanostructure according to the present invention. In FIG. 14, shows a case in which a promoter, to which RNA polymerase can bind, is present in the loop of a circular DNA template strand; shows a case in which a double strand is formed using a suitable primer that binds complementarily to the loop of a circular DNA template strand, so that RNA polymerase will recognize and bind to the double strand; and shows a case in which RNA polymerase recognizes and binds to the loop of a circular DNA template strand by recognizing the loop itself as a bubble structure.

FIG. 15 is a schematic view showing a process of synthesizing a helper-free Y RNA nanostructure according to the present invention.

FIG. 16 shows the results of PAGE analysis of various RNA nanostructures having an open or closed loop according to the present invention. Wherein, FIG. 16 (a) is the result of the PCT amplified product, and FIG. 16 (b) is the result of the self-assembled product after treatment with RNase H.

FIG. 17 shows the results of PAGE analysis of a helper-free Y RNA nanostructure.

FIG. 18 shows an AFM image of an amplified RCA product obtained using the circular DNA template strand of the present invention and an AFM image of a self-assembled RNA nanostructure.

FIG. 19 shows a portion of GFP mRNA, determined to be targeted by an RNAi sequence.

FIG. 20 depicts a graph showing the GFP gene expression inhibitory effects (left) and the gene silencing efficacy (right) of 19+2 siGFP, 23+2 siGFP, and 23+2 siGFP after Dicer, which are RNAi sequences against GFP.

FIG. 21 depicts a graph showing the GFP expression inhibitory effects (left) and the gene silencing efficacy (right) of 25+2Y-RNA, which is a Y-shaped RNA nanostructure comprising a RNAi sequence against GFP, and 25+2Y-RNA after Dicer obtained by previously treating the 25+2Y-RNA with a dicer in vitro.

FIG. 22 depicts graphs showing the results of FACS performed to examine the gene silencing efficacy of the RNA nanostructure of the present invention.

FIG. 23 depicts 3D FACS plots showing the gene regulatory effect of the RNA nanostructure of the present invention.

FIG. 24 compares dicer processing between the presence and absence of a 2-nt overhang.

FIG. 25 shows a schematic view showing expected mechanisms for an Y nanostructure serving as a dicer substrate and a non-dicer Y nanostructure (a), dicer processing using Y nanostructures having various arm lengths (b), the gene expression inhibitory effects of Y nanostructures having various arm lengths (c), and the long-term gene-silencing effect of a dicer substrate Y nanostructure (d).

FIG. 26 shows a schematic view of Linear DsiGDH, pre-Let7 siGDH and siY-GDH, used to identify the gene-silencing effect against a target gene GAPDH.

FIG. 27 depicts a graph showing the gene-silencing effect of Linear DsiGDH, pre-Let7 siGDH and siY-GDH against a target gene GAPDH.

Hereinafter, the present invention will be described in detail to help understand the present invention. However, the following description is provided for better understanding of the present invention and is not intended to limit the scope of the present invention.

RNA Nanostructure

Hereinafter, the structure of an RNA nanostructure according to the present invention will be described.

(1) Overall Structure

An RNA nanostructure according to the present invention has radially extending arms that are K in number. As used herein, the term “radially” refers to a shape by which the arms that are K in number extend radially (that is, a shape by which the arms extend from one central point in all directions, like a cobweb or a spoke in a two or three-dimensional view). K may be an integer ranging from 3 to 90. For example, where K is 3, the RNA nanostructure of the present invention has a Y shape. Where K is 4, the RNA nanostructure of the present invention has a “＋” shape. However, K is not limited to 3 or 4. Due to its structure, the RNA nanostructure of the present invention has an advantage in that it has increased resistance to in vivo nuclease such as endonuclease or exonuclease, indicating that it has increased stability in blood. Furthermore, because the size of the RNA nanostructure of the present invention can be controlled by controlling the length of each arm and the number of the arms, the RNA nanostructure of the present invention can exhibit an EPR (enhanced permeability and retention) effect. As used herein, the term “EPR effect” means that a molecule having a certain size tends to be accumulated in tumor tissue rather than normal tissue. In addition, because it is known that nanostructures are accumulated in different organs depending on their size (Nature Biotechnology 33, 941-951 (2015)), a target organ can be determined by controlling the length of each arm and the number of the arms.

Each of the arms in the RNA nanostructure of the present invention consists of a stem and a loop. The stem portion of each arm comprises a core portion and an RNAi sequence portion. The stem of each arm comprises one or more, preferably one RNAi sequence, and each RNAi sequence can be activated when the loop portion of each arm is open. The loop portion of each arm has one or more portions complementary to one or more DNA helpers, and thus a portion that forms RNA/DNA binding by binding with DNA helper can be cleaved by RNase H in the production process to open the loop. The RNA nanostructure of the present invention has advantages in that, because it includes a plurality of different RNAi sequences in a single structure, it can simultaneously silence expressions of multiple genes, and because it includes a plurality of RNAi sequences in a single structure, the local concentration of the RNAi sequence increases so that the binding affinity of a dicer will increase.

FIGS. 1 and 2 are schematic views showing examples of the RNA nanostructures of the present invention. FIG. 1 shows Y-shaped RNA nanostructures, each having three radially extending arms. Each of the arms may include a tag sequence that enables activation of the RNAi sequence to be confirmed based on color, fluorescence or the like. For example, an RNAi sequence against the fluorescent protein GFP, an RNAi sequence against RFP, an RNAi sequence against BFP, or the like, may be used, and in FIG. 1.

The portions indicated by ① to ⑤ on the left of FIG. 1 correspond to portions that are cleavable by RNase H. Each of the loops has one or more cleavable portions. Specifically, in the case of the RNA nanostructure shown in FIG. 1, the size of the RNAi sequence against GFP, included in the upper arm, is relatively large, and the upper loop is long so that RNA polymerase can bind thereto. Thus, the upper loop has two cleavable portions, and the two lower loops each have one cleavable portion. The number of the cleavable portions can be appropriately determined depending on the size of RNA nanostructure and the type of DNA helper. Furthermore, in production of the RNA nanostructure according to the present invention, at least one portion should be cleaved so that a product amplified by RNA polymerase will form an RNA nanostructure. Herein, the portion to be cleaved is not limited, and for example, as indicated by ⑤ in FIG. 1, the portion to be cleaved can be located in the stem portion of any one arm.

If only the ⑤ portion in the RNA nanostructure shown on the left of FIG. 1 is cleaved, an “ALL-OFF” state (that is, the RNA nanostructure will have a three-dimensional structure, but none of the RNAi sequences included in the arms will be activated) will be obtained. If the ① and ② portions are cleaved, a “GFP-ON” state (that is a state in which only the RNAi sequence of the arm having GFP inserted therein is activated) will be obtained. If only the ③ portion is cleaved, a “RFP-ON” state will be obtained, and if only the ④ portion is cleaved, a “BFP-ON” state will be obtained. If all the ①, ②, ③ and ④ portions are cleaved, an “ALL-ON” state (that is a state in which the RNAi sequences included in all the arms are activated) will be obtained, and if the ①, ② and ③ portions are cleaved, a “GFP, RFP-ON” state will be obtained, and if the ③ and ④ portions are cleaved, a “RFP, BFP-ON” state will be obtained, and if the ①, ② and ④ portions are cleaved, a “BFP, GFP-ON” state will be obtained.

The simplest Y-shaped RNA nanostructure having three arms as described above has an advantage in that activation of the RNAi sequence can be controlled to a total of 8 combinations. In the case of the RNA nanostructure having arms that are K in number, activation of the RNAi sequence can be controlled to 2^K.

(2) Structure of Each Arm

The stem of each arm of the RNA nanostructure according to the present invention may be 21-40 nucleotides (nt) in length. If the length of the stem is long, it is highly likely to induce an immune response. For this reason, the length of the stem may be, for example, 21-30 nt in length.

Each of the arms of the RNA nanostructure according to the present invention consists of a stem and a loop, wherein the stem is consists of a core and an RNAi sequence.

