Compositions for use in the treatment of CHD2 haploinsufficiency and method for identifying the same

Abstract

A method is provided for increasing the amount of chromodomain helicase DNA binding protein 2 (CHD2) in a neuronal cell, the method comprising introducing into the cell a nucleic acid agent that downregulates activity or expression of human Chaserr, the nucleic acid agent being directed to the final exon of human Chaserr, thereby increasing the amount of CHD2 in the neuronal cell.

Description

Technical Field

[0001] Cross - reference to Related Applications This application claims the benefit of priority of U.S. Provisional Patent Application No. 63 / 127,212, filed on December 18, 2020, the entire disclosure of which is incorporated herein by reference.

[0002] (Description of the Sequence Listing) An ASCII file named 89180SequenceListing.txt, created on December 19, 2021, containing 61,440 bytes and submitted simultaneously with the filing of this application, is incorporated herein by reference.

[0003] In some embodiments, the present invention relates to compositions for use in treating CHD2 haploinsufficiency and methods for their identification.

Background Art

[0004] The chromodomain helicase DNA - binding protein 2 (Chd2) gene encodes an ATP - dependent chromatin - remodeling enzyme that belongs to sub - family I of the chromodomain helicase DNA - binding (CHD) protein family, together with CHD1. Members of this sub - family are characterized by two chromodomains located in the N - terminal region and a central SNF2 - like ATPase domain [Tajul - Arifin, K., et al. Identification and analysis of chromodomain - containing proteins encoded in the mouse transcriptome. Genome Res. 13, 1416 - 1429 (2003)] and promote nucleosome disassembly, eviction, sliding, and spacing [Narlikar, G. J., Sundaramoorthy, R. & Owen - Hughes, T. M. Mechanisms and functions of ATP - dependent chromatin - remodeling enzymes. Cell 154, 490 - 503 (2013)].

[0005] In humans, CHD2 haploinsufficiency is associated with neurodevelopmental delay, intellectual disability, epilepsy, and behavioral problems [reviewed in Lamar, K.-MJ & Carvill, GL Chromatin remodeling proteins in epilepsy: lessons from CHD2-associated epilepsy. Front. Mol. Neurosci. 11, 208 (2018)]. Studies in mouse models and cell lines have also shown that Chd2 is involved in neuronal dysfunction.

[0006] In all the cases described, these individuals are haploinsufficient with respect to CHD2 and therefore possess an intact WT copy of CHD2. Thus, increasing CHD2 expression via Chaserr disruption, for example by using antisense oligonucleotides, may have therapeutic benefits.

[0007] Multiple lines of evidence indicate a strong correlation between the function of long non-coding RNAs (lncRNAs) and the function of chromatin modification complexes [Han, P. & Chang, C.-P. Long non-coding RNA and chromatin remodeling. RNA Biol. 12, 1094-1098 (2015)]. Numerous chromatin modifying factors have been reported to interact with lncRNAs [Han et al., see above]. Furthermore, lncRNAs in vertebrate genomes are enriched near genes encoding transcription-related factors, including numerous chromatin-related proteins [Ulitsky, I., Shkumatava, A., Jan, CH, Sive, H. & Bartel, DP Conserved function of lincRNAs in vertebrate embryonic development despite rapid sequence evolution. Cell 147, 1537-1550 (2011)], but the functions of most of these lncRNAs remain unknown.

[0008] Previous research by the inventors of this invention has revealed the existence of Chaserr, a conserved lncRNA located upstream of Chd2 (Rom et al. Nature Communications 2019 10:5092):1810026B05Rik (referred to as Chaserr, a CHD2 adjacent, suppressive regulatory RNA) in mice and LINC01578 / LOC100507217 (CHASERR) in humans. Chaserr is an lncRNA that is found upstream of the same strand as Chd2, is transcribed from the same strand as Chd2, and has not been characterized at all.

