Information provision method for cancer diagnosis or prognosis prediction using microrna containing 8-oxoguanine, cancer therapeutic agent and screening method thereof

Abstract

The present invention relates to an information provision method for cancer diagnosis or prognosis prediction using microRNA containing 8-oxoguanine (o8G), a cancer therapeutic agent and a screening method thereof. By using microRNA in which guanine (G) at the 2nd or 7th nucleotide from the 5' end is modified to 8-oxoguanine, the progression pattern of cancer can be analyzed and controlled such that the microRNA can be effectively used for diagnosis, prognosis analysis, and treatment of cancer.

Description

Method for providing information for cancer diagnosis or prognosis prediction using microRNA containing 8-oxoguanine, cancer treatment agent and screening method thereof

[0001] This patent application claims priority to Republic of Korea Patent Application No. 10-2023-0109171, filed with the Korean Intellectual Property Office on August 21, 2023, the disclosure of which is incorporated herein by reference.

[0002] The present invention relates to a method for providing information for cancer diagnosis or prognosis prediction using microRNA containing 8-oxoguanine, a cancer treatment agent and a screening method thereof, and more particularly, to a technology for analyzing the progression of cancer and controlling and treating it using microRNA in which the guanine at the 2nd to 7th nucleotide from the 5' end is modified to 8-oxoguanine.

[0003] Oxidative stress is linked to tumorigenesis, where cell proliferation under oxygen-deficient conditions requires glucose uptake via aerobic glycolysis, which continuously generates reactive oxygen species (ROS). Chronic inflammation or oxidative stress, a tumorigenic condition characterized by immune cell infiltration, is implicated, and chronic hepatitis induced by viral hepatitis and fatty liver disease is known to promote the progressive development of hepatitis, cirrhosis, and liver cancer.

[0004] ROS oxidize biomolecules in different ways, causing various effects on cell damage, mutations, and redox signaling. Among these oxidative modifications, the one that occurs on guanine residues in nucleic acids is 8-oxoguanine (o 8 G), which is paired with adenine (o 8 GA) and induces guanine-thymine (G>T) mutations. Importantly, RNA is more susceptible to these modifications than DNA, and o 8GA base pairing can alter regulatory RNA-RNA interactions.

[0005] MicroRNAs (miRNAs) are post-transcriptional regulatory factors that can silence target gene expression through base pairing between their seed region (positions 2 to 8) and mRNA. Any change affecting base pairing in the seed region redirects the miRNA to recognize a different set of target mRNAs. Recently, site-specific oxidation of miRNAs has been identified, where the o at position 7 of miR-1 8 G (7o 8 G-miR-1) is o 8 It sufficiently induces cardiac hypertrophy by inhibiting a new target through GA base pairing.

[0006] Several miRNAs have been reported to promote (so-called oncomiRs; e.g., miR-17 / 82 and miR-21) or suppress (e.g., miR-34, let-7, and miR-26a) tumor growth. Changes in miRNA expression have been linked to tumor malignancy, which is characterized by changes in miRNA-mediated target gene silencing, a hallmark of tumor initiation and progression. Although o 8 Although G variants may serve as a potential regulatory mechanism for oncogenic miRNAs, their general epitranscriptional function for redox regulation in cancer development has not been studied.

[0007] Accordingly, the present inventors confirmed that the progression of cancer can be analyzed and treated by controlling it using microRNA in which the guanine at the 2nd to 7th nucleotide from the 5' end is modified to 8-oxoguanine.

[0008] Accordingly, the purpose of the present invention is to provide a microRNA isolated from a subject in which at least one guanine (Guanine; G) among the 2nd to 12th nucleotides from the 5' end is replaced with 8-oxoguanine (o).8 It provides a method for providing information for diagnosing or predicting prognosis of cancer, including an identification step of determining whether the cancer has been transformed into G).

[0009] Another object of the present invention is to provide a pharmaceutical composition for treating cancer, comprising at least one selected from the group consisting of a microRNA in which at least one guanine among the 2nd to 12th nucleotides from the 5' end is modified to 8-oxoguanine; a modified nucleic acid that specifically binds to the modified microRNA; a recombinant vector comprising a gene encoding the modified nucleic acid; an analog in which at least one guanine among the 2nd to 12th nucleotides from the 5' end is substituted with uracil (U); and a recombinant vector comprising a gene encoding the analog.

[0010] Another object of the present invention is to provide a method for screening cancer therapeutic agent candidates, comprising the following steps:

[0011] A processing step of treating a candidate substance to tissue or cells separated from a subject;

[0012] An identification step of confirming whether one or more guanines among the 2nd to 12th nucleotides from the 5' end of the microRNA expressed in the tissue or cell are modified to 8-oxoguanine; and

[0013] A comparison step of comparing whether the frequency of guanine at the 4th nucleotide from the 5' end of the microRNA being modified into 8-oxoguanine increases or decreases, whether the frequency of guanine at the 3rd nucleotide from the 5' end of the microRNA being modified into 8-oxoguanine increases or decreases, or whether the frequency of guanine at the 2nd and 3rd nucleotides from the 5' end of the microRNA being modified into 8-oxoguanine increases or decreases.

[0014] Another object of the present invention relates to the use of 8-oxoguanine-modified microRNA for cancer diagnosis or prognosis analysis, cancer treatment, and cancer therapeutic agent screening.

[0015] The present invention relates to 8-oxoguanine (8-oxoguanine; o 8 The present invention relates to a method for providing information for cancer diagnosis or prognosis prediction using microRNA including G), a cancer treatment agent and a screening method thereof, wherein the microRNA including 8-oxoguanine according to the present invention exhibits functions as a cancer diagnosis or prognosis analysis agent, a cancer treatment agent, and a cancer treatment agent screening agent, depending on its location.

[0016] The present inventors have confirmed that the progression of cancer can be analyzed and treated by controlling it using microRNA in which the guanine at the 2nd to 7th nucleotide from the 5' end is modified to 8-oxoguanine.

[0017] Hereinafter, the present invention will be described in more detail.

[0018] One aspect of the present invention is a method for providing information for diagnosing or predicting prognosis of cancer, including an identification step of confirming whether one or more guanines (G) among the 2nd to 12th nucleotides from the 5' end of a microRNA isolated from a subject have been modified to 8-oxoguanine.

[0019] The above identification step may preferably be, but is not limited to, checking whether one or more of the 2nd to 9th nucleotides, one or more of the 2nd to 7th nucleotides, or one or more of the 2nd to 4th nucleotides from the 5' end of the microRNA has been modified to 8-oxoguanine.

[0020] In the present invention, the identification step may include the following steps:

[0021] A sequencing step of producing cDNA by reverse transcribing RNA containing 8-oxoguanine isolated from a subject and creating a sequencing library to identify the location of 8-oxoguanine; and

[0022] An analysis step for confirming the position where guanine is transformed into thymine (T) based on the above sequencing results, thereby confirming the position where guanine is transformed into 8-oxoguanine.

[0023] In the present invention, the microRNA may be at least one selected from the group consisting of miR-124, let-7, and miR-122.

[0024] In the present invention, the information providing method may additionally include a decision step of determining that the probability of developing glioma is low or the probability of having a good prognosis is high when the guanine at the 4th nucleotide from the 5' end of microRNA is modified to 8-oxoguanine.

[0025] In the present invention, the information providing method may additionally include a decision step of determining that the probability of developing liver cancer is low or the probability of having a good prognosis is high when the guanine at the 2nd and 3rd nucleotides from the 5' end of the microRNA is modified to 8-oxoguanine.

[0026] In the present invention, the information providing method may additionally include a determination step of determining that the probability of developing liver cancer or the probability of having a poor prognosis is high when either the guanine at the 3rd and 4th nucleotides from the 5' end of the microRNA is mutated to 8-oxoguanine.

[0027] In one embodiment of the present invention, the guanine at the 4th nucleotide from the 5' end of miR-124 (SEQ ID NO: 3) isolated from the subject is modified to 8-oxoguanine, thereby 8If identified as G-miR-124 (SEQ ID NO: 4), it can be determined that the probability of developing glioma is low or the probability of having a good prognosis is high.

[0028] In one embodiment of the present invention, the guanine at the 4th nucleotide from the 5' end of let-7 (SEQ ID NO: 5) isolated from the target is modified to 8-oxoguanine, thereby 4o 8 If identified as G-let-7 (SEQ ID NO: 6), it can be determined that the probability of developing glioma is low or the probability of having a good prognosis is high, and it can also be determined that the probability of developing liver cancer is high or the probability of having a poor prognosis is high.

[0029] In one embodiment of the present invention, the guanine at the 2nd and 3rd nucleotides from the 5' end of miR-122 (SEQ ID NO: 7) isolated from the subject is modified to 8-oxoguanine, thereby 2,3o 8 If identified as G-miR-122 (SEQ ID NO: 9), it can be determined that the probability of developing liver cancer is low or the probability of having a good prognosis is high.

[0030] In one embodiment of the present invention, the guanine at the 3rd nucleotide from the 5' end of miR-122 (SEQ ID NO: 7) isolated from the subject is modified to 8-oxoguanine, thereby 3o 8 If identified as G-miR-122 (SEQ ID NO: 8), it can be determined that the probability of developing liver cancer is high or the probability of having a poor prognosis is high.

[0031] Another aspect of the present invention is

[0032] One or more guanines (G) in the 2nd to 12th nucleotides from the 5' end are 8-oxoguanine (o 8 MicroRNA modified with G);

[0033] A modified nucleic acid that specifically binds to the modified microRNA;

[0034] A recombinant vector comprising a gene encoding the above modified nucleic acid;

[0035] An analogue in which one or more of the guanine positions at the 2nd to 12th nucleotides from the 5' end is substituted with uracil (U); and

[0036] A recombinant vector comprising a gene encoding the above analogue;

[0037] A pharmaceutical composition for treating cancer comprising at least one selected from the group consisting of:

[0038] In the present invention, the composition

[0039] A microRNA in which one or more guanines among the 2nd to 7th nucleotides from the 5' end are modified to 8-oxoguanine;

[0040] A modified nucleic acid that specifically binds to the modified microRNA;

[0041] A recombinant vector comprising a gene encoding the above modified nucleic acid;

[0042] An analogue in which one or more guanine positions among the 2nd to 7th nucleotides from the 5' end are substituted with uracil (U); and

[0043] A recombinant vector comprising a gene encoding the above analogue;

[0044] It may include at least one selected from the group consisting of .

[0045] In one embodiment of the present invention, the composition comprises

[0046] MicroRNA in which the guanine at the 4th nucleotide from the 5' end is modified to 8-oxoguanine;

[0047] An analogue in which the guanine position at the 4th nucleotide from the 5' end is replaced with uracil (U);

[0048] A recombinant vector comprising a gene encoding the above analogue;

[0049] MicroRNA in which the guanine at the 2nd and 3rd nucleotides from the 5' end is modified to 8-oxoguanine;

[0050] An analogue in which the guanine positions at the 2nd and 3rd nucleotides from the 5' end are substituted with uracil (U);

[0051] A recombinant vector comprising a gene encoding the above analogue;

[0052] A modified nucleic acid that specifically binds to a microRNA in which either the guanine at the 3rd and 4th nucleotides from the 5' end is modified to 8-oxoguanine; and

[0053] A recombinant vector comprising a gene encoding the above modified nucleic acid;

[0054] A pharmaceutical composition for treating cancer comprising at least one selected from the group consisting of:

[0055] The above analogue has uracil (U) substituted for guanine (G), and can mimic the complementary base arrangement of 8-oxoguanine and adenine (A). Therefore, by expressing the above analogue as a vector, a cancer therapeutic effect can be induced.

[0056] In the present invention, the modified microRNA may be at least one selected from the group consisting of miR-124, let-7, and miR-122.

[0057] In the present invention, the modified nucleic acid may include a polynucleotide complementary to six or more consecutive polynucleotides starting from either the second or third nucleotide from the 5' end of the modified microRNA.

[0058] The above modified nucleic acid may include adenine (A) as a nucleotide at a position that binds to one or more 8-oxoguanines among the 1st to 9th nucleotides from the 5' end of the modified microRNA.

[0059] In the present invention, the cancer may be glioma or liver cancer.

[0060] In one embodiment of the present invention, the pharmaceutical composition for treating glioma may be at least one selected from the group consisting of:

[0061] 4o where guanine at the 4th nucleotide from the 5' end of miR-124 (SEQ ID NO: 3) is modified to 8-oxoguanine 8 G-miR-124 (SEQ ID NO: 4), 4o in which the guanine position at the 4th nucleotide from the 5' end is replaced with uracil (U) 8 An analogue of G-miR-124, or a recombinant vector comprising a gene encoding said analogue; and

[0062] 4o where the guanine at the 4th nucleotide from the 5' end of let-7 (SEQ ID NO: 5) is modified to 8-oxoguanine 8 G-let-7 (SEQ ID NO: 6), 4o in which the guanine position at the 4th nucleotide from the 5' end is replaced with uracil (U) 8 An analogue of G-let-7, or a recombinant vector comprising a gene encoding said analogue.

[0063] In one embodiment of the present invention, the pharmaceutical composition for treating liver cancer may be at least one selected from the group consisting of:

[0064] 2,3o in which guanine at the 2nd and 3rd nucleotides from the 5' end of miR-122 (SEQ ID NO: 7) is modified to 8-oxoguanine 8 G-miR-122 (SEQ ID NO: 9), 2,3o in which the guanine positions at the 2nd and 3rd nucleotides from the 5' end are substituted with uracil (U) 8 An analogue of G-miR-122, or a recombinant vector comprising a gene encoding said analogue;

[0065] 4o where the guanine at the 4th nucleotide from the 5' end of let-7 (SEQ ID NO: 5) is modified to 8-oxoguanine 8 A modified nucleic acid that specifically binds to G-let-7 (SEQ ID NO: 6), or a recombinant vector comprising a gene encoding the modified nucleic acid; and

[0066] The guanine at the third nucleotide from the 5' end of miR-122 (SEQ ID NO: 7) is changed to 8-oxoguanine, resulting in 3o 8 A recombinant vector comprising a modified nucleic acid that specifically binds to G-miR-122 (SEQ ID NO: 8), or a gene encoding the modified nucleic acid.

[0067] When the pharmaceutical composition for treating liver cancer comprises the modified nucleic acid, it may be a modified nucleic acid that specifically binds to the modified microRNA, and the modified nucleic acid may include a polynucleotide complementary to at least six consecutive polynucleotides starting from any one of the 2nd and 3rd nucleotides from the 5' end of the modified microRNA. The modified nucleic acid may include an adenine as a nucleotide at a position that binds to at least one 8-oxoguanine among the 1st to 9th nucleotides from the 5' end of the modified microRNA.

[0068] The above modified nucleic acid is an RNA interference-inducing nucleic acid and can exert an effect of suppressing cancer cells or cancer tissues when injected into a subject, tissue or cell isolated from the subject.

[0069] Another aspect of the present invention is a method for screening a cancer therapeutic agent candidate, comprising the following steps:

[0070] A processing step of treating a candidate substance to tissue or cells separated from a subject;

[0071] An identification step of confirming whether one or more guanines among the 2nd to 12th nucleotides from the 5' end of the microRNA expressed in the tissue or cell are modified to 8-oxoguanine; and

[0072] A comparison step of comparing whether the frequency of guanine at the 4th nucleotide from the 5' end of the microRNA being modified into 8-oxoguanine increases or decreases, whether the frequency of guanine at the 3rd nucleotide from the 5' end of the microRNA being modified into 8-oxoguanine increases or decreases, or whether the frequency of guanine at the 2nd and 3rd nucleotides from the 5' end of the microRNA being modified into 8-oxoguanine increases or decreases.

[0073] The above identification step may preferably be, but is not limited to, checking whether one or more of the 2nd to 9th nucleotides, one or more of the 2nd to 7th nucleotides, or one or more of the 2nd to 4th nucleotides from the 5' end of the microRNA has been modified to 8-oxoguanine.

[0074] In the present invention, the microRNA may be at least one selected from the group consisting of miR-124, let-7, and miR-122.

[0075] In the present invention, the cancer may be glioma or liver cancer.

[0076] The above screening method may additionally include a selection step of selecting the candidate substance as a glioma treatment agent candidate when the frequency of guanine at the 4th nucleotide from the 5' end of the microRNA being modified to 8-oxoguanine increases in the comparison step.

[0077] In one embodiment of the present invention, the screening method is performed in the selection step by modifying the guanine at the 4th nucleotide from the 5' end of miR-124 (SEQ ID NO: 3) isolated from the subject to 8-oxoguanine, thereby 8 When the frequency of identification with G-miR-124 (SEQ ID NO: 4) increases, or the guanine at the 4th nucleotide from the 5' end of let-7 (SEQ ID NO: 5) is transformed into 8-oxoguanine, 4o 8 If the frequency of identification with G-let-7 (sequence number 6) increases, the candidate substance may be selected as a candidate substance for glioma treatment.

[0078] The above screening method may additionally include a selection step of selecting the candidate substance as a liver cancer treatment agent candidate substance when, in the comparison step, the frequency of guanine at the 2nd and 3rd nucleotides from the 5' end of the microRNA being modified to 8-oxoguanine increases, when the frequency of guanine at the 4th nucleotide from the 5' end of the microRNA being modified to 8-oxoguanine decreases, or when the frequency of guanine at the 3rd nucleotide from the 5' end of the microRNA being modified to 8-oxoguanine decreases.

[0079] In one embodiment of the present invention, the screening method comprises modifying the guanine at the 2nd and 3rd nucleotides from the 5' end of miR-122 (SEQ ID NO: 7) isolated from the subject in the selection step to 8-oxoguanine, thereby forming 2,3o 8 When the frequency of identification with G-miR-122 (SEQ ID NO: 9) increased, the guanine at the 4th nucleotide from the 5' end of let-7 (SEQ ID NO: 5) was transformed into 8-oxoguanine, resulting in 4o 8 When the frequency of identification with G-let-7 (SEQ ID NO: 6) is decreased, or the guanine at the 3rd nucleotide from the 5' end of miR-122 (SEQ ID NO: 7) is changed to 8-oxoguanine, 3o 8If the frequency of identification with G-miR-122 (SEQ ID NO: 8) decreases, the candidate substance may be selected as a candidate substance for liver cancer treatment.

[0080] The present invention relates to 8-oxoguanine (8-oxoguanine; o 8 The present invention relates to a method for providing information for diagnosing or predicting prognosis of cancer using microRNA including G, a cancer treatment agent and a screening method thereof, wherein the microRNA in which the guanine (G) at the 2nd to 7th nucleotide from the 5' end is modified to 8-oxoguanine can be used to analyze and control the progression of cancer, and thus can be effectively used for diagnosing, analyzing prognosis and treating cancer.

[0081] Figure 1a shows the reverse transcription (right) of o 8 o due to G:A pairing 8 Single nucleotide variant (SNV) analysis of tumor miRNA cDNA sequencing data (n=8,687; TCGA and GEO databases, left) focusing on G>T variants.

[0082] Figure 1b shows the distribution of positional SNVs (n=26,305) within high-confidence tumor miRNAs (n=852) classified by average read count per patient (avr, represented as heatmap intensity; top right panel); Top 100 vs. Bottom 100 SNVs; two-sided Kolmogorov-Smirnov (KS) test; and comparison of positional variant fractions between SNVs (n=10,000; right panel).

[0083] Figure 1c is a diagram showing the compositional analysis of the base types (A, T, G, and C; left panel) and mutation types of G (G>T, G>C, and G>A; right panel) within the top SNVs in the origin region.

[0084] Figure 1d is a hierarchical clustering of G>T SNVs (y-axis) in expressed miRNAs (n=85) for LGG patients (n=512, x-axis), displayed as a heatmap (SNV frequency; %) with location and clinical information of the G>T variants (“Survival”, dark gray line for early death; “Grade”, intense pink for high-grade tumors; “Oligodendro-”, oligodendroglioma; clusters of G>T SNVs highlighted with dotted lines; clusters of LGG patients shaded yellow (tumor suppressive) or red (tumor promoting)).

[0085] Figure 1e is a diagram showing the same hierarchical analysis as in Figure 1d, excluding LIHC, according to the sorted G>T positions (positions 2-18, y-axis) (“Alcohol”, alcohol consumption; “HBV”, hepatitis B virus infection; “HCV”, hepatitis C virus infection).

[0086] Figure 1f is a diagram showing a similar analysis to Figure 1e except that paired tumor (T1-T47) and normal samples (N1N47) from the same patient were used in LIHC (n=47; f), with the pairs displayed in order.

[0087] Figure 1g is a multiple sequence alignment diagram of the G>T cluster sequences of the 2nd to 4th nucleotides individually derived and identified from the results of Figures 1d-f (shown with information content (bit score, strength of sequence alignment) and consensus sequence pattern (Weblogo), G>T SNV positions highlighted by red dots).

[0088] Figure 1h is a multiple sequence alignment diagram of the G>T cluster sequences of the 5th to 6th nucleotides individually derived and identified from the results of Figures 1d-f.

[0089] Figure 1i is a multiple sequence alignment diagram of the G>T cluster sequences of the 7th to 8th nucleotides individually derived and identified from the results of Figures 1d-f.

[0090] Figure 1j is a multiple sequence alignment diagram of the G>T cluster sequences of the 11th to 12th nucleotides individually derived and identified from the results of Figures 1d-f.

[0091] Figure 1k is a diagram showing the G>T SNV levels of miRNAs calculated from the total read counts (log2[G>T SNV reads]) of LIHC and LGG patients (enriched G>T SNVs in miR-122, miR-124, and let-7 families are highlighted).

[0092] Figure 1l shows the in vitro oxidation of RNA (synthetic miR-122, let-7a and miR-124) after Fenton reaction. 8 This is a diagram showing the results of dot blot analysis of G (RNA (SYBR Gold), o 8 G(anti-o 8 G antibody)).

[0093] Figure 1m is a comparative diagram of the positional sequence variations in cDNA of synthetic miR-122 (left panel), let-7a (center panel) and miR-124 RNA (right panel) before (top panel) and after in vitro oxidation (Fenton reaction; bottom panel) (initial region, red shading indicates normalized mismatch (>T, %); in particular, o in the sequencing of in vitro oxidized RNAs 8 G>T frequency was high in dot blot analysis (i) 8 It was very low compared to the G signal (j), which is o 8 o based on G>T sequencing 8 (meaning an underestimation of the G transformation level).

[0094] Figure 1n is a plot of the results of the same analysis as in Figure 1j except that random RNA fragments of artificial transcripts (without GG to exclude the influence of positional bias) were used after in vitro oxidation (mock vs. Fenton reaction) (relative position analysis for G>T (%) 5' end (left graph) and 3' end (right panel)).