As shown in FIG. 2, the cores are located in the central portion of the RNA nanostructure of the present invention, in which the radially extending arms of the RNA nanostructure meet one another in the central portion. In each arm, the core may be any RNA sequence which is 2-20 nt in length, preferably 2-10 nt in length. In one embodiment of the present invention, the core portion may be any sequence having a high G/C content so as to increase the stability of the RNA nanostructure of the present invention. Meanwhile, the cores may consist of any double-stranded RNA sequences, wherein the double-stranded RNA sequence may be 50-100% complementary to each other.

In each of the arms, the loop may be 2-30 nt in length, for example, 12-24 nt in length, and the loop of each arm may be open or closed. The loop of each loop comprises one or more sequences complementary to one or more DNA helpers, and thus when DNA helper binds to the loop, a portion forming a RNA/DNA double strand can be cleaved by RNase H to open the loop. As shown in FIG. 2, the portion cleaved by RNase H has, at its 3'-end, a 2-10 nt single-stranded RNA, that is, an unpaired overhang. The presence of this overhang increases the likelihood of being recognized and cleaved by a dicer in vivo.

A dicer is an enzyme having two domains (a PAZ domain and a catalytic domain), in which the PAZ domain serves to recognize dsRNA, and the catalytic domain serves to cleave the dsRNA. It is known that the dicer cleaves a point approximately 21 nt (19 nt + 2 nt overhang) or 22 nt (20 nt+2 nt overhang) from the 5'-end or 3'-end of dsRNA. Namely, a dsRNA cleaved at 21 nt and a dsRNA cleaved at 22 nt may occur as a mixture. Meanwhile, it was reported that counting from the 5'-end is more predominant than counting from 3'-end.

FIG. 3 is a schematic view showing a case in which the arm of the RNA nanostructure of the present invention was cleaved by a dicer. As shown in FIG. 3, a 2-nt overhang is present at the 3'-end of each of the sense sequence and antisense sequence of the stem consisting of a 19 bp (base pair) RNAi sequence. Namely, when the loop of the stem in the RNA nanostructure according to the present invention is open, it will be easily recognized and cleaved by a dicer due to the presence of a 2-10 nt overhang at the 3'-end, and a 2-nt overhang will remain at the 3'-end of the portion cleaved by the dicer.

The stem of each arm comprises an RNAi sequence double-stranded with a sense sequence and an antisense sequence. The RNAi sequence is 19 nt or 20 nt in length.

The RNAi sequences included in the stems of different arms of the RNA nanostructures according to the present invention may be the same or different. Where the RNAi sequences included in the stems of different arms of the Y-shaped RNA nanostructure differ from each other, the RNAi sequences in the RNA nanostructures may be present at a ratio of 1:1:1. Where the RNAi sequences included in the stems of two of the three arms are the same and the RNAi sequence in the stem of the other one arm differs from the two RNAi sequences, the two same RNAi sequences and the different RNAi sequence may be present at a ratio of 2:1 in the RNA nanostructure. As described above, according to the present invention, the ratio of the RNAi sequences included in the RNA nanostructure can be precisely controlled as desired.

The 5'-end of the antisense strand of the RNAi sequence included in the stem of each arm may towards the loop (referred to as “forward structure”; see the left of FIG. 4), or the 5'-end of the sense strand may towards the loop (referred to as “reverse structure”; see the right of FIG. 4). Both the forward structure and the reverse structure can exhibit an excellent gene-silencing effect.

Whether the forward structure is obtained or whether the reverse structure is obtained may be determined depending on how the DNA template is designed in production of the RNA nanostructure according to the present invention.

In the present invention, the efficacy of the RNAi sequence may be increased by controlling the thermodynamic stability of the RNAi sequence included in the stem of each arm. Where the RNAi sequence of the each arm in the RNA nanostructure according to the present invention is in an activated state, when administered to a subject, it is loaded into a RISC after cleaving as a suitable length by a dicer or it is directly loaded into RISC without subjecting to a cleavage process by a dicer. Then, a thermodynamically strong strand in the loaded RNAi sequence is cleaved by Ago2 (Argonaute 2) which is a protein corresponding to a catalytic domain in the RISC, and a thermodynamically strong strand will remain. Thus, in the RNA nanostructure according to the present invention, the RNAi sequence may be designed such that the antisense sequence against a target gene in the RNAi sequence becomes a thermodynamically strong strand, thereby enhancing the gene-silencing efficacy of the RNAi sequence. As shown in the top of FIG. 5, the 5'-end of the antisense sequence is present in the portion to be cleaved by a dicer, and the 5'-end of the antisense sequence includes a nucleotide with an adenine-uracil base pair, and the 5'-end of the sense sequence includes a nucleotide with a thermodynamically strong guanine-cytosine base pair, and additionally, a nucleotide with an A-U base pair is present at four or more (about 57% or more) of seven nucleotides (i.e., a seed region) from the 5'-end of the antisense sequence. In this case, an RNAi sequence exhibiting excellent gene-silencing efficacy can be obtained. On the contrary, as shown in the bottom of FIG. 5, in the case in which 4 or more G-C base pairs from the 5'-end of the antisense sequence present in the portion to be cleaved by a dicer are present and in which an A-U base pair is present at the 5'-end of the sense sequence of the double-stranded RNA having an open loop, the antisense sequence against a target gene will be cleaved by Ago2, and only the sense sequence will remain, indicating that an RNAi sequence having low gene-silencing efficacy will be obtained.

FIGS. 6 to 9 show RNA nanostructures, each having arms that are K in number (K=3). Cleavage of the RNAi sequence by a dicer can result in 19+2 nt or 20+2 nt (that is, a 19-nt or 20-nt RNAi sequence and an 2-nt overhang at the 3'-end of the RNAi sequence), and both a forward structure and a reverse structure can be obtained depending on the direction of the RNAi sequence in the RNA nanostructure. Thus, FIG. 6 shows a 19+2 nt reverse structure having arms that are K in number; FIG. 7 shows a 19+2 nt forward structure; FIG. 8 shows a 20+2 nt reverse structure; and FIG. 9 shows a 20+2 nt forward structure. Meanwhile, considering that dicer cleavage can result in both 19+2 nt and 20+2 nt, the RNAi sequence can also be designed such that an A-U base pair is present at both 19 nt and 20 nt in the RNAi sequence in the RNA nanostructure. Namely, the RNAi sequence can also be designed such that an A-U base pair is present at both the 5'-end or the second nucleotide from the 5'-end of the antisense strand of the RNAi sequence.

An RNAi sequence that may be included in the RNA nanostructure of the present invention serves to induce the inhibition of complementary mRNA to thereby inhibit the expression of a certain gene, when it is delivered into a cell. Any RNAi sequence may be used without limitation, as long as it can exhibit an effect on the treatment of disease. Examples of such RNAi sequence include EGFR-siRNA, anti-VEGF, anti-GFP, anti-Luc, anti-EGF, anti-FVII, anti-ApoB, and the like, and such exemplary RNAi sequences are widely known in the art. In addition, in order to easily confirm whether RNAi sequence transcription, amplification and nanostructure formation effectively occurred, a fluorescent protein, such as RFP, BFP or GFP, or Luc (firefly luciferase), etc., may also be used.

The sense sequence and the antisense sequence in the RNAi sequence of the present invention are principally 100% complementary to each other. However, in order to reduce off-targeting, any portion of the sense strand of the RNAi sequence can be modified to make an un-matched portion, in such a manner that the sequences forming any one arm are at least 50% complementary to each other.

Gene-Silencing Mechanism of RNA Nanostructure

A mechanism, by which the RNA nanostructure of the present invention induces gene silencing when administered to a subject, is as follows (see a schematic view of FIG. 10).