[0009] Chaserr works in coordination with the CHD2 protein to maintain appropriate Chd2 expression levels. Loss of Chaserr in mice results in early postnatal lethality in homozygous mice and severe growth retardation in heterozygous mice. Mechanistically, Chaserr loss substantially increases the levels of Chd2 mRNA and protein, leading to transcriptional interference by inhibiting promoters found downstream of highly expressed genes. Chaserr production suppresses Chd2 expression only in the cis state and rescues the phenotypic consequences of Chaserr loss even when Chd2 is disrupted. Therefore, targeting Chaserr may be a strategy to increase CHD2 levels in haploinsufficient individuals.

[0010] Additional background technologies include the following: www(dot)iscb(dot)org / cms_addon / conferences / ismb2020 / postersdotphp?track=RegSys%20COSI&session=Bgithub(dot)com / lncLOOM / lncLOOM [Overview of the Initiative]

[0011] According to one aspect of several embodiments of the present invention, a method is provided for increasing the amount of chromodomain helicase DNA-binding protein 2 (CHD2) in nerve cells, comprising introducing a nucleic acid agent into cells that downregulates the activity or expression of human Chaserr, wherein the nucleic acid agent is directed to the terminal exon of human Chaserr, thereby increasing the amount of CHD2 in nerve cells.

[0012] According to one aspect of several embodiments of the present invention, a method for treating a disease or condition related to chromodomain helicase DNA-binding protein 2 (CHD2) haploinsufficiency in a subject requiring treatment comprises administering to the subject a nucleic acid agent that downregulates the activity or expression of human Chaserr in a therapeutically effective amount, wherein the nucleic acid agent is directed to the terminal exon of human Chaserr, thereby providing a method for treating a disease or condition related to CHD2 haploinsufficiency.

[0013] According to one aspect of several embodiments of the present invention, a nucleic acid agent is provided for use in the treatment of a disease or condition related to chromodomain helicase DNA-binding protein 2 (CHD2) haploinsufficiency in subjects requiring treatment of said disease or condition, the nucleic acid agent being directed to the terminal exon of human Chaserr.

[0014] According to some embodiments of the present invention, the human Chaserr includes alternatively spliced ​​variants selected from the group consisting of SEQ ID NO: 11 (NR_037600), SEQ ID NO: 12 (NR_037601), and SEQ ID NO: 13 (NR_037602).

[0015] According to some embodiments of the present invention, the nucleic acid agent hybridizes to a nucleic acid sequence element containing sequence number 2 (AUGG).

[0016] According to some embodiments of the present invention, the nucleic acid agent hybridizes to nucleic acid sequence elements selected from the group consisting of AAGAUG (SEQ ID NO: 5) and AAAUGGA (SEQ ID NO: 6).

[0017] According to some embodiments of the present invention, the nucleic acid agent hybridizes to nucleic acid sequence elements including AAGAUG (SEQ ID NO: 5) and / or AAAUGGA (SEQ ID NO: 6).

[0018] According to some embodiments of the present invention, the nucleic acid agent inhibits the binding of DHX36 to Chaserr.

[0019] According to some embodiments of the present invention, the nucleic acid agent is an antisense oligonucleotide.

[0020] According to some embodiments of the present invention, the antisense oligonucleotide has the nucleic acid base sequence described in SEQ ID NOs: 92-99 (where T is substituted with U).

[0021] According to some embodiments of the present invention, the nucleic acid agent is an RNA silencing agent.

[0022] According to some embodiments of the present invention, the nucleic acid agent is a genome editing agent.

[0023] According to some embodiments of the present invention, the nucleic acid agent is inducibly active.

[0024] According to some embodiments of the present invention, the nucleic acid agent is active in a tissue-specific or cell-specific manner.

[0025] According to some embodiments of the present invention, the disease or condition associated with chromodomain helicase DNA-binding protein 2 (CHD2) haploinsufficiency is selected from the group consisting of intellectual disability, autism, epilepsy, and Lennox-Gastaut syndrome (LGS).