[0095] Figure 2a is a diagram showing the results of hierarchical clustering analysis of LGG patients (n=512, y-axis) according to positional SNVs in the miR-124 sequence (frequency of SNVs (SNV reads / corresponding miRNA reads) is represented by heatmap intensity (%), 4G>T clusters are highlighted in red shading).

[0096] Figure 2b is a distribution diagram of the frequency (%) of 4G>T SNVs in miR-124 according to patient survival (years) analyzed by box plot (lines represent median, 1st and 3rd quartiles; whiskers represent minimum and maximum values; 'x' represents mean).

[0097] Figure 2c is a plot showing LGG patients with high 4G>T (upper quartile, 25%; red boxes, n=28) versus low 4G>T ratio (lower quartile, 25%; blue dots, n=28) (miR-124 (>100 RPM; gray boxes, low expression; left panel), Kaplan-Meier survival analysis (right panel); 5-year survival, P = 0.018, Wilcoxon signed-rank test (two-tailed)).

[0098] Figures 2d to 2f are diagrams showing the same analysis as in Figures 2a to 2c except for the let-7 family (wherein SNVs of let-7 family members were co-edited in the initiation region (positions 2-8); the 4G>T let-7 ratio of all LGG patients (n=512) was considered for Kaplan-Meier survival analysis (f, split quartiles, n=39; P=0.048, log-rank test, two-tailed)).

[0099] Figures 2g and 2h are diagrams showing hierarchical clustering of LIHC patients (n=47) based on positional SNVs of miR-122 in matched tumor-normal pairs (g) performed in Figure 2a (patient clusters with increased 3G>T miR-122 ratio in tumors are highlighted with red dashed boxes; SNVs at positions 2 and 3 of miR-122 (h) are highlighted with enlargements).

[0100] Figures 2i and 2j are drawings showing the same survival analysis as in Figures 2b and 2c except for 3G>T miR-122 in LIHC patients (n=41; P=4.6 x 10 -4 , log-rank test, two-tailed (j).

[0101] Figures 2k to 2m are drawings of the same analysis as Figures 2d to 2f except that matched tumor-normal pairs of LIHC patients (n=47) for the let7 series (Figures 2k and 2l) were used (survival analysis for 4G>T let-7(m); split quartiles, n=40; ​​P=0.017, log-rank test, two-tailed).

[0102] Figure 3a shows the o performed on HS683. 8 G immunoprecipitation (o 8 Schematic diagram of oxidized miRNA sequencing based on G-miSeq (o 8 The position of G is o during reverse transcription 8 Can be identified at single nucleotide resolution by G>T transitions, o 8 The G-miSeq method is o 8 G>T position based on immunoprecipitation and single nucleotide resolution compared to o 8 (proven to achieve 3,500-fold enrichment of G).

[0103] Figure 3b shows o in HS683 after antioxidant treatment. 8 The sequencing results of G-miSeq are plotted as a volcano plot (significance, -log 10(P-value); left panel) and MA plot (o 8 G immunoprecipitation (log2(o 8 G-IP)) and o 8 G enrichment (log2[o 8 G-IP / input], normalized by input lead-count o 8 Relative o based on read-count of G) 8 G enrichment; miR-124 (red) and let-7 family (orange, members also shown in orange) are shown).

[0104] Figure 3c is o 8 o Based on highly reliable G>T SNVs (n=60) derived from G-miSeq analysis 8 This is a location distribution diagram of G variants (only G>T SNVs in the origin region are indicated by cluster number; heatmap intensity: frequency (%) of G>T in miRNA).

[0105] Figure 3d shows o after antioxidant treatment. 8 This diagram shows positional SNVs in the let-7 family in G-miSeq (expressed as normalized mismatch (%); *P = 5.1 x 10 -7 ).

[0106] Figure 3e is 4o 8 G- or 4U-miR-124 (Fig. 3e, left panel) or 4o 8 Transfection of G- or 4U-let-7a (Fig. 3e, right panel) compared with non-targeting control miRNA (NT); cytometric analysis of total apoptosis (%, red, low values) measured by propidium iodide (PI) and annexin V staining (Fig. 3f) (*P = 3.5 x 10 each). -11 , 1.2 x 10 -9 , 2.7 x 10 -8 and 4.6 x 10 -8 ; In particular, 4o 8 G - Or 4U-let-7a has a significant number of PIs+ / Annexin V - This is a diagram showing the results of measuring proliferating cells by Ki67 staining (Fig. 3g) (blue, Ki67) that induces cell death, which is suspected to be a type of cell death other than apoptosis. - ; Red, Ki67 + ; * P = 1.5 x 10 each -6 , 3.0 x 10 -6 , 6.7 x 10 -6 and 2.7 x 10 -5 ; Quantitative results (Figs. 3f and 3g; right panel); n = 3, *P < 0.05).

[0107] Figure 3h shows 4o including tandem repeats of winter target sites (13 sites; 13x). 8 A diagram showing the competitive sponge inhibitors for G-modified miR-124 or let-7 (anti-4oxo) (anti-oxo is o 8 Effectively through competitive engagement through GA 8 (It can deplete G-miRNAs, preventing them from participating in RNA silencing).

[0108] Figure 3i is from HS683, endogenous 4o 8 G-miR-124 (left panel) or 4o 8 A diagram showing the luciferase reporter (4oxo site; 5x) assay to validate anti-4oxo (13x; pDual) against G-let-7 (right panel), reporter without site ('-'), control RNA transfection ('cont'); and reporter with initiation site (positions 2-8, 'seed'; 5x) ( * P = 0.0024 and 2.5 x 10 -5 ).

[0109] Figures 3j and 3k show the expression of miR-124 (Figure 3j, *P = 1.8 x 10 -6 ) or let-7 (Fig. 3k, * P = 1.4 x 10 -4) is a diagram showing the same proliferation assay as in Figure 3g, except for the transfection of anti-4oxo(5x).

[0110] Figure 3l shows invasion assay using Matrigel-based transwells after transfection of HS683 with anti-4oxo(13x) against miR-124 or let-7; whole Matrigel inserts (upper left panel) or enlarged images (lower left panel) stained with hematoxylin and eosin (H&E); and quantitative results (right panel) of enlarged replicate measurements (inner set, mean ± standard deviation (sd)). * P = 0.046 and 0.033, respectively).

[0111] Figure 3m is a drawing showing the same apoptosis assay as in Figure 3f, except that anti-4oxo(5x) against miR-124 or let-7 was transfected with temozolomide (TMZ) treatment. * P = 4.1 x 10 each -7 and 9.2 x 10 -7 ; Unless otherwise stated, all P values ​​were obtained from two-tailed t-tests and replicated in biologically independent samples (n≥3; data are presented as mean±standard deviation).

[0112] Figure 4a is 4o 8RNA-Seq analysis of HS683 cells transfected with G-miR-124, plotted as cumulative distribution function (CDF) plot (left panel) and volcano plot (right panel) compared to non-target control (NT) expression (4-oxo sites in 3'-UTR of miR-124, 7mer (positions 2-8); P-value of KS test compared to total mRNA (left panel); differentially expressed genes (DEGs) down-regulated using 4-oxo sites (red dotted line) selected as cutoffs (right panel, orange shaded) at statistical significance (>1.3, y-axis; P<0.05, two-tailed t-test) and fold change (log2 ratio <-0.8, x-axis) are shown; significant target genes are highlighted in red with labels).

[0113] Fig. 4b is 4o 8 A diagram showing the same RNA-Seq analysis as in Figure 4a except for G-let-7a transfection and a different fold change cutoff (log2 ratio <-0.4).

[0114] Figure 4c shows the endogenous 4o by anti-4oxo (13x; pDual plasmid) transfection. 8 RNA-Seq analysis of the inhibition of G-miR-124; it contains a 4-oxo site in its 3'-UTR, and 4o 8 A diagram showing the same CDF analysis as in Figure 4a except that derepression of putative 4oxo targets (left panel) showing G-miR-124-dependent downregulation (log2 ratio <0) was examined (same volcano analysis as in Figure 4a except using a different fold change cutoff (log2 ratio > 0.2).

[0115] Figure 4d shows endogenous 4o by anti-4oxo(5x) transfection. 8A diagram showing the same RNA-Seq analysis as in Figure 4c, except that G-let-7 was antagonized and a different fold change cutoff (log2 ratio > 0.8) was used.

[0116] Figure 4e is 4o 8 Gene ontology (GO) analysis diagram of high confidence 4oxo targets downregulated by G-miR-124 (log2 ratio <0 in Fig. 4a) and upregulated by anti-4oxo (log2 ratio >0 in Fig. 4c) (semantic space as cluster representatives; color and proportional size of circles are log 10 (P value); P value was calculated using DAVID (Database for Annotation, Visualization and Integrated Discovery).

[0117] Figure 4f is 4o 8 A diagram showing the same GO analysis as in Figure 4e, except for G-let-7.

[0118] Figures 4g and 4h show 4o using qPCR. 8 A diagram showing high-confidence 4oxo target validation in G-miR124- (Fig. 4g) or anti-4oxo-transfected HS683 (Fig. 4h) (relative levels are 4o 8 Transcript levels are shown in transfections of G-miR-124 and NT (Fig. 4g), or anti-4oxo and anti-NT ('cont') (Fig. 4h); * P = 7.0 x 10 each -3 , 1.6 x 10 -3 , 5.1 x 10 -5 , 3.4 x 10 -4 and 3.7 x 10 -3 (Fig. 4g); * P = 7.2 x 10 each -4 , 0.014, 0.037, 0.030, and 0.0472 (Fig. 4h)).

[0119] Figures 4i and 4j are 4o8 A diagram showing the same target validation analysis as in Figures 4g and 4h, except for G-let-7. * P = 0.012, 0.022, 0.002, 0.009, and 0.002, respectively (Fig. 4i); * P = 0.045, 0.005, 1.8 x 10 respectively -4 , 0.018 and 0.005 (Fig. 4j)).

[0120] Figures 4k and 4l are volcano plots for LGG (n= 507, TCGA); 4-oxo site in 3'-UTR, 7mer (positions 2-8); down-regulated targets (cutoffs (red dashed lines) for statistical significance (P < 0.05) and fold change (log2 ratio <-0.5); P < 0.05 or P < 0.1, log2 ratio < 0).

[0121] Figure 4m is a diagram showing the overlap of reproducible peaks of AGO RIP-Seq with miR-124 4-oxo sites in the 3'UTR of MEOX2 and EGFP, compared to human brain (BA4) from a glioblastoma (biological complexity, BC ≥ 2; red) patient (gray: 6-mer sites; red: 7-mer sites; AGO-mRNA reads for the 3'UTR of both EGFR isoforms are also shown as a heatmap due to their high peak heights).

[0122] Figure 4n is a diagram showing the same analysis as in Figure 4m except for the let-7 4-oxo moiety in CLTA and PUDP (all P values ​​are from biologically independent samples (n≥3) repeated; data are presented as mean ± sd unless otherwise stated).

[0123] Figure 5a is a diagram comparing the early (5 and 6 weeks) and late stages (17 and 18 weeks) of the establishment of a DEN-induced LIHC rat model by intraperitoneal (ip saeuaufi df,f; d,f;a ehdn.P.) injection of DEN (upper panel) (red arrows: tumor tissue; immunohistochemistry: hematoxylin and eosin (H&E), fibrosis: blue area; Ki67, proliferating cells: brown; 100 μm).

[0124] Figure 5b shows the quantification of ROS (n=12) in DEN-treated liver tissue based on relative fluorescence units (RFU) from CM-H2DCFDA (upper panel). * P = 3.1 x 10 -8 (Initial) and 8.3 x 10 -6 (Review)) and o of small RNA (<200 nucleotides) by ELISA 8 G quantification (bottom panel) ( * P = 1.8 x 10 -2 (Initial) and 2.5 x 10 -4 This is a drawing showing (review)) (o 8 G ratio, o per ng of total RNA 8 Amount of G(pg)) (ELISA method is to digest small RNA extracted from several random positions (n>3) in DEN-treated liver with nuclease to mononucleotide 8 Since G was quantified, taking into account the four base types, the average length of small RNA (~20 nucleotides), and the proportion of tumor cells in DEN-treated livers (~10%), o 8 0.05 in G ratio (late) is ~4% of total o 8 (which can be interpreted as G-miRNA).

[0125] Figures 5c and 5d are o 8 G-miSeq results are presented in MA plots (o) at different stages of DEN-induced LIHC. 8 G (log2(o 8G-IP)), top panel) and volcano plot in Fig. 5c (significance, -log 10 (P-value), lower panel) is a diagram showing miR-122 (red) and let-7 family (members are indicated in orange); location o 8 G distribution (based on G>T SNV ratio (%)) is o 8 o in the heatmap of the G>T miRNA cluster 8 Displayed with the amount of G-IP (number of readings, log2 ratio) (Fig. 5d)).

[0126] Figures 5e and 5f are o for tumor tissues derived from LIHC patients (n=2) performed in Figures 5c and 5d. 8 G-miSeq analysis, MA plot (o 8 G concentration vs o 8 A plot showing the amount of G (expressed as log2(o8G-IP); Fig. 5e) and the position frequency in replicates (Fig. 5f) (all P values ​​were obtained from two-tailed t-tests and replicates in biologically independent samples; n ≥ 3, data are means ± SD unless otherwise stated).

[0127] Figure 6a is o of Huh7 8 This is a schematic diagram of a G-miSeq experiment.

[0128] Figure 6b is o 8 o analyzed for G concentration 8 This is a diagram showing the G-miSeq results (log2(o) in Huh7 8 G-IP / input), MA plot (o 8 The amount of G, log2(o 8 (G-IP) are shown; miR-124 (red) and let-7 family (orange, members are also shown in orange) are highlighted with dotted lines).

[0129] Figures 6c and 6d show a heatmap for miRNAs in the G>T cluster in terms of positional G>T SNV ratio (%) (Figure 6c, left panel), and a heatmap for fold change (log2(antioxidant / mock)) in antioxidant-treated Huh7 (Figure 6c, right panel) (o 8 G-miR-122 and -let-7 are o 8 There was no significant decrease in G concentration, and the position o under antioxidant treatment 8 (Analysis showed that the G>T ratio was reduced); A diagram showing the positional SNV of miR-122 in antioxidant-treated Huh7 cells (Fig. 6d) (expressed as normalized mismatch (%); black; * P =4.83 x 10 -8 ).

[0130] Figures 6e and 6f show o in Huh7 between two mismatches (2MM) and one mismatch (1MM). 8 Comparison of G>T mismatch reads of G-miRNA (Fig. 6e); 2MM analysis of miR-122 sequences containing 2G>T (2MM with 2G>T, orange), 3G>T (2MM with 3G>T, red) or other mutations (2MM with other mutations, gray) (Fig. 6f) (o 8 Analyzed for the quantity of G (log2(o 8 G-IP); relative o 8 G ratio(o 8 Log2 ratio of G-IP, antioxidant vs. mock treatment, dotted line for no change; 2,3G>T highlighted by red arrows).

[0131] Figure 6g shows the construction of miR-122 KO in Huh7 cells using CRISPR / Cas9 (left panel) and o using a luciferase reporter with an oxo site (e.g., 2,3 oxo site). 8 G-miR-122 (e.g., 2,3o 8 This is a diagram that investigates the endogenous type of G-miR-122. *P = 3.2 x 10 compared to miR-122 KO Huh7 cells -4 , 3.6 x 10 -6 and 4.7 x 10 -6 (From left to right); Endogenous 2,3o in Huh7 8 G-miR-122 (~22% reduction) and 3o 8 The inhibitory activity of GmiR-122 (~18% reduction) was significant, and their effective concentration was estimated to be equivalent to transfecting cells with ~1-10 nM.

[0132] Fig. 6h is 3o 8 G- or 2,3o 8 Cytometric analysis of proliferating cells after transfection of HepG2 cells with G-miR-122; sequence (top panel); Ki67 staining (bottom panel); blue, Ki67. - ; Red, Ki67 + ; This is a diagram comparing transfection with non-target control miRNA (NT, gray) ( * P = 0.0002 and 0.015, respectively).

[0133] Figures 6i and 6j are 3o 8 G- or 2,3o 8 A diagram showing the invasion assay of miR-122 KO Huh7 cells after transfection of G-miR-122 (shown as H&E-stained Matrigel inserts (Fig. 6i); quantitative results (Fig. 6j, left graph) ( * P = 0.004 and 0.001, respectively, versus NT).

[0134] Figures 6k and 6l are 3o 8 G-miR-122 (anti-3oxo) and 2,3o 8 A diagram showing the same analysis as in Figures 6h and 6i, except that a competitive RNA inhibitor for G-miR-122 (anti2,3oxo) was transfected into Huh7 cells ( * P = 1.4 x 10 for Cont (Fig. 6k) -5 and 2.0 x 10 -5; Quantification of invasive cells (Fig. 6l) is shown in the right graph in Fig. 6j; Huh7 vs miR-122 KO Huh7; * P = 0.022; * P = 0.004 (anti-3oxo) and 0.022 (anti-2,3oxo) (cont) compared to control.

[0135] Figure 6m is a diagram showing the results of flow cytometry analysis using propidium iodide (PI) and annexin V staining for apoptosis assay after transfection of anti-3oxo or anti-2,3oxo RNA into Huh7 cells in the presence of sorafenib treatment (late apoptosis (PI + , Annexin V + ), red shading; early apoptosis (PI - , Annexin V + ), faint red shading; total apoptosis (%, red, low values); * P = 0.018 (anti-3oxo), vs NT; * P = 0.008 (anti-3oxo) and 3.3 x 10 -5 (anti-2,3oxo), NT + Compared to sorafenib; quantitative results (right panel)).

[0136] Figures 6n to 6p are 4o 8 Same cell proliferation as in Figures 6i and 6j, except for G-let-7a (Figure 6n; * P = 0.001) and cell invasion (Figs. 6o and 6p) analysis (quantification of invasive cells (Fig. 6p, left graph); *P = 3.2 x 10 for Huh7 -6 (NT)(cont); * P = 0.002 (4o 8 G-let-7a), NT vs.

[0137] Figures 6q to 6s are drawings showing the same analysis as Figures 6k to 6m except that let-7 anti-4oxo RNA was transfected into Huh7.* P = 0.001 (Fig. 6q); Quantification of invasive cells (Fig. 6r) is shown in the right graph of (Fig. 6p); * P = 3.3 x 10 -5 (anti-4oxo), for cont; * P = 0.038, for NT (Fig. 6s); * P = 0.001, compared with NT+sorafenib (Fig. 6s); all P values ​​were obtained from two-tailed t-tests; biologically independent sample replicates; n≥3; data are mean ± sd unless otherwise stated.

[0138] Figure 7a shows 4o in RNA-Seq analysis 8 This diagram shows the gene ontology (GO) analysis of high-confidence miR-122 3oxo targets in LIHC, which are downregulated by G-miR-124 (log2ratio<0) (log2ratio>0) and upregulated by anti-4oxo (log2ratio>0) (semantic space as cluster representative; color and proportional size of circles are log 10 (P value) is indicated; P value is obtained using DAVID).

[0139] Figures 7b and 7c show 3o using qPCR. 8 A diagram showing high confidence 3oxo target validation in G-miR-122 transfected HepG2 (Fig. 7b) or anti-3oxo-transfected Huh7 (Fig. 7c) (relative levels are 3o 8 Transcription levels are shown in G-miR-122 vs NT (Fig. 7b) or anti-3oxo vs anti-NT ('cont') transfection (Fig. 7c); * P = 1.8 x 10 each -5 , 0.001, 0.013, and 0.006 (Fig. 7b); * P = 0.003, 0.001, 0.001, and 8.2 x 10 respectively -5 (Fig. 7c).

[0140] Figures 7d and 7e are diagrams showing the validation of the miR-122 3-oxo site in AKT1S1 and IL6ST (Figure 7d), the site overlapping with the AGObound region (AGO CLIP) in LIHC patient tissue (red) (6mer site, red; 7mer site, dark red); and the luciferase reporter assay (Figure 7e) using the luciferase reporter assay (half maximal inhibitory concentration, IC50; target site inserted into the luciferase reporter, green shading in Figure 7d).

[0141] Figure 7f is a diagram showing DEG analysis of RNA-Seq data from LIHC (TCGA) patients for the 3G>T miR-122 cluster (n=82) compared to normal (n=50) (3-oxo region in 3'UTR, 7mer (positions 2-8); volcano plot, cutoff for fold change (log2 ratio <-0.5) for downregulated targets (red dashed line)).

[0142] Figure 7g shows the same GO analysis as in a except for the high-confidence miR-122 2,3oxo target (Figure 7g); the miR-122 2,3oxo site in the AGO binding region of ASPH and ZNF146 (Figure 7h) as shown in Figure 7d; the luciferase reporter assay (IC50) for the miR-122 2,3oxo site of ASPH and ZNF146 (Figure 7i) (same as Figure 7e except using SK-HEP-1 cells; target sites inserted into the luciferase reporter, green shading in Figure 7h); 2,3o 8 A diagram showing miR-122 2,3oxo target validation using qPCR in G-miR-122-transfected HepG2 (Fig. 7j) or anti-2,3oxo-transfected Huh7 (Fig. 7k) ( * P = 0.016, 0.001, 0.007, 0.0004, 0.007, 0.034, and 0.003, respectively (Fig. 7j); *P = 0.007, 0.009, 0.018, and 0.005, respectively (Fig. 7k)).

[0143] Figure 7l shows the same GO analysis as in Figure 7a except for the high-confidence let-7 4oxo target (Figure 7l); 4o 8 Same qPCR analysis as in Figures 7b and 7c except for Glet-7a (Figure 7m) and its competitive inhibitor anti-4oxo RNA (4x) (Figure 7n). * P = 0.001, 0.016, 0.04, and 0.011, respectively (Fig. 7m); * P = 0.035, 0.006, 0.014, and 0.003, respectively (Fig. 7n)); as in Fig. 7d; the let-7 4-oxo site in the AGO binding region of DDAH1 and GSR (Fig. 7o); and the same luciferase reporter assay as in Fig. 7e, except for the let-7 4-oxo site of DDAH1 (Fig. 7p).

[0144] Figure 7q is a diagram showing DEG analysis of LIHC (n=125) patient RNA-Seq data, same as Figure 7f, except for the 4G>T let-7 cluster (n=75).

[0145] Figures 7r and 7s show cytometry analysis of miR-122 oxo sites (3oxo or 2,3oxo) using the dFP reporter (Figure 7r, top) in individual Huh7 cells compared with NAC treatment (distribution of fold changes (log2[RFP / GFP]) is shown in parentheses in the violin plot (Figure 7r, bottom)); relative fold changes (3oxo or 2,3oxo sites vs no site, n≥3) are expressed as a function of dFP reporter expression level (log 10 (estimated to be GFP) was plotted (Fig. 7s) and analyzed for the lowest 25% of the RFP signal (within a similar range to GFP); * P < 0.05).