If the length of the stem of each arm in the RNA nanostructure of the present invention is longer than a certain length (e.g., 20-25 nt), when the RNA nanostructure is administered to a subject, the stem of each arm will be recognized as “long dsRNA” and cleaved by a dicer so that the length thereof will be adjusted to 20-25 nt (step , dicer processing). Because the RNA nanostructure of the present invention has a 2-nt overhang at the 3'-end of the RNAi sequence, which is exposed when the stem is cleaved, it will be easily recognized by a dicer. If the length of the stem of each arm is 20-25 nt or more, for example, 27 nt, the stem will be more easily recognized by a dicer. At the 3'-end of the RNAi sequence included in the stem of each arm, cleaved by the dicer as described above, a 2-nt overhang occurs. If the stem of each arm is sufficiently short in length, it will go directly to the next step without passing through step ① . The stem of each arm of the RNA nanostructure, cleaved by the dicer, is loaded into a RISC (RNA-induced silencing complex) via a RLC (RISC-loading complex) mediated by a dicer and the TRBP (the human immunodeficiency virus transactivating response RNA-binding protein) (step ②). In this process, vertebrates, including humans, asymmetry sensing occurs in which the TRBP is rearranged so that a thermodynamically weak strand in dsRNA can be recognized as guide RNA by Ago2.

The RISC is a ribonucleotide protein that recognizes and loads dsRNA therein. In the RISC, a thermodynamically strong strand in the loaded dsRNA is cleaved (passenger RNA) by Ago2 (Argonaute 2) that is a protein corresponding to a catalytic domain, and a thermodynamically weak strand remains (guide RNA) (step ③). Because the RNAi sequence included in the stem of each arm of the RNA nanostructure according to the present invention is designed such that the antisense sequence becomes a thermodynamically weak strand, the sense sequence is cleaved with high efficiency, and the antisense sequence remains. The antisense sequence recognizes the mRNA of a target gene (step ④) and binds complementarily to the mRNA to thereby form dsRNA which is then cleaved, thereby inducing gene silencing (step ⑤).

Method for Producing RNA Nanostructure

The RNA nanostructure of the present invention may be prepared by the following method.

(1) Summary of Method for preparation of RNA Nanostructure

FIG. 11 is a schematic view showing a method for preparing the RNA nanostructure of the present invention. First, a circular DNA template strand as shown in the center of the top of FIG. 11 is prepared. The circular DNA template strand has arms that are K in number (K is an integer ranging from 3 to 90), in which the stem of each of the arms comprises a core and an RNAi sequence. RNA polymerase may bind to one or more of the loops of each arms, and the loop comprises one or more sequences complementary to one or more DNA helpers.

When the circular DNA template strand is treated with RNA polymerase, RCT (rolling circle transcription) on the circular DNA template strand occurs to produce an amplified RNA product. When the amplified RNA product is treated with DNase I to remove the circular DNA template strand and is treated with one or more DNA helpers, the DNA helpers bind to a portion complementary to the amplified RNA product to form a RNA/DNA double strand, and the portion forming the double strand is cleaved by treatment with RNase H, whereby the amplified RNA product can form an RNA nanostructure and one or more loops in the RNA nanostructure can be opened. The 3'-end of the opened loop has a 2-nt overhang.

After the RNA nanostructure is prepared as described above, denaturing PAGE gel analysis may be performed in order to confirm whether the desired structure was accurately prepared.

(2) Method for Preparation of Circular DNA Template Strand

The circular DNA template strand described in (1) above has arms that are K in number, wherein K is an integer ranging from 3 to 90. The stem of each arm comprises a core and an RNAi sequence, and the loop of each arm comprises one or more sequences complementary to one or more DNA helpers, and one or more of the loops present in the arms comprise a portion to which RNA polymerase can bind. This circular DNA template strand may be used in the method for preparation of the RNA nanostructure according to the present invention, without regard to a method used to prepare the circular DNA template strand.

In one embodiment of the present invention, the circular DNA template strand may be prepared using the following DNA template strands that are K in number, wherein K is an integer ranging from 3 to 90.

In each of the DNA template strands, a sense sequence and an antisense sequence are present at both ends of the DNA template strand, respectively, while a portion for forming the loop of the circular DNA template strand is interposed between the sense sequence and the antisense sequence, and a portion for forming the core of the circular DNA template strand is present at the end of each of the sense sequence and the antisense sequence. For example, in any one DNA template strand, a loop portion may be present in the center of the DNA template strand, a sense sequence against an A gene may be present at one side of the loop portion, an antisense sequence against the A gene may be present at the other side of the loop portion, and a part of the core may be present at the end of each of the sense sequence and the antisense sequence. The loop-forming portion of each DNA template strand comprises one or more portions complementary to DNA helper. In addition, to one or more loop-forming portions of each DNA template strand, RNA polymerase may bind. Where each DNA template strand includes the sense sequence and antisense sequence forming an RNAi sequence as described above, when DNA template strands that are K in number are mixed with one another, nicks that are K in number will be present in the portion of stem of the circular DNA template strands, as shown on the right of FIG. 12 and FIG. 13. In this case, the distance between the 5'-end and the 3'-end in each of the nicks is close to each other, and thus the 5'-end and the 3'-end may be linked to each other by DNA ligase without using a separate primer, thereby forming the circular DNA template strand as shown in the center of the top of FIG. 11. In this case, in order to enable the structure to be stably formed, the structure may be designed such that the distance between the center of the structure and the nick is 9 nt or more in length and the distance between the nick and the start point of the loop is 9 nt or more in length (see FIG. 13; in FIG. 13, x≥9 nt and y≥9 nt). However, the position of the nick may be appropriately adjusted.

In another embodiment of the present invention, the circular DNA template strand may be prepared using the following DNA template strands that are K in number, wherein K is an integer ranging from 3 to 90. A sense sequence included in any one DNA template strand may bind complementarily to an antisense sequence included in another DNA template strand to thereby form an RNAi sequence. For example, any one DNA template strand may include a core portion in the center thereof, a sense sequence against an A gene may be present at one side of the core portion, an antisense sequence against the A gene may be present at the other side of the core portion, and a portion of a loop may be present at the end of each of the sense sequence and the antisense sequence. The loop-forming portion of each DNA template strand comprises one or more portions complementary to a DNA helper sequence. In addition, RNA polymerase may bind to one or more of the loops of the produced DNA template strand. If complementary portions between the DNA template strands that are K in number bind to each other as described above, nicks that are K in number will be present in the loop portions of the circular DNA template strand as shown on the left of FIG. 12. In this case, the distance between the 5'-end and the 3'-end in each nick is somewhat long, and for this reason, when the distance is shortened using a suitable primer, followed by treatment with DNA ligase, the 5'-end and the 3'-end may be linked to each other, thereby forming a circular DNA template strand as shown in the top of the center of FIG. 11.

The DNA template strand that is used in the method of the present invention may be designed such that the RNA nanostructure that is prepared by the method comprises a portion to which RNA polymerase binds, a portion complementary to one or more DNA helpers, sense and antisense sequences forming an RNAi sequence, etc. In this case, the portions may be included separately in the DNA template strands that are K in number, and these portions may be linked to each other by DNA ligase to form a complete sequence in the circular DNA template strand.

(3) Structure of Portion to Which RNA Polymerase Binds

In the method for preparing the RNA nanostructure of the present invention, a portion to which RNA polymerase can bind is present in one or more of the loops of the circular DNA template strands as described (1) above, and thus RNA polymerase binds to the portion, and RCT (rolling circle transcription) occurs to produce an amplified RNA product. Herein, the portion to which RNA polymerase can bind may have three different structures as shown in FIG. 14.

First, one or more of the loops may include a promoter for RNA polymerase ( in FIG. 14). For example, where T-7 RNA polymerase is used as RNA polymerase, the promoter sequence of the T-7 RNA polymerase may be, for example, 5'-TAATACGACTCACTATAGGGAT-3'. RNA polymerase and the promoter sequence thereof, which may be used in this RNA synthesis, are widely known to those skilled in the art.

Alternatively, when one or more of the loops are treated with a 12-30 nt (for example, 12-22 nt) complementary primer together with RNA polymerase, the RNA polymerase may bind to a double-stranded nucleic acid portion formed by the primer and the DNA of the loop (② in FIG. 14). In addition, RNA polymerase have the property of binding to a bubble formed by separation of double-stranded nucleic acids, and thus can recognize the loop itself as a bubble and bind to the loop, even in the absence of a promoter or a primer (③ in FIG. 14).

In all the ① to ③ structures, a sequence which is 8 nt or more in length is required for binding of RNA polymerase. Thus, one or more of the loops of the circular DNA template strand that is used in the method for preparing the RNA nanostructure of the present invention consist of 8-24 nt. Other loops may be 2-24 nt in length.

(4) Design of DNA Helper

In the method for preparing the RNA nanostructure of the present invention, the DNA helper may be a sequence having any length. Because each of the loops of the RNA nanostructure according to the present invention consists of 2-30 nt, the DNA helper may also be any DNA sequence which is 2-30 nt in length. For example, a 9-18-nt (preferably 9-12-nt) DNA helper may be used.

When 2'-O-methyl DNA is used as the DNA helper, a portion to be cleaved can be accurately specified, because a portion substituted with a 2'-O-methyl group is not cleaved by RNase H. The DNA helper that is used in the present invention may have nucleotides in which 4 or more DNAs are unmethylated, and other DNAs may be methylated or unmethylated. In one embodiment of the present invention, a 9-nt DNA helper, in which 5 nt from the 3'-end of the DNA helper are methylated, may be used. In another embodiment of the present invention, a 12-nt DNA helper, in which 4-7 nt from the 5'-end of the DNA helper are methylated, may be used. However, the degree of the methylation is not limited thereto, and the degree of methylation of the DNA helper may vary depending on the enzyme and DNA template used.

Because the target of RNase H is the RNA strand of the RNA/DNA double-strand, the DNA helper is reusable. Thus, the number of the DNA helpers that are used in the method of the present invention may be smaller than the number of amplified RNA products. In the method for preparing the RNA nanostructure according to the present invention, each DNA helper may be added at a concentration of 0.1 μM, but is not limited thereto.

Meanwhile, the length of the overhang in the prepared RNA nanostructure can be determined depending on the DNA helper. This is because the position that is cleaved by RNase H changes depending on the position at which the DNA helper binds to the amplified RNA product. In the present invention, the DNA helper may be designed such that the overhang in the prepared RNA nanostructure will be 2-10 nt, preferably 2 nt in length.

In the present invention, the term “RNAi sequence” refers to any double-strand RNA (dsRNA) sequence that induces RNA interference. The RNAi sequence in the present invention is meant to include small interfering RNA (siRNA). The mRNA of any target gene can be silenced by the RNAi sequence of the present invention, and thus the expression of the protein encoded by the mRNA can be inhibited.

In the present invention, “complementary sequences” may include, in addition to 100% complementary sequences, 60-100 complementary sequences, preferably 80-100% complementary sequences, more preferably 90-100% complementary sequences, even more preferably 95-100% complementary sequences, as long as these can retain the property of complementarily binding to each other.

It should be interpreted that the values described herein also include their equivalent ranges, unless otherwise specified.

Hereinafter, preferred examples will be given in order to help understand the present invention. It is to be understood, however, that these examples are for illustrative purposes only and are not intended to limit the scope of the present invention.

Example 1: Method 1 for Mass Production of RNA Nanostructures Using 2'-O-Methyl DNA Helper

1.1: Preparation of Circular DNA Template Strand

For mass production of RNA nanostructures, 80-nt, 64-nt and 64-nt strand 1, strand 2 and strand 3 (referred to as RV_Strand1, RV_Strand2 and RV_Strand3, respectively, in Table 1 below) were used as templates. In addition, five types of 2'-O-methyl 9-nt DNA helpers (H1(L)_9 nt, H2(G)_9 nt, H3(R)_9 nt, H4(B)_9 nt, and H5(stem)_9 nt) were used which could bind to each of the arms of produced RNA nanostructures. In order to specify a portion to be cleaved by RNase H, partially methylated 2'-O-methyl DNA helper was used.

Specific DNA sequences and 9-nt helper sequences are shown in Table 1 below.

(G: sense-GFP, G': antisense-GFP, R: sense-RFP, R': antisense-RFP, B: sense-BFP, B': antisense-BFP, Phos: phosphate group)

In Table 1 above, the loop portion in each of 80-nt, 64-nt and 64-nt single-stranded DNA (Integrated DNA Technologies, IDT) template strands having a phosphate group at the 5'-end is highlighted by a shadow, and sequences complementary to amplified RNA strand sequences, which are recognized and bound by 9-nt DNA helpers, are underlined. In the 9-nt DNA　helpers (Bioneer Corporation), the portions indicated by bold letters correspond to 2'-O-methylated portions. Each of the DNA template strands includes sense and antisense sequences of an RNAi sequence. Herein, Strand1 includes antisense-GFP (G') and sense-GFP (G) as index sequences, Strand2 includes antisense-RFP (R') and sense-RFP (R) as index sequences, and Strand3 includes antisense-BFP (B') and sense-BFP (B) as index sequences.

T4 DNA ligase, T7 RNA polymerase, pyrophosphatase, and RNase H, used in this Example, were all purchased from New England Biolabs (NEB).

A specific process for preparing the RNA nanostructures is as follows. First, 80-nt, 64-nt and 64-nt single-stranded DNA strand1, strand2 and strand3 (each 60 pmole) having a phosphate group at the 5'-end, as shown in Table 1 above, were added to 2 μL of 10x buffer (500 mM Tris-HCl (pH 7.5), 100 mM MgCl₂, 100 mM DTT, and 10 mM ATP) for T4 DNA ligase and 13 μL of DEPC-treated deionized water (DW) in a thermal cycler (Bio-Rad T100™), and hybridized to one another by reducing the temperature from 95℃ (2 min) to 10℃ at a rate of -1.0℃/sec. After hybridization, 1 μL of T4 DNA ligase (400 U/μL) was added to the hybridization product which was then incubated in the thermal cycler at 25℃ for 4 hours so that the 3'-end and 5'-end of each nick would be linked to each other, thereby producing a circular DNA template strand. Then, the product was heated at 65℃ for 10 minutes to inactivate the T4 DNA ligase. The product was used in the next reaction for rolling circle transcription of nucleic acid.

1.2: Process for Rolling Circle Transcription of Circular DNA Template Strand

2 μL of the circular DNA strand, produced in Example 1.1, was mixed with 2 μL of 10x RNAPol reaction buffer (400 mM Tris-HCl (pH 7.9), 60 mM MgCl₂, 20 mM spermidine, and 100 mM DTT), 0.8 μL of 25 mM rNTPs, 10.4 μL of DEPC-treated deionized water (DW), 0.8 μL pyrophosphatase (0.1 U/μL) and 4 μL of T7 RNA polymerase (50 U/μL). The mixture was incubated in a thermal cycler at 37℃ for 18 hours so that the T7 RNA polymerase would amplify an RNA strand complementary to the circular DNA template strand from the 18-nt loop portion (present in the loop of Strand1) of the circular DNA template strand. Next, the reaction product was heated at 65℃ for 10 minutes to inactivate the T7 RNA polymerase, thereby terminating the reaction.

1.3: Removal of DNA Template Strand by DNase I Treatment

Unless the DNA template strand was removed, the DNA template strand could act as a helper to destroy the amplified RNA in an RNase H treatment step. For this reason, the DNA template strand was removed in the following manner.

The RCA product (20 μL) was mixed with 4 μL of 10X DNase I buffer (100 mM Tris-HCl, 25 mM MgCl₂ and 5 mM CaCl₂), 15 μL of DEPC-treated deionized water (DW) and 1 μL of DNase I (2 U/μL). The mixture was incubated in a thermal cycler at 37℃ for 1 hour so that DNase I would degrade the DNA template strand. Next, the reaction product was heated at 65℃ for 10 minutes to inactivate the DNase I, thereby terminating the reaction.

1.4: 9-nt 2'-O-Methyl Hybridization and RNase H Cleavage, Followed by Self-Assembly

1.4.1: Synthesis of “All-ON” Y RNA Nanostructure

The DNase I product (40 μL), obtained in Example 1.3, was mixed with 8 μL of 10× RNase H reaction buffer (750 mM KCl, 500 mM Tris-HCl (pH 8.3), 30 mM MgCl₂, and 100 mM DTT), 0.8 μL of each of 10 μM of 9-nt 2'-O-

methyl DNA helper

1, 2, 3 and 4 (referred to as H1(L)_9nt, H2(G)_9nt, H3(R)_9nt and H4(B)_9nt, respectively, in Table 1 above), and 22.8 μL of DEPC-treated deionized water (DW). 6 μL of RNase H (5 U/μL) was added to the mixture which was then incubated in a thermal cycler at 37℃ for 18 hours, thereby cleaving the RNA strand of the RNA/DNA double strand. Next, the reaction product was heated at 65℃ for 20 minutes to inactivate the enzyme, and the temperature was reduced from 95℃ to 10℃ at a rate of -1.5℃/min, whereby the cleaved RNA strands were self-assembled.