[0146] Fig. 7t is a composite o 8Cellular heterogeneity and o in individual hepatoma cells based on flow cytometry analysis (Fig. 7s) and dFP reporter analysis using luciferase reporter assay under transfection with different concentrations of G-miR-122. 8 This is a diagram of the amount of G-miR-122 (liver cancer cells synthesized ~5-20 nM o 8 o equivalent to transfecting G-miR-122 8 It is assumed that effective concentrations of G-miR-122 can be endogenously produced).

[0147] Figure 7u is a diagram showing the reanalysis of AGO CLIP-Seq data from LIHC patients (n=9); positional SNVs of AGO-associated miR122 (relative site frequency (%) calculated (Figure 7v) with respect to the average frequency of control sites within mRNA peaks in AGO CLIPs from tumor (de novo tumor, red) or normal tissue samples (de novo normal, blue), according to reproducibility (biological complexity, BC) (GU, wobble pairing site; * P < 0.05, relative to control sites including GU sites; all P values ​​were obtained from two-tailed t-tests and data are presented as mean ± sd unless otherwise stated.

[0148] Figure 8a shows that miR-122 levels decrease during tumor promotion 3o during hepatocarcinogenesis. 8 Tumor suppression in G-2,3o 8 Schematic diagram of stepwise reprogramming of miR-122 by increased ROS oxidizing G-miR-122.

[0149] Figure 8b: Tumor size measurement in a xenograft mouse model; 3U-miR-122 (3 x 10 6 cells), 2,3U-miR-122 (8 x 10 6The diagram shows a photograph of the dissected tumor (right panel) after subcutaneous transplantation in parallel with miR-122 KO Huh7 cells transfected with either a control vector or a control vector (the scale bar marks 1 mm).

[0150] Figures 8c and 8d are drawings showing in vivo delivery of 2,3U-miR-122 (intravenous, IV) as a complex with polyethyleneimine (PE) (intraperitoneal injection, IP, postnatal day 14, p14) and non-targeting control miRNA (NT) (Figure 8c); and photographs of dissected livers at 10 months of age (10 M) (Figure 8d) (red arrows, tumor tissue; immunohistochemistry; hematoxylin and eosin (H&E), lipid droplets (circular pores); fibrosis (blue); Ki67, proliferating cells (brown)) to DEN-treated mice.

[0151] Figures 8e to 8h show o in the later stage of tumor formation (12 M). 8 In vivo experiments identical to those in Figure 8c (Figure 8e) except that competitive inhibitors for G-miR-122, anti-3oxo or anti-2,3oxo were delivered (in vivo derepression of corresponding target transcripts by anti-3oxo and anti-2,3oxo was investigated by RNA-Seq with CDF analysis of putative miR-122 oxo targets; P-values ​​of two-sided KS test compared to total mRNA (Figure 8f); liver tissue and H&E, trichrome and Ki67 staining photographs (Figure 8g) (shown in Figure 8d); and simultaneously showing delivery of 4U-let-7a (Figure 8h) (n≥3).

[0152] Figure 8i shows the results (Figure 8i, top) of hepatocyte-specific transgenic (TG; albumin enhancer / promoter) mice expressing 2,3U-miR-122 (Figures 8i and 8j, euthanized at 10M) or 3U-miR-122 (Figures 8k and 8l, euthanized at 8M) after DEN injection (p14) (WT, without transgenic cassette; TG, with transgenic cassette; TG +, high-expression transgenic cassette; WT mice are TG + DEN-injected TG were classified according to the expression levels of 2,3U-miR-122 (Fig. 8j) or 3U-miR122 transgene (Fig. 8l) in mice and high-fat diet (compare for response to DEN-induced hepatocarcinogenesis under high-fat diet and examined as in Fig. 8d). + A schematic diagram showing the mouse liver (measured by qPCR for albumin mRNA (Exp; %, relative to highest reading, normalized to WT); red arrows, tumor tissue).

[0153] Figure 8m shows the transgenic mouse line (TG) expressing competitive sponge inhibitors (12x oxo moiety) at 2,3o in hepatocytes. 8 G-miR-122 (Fig. 8m; anti-2,3oxo) or 3o 8 For G-miR122 (Fig. 8n; anti-3oxo) (albumin enhancer / promoter), DEN treatment with choline-deficient high-fat diet; TG injected with DEN expressing anti-2,3oxo (Fig. 8o, left panel; euthanized at 5.5 M) or anti-3oxo (Fig. 8p, left panel; euthanized at 5 M) as shown in Fig. 8d + Mouse liver photographs; Livers of TG and WT mice were also classified according to the expression levels of anti-2,3oxo (Fig. 8o, right panel) or anti-3oxo transgene (Fig. 8p, right panel) as measured by qPCR as in Fig. 8j and 8l; TG mice of different lines are separately indicated by the dotted line (Fig. 8o, right panel) (all scale bars, 100 μm; all P values ​​are obtained from two-tailed t-tests; repeated in biologically independent samples; data are mean ± sd unless otherwise stated).

[0154] The present invention relates to a microRNA isolated from a subject, wherein one or more guanines (G) among the 2nd to 12th nucleotides from the 5' end are 8-oxoguanine (o). 8 A method for providing information for diagnosing or predicting prognosis of cancer, including an identification step of determining whether the cancer has been transformed into G).

[0155] Hereinafter, the present invention will be described in more detail with reference to the following examples. However, these examples are only intended to illustrate the present invention, and the scope of the present invention is not limited by these examples.

[0156] Throughout this specification, "%" used to indicate the concentration of a particular substance is (wt / wt)% for solid / solid, (wt / vol)% for solid / liquid, and (vol / vol)% for liquid / liquid, unless otherwise stated.

[0157]

[0158] abstract

[0159] Oxidative stress is involved in tumorigenesis by altering gene expression. One such modification is 8-oxoguanine (o 8 G) is o 8 Although it alters RNA-RNA interactions through GA base pairing, its regulatory role remains elusive. Here, o 8 G-induced guanine-to-thymine sequencing variation and o of tumor microRNAs (miRNAs) 8 Based on the G sequence, the present inventors have identified a putative o that is mainly concentrated and oxidized in the 5'-terminal initiating sequence (positions 2 to 8) and is widely distributed and clinically associated with patients with lower-grade glioma (LGG) and liver hepatocellular carcinoma (LIHC). 8The G sequence location was discovered.

[0160] The present inventors have o at the 4th position of miR-124 8 G(4o 8 G-miR-124) and 4o 8 G-let-7 inhibits LGG, whereas 3o 8 G-miR-122 and 4o 8 G-let-7 promotes LIHC malignancy by redirecting the target transcriptome toward oncogenic regulatory pathways.

[0161] Tumor-promoting 3o 8 Tumor-suppressive 2,3o in G-miR-122 8 Stepwise oxidation of G-miR-122 occurs, and its specific inhibition or expression in mouse liver effectively alleviates diethylnitrosamine (DEN)-induced hepatocarcinogenesis. These findings provide o 8 Provides insights into the G-modified redox-regulated miRNA sequence code and its analogs and inhibitors.

[0162]

[0163] Example 1: Data Availability

[0164] In vitro oxidized RNA (SRP391033, SRP391030), small RNA-Seq (SRP353704), o 8 All raw sequencing data from G-miSeq (SRP353708, SRP353711, SRP353718, SRP353713, SRP353717), RNASeq (SRP353641, SRP353644, SRP353643, SRP353687, SRP353639, SRP391054), and whole genome (SRP390848) sequencing have been deposited in the Sequence Read Archive (SRA). All FASTQ files are available at clip.korea.ac.kr.

[0165]

[0166] Example 2: Cell lines and culture

[0167] SH-SY-5Y, HS683, U87MG, HeLa, HepG2, and Huh7 cells purchased from the Korea Cell Line Bank were cultured in DMEM (Dulbecco's modified Eagle's medium; Hyclone, #SH30243.01; SH-SY-5Y, HeLa, HS683, and Huh7) or MEM (Mmodified Eagle's medium; Hyclone, #SH30024.01; HepG2 and U87MG). SK-HEP-1 and Huh7 cells were obtained from HSRRB (Health Science Research Resources Bank) and cultured in DMEM.

[0168] All media were supplemented with 10% fetal bovine serum (FBS) (Gibco, #12483020), 100 U / ml penicillin, and 100 ug / ml streptomycin (Welgene, #LS 202-02) and cultured at 37°C under 5% carbon dioxide. Cell lines were tested for mycoplasma contamination based on PCR (Cosmogenetech, Korea). All cell lines were authenticated by short tandem repeat (STR) DNA profiling (Korea Cell Line Bank).

[0169] To induce oxidative stress, cells were treated with 100 μM hydrogen peroxide (H2O2; Sigma-Aldrich, #216763) for 6 h. Primary hepatocytes isolated from male C67BL / 6N mice (8-12 weeks old) were purified from a Percoll gradient and cultured in medium 199 (M199; Sigma-Aldrich) supplemented with 10% FBS, followed by treatment with 100 μM paraquat (Sigma-Aldrich, #36541) for 24 h.

[0170] When serum starvation was applied, the medium was replaced with the same medium without FBS for 24 h unless otherwise specified. To reduce the redox state, an antioxidant supplement suitable for cell culture (SigmaAldrich, #A1435) was added to the medium for at least 1 week, and optionally, 10 mM N-acetylcysteine ​​(NAC; Sigma-Aldrich, #216763) in 25 mM HEPES (pH 7.2–7.5; WelGENE, #BB001-01) was treated for 24 h before the experiment.

[0171] Transient transfections were performed using reverse transfection with Lipofectamine RNAiMAX (Invitrogen, #13778150) for RNA (miRNA or miRNA inhibitor; 30–50 nM), and Lipofectamine 2000 (Invitrogen, #11668019) or Lipofectamine 3000 (Invitrogen, #L3000015) for vector plasmids alone or together with RNA. The general protocols provided by the manufacturers were followed unless otherwise specified.

[0172] To achieve maximum efficiency, accurate cell numbers were counted using a Countess II Automated Cell Counter (Invitrogen, AMQAF1000) or a hemocytometer (Incyto, #DHC-N01) and plated before transfection. Cell morphology was examined after transfection using an inverted microscope (Leica DMi8).

[0173] For drug treatment, 100 uM temozolomide (TMZ) (Sigma-Aldrich, #T2577) was added to the medium at the time of transfection and harvested after 72 h. 24 h after reverse transfection, 10 uM sorafenib (Sigma-Aldrich, #SML2653) was treated and cultured for 48 h.

[0174] We prepared a miR-122 KO Huh7 cell line using the dual-gRNA system in CRISPR / Cas9. To this end, the miR-122 locus was knocked out of Huh7 cells using a miRNA CRISPR Knockout Kit (Canopy, #CPhsa-miR-122) according to the manufacturer's protocol.

[0175] Briefly, 3.5 x 10 in a 6-well plate 5 Huh7 cells were transfected with 2.5 μg of an all-in-one dual gRNA CRISPR-Cas9 plasmid targeting the genomic region of miR-122. Forty-eight hours after transfection, cells were grown in puromycin-containing selective medium for 2 months; lower concentrations of puromycin (1.6 μg / ml and 1 μg / ml) were used to minimize off-targeting compared to typical conditions (0.1–10 μg / ml).

[0176] Deletion of the intended miR-122 locus in puromycin-resistant cells was confirmed by genomic PCR using primers provided in the kit. MiR-122 knockout was confirmed by qPCR in miR-122 KO Huh7 cells (see Quantitative RT-PCR section).

[0177]

[0178] Example 3: Rat and mouse models

[0179] All animal experiments were approved by the Korea University Institutional Animal Care and Use Committee and were conducted in accordance with the Animal Care and Use Guidelines. In all cases, animals were housed under a standard 12-hour light cycle and had free access to food and water unless otherwise specified.

[0180] For the liver hepatocarcinoma (LIHC) mouse model, 5-week-old male SD rats (Koatech; Young Bio) were administered intraperitoneally (IP) with 70 mg / kg diethylnitrosamine (DEN) (Sigma-Aldrich, #N0756) weekly for a total of 5 or 6 weeks and were euthanized 24 hours after the last injection (“early”) or at the end of the 10-week experiment (weeks 17 or 18; “late”).

[0181] A mock injection control group (“mock”) was created using an equal volume of PBS (Welgene, #LB 004-02). In the case of death in the “mock” control group, comparable “mock” control groups were urgently prepared, including mice that did not receive DEN injection or were euthanized immediately (24 hours and 2 weeks) after a single DEN injection. To determine the antioxidant effect, 100 mg / kg NAC was administered intraperitoneally (IP) together with 70 mg / kg DEN and euthanized at 18 weeks.

[0182] Xenograft production was performed in 5- to 9-week-old male nude mice (nu / nu of BALB / c; Charles River Laboratories, Korea). 3 X 10 6 cells (miR-122 KO Huh7) or 8 X 10 6 Suspensions of cells (miR-122 KO Huh7 and Huh7) were transfected with expression vectors (3U-miR-122 or 2,3U-miR-122) or empty vector (control) on 50% thawed Matrigel (BD Biosciences, #354248), and then injected subcutaneously into both flanks of mice. Tumor growth was measured using a digital caliper and calculated using the following formula.

[0183] [Mathematical Formula 1]

[0184] Tumor volume (mm) 3 ) = length X width X height

[0185] To induce hepatocellular carcinoma in mice, 25 mg / kg DEN was administered intravenously to newborn P14 C57BL / 6 male mice (Koatech). 4 mg / kg RNA was injected into the tail vein (intravenous; IV) using in vivo-jetPEI (N / P ratio=6; Polyplus, #101000030) according to the manufacturer's protocol. Treatments were performed eight times weekly starting at 8.5 months (NT and 2,3UmiR-122) or 10 months (NT, miR-122 anti-3oxo, miR-122 anti-2,3oxo, and 4U-let-7a).

[0186] Mice were euthanized 10 months (NT and 2,3U-miR-122) or 13 months (NT, miR-122 anti-3oxo, miR-122 anti-2,3oxo, and 4U-let-7a) after additional RNA injections. When 2,3U-miR-122 (3.2 mg / kg) or 4U-let-7a (3.2 mg / kg) was injected, NT (0.8 mg / kg) was also injected to determine whether it was delivered to the liver. In particular, the duplex of 2,3U-miR-122 and 4U-let-7a used for injection contained fluorescein isothiocyanate (FITC) at the 5' end to confirm delivery.

[0187] Transgenic mice (TG) with hepatocyte-specific expression of 2,3U-miR-122, 3U-miR-122, anti-2,3oxo (miR-122), or anti-3oxo (miR-122) were generated. Transgene cassettes (~1.6 kb) were obtained from albumin::2,3U-miR-122, albumin::3U-miR-122, albumin::anti-2,3oxo, or albumin::anti-3oxo plasmids by SalI and BamHI treatment (o 8 (See Construction of G-miR-122 mimic and anti-oxo (miR-122) expression vectors) were injected into the pronucleus of mouse zygotes prepared from superovulated C57BL / 6 female mice after mating.

[0188] Modified single-cell zygotes were transplanted into pseudopregnant female mice (Macrogen, Korea). TG + Mice were identified by PCR analysis of genomic DNA using primer sets corresponding to hGH poly(A) as shown in Table 1 below.

[0189] Sequence number name sequence 5'-3'1hGH poly(A) Forward primerGGTGTTAAAGTACGCTGGAGCT2hGH poly(A) Reverse primerCATGCCTGCTATTGTCTTCCC

[0190] For 2,3U-miR-122 TG, two independent new lines (F0; #73 and #74) were generated, and their F2 lines were used for 25 mg / kg DEN injection at P14 and euthanized at 10 months of age. For 3U-miR-122 TG, two independent new lines (F0; #16 and #146) were generated, and their F2 lines were used for 25 mg / kg DEN injection at P14 and fed a high-fat diet (20% CHO, 60% fat and 20% protein; Research Diets, #D12492) at 6 months of age and euthanized at 8 months of age.

[0191] For anti-2,3oxo TG, two independent new lines (F0; #56 and #46) were generated, and the F2 line was used for 25 mg / kg DEN injection at P14, 1 month after birth, and choline-deficient high-fat diet (20% choline-free CHO, 60% fat and 20% protein, Research Diets, #D05010403), and euthanized at 5.5 months of age.

[0192] For anti-3oxo TG, F2 were used for 25 mg / kg DEN injection at P14, 1 month after birth they were put on a choline-deficient high-fat diet (20% CHO without choline, 60% fat and 20% protein, Research Diets, #D05010403) and euthanized at 5 months of age.

[0193]

[0194] Example 4: Human samples

[0195] Primary human liver cancer tissue was obtained from the Department of Surgery, Hanyang University College of Medicine. Donors obtained their consent in accordance with an approved IRB protocol (# HYI17-182-2). After obtaining tissue during liver cancer surgery, it was rapidly frozen and stored at -80°C until use in experiments.

[0196]

[0197] Example 5: Statistical Data Analysis and Bioinformatics

[0198] Standard laboratory practice randomization procedures were applied to cell line groups and mice of the same age and sex. The experimenter had access to subject assignment during the experiment and during outcome assessment. All statistical tests, including the Wilcoxon signed-rank test (two-tailed), the log-rank test (two-tailed), the Kolmogorov-Smirnov test (KS test, two-tailed), the t-test (unpaired two-tailed), and Fisher's exact test (two-tailed), were performed in Scipy (www.scipy.org / ), Bioconductor (www.bioconductor.org / ), and Excel. Unless otherwise stated, values ​​are means ± SD, statistical significance was set at P < 0.05 (vs. control, equal variances), and data were replicated in biologically random, independent samples (n ≥ 3).

[0199] For bioinformatics analysis, we generally used Python scripts through Biopython (biopython.org / ), R scripts through Bioconductor (www.bioconductor.org / ), and the UCSC Genome Browser (genome.ucsc.edu / ).

[0200] Sequencing read mapping was performed using Bowtie (bowtiebio.sourceforge.net / ), Bowtie2 (http: / bowtie-bio.sourceforge.net / bowtie2 / ), STAR (github.com / alexdobin / STAR / ), or TopHat2 (ccb.jhu.edu / software / tophat).

[0201] RNA-seq analysis was performed with Cufflinks (cole-trapnell-lab.github.io / cufflinks / ) including Cuffdiff (cole-trapnell-lab.github.io / cufflinks / cuffdiff / ).

[0202] Hierarchical clustering analysis was performed with the Cluster3.0 program (bonsai.hgc.jp / ~mdehoon / software / cluster / ) and visualized as heatmaps using the Java TreeView program (jtreeview.sourceforge.net / ) or Morpheus (software.broadinstitute.org / morpheus / ).

[0203] BioEdit 7.2 (bioedit.software.informer.com / 7.2 / ) was used for displaying information content and aligning multiple sequences.

[0204]

[0205] Example 6: SNV Analysis of Tumor MiRNA Sequencing Data

[0206] Small RNA sequencing data sets from patients with various tumors (n=8,687) were retrieved from The Cancer Genome Atlas (TCGA; n=8,253; portal.gdc.cancer.gov / ) and Gene Expression Omnibus (GEO; n=434, www.ncbi.nlm.nih.gov / geo / ).

[0207] Only high-quality reads (quality scores >30 for 90% of called bases) were selected, preprocessed, and mapped to human miRNA sequences in miRbase (www.mirbase.org / ) using Bowtie (--norc -a -v 1 -S -f) in the same way as previously done for YM500v2 (ngs.ym.edu.tw / ym500v2 / ).

[0208] We further refined high-fidelity miRNAs (human, n=852, annotated in miRBase database) and identified a total of 26,305 single-nucleotide variants (SNVs), the frequency of which was calculated as the average number of reads per patient sample (avr) and represented in a heatmap, and analyzed for positional and compositional biases in bases (A, T, G, and C) and mutation types in G (G>T, G>C, and G>A).

[0209] For focused SNV analysis, FASTQ files were retrieved from the downloaded BAM files of patients with low-grade glioma (LGG; 512 tumor samples) and hepatocellular carcinoma (LIHC; 372 tumors; 47 tumor-normal paired samples) using BEDTools (bedtools.readthedocs.io), realigned using Bowtie (--norc -l 7 -n 2 -a -f), and the collapsed sequencing reads were classified (using the bowtie-build indexer with default parameters) and mapped to human miRNA sequences in miRbase.

[0210] The results were further parsed by filtering out reads with mismatch errors (quality scores less than 30). After selecting reads with a single mismatch, the SNV levels in miRNAs were estimated by calculating reads per million mapped reads (RPM), focusing specifically on G>T, and the sum was expressed as a log2 value across all patients. miRNA expression levels were also measured in RPM.

[0211] The frequency of SNVs (SNV reads per total mapped reads of that miRNA; % or proportion) in expressed miRNAs was calculated for all patients in LGG (top 85 miRNAs) and LIHC (top 85 miRNAs, top 70 miRNAs for paired samples) and used for hierarchical clustering performed in the Cluster 3.0 program (average linkage using default parameters).

[0212] For adjacent sequence motif analysis of the identified clusters, WebLogo 3 (weblogo.threeplusone.com) was used. For analysis of let-7, SNVs in the initiating region (positions 2–8) were condensed by summing their read counts across let-7 family members (let-7a, b, c, d, e, f, g, and i).

[0213]

[0214] Example 7: Survival analysis

[0215] The results presented here are based on data generated from the TCGA Research Network (www.cancer.gov / about-nci / organization / ccg / research / structural-genomics / tcga). Clinical data within the LGG and LIHC datasets were downloaded to match miRNA sequencing results to obtain information on patient survival and tumor grade. Therefore, some patients were excluded, including those with missing miRNA sequencing results, those without follow-up information, and those without clinical information.

[0216] The frequencies of specific SNVs (4G>T miR-124, 4G>T let-7, 3G>T miR-122, and 2G>T miR-122) in LGG or LIHC were analyzed by box plots according to patient survival time, and the frequencies were examined in heatmaps. In the box plots, lines represent the median, 1st quartile, and 3rd quartile; bars (whiskers) represent the minimum and maximum; and 'x' represents the mean.

[0217] For Kaplan-Meier analysis, the R program was used with the "survival" and "KMsurv" packages to compare patient survival (overall or within a given patient population) with the frequency of specific G>T SNVs at high versus low rates (in quartiles), unless otherwise indicated. To exclude the effect of miRNA expression, survival analyses comparing SNV frequencies were further refined to consider only patients with high or low expression of a given miRNA, and this was analyzed by plotting miRNA expression (in RPM) versus SNV frequency.