The resulting produced “All-ON” Y RNA nanostructure is schematically shown in FIG. 1. As shown on the left of FIG. 1,

DNA helper

1, 2, 3 and 4 did bind to the ①, ②, ③ and ④ portions, respectively. Herein, the loop of the arm comprising the RNAi sequence against GFP was opened using two DNA helpers (helper 1 and 2), because the length of the loop was long. The remaining two loops were each opened using one DNA helper, thereby producing an “All-ON” Y RNA nanostructure.

1.4.2: Synthesis of “GFP-ON”, “RFP-ON”, and “BFP-ON” Y RNA Nanostructures

A “GFP-ON” Y RNA nanostructure was prepared in the same manner as described in Example 1.4.1, except that the DNase I product was treated with 0.8 μL of each of 9-nt 2'-O-

methyl DNA helper

1 and 2. Specifically, a Y RNA nanostructure was prepared in which the DNA helpers did bind only to the ① and ② portions on the left of FIG. 1 to open the loop to thereby make a “GFP-ON” state, and in which the BFP and RFP portions were in an “OFF” state because the loops were maintained in a closed state.

Furthermore, an “RFP-ON” Y RNA nanostructure was prepared in the same manner as described in Example 1.4.1, except that the DNase I product was treated with 0.8 μL of 9-nt 2'-O-methyl DNA helper.

In addition, a “BFP-ON” Y RNA nanostructure was prepared in the same manner as described in Example 1.4.1, except that the DNase I product was treated with 0.8 μL of 9-nt 2'-O-methyl DNA helper 4.

1.4.3: Synthesis of “All-OFF” Y RNA Nanostructure

An “ALL-OFF” Y RNA nanostructure was prepared in the same manner as described in Example 1.4.1, except that the DNase I product was treated with 0.8 μL of 9-nt 2'-O-methyl DNA helper 5. Specifically, a Y RNA nanostructure was produced in which the DNA did bind only to the ⑤ portion on the left of FIG. 1 and all the loops were closed (“OFF” state).

The experimental parameters and conditions used in each step of Example 1 are summarized in Table 2 below.

Example 2: Method 2 for Mass Production of RNA Nanostructures Using 2'-O-Methyl DNA Helpers

2.1: Preparation of Circular DNA Template Strand

For mass production of RNA nanostructures, 72-nt strand 1, 72-nt strand 2 and 72-nt strand 3 were used. In addition, three types of 2'-O-methyl 32-nt open DNA helpers (Open DNA helper (G)_32 nt, Open DNA helper (R)_32 nt, and Open DNA helper (B)_32 nt) and three types of protect RNA helpers (Protect RNA helper (G)_30 nt, Protect RNA helper (R)_30 nt, and Protect RNA helper (B)_30 nt) were used which could bind to each arm of the produced RNA nanostructure.

Specific DNA sequences and 32-nt open DNA helper and 30-nt protect RNA helper sequences are shown in Table 3 below.

In Table 2 above, the loop portion in each of the 72-nt single-stranded DNA (Integrated DNA Technologies, IDT) template strands having a phosphate group at the 5'-end is highlighted by a shadow, and sequences complementary to amplified RNA strand sequences, which are recognized and bound by the DNA helpers, are underlined.

The portions indicated by bold letters in the 32-nt open helpers (Integrated DNA Technologies, IDT) correspond to 2'-O-methylated portions, and the 30-nt protect helpers (Bioneer) are complementary sequences that protect the entire loop portion of amplified RNA and 5-nt arm portions at both sides with respect to the loop. The DNA template strands include antisense-GFP (G'), sense-GFP (G), antisense-RFP (R'), sense-RFP (R), antisense-BFP (B') and sense BFP (B') as index sequences.

EXO3 exonuclease, T4 DNA ligase, T7 RNA polymerase, pyrophosphatase and RNase H, used in this Example, were all purchased from New England Biolabs (NEB), and a RNase inhibitor, used in this Example, was purchased from Thermo Fisher.

The DNA templates, used in this Example, had three nicks as shown on the right of FIG. 12, in which 5'-phosphate and 3'-OH ends were present in the nicks so that they could be linked to each other by T4 DNA ligase to form a circular shape. Furthermore, the three formed loop portions were 20 nt in length, and each of the loops would be recognized by T7 RNA polymerase so that rolling circle transcription of nucleic acid could occur. In addition, the stem portion of each of the three arms included a siRNA sequence so that a structure itself formed by amplification without loading of a separate siRNA could function as an siRNA carrier. A specific process for producing the RNA nanostructure is as follows.

First, 72-nt single-stranded template DNA strand1, strand2 and strand3 (each 60 pmole) having a phosphate group at the 5'-end, as shown in Table 3 above, were added to 2 μL of 10x buffer (500 mM Tris-HCl (pH 7.5), 100 mM MgCl₂, 100 mM DTT, and 10 mM ATP) for T4 DNA ligase and 13 μL of DEPC-treated deionized water (DW) in a thermal cycler (Bio-Rad T100™), and hybridized to one another by reducing the temperature from 95℃ (3 min) to 4℃ at a rate of -1.0℃/sec. After hybridization, 1 μ of T4 DNA ligase (400 U/μL) was added to the hybridization product which was then incubated in the thermal cycler at 25℃ for 12 hours so that the 3'-end and 5'-end of each nick would be linked to each other, thereby producing a circular DNA template strand. Then, the product was heated at 65℃ for 10 minutes to inactivate the T4 DNA ligase, thereby terminating the reaction. The product was used in the next reaction for rolling circle transcription of nucleic acid. Next, in order to remove the remaining strands which were not circularly linked, the product was treated with EXO3 nuclease at 37℃ for 2 hours.

2.2: Process for Rolling Circle Transcription of Circular DNA Template Strand

3.5 μL of the circular template strand, prepared in Example 2.1, was mixed with 2 μL of 10x RNA Pol reaction buffer (400 mM Tris-HCl (pH 7.9), 60 mM MgCl₂, 20 mM spermidine, and 100 mM DTT), 0.8 μL of 25 mM rNTPs, 0.8 μL of pyrophosphatase (0.1 U/μL), 0.5 μL of RNase inhibitor, and 4 μL of T7 RNA polymerase (50 U/μL). The mixture was incubated in a thermal cycler at 37℃ for 16 hours so that the T7 RNA polymerase would amplify an RNA strand complementary to the circular RNA template strand from the 20-nt loop portion of the circular DNA template strand. Next, the product was heated at 65℃ for 10 minutes to inactivate the T7 RNA polymerase, thereby terminating the reaction.

2.3: Removal of DNA Template Strand by DNase I Treatment

The RCA product (20 μL) was mixed with 1 μL of DNase I (2 U/μL). The mixture was incubated in a thermal cycler at 37℃ for 20 minutes so that the DNase I would degrade the DNA template strand. Next, EDTA was added to the product to a concentration of 5 mM, and then the product was heated at 95℃ for 3 minutes to inactivate the DNase I, thereby completing the reaction.

2.4: Cleavage with RNase H, Followed by Self-Assembly

2.4.1: Synthesis of “All-ON” Y RNA Nanostructure

The DNase I product (20 μL), obtained in Example 2.3, was mixed with 4 μL of 10× RNaseH reaction buffer (750 mM KCl, 500 mM Tris-HCl (pH 8.3), 30 mM MgCl₂, and 100 mM DTT), 4 μL of 2 μM 32-nt 2'-O-methyl DNA helper G, R and B (referred to as Open DNA helper (G)_32nt, Open DNA helper (R)_32nt, and Open DNA helper (B)_32nt, respectively, in Table 3 above), and 14.5 μL of DEPC-treated deionized water. 2 μL of RNase H (5 U/μL) was added to the mixture which was then incubated in a thermal cycler at 37℃ for 2 hours, thereby cleaving the RNA strand of the RNA/DNA double strand. Then, the product was heated at 65℃ for 20 minutes to inactivate the enzyme, and the temperature of the product was reduced from 95℃ to 4℃ at a rate of -1.2 ℃/min, whereby the cleaved RNA strands were self-assembled.

2.4.2: Synthesis of “GFP-ON”, “RFP-ON” and “BFP-ON” Y RNA Nanostructures

A “GFP-ON” Y RNA nanostructure was prepared in the same manner as described in Example 2.4.1, except that the DNase I product was treated with 4 μL of a mixture of 2 μM 32-nt 2'-O-methyl DNA open helper (G) and 10 μM 30-nt RNA protect helper (R and B). Specifically, a Y RNA nanostructure was prepared in which the DNA open helper did bind only to the GFP portion to open the loop to thereby make a “GFP-ON” state, and in which the RNA protect helper did bind to the BFP and RFP portions so that the loops would be maintained in a closed state (“OFF” state).

Furthermore, a “RFP-ON” Y RNA nanostructure was prepared in the same manner as described in Example 2.4.1, except that the DNase I product was treated with 4 μL of a mixture of 2 μM 32-nt 2'-O-methyl DNA open helper (R) and 10 μM 30-nt RNA protect helper (G and B).

In addition, a “BFP-ON” Y RNA nanostructure was prepared in the same manner as described in Example 2.4.1, except that the DNase I product was treated with 4 μL of a mixture of 2 μM 32-nt 2'-O-methyl DNA open helper (B) and 10 μM 30-nt RNA protect helper (G and R).

2.4.3: Synthesis of “GFP/RFP-ON”, “GFP/BFP-ON” and “BFP/RFP-ON” Y RNA Nanostructures

A “GFP/RFP-ON” Y RNA nanostructure was prepared in the same manner as described in Example 2.4.1, except that the DNase I product was treated with 4 μL of a mixture of 2 μM 32-nt 2'-O-methyl DNA open helper (G and R) and 10 μM 30-nt RNA protect helper (B). Specifically, a Y RNA nanostructure was prepared in which the DNA open helper did bind to the GFP and RFP portion to open the loops to thereby make a “GFP/RFP-ON” state, and in which the RNA protect helper did bind to the BFP portion to close the loop (“OFF” state).

Furthermore, a “GFP/BFP-ON” Y RNA nanostructure was prepared in the same manner as described in Example 2.4.1, except that the DNase I product was treated with 4 μL of a mixture of 2 μM 32-nt 2'-O-metrhyl DNA open chamber (G and B) and 10 μM 30-nt RNA protect helper (R).

In addition, a “BFP/RFP-ON” Y RNA nanostructure was prepared in the same manner as described in Example 2.4.1, except that the DNase I product was treated with 4 μL of a mixture of 2 μM 32-nt 2'-O-methyl DNA open helper (B and R) and 10 μM 30-nt RNA protect helper (G).

2.4.4: Synthesis of “Helper-Free” Y RNA Nanostructure

A helper-free Y RNA nanostructure was prepared using the circular DNA template strand of Example 2.1 in the same manner as described in Example 2.2, except that 0.5 μL of RNase H was further added to the reaction mixture.

When RNase H is further added and reacted in the process of amplifying the RNA strand complementary to the circular DNA template strand of Example 2.2, the loop portion of the remaining DNA template strand can bind specifically to the amplified product so that it can act as a helper (FIG. 15). When the DNA template strand did bind to the RNA product, RNase H cleaves a certain RNA portion complementary to the loop sequence, and the cleaved RNA strands form a Y structure. This method for preparing the helper-free RNA nanostructure can produce a large amount of the Y structure without requiring an additional step, compared to the production method employing the 2'-O-methyl helpers.

2.4.5: Analysis of Y RNA Nanostructures

All the Y RNA nanostructures, prepared in Examples 2.4.1 to 2.4.3, were analyzed by 10% PAGE in 1X Tris-borate-EDTA (TBE) buffer at room temperature. The gel was stained with gel red, and visualized using Gel Doc™ EZ (Bio-Rad). The results are shown in FIG. 16.

As shown in FIG. 16, the Y RNA nanostructures had different sizes depending on the opening or closing of each loop.

In addition, a gel image of the helper-free Y RNA nanostructure, produced in Example 2.4.4, was acquired in the same manner as described above, and the results of comparison performed using G (+0) helper are shown in FIG. 17.

As shown in FIG. 17, the Y nanostructure was produced regardless of the presence or absence of the helper.

2.5: AFM Image of Self-Assembled Y Nanostructure

3.75 μL of the Y RNA nanostructure (0.2 μM), prepared in Example 2.4.1, was diluted in 21.25 μL of 1x reaction buffer (50 mM Tris-HCl (pH 7.5), 10 mM MgCl₂, 10 mM DTT, and 1 mM ATP), and deposited onto freshly cleaved mica (Pelco Mica sheets, Ted Pella Corp.). After 30 min incubation, the mica surface was rinsed with DEPC-treated deionized water and immediately dried using nitrogen gas. The resultant sample was scanned in non-contact mode on a Park NX-10 ADM (Park Systems Corp. Korea) with NC-NCH tips (Park Systems Corp), thereby obtaining an AFM image of the Y nanostructure. The results are shown in FIG. 18.

The left of FIG. 18 is an image of the amplified RCT product obtained using the circular DNA template strand. As can be seen therein, long extending single-stranded RNA was obtained by amplification with T7 RNA polymerase. The right of FIG. 18 is an image of self-assembled Y nanostructures obtained by RNase H treatment. As can be seen therein, it was observed that spherical nanostructures uniformly dispersed through the cleavage/annealing process were obtained.

Example 3: Gene Silencing by RNAi Sequence

3-1: GFP Gene Silencing by RNAi Sequence

As described in “(2) Structure of Each Arm” of the section “RNA Nanostructure” above, an RNAi sequence against GFP gene was designed. Herein, from about 700-nt GFP mRNA, a 19-20-nt continuous portion suitable for inclusion in the stem of the RNA nanostructure according to the present invention was selected, in which the 5'-end of the sense sequence (or the 3'-end of the antisense sequence) would be G or C, and the 3'-end of the sense sequence (or the 5'-end of the antisense sequence) would be A or U, and, at the same time, 4 or more of 7 nucleotides from the 5'-end of the antisense sequence were A or U. As a result, it was found that the portion highlighted by box in FIG. 19 is suitable as the target of the RNAi sequence of the present invention.

Two different RNAi sequences against GFP were prepared by Bioneer Corp. The positive control “19+2 siGFP” is an RNAi sequence (linear dsRNA that cannot serve as a dicer substrate) against GFP, which is 19 nt in length and has a 2-nt overhang, and “23+2 siGFP” is an RNAi sequence (a linear dsRNA capable of serving as a dicer substrate) against GFP, which is 23 nt in length and has a 2-nt overhang. The sense and antisense sequences are shown in Table 4 below.

The “23+2 siGFP” sequence was treated with a dicer in vitro to obtain “23+2 siGFP after Dicer”. A negative control (NC) was not treated with anything.

An experiment on gene-silencing efficacy was performed. KB-cells overexpressing GFP was dispensed in a 12-well plate at a density of 1x10⁵ cells/well. After 1 day, the cells were treated with 3 nM “19+2 siGFP”, “23+2 siGFP” “23+2 siGFP after Dicer” or NC together with Lipofectamine® RNAiMAX Reagent (Invitrogen). After 48 hours, the intensity of fluorescence was measured using a FACSCalibur system (BD Biosciences), and then the gene-silencing efficacy was measured using flowJo program. The results of the measurement are shown in FIG. 20.

The left of FIG. 20 is a graph showing GFP expression levels relative to the GFP expression level of the negative control taken as 100%. As can be seen therein, because the 23+2 siGFP served as a dicer substrate, it showed an enhanced gene-silencing inhibitory effect compared to the 19+2 siGFP, and the “23+2 siGFP after Dicer” showed a gene expression inhibitory effect similar to that of the 19+2 siGFP.