[0218] To determine statistical significance considering multiple testing correction, the Wilcoxon signed-rank test (two-tailed) or log-rank test (two-tailed) was performed using Bioconductor (www.bioconductor.org / ).

[0219]

[0220] Example 8: RNA synthesis and modification

[0221] Using the custom synthesis and modification services of TriLink Biotechnologies (USA), Integrated DNA Technologies (IDT) (USA), and Bioneer (Korea), RNA with various modifications was produced and quality monitored, reported, and confirmed through mass spectrometry. 8-oxoguanine (o) in ribonucleotides 8 G) was introduced at the indicated position of miRNA.

[0222] Sequence number name sequence 5'-3'3miR-124UAAGGCACGCGGUGAAUGCC44o 8 G-miR-124UAAo 8 GGCACGCGGUGAAUGCC5let-7aUGAGGUAGGUAGGUUGUAUAGUU64o 8 G-let-7aUGAo 8 GGUAGUAGGUUGUAUAGUU7miR-122UGGAGUGUGACAAUGGUGUUUG83o 8 G-miR-122UGo 8 GAGUGUGACAAUGGUGUUUG92,3o 8 G-miR-122Uo 8 Go 8 GAGUGUGACAAUGGUGUUUG

[0223] o 8 Mimetics of G-miRNA were synthesized by replacing G with U (underlined) at the indicated positions.

[0224] Sequence number name sequence 5'-3'104U-miR-124UAAUGCACGCGGUGAAUGCC114U-let-7aUGAUGUAGUAGGUUGUAUAGUU123U-miR-122UGUAGUGUGACAAUGGUGUUUG132,3U-miR-122UUUAGUGUGACAAUGGUGUUUG

[0225] As a control miRNA, a non-targeting miRNA (NT) derived from cel-miR-67 (a C. elegans-specific miRNA provided as a negative control by Dharmacon) was synthesized.

[0226] Sequence number name sequence 5'-3'14cel-miR-67UCACAACCUCCUAGAAAGAGUA15cel-miR-67(duplex 1)UCACAØCCUCCUAGAAGAGUAdTdT16cel-miR-67(duplex 2)uaCAACCUCCUAGAAAGAGUAdTdT

[0227] The non-targeting miRNA derived from the above cel-miR-67 was used as a duplex: 'dT' is thymidine deoxynucleotide, and was additionally modified to include 'Ø' (dSpacer) at position 6 or 2'-O-methyl in the part indicated by lowercase letters. In addition, o 8 Synthesize G control RNA, (STpharm) o 8 Control RNA containing G modification was used; 5'- -3'.

[0228] Sequence number name sequence 5'-3'17STpharmACUAAAGAAU(o 8 G)CUUGACAGUUAUCUCGUAUGCCGUCUUC

[0229] Duplexes of NT, miR-124, let-7a and miR122 (including their oxidized derivatives and analogs) were generated by annealing (90 °C for 2 min, 30 °C for 1 h and 4 °C for 5 min) to the corresponding synthetic passenger strands (ps).

[0230] Sequence number name sequence 5'-3'18NT-ps(1)uaCUCUUUCUAGGAGGUUGUGAdTdT19NT-ps(2)UACUCØUUCUAGGAGGUUGUGAdTdT20miR124-psCA UUCACCGGCGUGCCUUAdTdT21let-7-psCUAUACAACCUACUACCUCAdTdT22miR-122-psAACACCAUUGUCACACUCCAdTdT

[0231] To confirm delivery of the duplex miRNA, fluorescein isothiocyanate (FITC) was attached to the 5' end of the passenger strand. Note that, as previously reported, the dSpacer (basic deoxynucleotide) at position 6 or the 2-O-methyl groups at positions 1 and 2 were intentionally modified in the passenger strand to prevent seed-mediated off-target effects.

[0232] Competitive o 8 For G-miRNA inhibitors, let-7 anti-4oxo RNA (4x), miR-122 anti-3oxo (4x), miR-122 anti-2,3oxo (4x), and control (cont) (anti-NT; 4x) were synthesized. ('dA' is adenosine oxynucleotide)

[0233] Sequence number name sequence 5'-3'23let-7 anti-4oxo RNA(4x)dTCUACAUCUCUACAUCACUACAUCAUCUACAUCAdT24miR-122 anti-3oxo(4x)dTCACACUACACACUACAACACACUACACACUACAdT25miR-122 anti -2,3oxo(4x)dTCACACUAAACACACUAAACACACUAAACACUAAdA26miR-122 anti -2,3oxo(4x)(cont)dTGGUUGUGAGGUUGUGUGGUUGUGACGGUUGUGAdT

[0234] Note that 'dT' or 'dA' at both the 5' and 3' ends of the inhibitor were intentionally added to contribute to increased RNA stability. The underlined portions in the above sequence indicate the corresponding oxo sites.

[0235] For reference, the RNA sequence of Extended Data Figure 7c was synthesized with a 5' phosphate to generate a miR-122 duplex mutant.

[0236]

[0237] Example 9: RNA extraction

[0238] To purify RNA by size, the miRNeasy Mini Kit (Qiagen, #217084) was used for isolation of small RNA fractions (<~200 nucleotides) and large RNA fractions (>~200 nucleotides), supplemented with 2.5 mM deferoxamine mesylate (DFOM) (Sigma-Aldrich, #D9533) in Qiazol lysis reagent (Qiagen, #79306) to prevent in vitro oxidation.

[0239] For liver tissue samples (30-50 mg), a Minilys Personal Homogenizer (Bertin, #EQ06404) was used in the presence of 1 ml of quiazole lysis reagent (containing 2.5 mM DFOM). For total RNA purification, homogenized samples in quiazole lysis reagent were used according to the manufacturer's protocol, with the addition of only 20% chloroform (Sigma-Aldrich, #C2432).

[0240] To increase the yield of small RNA, 200 μl of water was added to the organic layer, followed by 700 μl of RLT (Qiagen, #79216) to obtain an additional aqueous phase of qazole lysis reagent for subsequent RNeasy column purification. Both a spectrophotometer (Denovix, #DS-11) and Qubit fluorometric quantification (Invitrogen, #Q32866) were used for accurate quantification of RNA.

[0241] To purify miRNA, gel extraction of ~20 nucleotides from total RNA was performed. Total RNA denatured in gel loading buffer II (90°C for 2 min, Ambion, #AM8546G) was separated on a 15% Novex TBE Urea gel (Invitrogen, #EC6885) labeled with a 20-nucleotide miRNA marker (synthetic miR124), and its size was cross-checked with a 14-30 ssRNA Ladder Marker (Takara, #3416).

[0242] After staining with SYBR Gold (Invitrogen, #S11494), ~20 nucleotides in size were excised from the gel and RNA was further purified using the ZR small-RNA PAGE Recovery Kit (Zymo Research, #R1070) according to the manufacturer's protocol.

[0243]

[0244] Example 10: Fenton reaction

[0245] In vitro oxidation of RNA was induced via the Fenton reaction using a modified conventional method. Briefly, synthetic miRNA (1 μg) or fragmented RNA (2 μg) was incubated with 0.5 mM H2O2 (Sigma-Aldrich, #216763) at 37°C for 1 h (Eppendorf ThermoMixer C) in the presence of ferric citrate (III) (Sigma-Aldrich, #F6129), 5 mM ascorbic acid (Sigma-Aldrich, #A92902), and 10 mM sodium phosphate [pH 7.4] (Biosesang, #SR2154-100-74) in a total volume of 30 μl.

[0246] The reaction was stopped by adding 10 μl of 40 mM DFOM (Sigma-Aldrich, #D9533) and incubated on ice for 5 min. Oxidized RNA was ethanol precipitated (1 / 10 volume of 3 M sodium acetate (pH 5.5) [Invitrogen, #AM9740] and 1 μl of glycogen [5 mg / ml; Thermo Scientific, #R0551]; total volume 50 μl) and then washed with 75% ethanol. 500 ng of oxidized RNA was used to construct a small RNASeq library.

[0247] For in vitro oxidation of miR-122 and let-7, miR-122 (SEQ ID NO: 7) or let-7a (SEQ ID NO: 5) RNA (4 μg) was used. Purification was performed using a MicroSpin G-25 column (Cytiva, #27532501) and separately subjected to the Fenton reaction. For in vitro oxidation of miR-124, 1 μg of synthetic miR-124 (SEQ ID NO: 27) (Bioneer) was used in the Fenton reaction together with miR-30a (1 μg), a brain-enriched miRNA, in the presence of another void miRNA (7 μg).

[0248] Sequence number name sequence 5'-3'27synthetic miR-124UAAGGCACGCGGUGAAUGCCAA

[0249] For this Fenton reaction with 9 μg total RNA, 9 μl of recombinant RNase inhibitor (40 U / ul, Takara, #2313A) was added, and then 9-fold more volume was used by varying the concentrations of H2O2 (0.5, 2, and 10 mM). As a control, the same reaction was performed without H2O2 and used to construct a small RNA-Seq library.

[0250] Note that the incubation time was short (30 min) when high concentrations of H2O2 (10 mM) were used. To produce an in vitro oxidized control RNA, an artificial sequence containing only G in the middle (SEQ ID NO: 28) was synthesized and 3 μg of it was applied to the Fenton reaction (using 2 mM H2O2).

[0251] Sequence number name sequence 5'-3'28 artificial sequenceUUUCCCCUGUUCCCUUCCUU

[0252] After ethanol precipitation, in vitro oxidized control RNA was used in transfection experiments.

[0253]

[0254] Example 11: o 8 G dot blot analysis

[0255] For dot blot, in vitro oxidized RNA (500 ng) was spotted onto Zeta-Probe blotting membrane (Bio-Rad, #1620153) via Fenton reaction and analyzed using a Spectrolinker XL-1000 (Spectroline, #XLE-1000) at optimal crosslinking (120 mJ / cm). 2 ) was crosslinked by UV irradiation as an option.

[0256] The above membranes were blocked using Blocker FL Fluorescent Blocking Buffer (ThermoFisher, #37565) for 1 h at room temperature (RT), and anti-o in blocking buffer for 1 h at RT. 8 Incubated with G antibody (15A3, 1:5000; QED Bioscience, #12501).

[0257] Afterwards, the membranes were washed three times for 10 min each in TBST, and fluorophore-conjugated secondary antibodies (Alexa Fluor 680 Goat anti-mouse IgG, 1:15000 in 1 x TBST, Invitrogen, #A32729) were used in TBST for 1 h at RT. After washing the membranes three times for 10 min each in TBST to remove unbound secondary antibodies, the fluorescence signals were directly quantified with an iBright FL100 imaging system (Invitrogen, #A32752) and calculated as the relative intensity to the total amount of spotted RNA stained with SYBR Gold (1:10000, Invitrogen).

[0258]

[0259] Example 12: In vitro production of random RNA fragments

[0260] To generate artificial RNA transcripts containing a stretch of all possible combinations of flanking sequences (three nucleotides) on the G side, the sequences were designed by randomly assigning all possible three nucleotides without G. In the process of determining the order of these flanking sequences, the sequences were randomly assigned to five different artificial transcripts to avoid generating more than three consecutive nucleotides (of the same base with or without a G gap).

[0261] For reference, we designed these artificial transcripts to investigate the site bias of oxidation while excluding the reported effect of GG on the redox potential of nucleic acids and to avoid continuous stretches of G. For in vitro transcription, we added the T7 promoter sequence followed by a random spacer sequence (50 bp, G-free) to the 5' end of these artificial transcripts. Then, we added a 3' spacer consisting of a random spacer (30 bp, G-free) and the SP6 promoter sequence to the 3' end, followed by the addition of poly(A).

[0262] These five transcript sequences were synthesized as DNA and cloned into the pBHA vector (Bioneer). The DNA templates were amplified by PCR (using primers of SEQ ID NO: 29 and 30) using Q5 High-Fidelity 2X Master Mix (NEB, #M0492), and further purified using the QIAquick Gel extraction kit (QIAGEN, #28706).

[0263] Sequence number name sequence 5'-3'29Forward primerGTCTCATGAGCGGATACATAT30Reverse primerCTGTGGATAACCGTATTACCG

[0264] DNA templates (244 ng; 5 different template combinations) were subjected to in vitro transcription using the HiScribe T7 High Yield RNA Synthesis Kit (NEB, #E2040S) according to the manufacturer's protocol. RNA products were then further purified using the RNeasy Mini Kit (QIAGEN, #74106) with DNase I treatment.

[0265] To fragment the artificial transcripts into smaller fragments, 4 μg of the transcripts were applied to the NEBNext Magnesium RNA Fragmentation Module (NEB, #E6150) according to the manufacturer's protocol. Further purification was performed using the RNeasy Mini Kit (QIAGEN, #74106), followed by G-25 filtration using a MicroSpin G-25 column (Cytiva, #27532501).

[0266] Of these, 1 μg of fragments were further oxidized in vitro via the Fenton reaction, and 100 ng of these fragments were subjected to the random priming step of the NEBNext Ultra II RNA Library Prep Kit for Illumina (NEB, #E7760S) to construct RNA-Seq libraries. Simultaneously, unoxidized RNA fragments (100 ng, without Fenton reaction) were also used to construct a control RNA-Seq library.

[0267]

[0268] Example 13: Sequencing analysis of in vitro oxidized RNA

[0269] Sequencing libraries prepared from in vitro oxidized RNA were sequenced with 150 paired-end reads (PE150) on a MiniSeq System (Illumina). For in vitro oxidized miRNA analysis, only 150 single-end reads (SE150; miR-124) or SE50 (miR-122 and let-7a) were used for further analysis.

[0270] Demultiplexed sequencing reads (FASTQ files) were aligned using Bowtie, allowing two mismatches (Bowtie --norc -l 7 -n 2 -a -f), and the collapsed sequencing reads were classified (using the bowtie-build indexer with default parameters) and mapped to mature miRNA sequences (miR-124, miR122, and let-7a).

[0271] Only mismatches with a high confidence base calling score (Q=37) were considered, which was used to calculate the mismatch rate (%) for all SNVs.

[0272] To analyze in vitro oxidized random RNA fragments, FASTQ files were processed to trim adapter sequences using Cutadapt (version 1.15, cutadapt.readthedocs.io / en / stable / ) and aligned to designed artificial transcript sequences using Bowtie2 (bowtie2 -p 12 --rf; version 2.3.4, bowtiebio.sourceforge.net / bowtie2).

[0273] Sequence number name sequence 5'-3'31cutadaptcutadapt -a AGATCGGAAGAGCACACGTCTGAACTCCAGTCAC -A AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGTAGATCTCGGTGGTCGCCGTATCATT -m 1

[0274] Then, appropriately aligned reads were selected using SAMtools (SAMtools view -F 1804 -f 2; version 1.9, samtools.sourceforge.net / ) and duplicates were removed using Picard's Markduplicates program (Picard MarkDuplicates ASSUME_SORTED=TRUE REMOVE_DUPLICATES=true; version 2.20.6, broadinstitute.github.io / picard / ). After that, the mapping results of only single-end reads (SE150) were parsed using Python's Pysam module (pysam.readthedocs.io / en / latest / ), and analyzed with the pileup function to determine the frequency and position of the mutations, and if the quality score was ≥30, the positional G>T ratio (%) was analyzed for the 5' or 3' end of the RNA fragments.

[0275]

[0276] Example 14: o by ELISA 8 Quantification of G

[0277] OxiSelect Oxidative RNA Damage ELISA kit (Cell Biolabs, #STA-325) was used according to the manufacturer's protocol. 8 Colorimetric detection and quantification of G were performed. Sample preparation included treating purified small RNA (1.5 ug) with nuclease P1 (Wako, #145-08221) at 37°C for 1 hour, followed by alkaline phosphatase (Roche, #10713023001), and anti-o 8 Pre-coated o on the plate for binding of G antibody 8 A HRP-conjugated secondary antibody was added for competition with the G / BSA conjugate and for detection.

[0278] The absorbance of the sample was measured using a spectrophotometer (450 nm, GloMax-Multi Detection System; Promega). 8 The concentration of G is predetermined o 8 G is estimated by comparison with the standard curve, and further the relative amount, o 8 G ratio(o 8 It was expressed as G RNA / total RNA.

[0279]

[0280] Example 15: ROS measurement

[0281] To quantitatively measure cellular ROS levels, a dihydroethidium (DHE)-based flow cytometry analysis, a ROS fluorescent dye, was applied using a Muse Cell Analyzer (Luminex, #MCH100106) together with the Muse Oxidative Stress Kit (Luminex, #MCH100111). Briefly, HS683 and Huh7 cells (10 6 After harvesting, fixing, and staining, cells (6-well plate) were measured in replicates (n≥3) according to the manufacturer's protocol. For ROS levels, the same number of cells (n≥3,000) were evaluated.

[0282] To visualize cellular ROS levels, harvested SH-SY-5Y, HS683, and Huh7 cells were washed with serum-free DMEM and incubated in the dark at 37°C for 15 min with 5 μM CM-H2DCFDA (Invitrogen, #C6827). After incubation, the cells were washed again with serum-free DMEM, and ROS images were acquired using an inverted fluorescence microscope (Leica DMi8).

[0283] CM-H2DCFDA was used to measure the amount of ROS in rat liver tissue lysates. Briefly, homogenized lysates were prepared using 1 x RIPA lysis buffer (DyneBio, #CBR3040). The lysates were supplemented with protease inhibitors (cOmplete, Mini, EDTA-free; Roche, #04693159001) and 2.5 mM DFOM, and proteins were quantified using the BCA protein assay kit (Pierce, #23227).

[0284] Afterwards, the lysate containing 100 μg of protein was incubated in the dark at 37°C for 30 min with 2 μM CM-H2DCFDA. The fluorescence signal was detected using a Qubit 2.0 Fluorometer (Invitrogen).

[0285]

[0286] Example 16: Construction of a Luciferase Reporter

[0287] To measure miRNA-mediated gene silencing activity, a luciferase reporter targeting a 4oxo moiety was constructed using the psiCheck-2 (Promega, #C8021) vector. This target region is composed of tandem repeats (n=5) and is inserted into the 3'UTR of Renilla luciferase (hRLuc; psiCheck-2). For miR-124, synthetic duplex oligos containing the miR-124 seed or the 4oxo moiety (Bionics) were cloned into the psiCheck-2 plasmid via the XhoI and NotI sites, as indicated.

[0288] Sequence number name sequence 5'-3'32mir-124 seed sites (5x) forwardTCGAGGTGCCTTAGTGCCTTAGTGCCTTAGTGCCTTAGTGCCTTACG33mir-124 seed sites (5x) reverseGGCCGCTAAGGCACTAAGGCACTAAGGCACTAAGGCACTAAGGCACC344oxo site (5x) forwardTCGAGGTGCATTAGTGCATTAGTGCATTAGTGCATTAGTGCATTACG354oxo site(5x) reverseGGCCGCTAATGCACTAATGCACTAATGCACTAATGCACTAATGCACC364oxo site(13x) forwardTCGAGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTAGC374oxo site(13x) reverseGGCCGCTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCAC

[0289] For reference, this luciferase reporter vector with miR-124 4oxo sites (5x and 13x) acts as a sponge suppressor, 4o 8 It was also used as a miR-124 anti-4oxo (5x and 13x) due to its competitive activity in inhibiting G-miR-124.

[0290] Because the luciferase gene of psiCheck-2 harbors a let-7 target site, a novel dual luciferase reporter plasmid (psi-RL2-luc2) with increased sensitivity to let-7-mediated repression was constructed. The modified psiCheck-2 vector (psi-RL2) was further modified to remove the inserted target and XhoI sites by self-ligation of NotI, in which the humanized Renilla luciferase gene (hRL) was replaced (via NheI and XhoI sites) with the Renilla luciferase gene (RL) from the pRL-TK plasmid (Promega, #E2241) from a previous study.

[0291] Subsequently, the humanized firefly luciferase gene (hluc+) was converted to an engineered firefly luciferase gene (luc2) with a multiple cloning site in the pmirGLO plasmid (Promega, E1330) through partial digestion with ApaI and XbaI. Using psi-RL2-luc2, a synthetic duplex oligo (Bionics) containing the let-7 seed sequence and the 4oxo site was cloned into the 3'UTR of luc2 via the SacI and XhoI sites.

[0292] Sequence number name sequence 5'-3'38let-7 seed sites (5x) forwardCACTACCTCAACTACCTCTACTACCTCGACTACCTCCACTACCTCC39let-7 seed sites (5x) reverseTCGAGGAGGTAGTGGAGGTAGTCGAGGTAGTAGAGGTAGTTGAGGTAGTGAGCT40let-7 4oxo site (5x) forwardCACTACATCAACTACATCTACTACATCGACTACATCCACTACATCC41let-7 4oxo site(5x) reverseTCGAGGATGTAGTGGATGTAGTCGATGTAGTAGATGTAGTTGATGTAGTGAGCT

[0293] It is noteworthy that the luciferase reporter vector with the let-7 4oxo moiety (5x) acts as a sponge suppressor, 4o 8 It was also used as a let-7 anti-4oxo (5x) due to its competitive activity in inhibiting G-let-7.

[0294] Since the hRL gene contains the miR-122 3oxo site, the hRL gene of psiCheck-2 was also replaced with the RL gene of the pRL-TK plasmid, and the fragment was amplified to include the NheI and XhoI sites.

[0295] Sequence number name sequence 5'-3'42hRL forwardCTCACTATAGGCTAGCCACC43hRL reverseATCGCACTCGAGCCGCTCTAGAATTATTGTTC

[0296] Using the psi-RL4 plasmid, a synthetic duplex oligo (Bionics) containing the miR-122 origin sequence, 2oxo, 3oxo and 2,3oxo sites was cloned into the 3'UTR of RL via XhoI and NotI sites.

[0297] Sequence number name sequence 5'-3'44miR-122 seed sites (5x) forwardTCGAGTCACTCCTCACTCCTCACTCCTCACTCCTCACTCCGC45miR-122 seed sites (5x) reverseGGCCGCGGAGTGAGGAGTGAGGAGTGAGGAGTGAGGAGTGAC462oxo site (5x) forwardTCGAGCACTCAATCACTCAATCACTCAATCACTCAATCACTCAAGC472oxo site(5x) reverseGGCCGCTTGAGTGATTGAGTGATTGAGTGATTGAGTGATTGAGTGC483oxo site(5x) forwardTCGAGCACACTACACACACTACACACACTACACACACTACACACACTACAGC493oxo site(5x) reverseGGCCGCTGTAGTGTGTGTAGTGTGTGTAGTGTGTGTAGTGTGTGTAGTGTGC502,3oxo site(5x) forwardTCGAGCACACTAAACACACTAAACACACTAAACACACTAAACACACTAAAGC512,3oxo site(5x) reverseGGCCGCTTTAGTGTGTTTAGTGTGTTTAGTGTGTTTAGTGTGTTTAGTGTGC

[0298] For validation of the miR-124 4oxo target site, the 3'UTR region of MEOX2 (461 bp) and the target site of EGFR (EGFR #2; 401 bp) were amplified from cDNA of HS683 (reverse transcribed using SuperScript III First-Strand Synthesis System; Invitrogen, #34580) using Ex Taq (Takara, #RR003) with primer pairs containing restriction enzyme sites (XhoI and NotI) and cloned into the psiCheck-2 vector.