The right of FIG. 20 is a graph showing gene silencing efficacy relative to the gene silencing efficacy of the positive control 19+2 siGFP taken as 100%. As can be seen therein, the gene silencing efficacy of the 23+2 siGFP was higher than that of the “23+2 siGFP after Dicer”. This is thought to be because the “23+2 siGFP after Dicer” was previously cleaved by a dicer to a length of 19+2 nt or 20+2 nt and was used to treat the cells, and thus dicer-mediated RISC loading did not occur so that the gene silencing efficacy would decrease.

3-2: GFP Gene Silencing by RNA Nanostructure Comprising RNAi Sequence

An RNAi sequence against GFP, which is 19 nt in length and has a 2-nt overhang, was designed. Using the RNAi sequence, an RNA nanostructure (“25+2Y-RNA”) was prepared according to the method described in Example 2. Furthermore, the 25+2Y-RNA was treated with a dicer in vitro to obtain “25+2Y-RNA after Dicer”. Specific 25+2Y-RNA sequences are shown in Table 5 below.

An experiment on gene silencing efficacy was performed. KB-cells overexpressing GFP were dispensed in a 12-well plate at a density of 1x10⁵ cells/well. After one day, the cells were treated with 1.5 nM “25+2Y-RNA” or “25+2Y-RNA after Dicer” together with Lipofectamine® RNAiMAX Reagent (Invitrogen). A negative control (NC) was not treated with anything. After 48 hours, the intensity of fluorescence was measured using a FACSCalibur system (BD Biosciences), and then the gene-silencing efficacy was measured using flowJo program. The results of the measurement are shown in FIG. 21.

The left of FIG. 21 is a graph showing GFP expression levels relative to the GFP expression level of the negative control taken as 100%. As can be seen therein, the gene expression inhibitory effect of the 25+2Y-RNA after Dicer was lower than the gene expression inhibitory effect of the 25+2Y-RNA serving as a dicer substrate.

The right of FIG. 20 is a graph showing gene silencing efficacy relative to the gene silencing efficacy of the 25+2Y-RNA taken as 100%. As can be seen therein, the gene silencing efficacy of the “25+2Y-RNA after Dicer” was significantly lower than that of the 25+2Y-RNA. This is thought to be because the “25+2Y-RNA after Dicer” was previously cleaved by a dicer to a length of 19+2 nt or 20+2 nt and was used to treat the cells, and thus dicer-mediated RISC loading did not occur so that the gene silencing efficacy would decrease.

3-3: GFP Gene Silencing by ON/OFF of RNA Nanostructure Comprising RNAi Sequence

In order to examine the gene-silencing effects by ON/OFF of nanostructures, GFP-KB cells were dispensed in a 12-well plate at a density of 1.0 × 10⁶ cells/well. After 24 hours of culture, 1 nM of each of a 25+2 Y nanostructure synthesized by IDT, the RFP-ON Y nanostructure obtained in Example 2.4.2, the R/GFP-ON nanostructure obtained in Example 2.4.3, and the All-ON nanostructure obtained in Example 2.4.1, was mixed with 3 μl of Lipofectamine® RNAiMAX (Invitrogen) at room temperature for 20 minutes, and then the cells were treated with each of the mixtures. After 48 hours, the expressions of GFP in the GFP-KB cells were measured using a FACS Calibur system (BD biosciences). The GFP gene silencing effects were analyzed using flowJo program. The results are shown in FIG. 22.

As can be seen in FIG. 22, the RFP-ON Y nanostructure showed no gene silencing effect, because the GFP loop thereof was closed. Unlike this, the R/GFP-ON and R/G/BFP-ON Y nanostructures showed the GFP gene-silencing effect in the same manner as the synthesized Y nanostructure, because the GFP loop thereof was open.

3-4: Programmable Gene Regulation Using ON/OFF RNA Nanostructures

Hela cells were dispensed in a 12-well plate at a density of 1.0 × 10⁶ cells/well with DMEM medium (10% FBS, 1% penicillin/streptomycin). After 24 hours of culture, the medium was replaced with 900 μl medium. For co-expression of RFP, BFP and GFP in the Hela cells, each in vitro transcribed mRNA together with stemfect was transfected into the cells. Next, 10 pmole individual RFP, GFP and BFP siRNA mixture, and the All-ON Y nanostructure sample obtained in Example 2.4.1. were mixed with 3 μl Lipofectamine® RNAiMAX (Invitrogen) in 50 μl DPBS at room temperature for 20 minutes, and then the cells were treated with the mixture. After 48, the expression of each fluorescent protein was measured using BD LSRFORTESSA (BD Biosciences, USA), and analyzed using flowJo program. Fluorescence images were acquired using a confocal microscope (Nikon A1, Japan). The results are shown in FIG. 23.

As shown in FIG. 23, in the group treated with the individual siRNA mixture (FIG. 23b), the ratio of siRNA in the cells was different, but in the group treated with the Y nanostructure (FIG. 23c), the ratio of RFP, GFP and BFP siRNA was 1:1:1.

Example 4: Effect of Y Nanostructure as Dicer Substrate RNA

4-1: Dicer Processing in the Presence or Absence of 2-nt Overhang

2-nt overhang-free 19+0 nt (”19+0Y-RNA”) and 23+0 nt (”23+0Y-RNA”) and 2 nt overhang-containing 19+2 nt (”19+2Y-RNA”) and 23+2 nt (”23+2Y-RNA”) Y nanostructures were synthesized by ordered from IDT (Intergrated Device Technology, USA). Specific 19+0Y-RNA, 23+0Y-RNA, 19+2Y-RNA and 23+2Y-RNA sequences are shown in Table 6 below.

0.3 μg of each of the 19+0 nt, 19+2 nt, 23+0 nt and 23+2 nt Y nanostructures was incubated in a recombinant human dicer enzyme kit (Genlantis, USA) at 37℃ for 6 hours, and then all the products were analyzed by 15% PAGE in 1X Tris-borate-EDTA (TBE) at room temperature. The gel was stained with gel red (Biotium, USA), and visualized using Gel Doc™ EZ (Bio-Rad). The results are shown in FIG. 24.

As shown in FIG. 24, the 2-nt overhang-containing Y nanostructures were more easily processed with the dicer than the 2-nt overhang-free Y nanostructures. This suggests that, in the presence of the overhang, the possibility for the Y nanostructures to be recognized and cleaved by the dicer increases.

4-2: Gene Silencing Effects of Nanostructures Having Various Arm Lengths

2-nt overhang-containing 19+2 nt (SEQ ID NOs: 31 to 33), 23+2 nt (SEQ ID NOs: 34 to 36) and 25+2 nt (SEQ ID NOs: 22 to 24) Y nanostructures were synthesized according to the method of Example 1. 0.3 μg of each of the 19+2 nt, 23+2 nt and 25+2 nt Y nanostructures was incubated in a recombinant human dicer enzyme kit (Genlantis, USA), and gel images of the Y nanostructures were acquired in the same manner as described in Example 4.1. The results are shown in FIG. 25b.

As shown in FIG. 25b, the 19 nt of the Y nanostructure was not cleaved by the dicer.

Next, the gene silencing effects of Y nanostructures having various arm lengths were examined. First, GFP-KB cells were seeded in RPMI medium (10% FBS, 1% penicillin/streptomycin) in a 96-well assay plate at a density 0.1 × 10⁶cells/well at concentrations of 100%, 70% and 50%. The cells were incubated for 24 hours, after which each of 0.05 pmole, 0.5 pmole and 2.5 pmole of the sample was mixed with 1.5 ul of Lipofectamine® RNAiMAX (Invitrogen) at room temperature for 20 minutes and adjusted to 500 μl by addition of 450 μl of RPMI medium, and then each well was treated with 100 μl of each of the three concentrations of the sample (N=4). Using Infinite® 200 PRO NanoQuant (Tecan) in bottom reading mode, GFP expressions in the cells after 48 hours (2 days), 96 hours (4 days) and 144 days (6 days) were measured. Fluorescence was analyzed by two-way ANOVA analysis (GraphPad Prism 5), and significance was tested based on the correlation between the dicer substrate group and the non-dicer substrate group (19+2 dsGFP). The results are shown in FIGS. 25c and 25d.