[0299] For another EGFR target site (EGFR #1), synthetic duplex oligos (Bionics) spanning the target sites were also cloned into the psiCheck-2 vector. The same cloning strategy using synthetic oligos was applied to construct luciferase reporters for let-7 and miR-122 oxo sites, utilizing tandem repeats for sensitive detection: let-7 4oxo sites of CLTA, TIA1, PUDP (3 repeats), and DDAH1 (3 repeats) cloned into the psiCheck-2 vector (XhoI and NotI sites); miR-122 3oxo sites of AKT1S1 and IL6st (3 repeats) cloned into psi-RL4 (XhoI and NotI sites).

[0300] The 3'UTR regions of ASPH (388 bp) and ZNF146 (389 bp) were amplified from genomic DNA (extracted using AccuPrep Genomic DNA Extraction Kit [Bioneer, #K-3032]) using Q5 High-fidelity 2x Master Mix (NEB, #M0492) with primer pairs containing restriction enzyme sites (XhoI and NotI) at the 2,3-oxo site of miR-122 and cloned into the psi-RL4 vector.

[0301]

[0302] Example 17: Construction of a dual fluorescent reporter

[0303] The dual fluorescent protein (dFP) reporter vector (pDual) was constructed based on psiCheck-2 (Promega), in which the luciferase gene was replaced with a fluorescent protein gene. The eGFP gene from pUlta (Addgene, #24129; Malcolm Moore) was amplified and replaced with hluc+ via the ApaI and XbaI sites, and the TurboRFP gene from pTRIPZ (Thermo Scientific) was amplified and replaced with hRluc via the NheI and XhoI sites.

[0304] Sequence number name sequence 5'-3'52eGFP forwardGAAGGGCCCTATGGTGAGCAAGGGCG53eGFP reverseGACTCTAGAATTATTAGTACAGCTCGTCCATGCC54TurboRFP forwardTAGGCTAGCCACCCATGAGCGAGCTGATCAAGGA55TurboRFP reverseGGGCTCGAGCGATCGCCTAGAATTATCTGTGCCCCCAGTTTGCTA

[0305] The miR-122 3oxo (5x), miR-122 2,3oxo (5x), miR-124 4oxo (13x; forward: SEQ ID NO: 51, reverse: SEQ ID NO: 52) and let-7 4oxo (13x; forward: SEQ ID NO: 53, reverse: SEQ ID NO: 54) sites were inserted into pDual following the same subcloning method used in psiCheck-2 (XhoI and NotI sites).

[0306] Sequence number name sequence 5'-3'56miR-124 4oxo(13x) forwardTCGAGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTACGTGCATTAGC57miR-124 4oxo(13x) reverseGGCCGCTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCACGTAATGCAC58let-7 4oxo(13x) forwardTCGAGCTACATCATCTACATCACCTACATCAGCTACATCAACTACATCATCTACATCACCTACATCAGCTACATCAACTACATCATCTACATCACCTACATCAGCTACATCAACTACATCAGC59let-7 4oxo(13x) reverseGGCCGCTGATGTAGTTGATGTAGCTGATGTAGGTGATGTAGATGATGTAGTTGATGTAGCTGATGTAGGTGATGTAGATGATGTAGTTGATGTAGCTGATGTAGGTGATGTAGATGATGTAGC

[0307] For reference, pDual with miR-124 4oxo site (13x) or let-7 4oxo site (13x) acts as a sponge inhibitor 8 G-miR-124 or 4o 8 Due to its competitive activity in inhibiting G-let-7, it was also used as miR-124 or let7 anti-4oxo(13x).

[0308]

[0309] Example 18: Luciferase reporter assay

[0310] Various luciferase reporter plasmids (psiCheck-2 or modified plasmids containing the site of interest) were transfected into SH-SY-5Y, HS683, Huh7, and miR-122 KO Huh7.

[0311] For cotransfection, 30 nM RNA (synthetic duplex miRNA, miRIDIAN microRNA hairpin inhibitor [miR-124; Dharmacon, #IH300592-06], or anti-oxo RNA) or 2 µg / 96-well anti-oxo vector plasmid was cotransfected with luciferase reporter plasmid (2 µg / 96-well).

[0312] Luciferase activity was typically measured 24 h after transfection. For H2O2 treatment (100 μM), treatment was initiated 20 h after transfection and incubated for 4 h. Serum starvation was initiated at the time of transfection. Relative activities (Renilla luciferase vs. firefly luciferase) were measured using the Dual Luciferase Reporter Assay System (Promega, #E1910) with the GloMax-Multi Detection System (Promega) in replicates (n≥6) according to the manufacturer's protocol.

[0313] To evaluate the efficiency of miRNA-mediated inhibition, various concentrations of synthetic duplex miRNAs (from 0 to 75 nM) were transfected into HeLa or SK-HEP-1 cells. Then, nonlinear least squares fitting was performed using Scipy (scipy.optimize.curve_fit()) to fit a sigmoid function to obtain IC 50 was calculated. If the least squares does not fit the function, the approximate IC from the regression line 50 was calculated.

[0314]

[0315] Example 19: o 8G-miSeq

[0316] o 8 G-miSeq was performed in the conventional manner. Antibody-conjugated magnetic beads were prepared by incubating the following reaction for 1 hour at room temperature (RT) with rotation: 12.5 μg anti-o in PBS (total volume 500 μl) 8 125 μl of G antibody (15A3; QED Bioscience, #12501) and Dynabeads Protein G (Invitrogen, #10004D) were supplemented with 1.2 mg of deoxyguanosine (dG; Sigma, #D7145) to mask nonspecific interactions with unoxidized G.

[0317] After two PBS washes, o performed in HS683 8 Except for G-miSeq, usually one o 8 Only half of the beads prepared for the G-miSeq reaction were used. 10 μg of small RNA sample purified using the miRNeasy Mini Kit (Qiagen) was applied to beads prepared in 200 μl of PXL (PBS, 0.1% SDS [Invitrogen, #AM9820], 0.5% deoxycholate [Sigma-Aldrich, #D6750], 0.5% NP-40 [Sigma-Aldrich, #74385]) containing 2.5 mM DFOM (Sigma) and 40 U of recombinant RNase inhibitor (Takara, #2313A).

[0318] When using a spike-in control, the synthetic o that is not expressed in the sample 8 G-miRNA (2o for HS683 8 3o for G-miR-122, Huh7 80.1 pg of GmiR-1) was added. After 2 h of incubation with rotation at 4°C, a series of washing steps were performed; TE (15 mM Tris-HCl pH 7.5 [Invitrogen, #15567027], 5 mM EDTA [Invitrogen, #AM9260G]); high detergent and TE (1% Triton X-100 [Sigma-Aldrich, #T1020], 1% deoxycholate, 0.1% SDS, and 2.5 mM EGTA [Biosesang, #ER2024-001-80]) and high salt (1 M NaCl [Sigma-Aldrich, #AM9760G]); high detergent and extreme salt and TE (2 M NaCl); high detergent and salt and TE (120 mM NaCl); Wash buffer (50 mM Tris-HCl pH 7.5, 1 mM MgCl [Sigma-Aldrich, #AM9530G], 150 mM NaCl, 0.05% NP-40).

[0319] Small RNA on beads (o 8 G-IP) was extracted using 700 µl of quiazole lysis reagent supplemented with 2.5 mM DFOM, and purified using RNA Clean & Concentrator-5 kit (Zymo, #R1016) or miRNeasy Mini Kit (for clinical samples) after adding 20% ​​chloroform. Then, purified small RNA (o) was extracted according to the small RNA-Seq library construction method. 8 A sequencing library was constructed using G-IP.

[0320]

[0321] Example 20: Construction of a Small RNA-Seq Library

[0322] Synthetic RNA or purified small RNA from in vitro oxidation (500 ng) was extracted directly from the lysate (input; 450-500 ng) or using the miRNeasy Mini Kit (Qiagen).8 G antibody (o 8 cDNA sequencing libraries were constructed using immunoprecipitation using G-IP. Specifically, 0.385 μM of 3' adapter (rApp: adenylate, PMe: methylphosphonate, ddC: dideoxy-cytosine; Trilink) (SEQ ID NO: 52) was prepared with RNA samples on ice after incubation at 70°C for 2 min.

[0323] Sequence number name sequence 5'-3'603' adapter: 5'-adenylation(rApp)T(PMe)GGAATTCTCGGGTGCCAAGG(ddC)

[0324] Afterwards, 1 ul of T4 RNA ligase 2 cleavage KQ (NEB, #M0373) was added to a total of 13 ul of T4 RNA ligase buffer supplemented with 20% PEG 8000 and 20 U recombinant RNase inhibitor and incubated at 28°C for 1 hour and then inactivated at 65°C for 20 minutes.

[0325] Afterwards, 0.385 uM 5' adapter (Trilink) (SEQ ID NO: 61) was added to 2 ul of T4 RNA ligase (Thermo, #EL0021), 40 U of recombinant RNase inhibitor and 1 ul of BSA (1 mg / ml); incubated at 28°C for 1 hour and inactivated at 65°C for 20 minutes.

[0326] Sequence number name sequence 5'-3'615' adaptorGUUCAGAGUUCUACAGUCCGAUCAUCNNNN(2'-methoxy-C)

[0327] Ligated RNA samples were denatured at 70°C for 2 minutes with 1 μl of 10 uM RT primer (SEQ ID NO: 62), placed on ice for 1 minute, and reverse transcribed as follows: 44 μl of 0.86 x first strand buffer containing 1 μl of SuperScript III reverse transcriptase (Invitrogen, #18080051), 398 uM dNTPs, 7.95 mM DTT, and 4 U recombinant RNase inhibitor; incubated at 50°C for 1 hour and 70°C for 15 minutes.

[0328] Sequence number name sequence 5'-3'62RT primerGCCTTGGCACCCGAGAATTCCA

[0329] The prepared cDNA was further processed to include Truseq index adapter sequences (Illumina). Typically, cDNA was pre-amplified by PCR using Q5 High-fidelity 2x Master Mix (NEB, #M0492) with 158 nM universal forward primer (SEQ ID NO: 63) and 158 nM RT primer: 98 °C for 30 s; 5 cycles of 98 °C for 10 s, 60 °C for 30 s, and 72 °C for 15 s; and 72 °C for 10 min; and purified using the QIAquick PCR Purification Kit (Qiagen, #28106).

[0330] Sequence number name sequence 5'-3'63universal forward primerAATGATACGGGCGACCACCGAGATCTACACGTTCAGAGTTCTACAGTCCGA

[0331] Subsequently, qPCR (Rotor-Gene Q; Qiagen, #9001550) was performed using Q5 High-fidelity 2x master mix with SYBR Green I (1:10000, Invitrogen) to determine the cycle number required to reach the linear amplification range using a portion of the sample. PCR was performed on the remaining samples using Q5 High-fidelity 2x master mix with a primer set to generate a multiplexing barcode (333 nM universal forward primer, 333 nM barcode primer: SEQ ID NO: 64) with the defined optimal cycle number.

[0332] Sequence number name sequence 5'-3'64 barcode primerCAAGCAGAAGACGGCATACGAGAT-6mer barcode-GTGACTGGAGTTCCTTGGCACCCGAGAATTCCA

[0333] The expected sizes of PCR products containing ~20 nt inserts were quantified using Pippin Prep (Sage Science, #PIP0001) and 3% agarose gel cassettes, and cross-checked using the Qubit dsDNA HS Assay Kit (Invitrogen, #Q32854) and a fragment analyzer (Advanced Analytical, #M5310AA). Finally, all prepared multiplexed libraries were accurately pooled and sequenced with 50 or 75 single-end (SE75) on a MiniSeq System (Illumina). If necessary, libraries were further analyzed on a HiSeq X Ten System (Illumina) and sequenced with 150 pair-end reads (PE150). All sequence data (FASTQ) are deposited in the Sequence Read Archive (SRA) database (SRP353704, SRP353708, SRP353711, SRP353718, SRP353713, SRP353717).

[0334]

[0335] Example 21: o 8 Analysis of G-miSeq data

[0336] Demultiplexed sequencing reads (FASTQ files) were aligned using Bowtie, allowing two mismatches (Bowtie --norc -l 7 -n 2 -a -f). The collapsed sequencing reads were sorted (using the bowtie-build indexer with default parameters) and mapped to mature miRNA sequences of the same species in miRBase. The results were further parsed by filtering out reads with mismatch errors, with quality scores (Q) lower than a threshold for each sample (Q<30: HepG2; Q<37: SHSY-5Y; Q<38, HS683, Huh7, rat liver, human liver).

[0337] The number of reads per miRNA is normalized to the total number of reads (reads per million total reads, RPM) and the log2 values ​​(log2[input] and log2[o 8 G-IP]) was used. In particular, o 8 G concentration is o 8 G-IP sequencing (o 8 The RPM of G-IP was estimated by specifying the RPM of small RNA-seq (input) and the log2 ratio (log2[o 8 It was calculated as [G-IP / input].

[0338] Differential o according to redox status (antioxidant treatment) or cancer progression (DEN-treated LIHC model) 8 To analyze G enrichment, o between different conditions (e.g. antioxidant vs. mock treatment) 8 Relative o from G concentration ratio 8G enrichment values ​​were calculated. The percentage of SNVs occurring at each position in the miRNA sequence with the highest Phred score (base calling) was calculated relative to the total match frequency; mismatch (%). Among the reads mapped to miRNA sequences with two mismatches, reads aligned to miR-122 were analyzed focusing on the two mismatches. 8 The number of reads containing two mismatches in G-IP (log2RPM) was investigated along with the change in frequency after antioxidant treatment (relative o 8 G ratio, o 8 Log2 ratio of G-IP, antioxidant vs. mock treatment).

[0339]

[0340] Example 22: Apoptosis assay

[0341] For quantitative measurement of apoptosis, annexin V apoptosis detection kit (Invitrogen, BMS500FI) was used according to the manufacturer's protocol. Briefly, after washing with PBS, 0.4-1 X 10 5 Cells were resuspended in 195 μl of binding buffer containing 5 μl of annexin V and incubated at RT for 10 min. After washing with binding buffer, cells were resuspended in 200 μl of binding buffer. 10 μl of 20 μg / mL propidium iodide (PI) was added, and cells were analyzed using an Attune NxT flow cytometer (Thermo Fisher Scientific).

[0342] For hepatoma cell line experiments, apoptosis analysis was performed using the Muse annexin V & dead cell kit (Luminex, #MCH100105) and the Muse cell analyzer (Luminex, #0500-3115) according to the manufacturer's protocol. Apoptosis measurements were generally performed 72 h after transfection unless otherwise specified. The MTT assay was performed to quantify live cells (miR-122 KO Huh7 cells treated with low concentrations [100 nM] of sorafenib).

[0343] Specifically, cells were cultured in medium supplemented with 500 μg / ml MTT reagent (Duchefa, #M1415.0001) at 37°C for 3 h. After incubation, cells were treated with DMSO (Sigma-Aldrich, #D8418) for 5 min at room temperature. Absorbance (590 nm) was measured using a GloMax-Multi Detection System (Promega).

[0344]

[0345] Example 23: Ki67 proliferation assay

[0346] To quantify the degree of proliferation, the percentage of proliferating cells was measured based on Ki67 expression. For this analysis, the Muse Ki67 Proliferation Kit (Luminex, #MCH100114) was used according to the manufacturer's protocol. Briefly, cell suspension (1 X 10 5Cells) were prepared, fixed, and incubated with permeabilized fluorescence-conjugated Ki67 antibody (Muse Hu Ki67PE antibody; Luminex, #4700-1667) 48 h (HepG2, Huh7, and NCTC clone 1469) or 72 h (HS683, U87MG, and SH-SY5Y) after transfection, and analyzed on a Muse cell analyzer (Millipore, #0500-3115) using Muse Hu IgG1-PE antibody (Luminex, #4700-1669) as a negative control experiment.

[0347]

[0348] Example 24: Cell cycle analysis

[0349] For cell cycle analysis, cells were harvested 48 h (HepG2, Huh7, and SK-HEP-1) or 72 h (HS683 and U87MG) after transfection and resuspended (1 × 10 6 Cells) were then fixed with 700 μl of ethanol (Biosesang) at 4°C for 1 hour. Afterwards, 10 μl of RNase A (10 μg / ul, Intron Biotechnology, #IBS-BR003) was treated at 37°C for 30 minutes, and then 10 μl of 1 μg / ul of propidium iodide (PI; Sigma-Aldrich) was added. The cells were then analyzed using an Attune NxT flow cytometer (Thermo). Cell cycle analysis was also performed using a Muse cell analyzer (Luminex) with the Muse Cell Cycle Kit (Luminex, #MCH100106) according to the manufacturer's protocol.

[0350]

[0351] Example 25: Wound-Regeneration Assay

[0352] For wound regeneration assay, 1 X 10 per well in a 24-well plate 5Dog cells (SK-HEP-1 and Huh7) or 4 X 10 5 Dog cells (HepG2) were transfected with 50 nM RNA duplex, 250 ng of plasmid using RNAiMax (Invitrogen) or Lipofectamine 3000 (Invitrogen), and cultured for 24 h to achieve 90% confluence.

[0353] Vertical or horizontal wounds were created using a SPLScar Scratcher (SPL, #201924). Wounded cells were washed once and grown in medium supplemented with 10% FBS for 24–72 h (SK-HEP-1: 24 h, HepG2: 72 h, Huh7: 48 h). Cell images (n≥3, biologically independent samples) were captured with an inverted light microscope (Leica DMi8) and quantitatively analyzed using ImageJ software (imagej.nih.gov / ) equipped with a macro (edge-detection / thresholding method) for automated analysis of scratch wounds.

[0354]

[0355] Example 26: Cell invasion assay

[0356] HS683 or Huh7 cells were seeded in 24-well plates (HS683: 2.5 x 10 per well 4 Cells, Huh7: 5 x 10 per well 4 (Dog cells) and reverse transfected using RNAiMax (Invitrogen) containing 50 nM RNA duplex according to the manufacturer's protocol. After 24 h, cells in the transwells were replaced to grow in medium without FBS. The lower chamber contained medium containing 10% FBS as a chemoattractant. After 48 h, non-invading cells were removed with a cotton swab.

[0357] Cells that had invaded the PET membrane pre-coated with Matrigel were then visualized by hematoxylin and eosin (H&E) staining: After washing with PBS (WelGENE, #LB-004-02), the membrane was fixed with 100% methanol (Biosesang, #MR1066-500-1) for 5 min. After a quick rinse with distilled water, the membrane was stained with hematoxylin (YD Diagnostics, #S2-5) for 5 min, washed again with distilled water, and stained with eosin (Sigma-Aldrich, #E6003) for 1 min.

[0358] Membranes (at least three membranes from biologically independent replicates) were washed twice with distilled water and mounted on glass slides. Stained cells were then randomly sampled from at least three different areas on the same membrane using a Leica DMi8 inverted light microscope. The average cell count was quantified and represented as a bar graph with standard deviation.

[0359]

[0360] Example 27: RNA-Seq

[0361] RNA-seq libraries were prepared using the strand-displacement stop / ligation method, o 8 G-miRNA or o 8 It was generated from large RNA derived from the expression of a G-miRNA inhibitor. Polyadenylated mRNA was purified from 1.5 μg of large RNA (extracted with the RNeasy mini kit; QIAGEN) using 10 μl of Dynabead oligo(dT)25 (Invitrogen, #61002) according to the manufacturer's protocol. Then, 10 ng of purified mRNA was used for RNA-seq library preparation using the CORALL total RNA-seq library prep kit (Lexogen, #095) according to the manufacturer's protocol.

[0362] For analysis of the SK-HEP-1 cell line, the SENSE total RNA-seq library prep kit (Lexogen, #001) was used. After cDNA construction, the library was amplified by PCR using the lowest optimal cycle determined by qPCR. The quality of the amplified library was confirmed using a Fragment Analyzer (Agilent), which estimates size distribution and quantity. After cross-checking the library concentration using the Qubit dsDNA HS assay kit (Thermo, #Q32854), all prepared multiplexed libraries were accurately pooled and sequenced using SE50 or SE75 on a MiniSeq System (Illumina). If necessary, the libraries were sequenced using PE150 on a HiSeq X Ten System (Illumina).

[0363] All FASTQ files are archived in the SRA database (SRP353641, SRP353644, SRP353643, SRP353687, SRP353639).

[0364]

[0365] Example 28: RNA-Seq Data Analysis

[0366] Demultiplexed sequencing reads (FASTQ files) were processed to the same restricted read length (e.g., SE50, SE75, or PE150) within the same experimental batch and aligned to the human genome (hg19) using TopHat2 or STAR, supplying RefSeq gene annotations.

[0367] (tophat2 -G --no-novel-juncs -a 4 -g 1 --b2-sensitive --segment-length 20 --librarytype fr-secondstrand)

[0368] (SE50의 경우: star --outSAMattrIHstart 0 --outSAMstrandField intronMotif --outFilterType BySJout --outFilterMultimapNmax 20 --outFilterIntronMotifs RemoveNoncanonicalUnannotated --outMultimapperOrder Random --alignSJoverhangMin 8 -alignSJDBoverhangMin 1;

[0369] PE150의 경우: star --outSAMattrIHstart 0 --outSAMstrandField intronMotif --outFilterType BySJout --outFilterMultimapNmax 20 --outFilterIntronMotifs RemoveNoncanonicalUnannotated --outMultimapperOrder Random --alignSJoverhangMin 8 -alignSJDBoverhangMin 1 --alignIntronMax 1000000 --alignMatesGapMax 1000000 --outFilterMismatchNoverLmax 0.1 --outFilterMatchNminOverLread 0.2 --outFilterScoreMinOverLread 0.2 --seedSearchStartLmax 20)

[0370] The first 12 nucleotides from the beginning of the read were trimmed. Note that for PE150, the first 9 nucleotides from the beginning of the R2 read were also trimmed before mapping, following the instructions of the RNA-Seq library kit manufacturer (Lexogen). For RNA-Seq data (SK-HEP-1) from the SENSE total RNA-seq library prep kit (Lexogen, #001), the first 9 nucleotides from the beginning of the read were trimmed and aligned to the human genome (hg19) using TopHat2 with the same parameters except for --librarytype fr-firststrand .