As shown in FIG. 25c, the gene inhibitory effects of the 19+2 Y nanostructure, 23+2 Y nanostructure and 25+2 Y nanostructure significantly increased in a concentration-dependent manner, compared to the 19+2 dsGFP.

In addition, as shown in FIG. 25d, the 25+2 Y nanostructure showed significantly increased gene inhibitory effects at 0.2 nM and 1 nM compared to the 19+2 dsGFP, and such increased inhibitory effects were long-lasting.

4-3: Gene-silencing effects depending on RNA structure

The gene silencing effects depending on a structure of nanostructure were examined by using GAPDH gene (Glyceraldehyde 3-phosphate dehydrogenase gene).

TRBP (the human immunodeficiency virus transactivating response RNA-binding protein) and PACT (protein kinase RNA activator) are a cofactor of Dicer. It was reported that in case of miRNA present in the body, dicer processing happens well regardless of whether to form Dicer/TRBP complex or Dicer/PACT, but in case of linear dicer substrate dsRNA, dicer processing normally happens when Dicer/TRBP complex is formed, whereas dicer processing is inhibited when Dicer/PACT complex is formed (Lee et al., Nucleic Acids Research, 2013, Vol. 41, No. 13).

As a RNA structure having a different structure, a total of three RNA structures of “Linear DsiGDH” which is a linear dicer substrate dsRNA, “Pre-let7 siGDH” which is a siRNA mimicking pre-miRNA structure, and “25+2Y-RNA” (hereinafter, referred to as “siY-GDH”) which is an RNA nanostructure according to the present invention were ordered from IDT (Intergrated Device Technology, USA). Specific RNA structure sequences are shown in Table 7 below.

1 μM of One strand of each complementary RNA strand pair were added to 1xPBS (Gibco, #10010023) and hybridized by reducing the temperature to 4℃ after incubating at 95℃ for 3 min.

Next, Normal HeLa cells, TRBP K/O cells and PACT K/O cells were dispensed in a 6-well plate at a density 0.3 × 10⁶ cells/well density with DMEM medium (10% FBS, 1% penicillin/streptomycin). After 24 hours, each cells were treated with 1 nM of each samples. After 24 hours, mRNA were extracted by using TRIzol (Invitrogen), and then cDNA were synthesized according to the protocol of SuperScriptTM II Reverse Transcriptase (Invitrogen) based on 25 μg of mRNA. RT-PCR were performed with the synthesized cDNA using TOPrealTM qPCR 2X PreMIX (Enynomics), GAPDH forward/reverse primer and b-actin forward/reverse primer. Cq values for GAPDH were corrected to b-actin, and ΔΔCq value were then calculated to analyze the degree of gene expression. The results are shown in FIG. 27.

As shown in FIG. 27, in the case of linear DsiGDH, it showed approximately 20% of gene-silencing efficacy in the normal HeLa cells and it showed a sharp gene-silencing inhibitory effect in TRBP K/O cells, whereas it showed an enhanced gene-silencing efficacy in PACT K/O cells because dicer processing happens only by Dicer/TRBP.

And, in the case of pre-Let7 siGDH, it is confirmed that inhibition of dicer processing by Dicer/PACT complex does not happen because gene-silencing efficacy has hardly changed in TRBP K/O cells unlike linear DsiGDH.

Also, in the case of siY-GDH, unlike linear DsiRNA, it is confirmed that dicer processing was not inhibited by Dicer/PACT complex and showed a similar behavior with pre-let7 miRNA

Claims

An RNA nanostructure having radially extending arms which are K in number,

wherein K is an integer ranging from 3 to 90

each of the arms consists of a stem and a loop,

wherein

1) the stem of each arm consists of a core and an RNAi sequence,

i) the core is 2-20 nucleotides (nt) in length, and

ii) the RNAi sequence is double-stranded which consists of sense and antisense sequences for any gene, and is 19 or 20 nt in length,

wherein the antisense sequence contains a nucleotide in which a U (uracil)-A (adenine) base pair is present at 5'-end, and the sense sequence contains a nucleotide in which a G (guanine)-C (cytosine) base pair is present at 5'-end,

the antisense sequence contains a nucleotide in which 4 or more U (uracil)-A (adenine) base pairs are present within the 7th nucleotides from the 5'-end (seed region),

2)　the RNAi sequences included in the stem of each arm are the same or different, and

3) the loop of each arm consists of 2-30 nucleotides and is open or closed, respectively, and where the loop is open, it has a single-stranded RNA (overhang) of 2-10 nt in length at the 3'-end of the RNAi sequence.
The RNA nanostructure of claim 1, wherein the 5'-end of the antisense sequence of the RNAi sequence included in the stem of each arm towards the loop, or the 5'-end of the sense sequence towards the loop.
The RNA nanostructure of claim 1, wherein the RNAi sequence has a 2-nt single-stranded RNA overhang at 3'-end, when the loop of each arm is open.
The RNA nanostructure of claim 1, wherein the core included in the stem of each arm has a G (guanine)-C (cytosine) content (G-C content) of 50% or more.
The RNA nanostructure of claim 1, wherein K is 3 or 4.
A method for preparing an RNA nanostructure having radially extending arms which are K in number, the method comprising the steps of:

(A) preparing a circular DNA template strand having radially extending arms which are K in number, wherein K is an integer ranging from 3 to 90, and each of the arms consists of a stem and a loop,

wherein

1) the stem of each arm consists of a core and an RNAi sequence,

i) the core is 2-20 nucleotides (nt) in length, and

ii) the RNAi sequence is double-stranded sequence which consists of sense and antisense sequences for any gene, and is 19 or 20 nt in length,

wherein the antisense sequence contains a nucleotide in which a U (uracil)-A (adenine) base pair is present at 5'-end or the second nucleotide from the 5'-end, and the sense sequence contains a nucleotide in which a G (guanine)-C (cytosine) base pair is present at 5'-end,

the antisense sequence contains in which 4 or more U (uracil)-A (adenine) base pairs are present within the 7th nucleotides from the 5'-end (seed region),

2)　the RNAi sequences included in the stem of each arm are the same or different, and

3) the loop of each arm consists of 2-30 nucleotides, and at least one of the loops of each arm consists of 8-30 nucleotides;

(B) treating the circular DNA template strand, prepared in step (A), with RNA polymerase, and obtaining an amplified RNA product by RCT (rolling circle transcription);

(C) treating the amplified RNA product with DNase I to remove the circular DNA template strand;

(D) treating the amplified RNA product, from which the circular DNA template strand was removed, with at least one DNA helper, and treating the amplified RNA product with RNase H to cleave a portion forming a RNA/DNA double strand with the DNA helper; and

(E) obtaining the RNA nanostructure by self-assembly of the cleaved amplified RNA product.
The method of claim 6, wherein step (A) comprises the steps of:

a) preparing DNA template strands which are K in number, in which sense and antisense sequence for any gene are present at both ends of the DNA template strand, respectively, while a portion for forming the loop of the circular DNA template strand is interposed between the sense sequence and the antisense sequence, and a portion for forming the core of the circular DNA template strand is present at the end of each of the sense sequence and the antisense sequence;

b) hybridizing the DNA template strands, which are K in number, to each other to allow complementary sequences to bind to each other, thereby preparing a DNA template having nicks which are K in number; and

c) treating the DNA template having nicks, which are K in number, with DNA ligase, thereby preparing the circular DNA template strand.
The method of claim 6, wherein the DNA helper in step (D) is a DNA helper wherein the DNA helper is 2-30 nt in length and is 2'-O-methylated at 3'-end, 5'-end or any middle part of the nucleotide, including at least 4 of un-modified parts.
The method of claim 6, wherein the RNA polymerase in step (B) is T7-RNA polymerase.
The method of claim 6, wherein step (B) comprises treating the circular DNA template strand with a primer together with the RNA polymerase, and producing the amplified RNA product by RCT using the treated circular DNA template, wherein the primer has a 12-30 nt RNA sequence complementary to an 8-30 nt DNA template strand loop.
The method of claim 6, wherein one or more of the loops of the circular DNA template strands in step (A) comprises an RNA polymerase promoter which is recognized by the RNA polymerase.
The method of claim 6, wherein K is 3 or 4.