[0371] Transcript abundance was quantified as reads per kilobase of transcript per million mapped reads (RPKM) using Cufflinks (Cufflinks -G -b --compatible-hits-norm --library-type frsecondstrand). For RNA-Seq data (SK-HEP-1) from SENSE total RNA-seq library prep kit (Lexogen, #001), the same parameters were used except for changing the library type option in Cufflinks (--library-type fr-firststrand).

[0372] Statistical significance of differential expression (determined by P-value) was analyzed by Cuffdiff (Cuffdiff --FDR 0.1 -b --compatible-hits-norm --library-type frsecondstrand) providing mapping results in groups with the same conditions. Only values ​​with valid status were selected. P-values ​​were calculated from all replicates (t-test, 2-tailed) and fold change (log2 ratio) using the mean or median RPKM within samples within the same experimental batch; o 8 G-miRNA or o 8G-miRNA inhibitor (anti-oxo) expression samples vs. control experiments (NT or control).

[0373] Given o 8 G-miRNA or o 8 To determine the propensity of putative target transcripts to be repressed or derepressed by expression of G-miRNA inhibitors (anti-oxo), the cumulative distribution function (CDF) was analyzed according to the fold change (log2 ratio) with RPKM values ​​derived from Cufflinks and Cuffdiff. When linear normalization was required, median centering was applied to the samples before calculating the fold change. The predicted o 8 G-miRNA targets were selected as transcripts containing a cognate oxo site in their 3'UTR (defined by RefSeq downloaded from the UCSC genome browser) and, unless otherwise indicated herein, were generally o 8 All 6mer matches for positions 2-8 of G-miRNA (o 8 (including matches of G:A) were searched.

[0374] “No site” indicates transcripts that do not contain a cognate 6-mer site at the start (positions 2-8) of the mRNA sequence. “Total” indicates total mRNA expressed above a threshold (RPKM >0, 0.1, or 1; determined by scatter plot and MA plot). Putative oxo targets are transcripts with an oxo site in the 3'UTR that was not detected in the RNA-Seq experiment. 8 It is also downregulated in the presence of GmiRNA (log ratio < 0).

[0375] High-confidence oxo targets refer to transcripts with an oxo moiety in their 3'UTR, and the corresponding o 8 Downregulated in the presence of G-miRNA (log ratio <0), but o in RNA-Seq experiments 8Up-regulated in the absence of G-miRNA (log ratio > 0). KS test was performed using Scipy (scypy.stats.ks_2samp()) with respect to the distribution of total expressed mRNA.

[0376] To identify differentially expressed genes (DEGs), the distribution of fold change (log2 ratio) according to the magnitude of RPKM values ​​(log2[average volume]) was examined, and the RPKM cutoff was determined to minimize the influence of experimental fluctuation, and MA plot analysis was performed early. Then, fold change (log2 ratio) and significance (-log 10 [P-value]) was calculated and volcano plot analysis was performed with the selected RPKM values ​​(log2(average volume)>0, 1.5, 2; determined by MA plot analysis).

[0377]

[0378] Example 29: Gene ontology analysis

[0379] Gene ontology (GO) analysis is a highly reliable o 8 It was performed using DAVID (https: / david.abcc.ncifcrf.gov / ) with G-miRNA target supply, and o in RNA-Seq analysis 8 G-miRNA was downregulated (log2ratio<0) and anti-4oxo (log2ratio>0). The transcript list expressed in the matched control set was used as the basis, and log2(average volume) > 1 (Figs. 4g, 4m, and 4r) or > 2 (Figs. 2m and 2n) was selected as the default parameter in DAVID unless otherwise specified.

[0380] GO analysis results were visualized for significantly enriched GO terms with EASE scores <0.1 or <0.2 in biological process networks using REVIGO ( / revigo.irb.hr / ); semantic space as cluster representative; circle color and proportional size are log 10 (P-value) is shown; the P-value was obtained from DAVID's EASE score. Interconnected graphs were also analyzed, and GO terms with high similarity were linked to indicate similarity with breadth.

[0381]

[0382] Example 30: Quantitative real-time PCR

[0383] o 8 To validate G-miRNA targets, quantitative RT-PCR (qPCR) was performed to determine whether the target was o 8 G-miRNA or o 8 Expression of G-miRNA inhibitor (anti-oxo) was performed. Large RNAs (>200 nt) were isolated using the RNeasy Mini Kit (QIAGEN) via on-column DNA digestion with an RNase-free DNase set (QIAGEN, #79254).

[0384] RT was then performed using SuperScript III RT (Invitrogen, #18080051) with fixed oligo(dT) primers. qPCR was performed using the QuantiNova SYBR Green RT-PCR kit (QIAGEN, #20805) and o 8 Custom primers for G-miRNA target mRNA (4o 8 G-miR-124 in LGG: EFGR, ATXN10, IGF2BP3, BCAT1, and MEOX2; 4o 8 G-let-7 in LGG: CLTA, TIA1, PUDP, HOXB3, and AZIN1; 3o 8G-miR-122 in LIHC: AKT1S1, IL6st, TMEM170B, and NHRLC2; 2,3o 8 G-miR-122 in LIHC: ASPH, ZNF146, ADAM9, FAM168B, STEAP2, LRP10, and SMOC1; 4o 8 G-let-7 in LIHC: DDAH1, CD164, PAWR, and GSR) were performed using Rotor-Gene Q (QIAGEN).

[0385] For reference, the qPCR results for MEOX2 in Figure 2o were combined from two independent qPCR experiments using two different primer sets. All reactions were performed in triplicate using a standard two-step cycling protocol. Relative quantification was calculated using the △CT method, using GAPDH as a control. Expression of the transgene (under the albumin promoter) in 2,3U-miR-122 or 3U-miR-122 TG mice was also quantified using qPCR compared to endogenous mouse albumin mRNA.

[0386] SEQ ID NO: Name Sequence 5'-3'65endogenous mouse albumin mRNA forward primerCAAGAGTGAGATCGCCCATCG66endogenous mouse albumin mRNA reverse primerTTACTTCCTGCACTAATTTGGCA67qPCR forward primerCGTCTGTTCTCTCCGAGCAA68qPCR reverse primerGGGCAAACAACAGATGGCTG

[0387] To quantify miRNAs, TaqMan MicroRNA Assay (Applied Biosystems, #4427975) was used according to the manufacturer's protocol. Briefly, 50 ng of small RNAs were reverse transcribed using a MicroRNA Reverse Transcription Kit (Applied Biosystems, #4366596) containing specific RT probes that recognize each miRNA sequence. Using the RT product, qPCR reactions were performed in TaqMan Universal PCR Master Mix (No AmpErase UNG; Invitrogen, #4440040) using specific qPCR primers. The RT probes and qPCR primers used to detect specific miRNAs were provided in the kit as follows: miR122, hsa-miR-122 (ID:002245); NT, cel-miR-67-3p (ID:000224); U6, U6 snRNA (ID:001973); qPCR analyses were performed in technical replicates (n ≥ 3) and in parallel reactions without reverse transcriptase (negative control) on a Rotor-Gene Q (Qiagen).

[0388]

[0389] Example 31: Analysis of AGO RIP, AGO CLIP, and CLEAR-CLIP sequencing data.

[0390] For total RNA (n=5), AGO RIP-Seq data (FASTQ files) generated from glioblastoma patients were retrieved from the GEO database (GSE106810) and aligned to the human genome (hg19) using Bowtie2 (http: / bowtie-bio.sourceforge.net / bowtie2 / ) with default parameters. After converting BAM files to BED format using BedTools (bedtools.readthedocs.io), the mapped reads were further analyzed to identify AGO-binding regions using the CLIPick program (clip.korea.ac.kr / clipick) with the function of selecting reads from reproducible clusters (CLIPick.reproducible_peaks), where reads overlapped within at least two biological replicates (biological complexity, BC≥2). Based on AGO RIPSeq data analysis, 189,582 clusters (peak width, -30 / +30) were defined as AGO binding regions in glioblastoma and used in this study.

[0391] To investigate AGO binding regions in the normal human brain, AGO CLIP-Seq reads (FASTQ files) generated from the prefrontal cortex (BA4 area, n=5) were downloaded from the GEO database (GSE52084) and mapped using Bowtie2 with the local alignment option, allowing one mismatch in each seed (-local -N1), and then duplicate reads were removed.

[0392] Final read counts from glioblastoma patients (BC≥2) or normal brains (BA4; all reads) were compiled into mapped reads (bedGraph) using BedTools and visualized by the UCSC Genome Browser (genome.ucsc.edu / ). To identify 4oxo target sites of miR-124 and let-7 within the AGO binding region, 7mer (positions 2-8), 6mer (positions 2-7), and offset 6mer (positions 3-8) matches were searched in the edited read regions.

[0393] AGO CLIP-Seq data (FASTQ files) from human liver cancer patients (n=9, tumor and patient-matched normal tissue pairs with technical replicates) were retrieved from the GEO database (GSE97056). Reads containing miRNA sequences were identified by Bowtie's 'reverse' mapping (alignment of mature miRNA sequences to sample reads) allowing two mismatches (Bowtie --norc -l 7 -n 2 -a -f).

[0394] The positional mismatch (%) of miRNA reads was analyzed against the miR-122 seed sequence. To identify the miR-122 3oxo site, miR-122 2,3oxo site, and let-7 4oxo site in liver cancer patients, raw sequencing data (FASTQ) were also preprocessed using CLIP Tool Kit (github.com / chaolinzhanglab / ctk) and FASTX-Toolkit (hannonlab.cshl.edu / fastx_toolkit / ).

[0395] Initially, reads were filtered based on quality score (fastq_filter.pl -if sanger -f min:0-3:20 avg:2040:20 -maxN -1 -v -of fasta). Reads were then further collapsed (fasta2collapse.pl), trimmed for 5' adapter sequences (fastx Trimmer -f 21), removed 5' degenerate linkers (stripBarcode.pl -linker 5), and removed for 3' adapter sequences (fastx Clipper; -l 20 -a GTGTCAGTCACTTCCAGCGG -n) as instructed in the data processing description of GSE97056.

[0396] Preprocessed reads were aligned to the human genome (hg19) using Bowtie2 with default parameters. The aligned AGO CLIP reads were finally generated as BED files using BedTools and SAMtools (samtools.sourceforge.net / ) and used in the CLIPick program, which has a function to select reads from reproducible clusters (CLIPick.reproducible_peaks), to generate AGO binding regions (all) with reproducibility (BC≥2).

[0397] Processed AGO CLIP reads were edited and matched to miR-122 3oxo, miR-122 2,3oxo and let-7 4oxo positions, 7mer (positions 2-8), 6mer (position 27) hematoxylin and offset 6mer (positions 3-8) were retrieved and applied to the UCSC genome browser.

[0398] To analyze specific AGO interactions specific to liver cancer, we exclusively defined tumor-specific AGO clusters (de novo tumor clusters) using BedTools by subtracting AGO interactions from patient-matched normal tissue (paired subtraction) or by subtracting all AGO interactions from normal tissue (joint subtraction). As a control set, normal-specific AGO clusters (de novo normal clusters) were also defined by paired subtraction.

[0399] AGO CLIP then reads overlapping de novo tumor clusters (paired subtractions), which are further processed by CLIPick to identify de novo tumor peaks with refined widths for various levels of reproducibility (11 / +10, all; -12 / +11, BC≥2; -13 / +13, BC≥3). The same method was used to process de novo normal peaks (11 / +10, all; -12 / +11, BC≥2; -13 / +13, BC≥3) and de novo tumor peaks from combined subtractions (-10 / +10, all; -11 / +11, BC≥2; -13 / +13, BC≥3). Based on various reproducibility, de novo tumor peaks and de novo normal peaks (paired subtractions) were analyzed for putative oxo sites of miR-122 (GA; including 2,3oxo sites) and compared with the seed sites and the GU wobble base pair positions (GU) of the seed (Fig. 5d); miR-122 seed sites mediated by each GG mismatch (GG) were used as negative controls along with the calculated expected frequency (Exp). The same analysis was performed for de novo tumor peaks from the paired subtractions.

[0400] Hepatoma cell line Huh-7.5 59CLEAR-CLIP (Covalent Affinity Identification of Endogenous AGO-Binding RNAs via Cross-linking and Immunoprecipitation) data were retrieved from the GEO database (GSE73059). Sequencing reads (FASTQ; GSM1881504-GSM1881515, n=12) were preprocessed and mapped as described previously with minor modifications. After adapter sequences were removed, potential chimeric reads containing miRNA sequences were identified by 'reverse' mapping using Bowtie, where mature miRNA sequences were aligned against the sample library.

[0401] The remaining sequences were then mapped to the human genome (hg38), only one overlapping exon was further selected, all copies were joined together, and ultimately analyzed only for miR-122 containing chimeric reads. miR-122 oxo sites (3oxo and 2,3oxo sites; o 8 To search for 7mer sites (via GA base pairs), upstream (5', -) and downstream (3', +) positions were extended (- / + 0, - / + 50, - / + 100) from mapped chimeric reads, and the 7mer site frequency (positions 2-8; %) from miR-122 seed and control sites (2GU and 2,3GU sites; via GU wobble pairs) was compared.

[0402]

[0403] Example 32: Immunohistochemistry of liver tissue

[0404] Dissected rat or mouse liver tissues were fixed (10% Neutral Buffered Formalin [Biosesang, F2013100-00], RT for overnight), paraffin-embedded, sectioned (5 μm deep), and stained with hematoxylin and eosin (H&E), Masson's trichrome, and Ki67 stain (Ki67 antibody [D3B5]; Cell Signaling Technology, #12202; secondary antibody HRP Plus DAB; GBI Labs, #D41-6). Paraffin-embedded sections, H&E staining, Masson's trichrome staining, and Ki67 staining were performed at the Pathology Core Facility, Korea University College of Medicine, through service requests. Slides were scanned at the Pathology Core Facility, Seoul National University College of Medicine. Ki67-positive cells in tissue sections were manually counted within randomly selected areas (n=9) from each slide of biological replicates (n≥3) and calculated as a percentage of the total number of cells quantified by the ImageJ program (imagej.nih.gov / ).

[0405]

[0406] Example 33: Dual Fluorescence Reporter Assay Using Flow Cytometry

[0407] Fluorescent protein reporter vectors were used to measure miRNA-mediated gene repression at the single-cell level using flow cytometry. First, Huh7 cells were transfected with the pDual vector (GFP:RFP-site) containing miR-122 3oxo or 2,3oxo sites, and cells were harvested 48 h after transfection.

[0408] For antioxidant treatment, 2 mM NAC was treated for 24 h from the time of transfection. Cells were harvested in PBS (Biosesang) containing 4% BSA (Bovogen) and 5 mM EDTA (Sigma) and analyzed on an Attune NxT flow cytometer equipped with blue (488 nm) and yellow (561 nm) lasers to achieve full excitation (a state higher in energy than the ground state, the lowest energy quantum mechanical state) of both GFP and RFP.

[0409] After gating on single live cells (based on FSC and SSC), cells without reporter expression were further filtered out based on the basal level of autofluorescence estimated by psi-Check2 (no position) transfected cells. Scatterplots of RFP vs. GFP signals were then analyzed and selected for further analysis. The distribution of the log ratio between RFP and GFP signals (log 10 [RFP / GFP]) was analyzed using box plots using Excel or violin plots using seaborn (seaborn.pydata.org / ,seaborn.violinplot()).

[0410] RFP and GFP signals are reporter values ​​(log 10 [RFP]) was further analyzed as a relative ratio by averaging (log 10 (estimated as GFP; equally binned). Ultimately, relative fold changes were calculated by normalizing the relative ratios to values ​​derived from the same reporter without position; the mean of the RFP values ​​in each bin of GFP values ​​was expressed as the value without position control. To narrow the analysis to only the lowest 25% of reporter expression, relative fold changes were recalculated using only the cells with the lowest 25% of reporter values ​​in each bin of vector expression values.

[0411]

[0412] Example 34: o 8 Construction of G-miR-122 mimic and anti-oxo (miR-122) expression vectors

[0413] 3o 8 G-miR-122 and 2,3o 8 To construct expression vectors for 3U-miR-122 and 2,3U-miR-122 as mimics of G-miR-122, mutagenesis was introduced into the pcDNA3.2 / V5 mmu-mir-122 plasmid (Addgene, #26325), a gift from Dr. David Bartel. A 498-bp region spanning mmu-miR-122 was amplified by PCR (forward1: SEQ ID NO: 69, reverse1: SEQ ID NO: 70) and containing point mutations from two PCR products joined together to generate 3U-miR-122 (forward1: SEQ ID NO: 71, reverse: SEQ ID NO: 72; forward: SEQ ID NO: 73, reverse1: SEQ ID NO: 74) or 2,3UmiR-122 (forward1: SEQ ID NO: 75, reverse: SEQ ID NO: 76; forward: SEQ ID NO: 77, reverse1: SEQ ID NO: 78), and cloned back into pcDNA3.2 / V5 plasmid via AseI and XbaI sites.

[0414] Sequence number name sequence 5'-3'69mmu-miR-122 forward primer(1)GAACCCACTGCTTACTGGCT70mmu-miR-122 reverse primer(1)GCGTAGAATCGAGACCGAGG713U-miR-122 forward primer(1)GAACCCACTGCTTACTGGCT723U-miR-122 reverse primer(1)TCACACTACACAGCTCTGCT733U-miR-122 forward primer(2)AGCAGAGCTGTGTAGTGTGA743U-miR-122 reverse primer(2)GCGTAGAATCGAGACCGAGG752,3UmiR-122 forward primer(1)GAACCCACTGCTTACTGGCT762,3UmiR-122 reverse primer(1)TCACACTAAACAGCTCTGCT772,3UmiR-122 forward primer(2)AGCAGAGCTGTTTAGTGTGA782,3UmiR-122 reverse primer(2)GCGTAGAATCGAGACCGAGG

[0415] These vectors were confirmed to express 3U-miR-122 and 2,3U-miR-122 in Huh7 and were used in xenograft experiments.

[0416] To selectively express 2,3U-miR-122 and 3U-miR-122 in hepatocytes, we constructed albumin::2,3U-miR-122 and albumin::3U-miR-122 plasmids and generated liver-specific transgenic mice.

[0417] A 317-bp region spanning mmu-miR-122 was amplified by PCR from pcDNA3.2 / V5 2,3U-miR-122 or 3UmiR-122 plasmid (forward2: SEQ ID NO: 79, reverse2: SEQ ID NO: 80) and further modified to have the loop sequence of human pre-miR122 using a combination of two PCR products (forward2: SEQ ID NO: 81, reverse: SEQ ID NO: 82; forward: SEQ ID NO: 83, reverse2: SEQ ID NO: 84).

[0418] Sequence number name sequence 5'-3'79mmu-miR-122 forward primer (2) ATCGATTCCCTATGCTCCAGTTTC80mmu-miR-122 reverse primer (2) GATATCTAAGCACAAAGAGGACCAG81human pre-miR122 forward primer (1) ATCGATTCCCTATGCTCCAGTTTC82human pre-miR122 reverse primer(1)GGCGTTTTGATAGTTTAGACACAAACACC83human pre-miR122 forward primer(2)GGTGTTTGTGTCTAAACTATCAAACGCC84human pre-miR122 reverse primer(2)GATATCTAAGCACAAAGAGGACCAG

[0419] Then, it was cloned into pBluescript2 SK(-) vector (ClaI and EcoRV sites) containing the promoter sequence and the mouse albumin enhancer (mm39: chr5: 90,597,504-90,598,315) fused with bGH poly(A) (bovine growth hormone polyadenylation), a gift from Dr. Seung-Hoe Koo (mm39: chr5: 90,608,588-90,608,782).

[0420] To construct hepatocyte-specific expression vectors for competitive sponge inhibitors against 2,3o8GmiR-122 (albumin::anti-2,3oxo) or 3o8G-miR-122 (albumin::anti-3oxo), we first amplified the entire albumin::3U-miR-122 plasmid by PCR, excluding the hairpin sequence of 3U-miR-122 (forward: SEQ ID NO: 85; reverse: SEQ ID NO: 86), and made the amplicons include two restriction enzyme sites, BstB1 and BglII. Then, a synthetic DNA duplex containing 12 repeats of the miR-122 2,3oxo site (forward: SEQ ID NO: 87; reverse: SEQ ID NO: 88) and 3oxo site (forward: SEQ ID NO: 89) was ligated to the plasmid amplicon via the BstB1 and BglII sites. The resulting albumin::anti-2,3oxo and albumin::anti-3oxo were used to generate transgenic mice.

[0421] Sequence number name sequence 5'-3'85albumin::3U-miR-122 plasmid forward primerCACTAAAAGATCTGTCCACTCCACGCGTGACTTG86albumin::3U-miR-122 plasmid reverse primerGGAAATTCGAACAGGCACAAGGGTATTAAC87miR-122 2,3oxo forward primerCGAACACACTAAAACACACTAAATCACACTAAAGCACACTAAACCACACTAAAACACACTAAACCACACTAAATCACACTAAAGCACACTAAAACACACTAAATCACACTAAAGCACACTAAA88miR-122 2,3oxo reverse primerGATCTTTTAGTGTGCTTTAGTGTGATTTAGTGTGTTTTAGTGTGCTTTAGTGTTGATTTAGTGTGGTTTAGTGTGTTTTAGTGTGGTTTAGTGTGCTTTAGTGTGATTTAGTGTGTTTTAGTGTGTT89miR-122 3oxo forward primerCGAACACACTACAACACACTACATCACACTACAGCACACTACACCACACTACAACACACTACACCACACTACATCACACTACAGCACACTACAACACACTACATCACACTACAGCACACTACA

[0422]

[0423] Example 35: Presumed O of the tumor 8 Configuring a web server for G-miRNA sequences

[0424] Putative O in tumor patients based on analysis of G>T SNVs (TCGA and GEO data; n = 8,687) 8To provide G-miRNA sequences, a web server (clip.korea.ac.kr / gt_db / ) was implemented using MySQL and set up on a 12-core server. The web interface was implemented using HTML, CSS, and JavaScript components, and the JavaScript component used the jQuery library (jquery.com / ). Within the server, users can automatically complete miRNA names to help formulate queries. All SNV information was sourced from a prebuilt MySQL table containing a ranked list of G>T frequencies, specifically focusing on the seed region of miRNAs.

[0425]

[0426] Example 36: Genomic DNA extraction

[0427] Genomic DNA was extracted using the sodium iodide method, which prevents in vitro oxidation. First, rat liver tissue (~120 mg) was homogenized using a Minilys Personal Homogenizer (Bertin, #EQ06404) and then applied to the DNA Extractor TIS Kit (WAKO, #296-67701) according to the manufacturer's protocol #2. The concentration of the extracted genomic DNA was measured by spectrophotometry (Denovix, #DS-11) and used to construct whole-genome sequencing libraries.

[0428]

[0429] Example 37: Whole genome sequencing (WGS)

[0430] Extracted genomic DNA (2 μg) was sheared in 100 μl of TE buffer (10 mM Tris-HCL pH 8.0, 1 M EDTA pH 8.0) using a Bioruptor Pico (Diagenode, B01060002) under 13 cycles of sonication (30 s on, 30 s off), resulting in an average fragment size of ~200 bp without inducing in vitro oxidation. 1 μg of sheared DNA was then applied to WGS library construction using the NEBNext Ultra II DNA Library Prep Kit for Illumina (NEB, #E7645) and NEBNext Multiplex Oligos from Illumina (NEB, #E7335).

[0431] After constructing adapter-ligated DNA fragments, the desired product insert size (~200 bp) was purified using SPRIselect reagent (Beckman Coulter, #B23318), used for PCR amplification (4 cycles), and further cleaned up using SPRIselect reagent (Beckman Coulter) according to the manufacturer's protocol. The Qubit dsDNA HS assay kit (Invitrogen, #Q32854) and Fragment Analyzer (Advanced analytical, #M5310AA) were used to verify the quantity and quality of the final library.

[0432] For quality verification, the pooled libraries were sequenced using a Miniseq system (Illumina) in a 50-bp single-end run (SE50), and then sequenced on a HiSeq X Ten System (Illumina) in a 150-bp paired-end run (PE150) for in-depth analysis. Raw sequence data (FASTQ) were uploaded to the Sequence Read Archive (SRA) database.

[0433]

[0434] Example 38: WGS analysis for SNVs in miRNA regions

[0435] Paired-end sequencing reads from WGS libraries (FASTQ) with quality passing FastQC were initially trimmed of adapter sequences using Cutadapt (version 1.15, cutadapt.readthedocs.io / en / stable / ) and aligned to the mouse genome (rn6) using Bowtie2 (bowtie2 -p 12 --score-min L,-0.6,-0.1; version 2.3.4, bowtie-bio.sourceforge.net / bowtie2). Appropriately aligned reads were then selected using SAMtools (SAMtools view -F 1804 -f 2; version 1.9, samtools.sourceforge.net / ) and deduplicated by Picard's Markduplicates program (Picard MarkDuplicates ASSUME_SORTED=TRUE REMOVE_DUPLICATES=true; version 2.20.6, broadinstitute.github.io / picard / ).

[0436] Reads containing the genomic region of mature miRNAs (>1 nucleotide overlap) from the processed BAM files were selected based on the following miRBase annotation. Note that for this analysis, we used genomic coordination of mature miRNAs truncated by three nucleotides from the 3' end, which allows for the consideration of isomirs with different 3' ends and also excludes the high rate of variants arising from 3'-end modifications in small RNA-Seq analyses.

[0437] We then used Python's Pysam module (pysam.readthedocs.io / en / latest / ) to analyze paired-end reads, split them into stacked functions, and then combined them to calculate the percent variation (%, relative to the total count, combining variants and matches). Note that we analyzed variation only in reads with a quality base calling score (Q score > 30). At the same time, o 8 We calculated the percent variation using the small RNA-Seq results mapped and analyzed up to 2 mismatches (Q score > 30) in G-miSeq and compared them with the WGS results. Because of the double-stranded nature of DNA, G>T SNVs can be caused by C>A in DNA, so we calculated SNVs for both DNA and RNA based only on G (G>T, G>A, and G>C) and A (A>T, A>G, and A>C), which have been reported to have the lowest redox potential.

[0438]

[0439] Test Example 1: Presumptive distribution of miRNA initiation sequences in tumor patients 8 G>T variant

[0440] o in cDNA 8 Motivated by the nature of G-induced G>T mutations, we comprehensively analyzed single-nucleotide variants (SNVs) identified in miRNA sequencing (miRNA-Seq) data of cancer patients (The Cancer Genome Atlas (TCGA; n=8,253) and Gene Expression Omnibus (GEO; n=434)) (Fig. 1a).

[0441] Frequent SNVs showed a significant tendency to be located in the miRNA origin (positions 2 to 7) (top 100 vs. bottom 100; P = 0.024, KS test; Fig. 1b). This positional bias was primarily observed at G bases (45% of the top 20 SNVs, compared to other bases), and most of them represented G-to-T base transversions (G>T, 89% of the top 20 Gs; Fig. 1c).

[0442] According to their frequency in patients, G>T SNVs of miRNAs were distinctly positionally clustered in the 2nd, 3rd, and 4th positions with a subdivided pattern (clusters 2-1, 2-2, 3-1, 3-2, 4-1, and 4-2), and the subdivided pattern grouped patients with different clinical features in both lower-grade glioma (LGG; n=512; Fig. 1d) and liver hepatocellular carcinoma (LIHC; n=372; Fig. 1e).

[0443] Patients with the G>T cluster generally had worse survival and tumor grade, especially in groups that tended to be associated with hepatitis B virus (HBV) infection and alcohol consumption in LIHC (e.g., clusters 2-2, 3, and 4-2), whereas patients with cluster 4-1 had a milder tumor phenotype in LGG (Fig. 1d, e).

[0444] A similar pattern of clustered metastases during tumor progression was observed in LIHC patient-matched tumors compared to adjacent normal-appearing paired tissues (n=47, Fig. 1f), where SNVs were also identified in the origin region, but at a lower frequency than in tumor tissue data.

[0445] The identified clusters shared a sequence composition pattern according to G>T positions, more distinctly at the initiation region (positions 2 to 8; Figs. 1g to 1i) than at other positions (e.g., positions 11 and 12; Fig. 1j); this finding included G>T at position 4 of let-7a or miR-124 (4G>T), and positions 2 and / or 3 of miR-122 (2G>T and / or 3G>T), depending on the tissue-specific expression of miRNAs (Fig. 1k).

[0446] Consistent with this positional occurrence, o 8 In vitro oxidation (Fenton reaction) of synthetic miR-122, miR-124, and let-7, as verified by dot blot using G-specific antibodies (Fig. 1l), revealed similar positional o enrichment in their initiating sequences. 8 The G>T pattern was shown (Fig. 1m). Since the 5'-terminal region of DNA has been reported to have a lower redox potential, the present inventors randomly generated in vitro oxidized RNA fragments from artificial transcripts containing all possible combination sequences (three nucleotides) spaced by one G. 8 This trend in RNA was investigated and confirmed by detecting a 5'-end bias of G>T (Fig. 1n). Therefore, o of miRNA starting sequences 8 The 5' end bias of G>T is thought to follow the intrinsic sensitivity of nucleic acids to oxidation and is positionally influenced by the adjacent sequence composition.

[0447]

[0448] Test Example 2: Putative o of miRNA origin associated with tumor patients 8 G>T variant

[0449] In LGG, 4G>T miR-124 was significantly enriched in advanced tumors (G2 vs. G3 grade), identified as a prominent cluster in LGG patients (Fig. 2a), and enriched in long-term survivors (Fig. 2b). Excluding patients with low miR-124 expression (Fig. 2c, left panel), survival analysis revealed that higher frequencies of 4G>T miR-124 (top 25% vs. bottom 25%) were significantly associated with improved 5-year survival (P=0.018, Wilcoxon signed-rank test; Fig. 2c, right panel). The same analysis of the let-7 family using SNVs identified (compiled) in the shared origin sequence also revealed a 4G>T let-7 cluster in LGG (Fig. 2d), increased 4G>T let-7 in long-term survivors (Fig. 2e), and significantly prolonged survival in patients with high 4G>T let-7 frequency (P=0.048, log-rank method).

[0450] In let-7a, which has the most 4G>T variants among the let-7 family, 4G>T was also clustered, and its high frequency was associated with improved survival. Notably, 2G>T was additionally clustered in let-7a, and its high frequency was associated with decreased patient survival, although only in a minority of patients. These results are o 8 G is at the fourth position (4o) of miR-124 8 G-miR-124) and the fourth position of let-7 (4o 8 It occurs in G-let-7) and indicates that it can inhibit LGG.

[0451] In LIHC, as previously reported, miR-122 expression was suppressed and associated with increased survival. Nevertheless, the degree of 2G>T or 3G>T in miR-122 was increased in proportion to tumor grade.

[0452] When comparing patient-matched tumors with adjacent normal-appearing tissue (n=47), cluster analysis revealed that 2G>T in miR-122 was widely distributed in normals (~90%; likely benign tumors), but that approximately 25% of these were changed to 3G>T in malignant tumors (Fig. 2g, h).

[0453] The same pattern was observed in the analysis of all LIHC patients. 3G>T miR-122 tended to be restricted in long-term survivors (Fig. 2i), and the high proportion of 3G>T miR-122 (excluding patients with low miR-122 expression; Fig. 2j, left panel) was reduced in long-term survivors (top 25% vs. bottom 25%; P=4.6x10 -4 , log-rank method; Fig. 2j, right panel).

[0454] The same analysis for 2G>T in miR-122 revealed a bimodal distribution for patient survival, with high 2G>T being associated with decreased survival, whereas intermediate 2G>T had the opposite effect. Despite the limited detection of two mismatches, this could be explained by the abundance of 2,3G>T in the intermediate 2G>T group in miR-122 (7 out of 8) (2G>T and 3G>T in the same miR-122 sequence), suggesting that prolonged patient survival was due to 2,3G>T miR-122.

[0455] For the let-7 family, where overall expression levels were consistent across tumor grades, 4G>T SNVs were significantly enriched in LIHC. In contrast to the suppressed overall 2G>T frequency, the frequency of 4G>T in let-7 was significantly increased in tumors compared to patient-matched normal tissues (Fig. 2k).

[0456] Indeed, 4G>T let-7 was predominant in advanced tumor stages and tended to be restricted to patients with long-term survival (Fig. 2l), and its high frequency (within the same let-7 expression range) significantly decreased patient survival compared to its low frequency (P=0.017, log-rank method; Fig. 2m). Notably, 2G>T let-7 showed a constant frequency regardless of patient survival.

[0457] In summary, the present inventors have 3o 8 G-miR-122 and 4o 8 We inferred that G-let-7 is produced in LIHC and may promote tumor progression.

[0458]

[0459] Test Example 3: Positional o in LGG 8 Validation and sequencing of G-miRNA

[0460] o 8 To validate the G>T miRNA variant, we first 8 We confirmed the altered target recognition of G-miR-124. A new 4oxo recognition site (o) with verified specificity was identified using synthetic RNA. 8 Using a luciferase reporter with a GA base pair, we observed significant inhibition of its activity under H2O2 treatment in human neuroblastoma cell line SH-SY5Y, but relief of its inhibition by miR-124 inhibitor.

[0461] Serum starvation, a pathologically relevant condition in tumors, elevated ROS levels in SH-SY5Y, thereby increasing the frequency of 4G>T miR-124 (approximately 40%) in small RNA-Seq and silencing the miR-124 4oxo reporter in both SH-SY5Y and human LGG cell lines.

[0462] ROS + After antioxidant treatment, which significantly reduced the number of cells, the inventors o 8Applying the G-miSeq method to HS683, anti-o 8 o using G antibody 8 Isolate G-miR and o 8 Based on G>T base transition o 8 G was sequenced at the single nucleotide level (Fig. 3a).

[0463] o 8 G immunoprecipitation (o 8 G immunoprecipitation; o 8 In the G-IP analysis, we found that miR-124 was the most differentially oxidized, but let-7 family members were generally oxidized to the greatest extent; this finding was o 8 G abundance (o compared to miRNA expression) 8 It was further confirmed by analyzing that G amount) was significantly reduced after antioxidant treatment (relative o 8 G abundance, log2 ratio; Fig. 3b).

[0464] The same positional G>T cluster in LGG patients is o 8 It was identified in G-miRNA, where antioxidant treatment decreased its G>T SNV frequency (Fig. 3c). Specifically, o in miR-124 8 Oxidation to G occurred only at the 4th position (Fig. 3d, left panel) and the let-7 family also had a total o at the 4th position. 8 The G variant was shown (Fig. 3d, right panel).

[0465] Overall, we have demonstrated that G>T SNVs in miRNA sequences of LGG patients are derived from collective mutations.

[0466]

[0467] Test Example 4: 4o 8 G-miR-124 and 4o 8 G-let-7 suppresses the malignancy of glioma cells.

[0468] o from LGG 8To investigate the function of G-miRNA, the present inventors synthesized 4o 8 G-miR-124, 4o 8 G-let-7a or its analogue (U is o 8 GA was replaced; 4U-miR-124 and 4U-let-7a) were introduced into HS683 (Fig. 3e) and a decrease in cell number was observed.

[0469] Actually, 4o 8 G-miR-124 and 4o 8 G-let-7a has significant tumor-suppressive activity, i.e., induction of apoptotic cell death (Annexin 5 + ; Figure 3f), reduction of proliferating cells (Ki67 + ; Fig. 3g), increased cell cycle arrest, and reduced invasive activity (Matrigel invasion). In particular, expression of intact miR-124 induced morphological changes in HS683 fibroblasts into epithelial-like cells with increased apoptosis and cell cycle arrest, and decreased proliferation as part of the differentiation process, whereas expression of let-7a did not show any noticeable changes.

[0470] The same results were observed in the U87MG glioblastoma cell line, 4o 8 G-miR-124, 4o 8 The tumor-suppressive activity of G-let-7a and its analogs was confirmed.

[0471] By introducing the concept of miRNA sponge, o 8 The physiological relevance of G-miRNAs has been investigated through loss-of-function studies, where 4o 8 G-miR-124 and 4o 8A competitive inhibitor (anti-4oxo) for G-let-7 was prepared containing tandem repeats of the cognate target site (4oxo moiety; Fig. 3h) and was validated for specific inhibition by luciferase reporter assay (Fig. 3i).

[0472]

[0473] When anti-4oxo (against miR-124 or let-7) was transfected into HS683 cells, it significantly enhanced cell proliferation (Figs. 3J and 3K) and tumor cell invasion (Fig. 3L). Increased proliferation was also observed in anti-4oxo-transfected U87MG cells.

[0474] 4o 8 G-miR-124 or 4o 8 The progressive tumor behavior achieved through inhibition of G-let-7 also exhibited resistance to temozolomide (TMZ), a widely used drug for the treatment of malignant gliomas (Fig. 3m). Consistent with this, 4o 8 G-miR-124 or 4o 8 TMZ-induced cancer cell death was synergistically promoted by delivering G-let-7.

[0475] In summary, 4o 8 G-miR-124 and 4o 8 G-let-7 has been shown to suppress the malignancy of glioma cells.

[0476]

[0477] Test Example 5: 4o in LGG 8 G-miR-124 and 4o 8 Extensive redirection of G-let-7 targets

[0478] 4o in HS683 cells 8 G-miR-124 or 4o 8RNA sequencing was performed on transfected G-let-7, and 4o containing a cognate 4oxo site in the 3' untranslated region (3'UTR) 8 G-miR-124 or 4o 8 A significant trend toward downregulation (Fig. 4a) was identified in the putative target mRNA of G-let-7 (Fig. 4b). RNA-Seq analysis was performed using a competitive sponge inhibitor (anti-4oxo) and the 4oxo target (4o 8 Finally, we defined high-confidence 4oxo targets that were downregulated by overexpression (log ratio <0, Figures 4a and 4b) and upregulated by competitive inhibition (log ratio >0, Figures 4c and 4d).

[0479] High-confidence 4oxo targets are involved in various carcinogenic pathways. 8 Epidermal growth factor receptor (EGFR) signaling, apoptosis signaling, cell cycle regulation, and angiogenesis for G-miR-124 (Fig. 4e), and 4o 8 Gene ontologies such as apoptosis signaling, cell division, and cell migration (Fig. 4f) for G-let-7 were significantly enriched (P<0.05).

[0480] 4o 8 Among the G-miR-124 targets, the proto-oncogene EGFR24, the TMZ resistance gene ATXN1025, the glioblastoma-specific oncogenic post-transcriptional regulator IGF2BP3, and the glioblastoma-promoting catabolic factor BCAT1 were selected and validated in HS683 (qPCR, gain-of-function, Fig. 4g; loss-of-function, Fig. 4h).

[0481] Also included are the tumor-promoting integrin-based cell migration regulator CLTA, the posttranscriptional regulator TIA1, the metabolic anticancer drug resistance gene PUDP (also known as HDHD1A), the glioblastoma-promoting transcription factor HOXB3, and the oncogenic RNA editing target AZIN1. 8 G-let-7 targeting was validated (gain of function; Fig. 4i; loss of function; Fig. 4j).

[0482] Because patient-matched RNA-Seq data were available for LGG patients in the TCGA data (n=507), we performed DEG analysis on patients with the 4G>T miR-124 cluster (n=74) compared to other patients (P<0.01), including glioblastoma-promoting MEOX233,34 (validated; Figs. 4g and 4h) and EGFR. 8 We confirmed that the G-miR-124 target was downregulated (Fig. 4k).

[0483] We also performed the same RNA-Seq analysis on patients with the 4G>T let-7 cluster (n=144), and found that 4o genes such as CLTA, HOXB3, and AZIN1 8 Additional G-let-7 targets were identified (Fig. 4l).

[0484] We also retrieved and reanalyzed AGO-linked sequencing data (AGO RIP-Seq) from glioblastoma patient tissues and normal human brain (AGO CLIP-Seq; cortical BA4 region) to comparatively map the 4-oxo sites in AGO. 4o 8 MEOX2 and EGFR, 4o for G-miR-124 (Fig. 4m) and CLTA, PUDP (Fig. 4n) 8 G-let-7 was confirmed to be functional by luciferase reporter assay.

[0485] Overall, 4o 8 G-miR-124 and 4o 8 We confirmed that G-let-7 re-directed broad miRNA target repression to suppress LGG progression.

[0486]

[0487] Test Example 6: Stepwise o of miRNA in the process of liver cancer development 8 G variant

[0488] o while tumors develop in the liver 8 We investigated G-miRNA. First, we prepared a diethylnitrosamine (DEN)-induced LIHC mouse model that expressed progressive hepatocarcinogenesis (early and late time points, Fig. 5a) showing accumulation of hepatic lipid vacuoles (H&E staining), fibrosis (Masson's trichrome staining), and cell proliferation (Ki67 staining). DEN-induced LIHC significantly reduced the number of proliferating cells (Ki67 + ), ROS levels (CM-H2DCFDA) and small RNA o 8 G(o 8 G ELISA, <200 nucleotides) was significantly increased (Fig. 5b).

[0489] After confirming the redox dependence in the DEN-induced LIHC model, co-treatment with the antioxidant N-acetylcysteine ​​(NAC) attenuated tumor formation, o 8 G-miSeq was performed and miR-122 was found to have the highest o despite a decrease in total miR-122 levels after DEN treatment. 8 G enrichment value (o for expression 8 It was most oxidized (related to G-IP).

[0490] In relation to the mock processing, o 8 G-miR-122 and o 8 The G-let-7 family significantly increased from the early stage to the late stage after DEN treatment (relative o 8 G concentration, Figure 5c). o 8The G>T frequency of G-miRNAs also gradually increased, clearly confirming the stepwise pattern of positional G>T clusters during DEN-induced hepatocarcinogenesis (Fig. 5d).

[0491] o 8 G occurred more frequently in the initiation region with an increase in G>T clusters 2-1, 2-2, 3-1, 3-2, and 4-1, where positions 2 and 3 of miR-122 and position 4 of the let-7 family were preferentially oxidized (Fig. 5d). In the livers of LIHC patients, o 8 G-miSeq experiments showed the same localization pattern of G>T clusters (n=2; Fig. 5f), and predominant oxidation of miR-122 and let-7 families was consistently observed (Fig. 5e).

[0492] To exclude compounding effects in tissue heterogeneity, o was performed in the human LIHC-derived hepatocellular carcinoma cell line Huh7, which is the target cell line, because a subpopulation with particularly high ROS levels (ROS+) and miR-122 expression is observed. 8 G miRNA was sequenced (Fig. 6a).

[0493] Consistently, miR-122 and let-7 families were the most oxidized miRNAs in Huh7 (Fig. 6b), although antioxidant treatment reduced ROS + It reduced the number of Huh7 cells, but o 8 G concentration was only slightly reduced. However, 4o 8 G-let-7 and 2o 8 The G>T cluster pattern (Fig. 6c) containing G-miR-122 (Fig. 6d) was significantly reduced, with some o occurring at multiple locations within the same miRNA molecule. 8 This could be interpreted as a case of suppressing G.

[0494] Actually, o 8Two mismatch G>T variants were found in G-miR-122 (Fig. 6e), where antioxidant treatment resulted in a change in o in the read, mostly at positions 2 or 3. 8 Only one of G>T was reduced, but o in both positions at the same time 8 G also decreased (2, 3o 8 G, Fig. 6f).

[0495] These specific o 8 To confirm the existence of the GmiR-122 type, miR-122 was knocked out (KO) in Huh7 by CRISPR / Cas9 (Fig. 6g), and synthetic 2o 8 G-, 3o 8 G- and 2,3o 8 A luciferase reporter was used to quantify specific responses to different concentrations of G-miR-122 (2oxo, 3oxo, 2,3oxo moieties) to examine the inhibition target; endogenous 2,3o 8 GmiR-122 (~22% reduction) and 3o 8 The extent of reporter suppression for G-miR-122 (~18% reduction) was substantial, with their effective concentration estimated to be equivalent to transfecting cells with ~1 to 10 nM (Fig. 6g).

[0496] In summary, these results indicate that stepwise oxidation occurs positionally in G>T cluster miRNAs, including the most heavily oxidized 4o 8 G-let-7, 3o 8 G-miR-122 and 2,3o 8 G-miR-122 is included and is substantially oxidized during liver cancer development.

[0497]

[0498] Test Example 7: 3o 8 G-miR-122 and 4o 8 G-let-7 promotes liver cancer development and 2,3o 8 G-miR-122 inhibits liver cancer development

[0499] To investigate the oxidative function of miR-122 in liver cancer, synthetic 3o 8 G-miR-122 or 2,3o 8 G-miR-122 was transfected into HepG2, miR-122 KO Huh7, and SK-HEP-1 cells (Figs. 6h to 6j), which did not express miR-122.

[0500] 3o 8 G-miR-122 conferred tumor-promoting functions such as cell proliferation (Fig. 6h), cell cycle, cell migration (wound healing assay), and tumor invasion (Matrigel invasion; Figs. 6i and 6j), but 2,3o 8 G-miR-122 showed opposite effects on tumor suppression activity. Similar effects on cell proliferation, cell cycle, and cell migration were observed from its mimics (3U-miR-122 and 2,3U-miR-122).

[0501] In a normal mouse liver cell line (NCTC clone 1469), proliferation was 3o 8 Enhanced by G-miR-122, but tumor suppression 2,3o 8 No differences were detected by G-miR-122. Consistent results were obtained for proliferation (Fig. 6k), cell cycle, cell migration, and tumor invasion (Figs. 6j and 6l). 8 G-miR-122 (anti-3oxo) and 2,3o 8 Obtained from Huh7 using a competitive inhibitor for G-miR-122 (anti-2,3oxo).

[0502] Notably, gain-of-function studies using miR-122 precursor expression demonstrated delayed cell migration in hepatoma cell lines, as previously reported. However, when mature miR-122 was directly transfected with synthetic duplex RNA, multiple experiments using different miR122 duplex variants resulted in only negligible differences in wound healing assays.

[0503] Interestingly, expression of miR-122 precursor from a plasmid vector o 8 As described in the G-miSeq performance, 2,3o in HepG2 cells 8 G-miR-122 was produced. As in previous studies, not only whole miR-122 but also 2,3o 8 Treatment with a miR-122 inhibitor that can reduce the level of G-miR-122 increased cell migration activity in Huh7 cells. These results suggest that the tumor suppressive activity of miR-122, previously observed in hepatoma cell lines 21, 37-39, is due to the oxidized 2,3o during miRNA biogenesis. 8 This indicates that it may be derived from G-miR-122.

[0504] Next, o for sorafenib, a drug widely used in the treatment of advanced liver cancer 8 The function of G-miR-122 was tested. miR-122 o 8 As observed in G mimics (3U-miR-122 and 2,3U-miR-122), 3o 8 G-miR-122 was suppressed, but 2,3o 8 G-miR-122 promoted apoptosis, which was significantly enhanced in the presence of sorafenib. Apoptosis assays using anti-3oxo or anti-2,3oxo antibodies against miR-122 in sorafenib-treated Huh7 cells also supported these results (Fig. 6m).

[0505] 4o 8 Regarding G-let-7, tumor-promoting activity was also confirmed in hepatoma cell lines; it induced cell proliferation (Fig. 6n), cell cycle progression, cell migration, tumor invasion (Figs. 6o and 6p), and resistance to sorafenib-induced apoptosis. Consistent results were obtained in Huh7 cells for loss-of-function experiments (anti-4oxo against let-7) and for cell proliferation (Fig. 6q), cell cycle, tumor invasion (Figs. 6p and 6r), and sorafenib-induced apoptosis (Fig. 6s).

[0506] In addition, tumor-promoting activity of 4U-let-7 was observed in cell cycle and wound healing assays. Overall, 3o 8 G-miR-122 and 4o 8 Tumor-promoting function of G-let-7 and 2,3o 8 The tumor suppressive role of G-miR-122 was verified.

[0507]

[0508] Experimental Example 8: Extensive Retargeting of 3o8G-miR-122, 2,3o8G-miR-122, and 4o8G-let-7 Targets in LIHC

[0509] Afterwards, 3o in HepG2 cells (RNA-Seq) 8 G-miR-122, 2,3o 8 G-miR-122 or 4o 8 Transcriptome-wide changes affected by G-let-7a expression were assessed. In CDF analysis, we observed significant repression of putative target transcripts harboring cognate oxo sites (3oxo, 2,3oxo, or 4oxo sites; 7mer, 3'UTR), which were primarily downregulated in DEGs.

[0510] 3o 8 Similar results were obtained for SK-HEP-1 cells transfected with G-miR-122. 3o 8 G-miR-122, 2,3o 8 G-miR-122 or 4o 8 In the absence of G-let-7 (anti-3oxo, anti-2,3oxo or anti-4oxo of Huh7), the putative targets are upregulated, significantly activated in gene expression and defined as primarily upregulated DEGs.

[0511] 3o with considerable reliability 8G-miR-122 targets were selected (gain of function, log ratio <0; loss of function, log ratio >0) and found to be significantly overrepresented in gene ontology associated with VEGFR signaling pathway, cell proliferation, angiogenesis, cell migration, and invasion-related functions (Fig. 7a).

[0512] Among them, the tumor suppressor negative regulator mTOR kinase AKT1S140, the antifibrotic IL-6 signal transducer IL6ST (encoding gp130), the β-catenin pathway regulator TMEM170B42, and the caspase-8-mediated proapoptotic regulator NHRLC243 were validated (gain of function, Fig. 7b; loss of function, Fig. 7c); AKT1S1 and IL6ST were also confirmed in SK-HEP-1 cells.

[0513] We then reanalyzed high-throughput sequencing data of AGO-related RNAs derived from crosslinking immunoprecipitation (AGO CLIP) isolated from LIHC patient tissues. AGO-miR-122 was o 8 Because it showed the same G>T pattern (positions 2 and 3) as G-miR-122, it was verified as 3o 8 The function of the site was confirmed at AKT1S1 and IL6ST using G-miR-122 mRNA target (Fig. 7d) and luciferase reporter assay (Fig. 7e).

[0514] Using LIHC patient-matched RNA-Seq data (n=132) from TCGA, we performed DEG analysis on patients with 3G>T miR122 cluster (n=82) compared to normal patients, revealing preferential downregulation of 3oxo targets including IL6ST and NHLRC2 (Fig. 7f).

[0515] Highly reliable 2.3o 8The same strategy was applied to purify G-miR-122 targets, which were shown to positively regulate tumorigenesis, apoptosis signaling pathways, fibrosis, angiogenesis, cell migration, and invasion (Fig. 7g).

[0516] Analysis of AGO CLIP data from the LIHC22 patient also detected 2,3-oxo moieties in tumor-promoting targets, including the key hepatocellular carcinoma factor ASPH44, tumor-overexpressed genes ZNF14645 and SMOC146, tumor immune resistance regulator ADAM947, T cell regulatory exosomal mRNA FAM168B48, invasive tumor progression factor STEAP249, and patient survival-associated gene LRP1050 (Fig. 7h). The 2,3-oxo moieties in ASPH and ZNF146 were validated by luciferase reporter assays (Fig. 7i), and the other targets were confirmed by qPCR (gain-of-function, Fig. 7j; loss-of-function, Fig. 7k).

[0517] 4o 8 For G-let-7, high-confidence targets were purified and demonstrated functions in homeostasis, oxidative stress response, lipid metabolism, and vesicle trafficking (Fig. 4l). Among these targets, DDAH151, a metabolic regulator associated with nitric oxide production; CD16452, a signaling factor that inhibits stem cell proliferation; GSR53, an antioxidant and LIHC inhibitor; and PAWR54, a tumor suppressor that promotes apoptosis, were validated (gain-of-function, Fig. 7m; loss-of-function, Fig. 7n).

[0518] Some of these 4oxo sites overlap with the AGO-binding region of patient LIHC22 (Fig. 7o), and their association with DDAH1 was validated by luciferase reporter assays (Fig. 7p). RNA-Seq data from LIHC in TCGA (n=125) also showed that 4oxo targets, including DDAH1 and CD164, tended to be downregulated in patients with the 4G>T let-7 cluster (n=75; Fig. 7q).

[0519] In summary, we have 3o8 G-miR-122, 2,3o 8 G-miR-122 and 4o 8 We demonstrate that G-let-7 clearly re-directs broad-spectrum inhibition to control LIHC progression.

[0520]

[0521] Test Example 9: o in LIHC 8 Effective cellular levels and redirected target interactions of G-miR-122

[0522] Limit o 8 G detection has been explained by cellular heterogeneity, including a small number of driver cells prone to miRNA oxidation. To investigate the abundance of effective o8G-miR-122 in individual hepatoma cells, flow cytometry was performed using a dual fluorescent protein (dFP) reporter expressing the red fluorescent protein (RFP) gene and the green fluorescent protein (GFP) gene, which contain the target site.

[0523] Depending on redox status and miR-122 expression, silencing of the dFP reporter harboring miR-122 3oxo or 2,3oxo moieties was observed, which was prominent in a subset of Huh7 (Fig. 7r).

[0524] Notably, considering only the 25% of cells with the lowest RFP values, dFPs harboring miR-122 3oxo or 2,3oxo sites quantified effective repression (Fig. 7s), exhibiting a fold change of -1.25 to -0.25 (~30-80% in reporter activity) compared to the reporter without the site (log2 ratio).

[0525] So o 8 The concentration of G-miR-122 can be estimated to be effectively high at some individual cell levels (Fig. 7t), which corresponds to a synthetic o concentration of ~5-20 nM.8 Equivalent to that obtained by transfecting G-miR-122 (~30-80% in luciferase reporter activity).

[0526] o 8 To further confirm that G-miR-122 directly binds to oxo sites, we closely examined AGO CLIP-Seq data for miRNA target sites in LIHC patient tissues, where AGO-miR-122 displayed the same positional pattern as G>T mutations (positions 2 and 3, Figure 7u). Consistent with reproducible AGO binding (biological complexity, BC), the frequency of miR-122 3oxo or 2,3oxo sites was increased in tumor-specific AGO CLIP peaks (de novo tumors), selected exclusively (Figure 7v) or in combination with matched normal tissue data from each patient.

[0527] In contrast, miR-122 3oxo or 2,3oxo sites were not significantly detected in normal-specific AGO CLIP peaks (de novo normal), and miR-122 seed sites were decreased in de novo tumors but increased in de novo normals (Fig. 7v). Furthermore, analysis of CLEAR-CLIP (covalent binding of endogenous AGO-binding RNAs via cross-linking and immunoprecipitation) data in Huh-7.5 cells identified a significant number of 3oxo sites in chimeric reads linked between miR-122 and target sites in hepatoma cells.

[0528] Overall, the cellular level of o valid in LIHC 8 We confirmed changes in miR-122-target interactions mediated by G variants.

[0529]

[0530] Test Example 10: o in the development of liver cancer in vivo 8 Functional verification of G-miR-122

[0531] To verify hepatocarcinogenesis, tumor promotion 3o 8 Tumor suppression from G-miR-122 2,3o 8 Stepwise oxidation of miR-122 accumulated up to G-miR-122 (Fig. 8a) was attempted. First, 3o 8 G-miR-122 enhanced the colony-forming ability of liver cancer cells, but 2,3o 8 G-miR-122 was confirmed to alleviate. In addition, reduced colony formation was observed at 2,3o 8 Through vector transfection resulting in the production of G-miR-122, we observed that it could contribute to the expression of the miR-122 precursor.

[0532] Expanding to a xenograft mouse model, o in liver cancer cells 8 G-miR-122 mimics (3U-miR-122 or 2,3U-miR-122) were expressed and showed inhibitory activity against the cognate 3oxo or 2,3oxo sites (luciferase reporter assay) by subcutaneous injection into nude mice; 3U-miR-122 was also shown to be potent in a mouse xenograft model (miR-122 KO Huh7, Fig. 8b; Huh7). In addition, 2,3U-miR-122 sensitizes tumor cells to apoptosis even at low concentrations of sorafenib (10 nM) in vitro.

[0533] Next, 2,3U-miR-122 (conjugated with polyethyleneimine) was administered to DEN-injected mice at the time of tumor initiation (~9 months; Fig. 8c). After confirming liver delivery using qPCR, we examined the livers treated with 2,3U-miR-122 and found that DEN-induced tumorigenesis was dramatically attenuated, resulting in prevention of steatosis, fibrosis, and proliferation (Fig. 8d).

[0534] o in vivo 8To suppress G-miR-122, anti-3oxo or anti-2,3oxo was introduced into DEN-injected mice at the time of tumor progression (~12 months; Fig. 8e), when miRNA oxidation is assumed to be fully completed. At this time, miRNA oxidation is assumed to be fully completed. Delivery to the liver and corresponding o 8 We verified the function of G-miR-122 to induce global inhibition of targeting (Fig. 8f), and found that anti-3oxo alleviated, but anti-2,3oxo aggravated, DEN-induced steatosis, cirrhosis, and tumor formation (Fig. 8g).

[0535] o 8 For G-let-7, 4U-let-7a was administered to DEN-treated mice and aggravated tumor formation was observed (Fig. 8h), which was 4o during hepatocarcinogenesis. 8 Suggests the oncogenic activity of G-let-7.

[0536] We also generated transgenic mice (TG) expressing 2,3U-miR-122 or 3U-miR122 in hepatocytes (albumin enhancer / promoter; Fig. 8i). In two independent TG, we found that DEN-induced LIHC was significantly suppressed in 2,3U-miR-122 TG (Fig. 8i), and the degree of tumor formation was inversely correlated with the level of 2,3U-miR-122 expression (Fig. 8j).

[0537] We also observed that 3U-miR-122 TG promoted DEN-induced LIHC (including high-fat diet) (Fig. 8k), and the extent was correlated with the expression level of 3U-miR-122 in the TG (Fig. 8l). Finally, we generated liver-specific TG expressing anti-2,3oxo or anti-3oxo (the inhibitory activity of the expression vector was confirmed by qPCR) to inhibit endogenous o 8 Loss of function was investigated for G-level miR-122 in hepatocytes (Figs. 8m and 8n), by RNA-Seq analysis. 8Overall inhibition of G-miR-122 targets was verified.

[0538] Consistently, anti-2,3oxo promotes DEN-induced hepatocarcinogenesis, which correlates with the expression levels of competitive anti-oxo inhibitors (Fig. 8o), whereas anti-3oxo attenuates it (Fig. 8p). Overall, endogenous levels of o occurring positionally in miRNAs 8 We conclude that G oxidation acts as an epitranscriptional reprogramming mechanism that significantly controls tumor development.

[0539]

[0540] conclusion

[0541] Based on sequence analysis and functional studies, we suggest that the elevation of ROS during tumorigenesis is mediated by a sequence-specific o in the initiation region of miRNAs. 8 We demonstrated that G variants alter miRNA-mediated target repression and control tumor development. These o 8 The positional specificity of G appears to be determined by the nucleotide sequence composition and redox potential characteristics of the nucleic acid (Figs. 1i to 1k), where the 5' end, the GG repeat sequence, and adjacent oxidized bases are known to be preferentially oxidized due to the movement of radical cations.

[0542] Additionally, RNA-binding proteins such as components of the RNA-induced silencing complex (RISC), including AGO, associated with miRNA or its precursor forms, o 8 may further influence the selectivity of the G variant. o 8 Based on the G>T base transition of cDNA induced by G, a database web server (clip.korea.ac.kr) was constructed to identify all potential o in tumor miRNA sequences (n=550). 8 It can provide location information for G.

[0543] In contrast to the frequent G>T SNVs in miRNA sequencing, DEN-induced liver cancer did not show an increased frequency of G>T in the genomic region of miRNA sequences. G>T SNVs or o generated in the DNA encoding miRNAs 8 It shows that there is a non-negligible effect from G.

[0544] position-specific o 8 The effect of G miRNA is o 8 It was found to depend on GA base pairs o 8 Replacing G with U resulted in the same phenotype (Figs. 3f and 8) o 8 Transfecting the control sequence with the G variant does not result in any change, so this is o 8 This means that the G variant itself has little effect on tumor cells.

[0545] 7o 8 As previously reported for G-miR-1, 3o 8 G-miR-122 is maintained o 8 Although it can suppress the initiation site of the luciferase reporter through GC base pairing, it is less potent than unoxidized miR-122. 8 Although G-miRNA appears to partially retain the function of non-oxidized miRNA, o 8 G may also affect the stability of existing GC base pairs, which are particularly important for miRNA regulation (e.g., hairpin structures of pre-miRNA or mature miRNA duplexes). 8 Further studies considering the biophysical and structural effects of the G variant would be worth investigating this alternative mechanism.

[0546] o 8 Most tumor cell death regulated by G-miRNA undergoes apoptosis (Figs. 3f and 3m), but some of them, particularly in LGG, undergo 4o8 It appears to follow apoptosis with intact cell morphology, along with expression of G-let-7, and does not display characteristics of apoptosis (e.g., shrinkage, blebbing, and chromatin condensation) or necrosis (e.g., swelling of the cytoplasm and organelles, and rupture of the cell membrane).

[0547] To support this, 4o 8 G-let-7 expression is significantly increased in a number of PIs + / Annexin V - We observed that it induces apoptosis, a type of cell death other than apoptosis (Fig. 3f). 4o 8 G-let7 functions differentially depending on the cancer type (LGG or LIHC), but 4o 8 G-let-7 exhibited distinct target genes that function in lipid and antioxidant metabolism (Fig. 7l), one of which is glutathionedisulfide reductase (GSR; Fig. 7o), which produces the cellular antioxidant GSH, which activates GPX4 to protect against ferroptosis. Therefore, it may be involved in ferroptosis, a type of programmed cell death associated with oxidative stress induced by iron-dependent lipid peroxidation.

[0548] o 8 Sequencing of G>T mutations is generally inefficient in reverse transcription. 8 G:A base pairing and o 8 It was detected at a low rate due to G processing, and multiple o 8 For miRNAs with G variants, this was exacerbated. These marginal signals are due to low levels of endogenous o 8 Although it may be misinterpreted as G-miRNA, the low o of fully oxidized RNA in vitro 8 As explained in the G>T frequency (o 8G dot blot, Fig. 1i, RNA sequencing results after Fenton reaction, Figs. 1j and 1k), rather o 8 This likely reflects methodological limitations in G-position detection.

[0549] Additionally, cellular heterogeneity is o 8 The G detection signal can be further diluted (Fig. 7t). In fact, the heterogeneity o 8 We observed that a physiological mode of G regulation can be imparted, and endogenous o 8 Intracellular target inhibition mediated by G-miR-122 became substantial in some populations at the individual tumor cell level (Figs. 7r to 7t), suggesting o in loss-of-function studies. 8 Describe the physiological effects of G-miR-122 and its competitive o 8 Hepatocyte-specific expression of G-miR-122 inhibitor was used (Figs. 8g, 8m to 8p).

[0550] These findings suggest that ROS can positionally reprogram gene expression through a consensus sequence code that can be used posttranscriptionally. 8 G oxidation epitranscriptionally re-directs miRNAs, thereby 8 These results indicate that G-miRNAs may serve as potent endogenous regulators of tumor development.

[0551] The present invention relates to a method for providing information for cancer diagnosis or prognosis prediction using microRNA containing 8-oxoguanine, a cancer treatment agent and a screening method thereof, and more particularly, to a technology for analyzing the progression of cancer and controlling and treating it using microRNA in which the guanine at the 2nd to 7th nucleotide from the 5' end is modified to 8-oxoguanine.

Claims

1. One or more guanines (Guanine; G) among the 2nd to 12th nucleotides from the 5' end of the microRNA isolated from the subject are replaced with 8-oxoguanine (8-oxoguanine; o 8 A method for providing information for diagnosing or predicting prognosis of cancer, comprising an identification step of determining whether the cancer has been transformed into G).

2. In the first paragraph, the identification step comprises the following steps: A sequencing step of producing cDNA by reverse transcribing RNA containing 8-oxoguanine isolated from a subject and creating a sequencing library to identify the location of 8-oxoguanine; and An analysis step for confirming the position where guanine is transformed into thymine (T) based on the above sequencing results, thereby confirming the position where guanine is transformed into 8-oxoguanine.

3. An information providing method in claim 1, wherein the microRNA is at least one selected from the group consisting of miR-124, let-7, and miR-122.

4. In the first paragraph, the information providing method further includes a decision step of determining that the probability of developing glioma is low or the probability of having a good prognosis is high when the guanine at the 4th nucleotide from the 5' end of the microRNA is changed to 8-oxoguanine.

5. In the first paragraph, the information providing method further includes a determination step of determining that the probability of developing liver cancer is low or the probability of having a good prognosis is high when the guanine at the second and third nucleotides from the 5' end of the microRNA is transformed into 8-oxoguanine.

6. In the first paragraph, the information providing method further comprises a determination step of determining that the probability of developing liver cancer or the probability of having a poor prognosis is high when the guanine at either the 3rd or 4th nucleotide from the 5' end of the microRNA is transformed into 8-oxoguanine. 7.5' At least one guanine (G) among the 2nd to 12th nucleotides from the end is 8-oxoguanine (o 8 MicroRNA modified with G); A modified nucleic acid that specifically binds to the modified microRNA; A recombinant vector comprising a gene encoding the above modified nucleic acid; An analogue in which one or more of the guanine positions at the 2nd to 12th nucleotides from the 5' end is substituted with uracil (U); and A recombinant vector comprising a gene encoding the above analogue; A pharmaceutical composition for treating cancer comprising at least one selected from the group consisting of:

8. In the 7th paragraph, the composition A microRNA in which one or more guanines among the 2nd to 7th nucleotides from the 5' end are modified to 8-oxoguanine; A modified nucleic acid that specifically binds to the modified microRNA; A recombinant vector comprising a gene encoding the above modified nucleic acid; An analogue in which one or more guanine positions among the 2nd to 7th nucleotides from the 5' end are substituted with uracil (U); and A recombinant vector comprising a gene encoding the above analogue; A pharmaceutical composition for treating cancer, comprising at least one selected from the group consisting of:

9. A pharmaceutical composition according to claim 7, wherein the modified microRNA is at least one selected from the group consisting of miR-124, let-7, and miR-122.

10. A pharmaceutical composition according to claim 7, wherein the modified nucleic acid comprises a polynucleotide complementary to at least 6 consecutive polynucleotides starting from either the 2nd or 3rd nucleotide from the 5' end of the modified microRNA.

11. A pharmaceutical composition according to claim 7, wherein the modified nucleic acid comprises adenine (A) as a nucleotide at a position that binds to one or more 8-oxoguanines among the 1st to 9th nucleotides from the 5' end of the modified microRNA.

12. A pharmaceutical composition according to claim 7, wherein the cancer is glioma or liver cancer.

13. A method for screening cancer therapeutic agent candidates, comprising the following steps: A processing step of treating a candidate substance to tissue or cells separated from a subject; One or more guanines (Guanine; G) among the 2nd to 12th nucleotides from the 5' end of the microRNA expressed in the tissue or cell are replaced with 8-oxoguanine (8-oxoguanine; o 8 An identification step to determine whether it has been transformed into G); and A comparison step of comparing whether the frequency of guanine at the 4th nucleotide from the 5' end of the microRNA being modified into 8-oxoguanine increases or decreases, whether the frequency of guanine at the 3rd nucleotide from the 5' end of the microRNA being modified into 8-oxoguanine increases or decreases, or whether the frequency of guanine at the 2nd and 3rd nucleotides from the 5' end of the microRNA being modified into 8-oxoguanine increases or decreases.

14. A screening method according to claim 13, wherein the microRNA is at least one selected from the group consisting of miR-124, let-7, and miR-122.

15. A screening method according to claim 13, wherein the cancer is glioma or liver cancer.

Images