[0026] According to one aspect of several embodiments of the present invention, a method for analyzing a set of sequences describing a plurality of homologous polynucleotides, Construct a graph having multiple nodes arranged in layers and multiple edges connecting nodes in consecutive layers, where the first layer represents a sequence describing a query polynucleotide, each node represents a k-mer within that sequence, each edge connects nodes representing the same or homologous k-mers, and each layer represents a sequence of the set such that k is between 6 and 12; Searching the graph by finding continuous non-crossing paths along the edges of the graph; and A method is provided which includes generating an output that identifies a kmer corresponding to at least one pathway as a nucleic acid sequence of interest in function.

[0027] According to some embodiments of the present invention, the method includes iteratively repeating the construction and the search for k-mers that are shorter each time, before generating the output.

[0028] According to some embodiments of the present invention, the method includes applying, in each iterative cycle, the path obtained in the previous iterative cycle as a constraint for the search.

[0029] According to some embodiments of the present invention, the search includes applying a criterion based on path depth as a constraint for the search, such that the search is prioritized over shallower paths.

[0030] According to some embodiments of the present invention, the search involves applying an integer linear programming (ILP) method to the graph.

[0031] According to some embodiments of the present invention, the homologous polynucleotides are DNA sequences.

[0032] According to some embodiments of the present invention, the homologous polynucleotide is an RNA sequence.

[0033] According to some embodiments of the present invention, the method comprises aligning the sequences in the set in a predetermined order to provide a multiple alignment having a plurality of alignment layers, wherein the first layer is the query polynucleotide from the plurality of homologous polynucleotides, and the plurality of alignment layers correspond to each layer of the graph.

[0034] According to some embodiments of the present invention, the predetermined order is evolution-dictated, and optionally, the query is the most evolved among the homologous polynucleotides.

[0035] According to some embodiments of the present invention, the homology between the homologous k-mers is 70% or more.

[0036] According to some embodiments of the present invention, the homologous polynucleotides include a partial sequence.

[0037] According to some embodiments of the present invention, the homologous polynucleotide is selected from the group consisting of 3'UTR, lncRNA, and enhancer.

[0038] Unless otherwise defined, all technical and / or scientific terms used herein have the same meaning as those generally understood by those skilled in the art to which the present invention pertains. Similar or equivalent methods and materials to those described herein may be used in carrying out or testing embodiments of the present invention, but exemplary methods and / or materials are described below. In case of any conflict, the patent specification, including the definitions, shall prevail. Furthermore, materials, methods and examples are illustrative and not necessarily intended to limit the scope.

[0039] Some embodiments of the present invention are described herein merely as examples with reference to the accompanying drawings. With further detailed and specific reference to the drawings, it is emphasized that the details shown are illustrative and for illustrative purposes only, for the purpose of illustrating embodiments of the present invention. In this regard, the description made with reference to the drawings will make it clear to those skilled in the art how embodiments of the present invention may be carried out. [Brief explanation of the drawing]

[0040] [Figure 1A] Figures 1A and 1B provide an overview of an embodiment for discovering nucleic acid sequence elements called the "LncLOOM" framework. (A) Overview of the LncLOOM method. LncLOOM processes an ordered list of sequences and recovers a set of ordered motifs conserved to various depths that can be further annotated as miRNA binding sites or RBP binding sites. [Figure 1B] (B) Schematic diagram of graph construction and motif discovery using integer linear programming (ILP) to find long non-intersecting paths. Sequences are ordered such that the evolutionary distance from the top layer (human) increases monotonically. BLAST high-scoring pairs (HSPs) (see "Method") that can be used to constrain the placement of edges are shown as pink and red blocks below each sequence. The graph is used to construct the ILP problem, and its solution is used to construct a set of long paths corresponding to the conserved synteny motif (sequences 29-32). [Figure 2A] Figures 2A-F show the discovery of conserved elements in Cyrano lncRNA. (A) Overview of the genomic composition of Cyrano exons in selected species. [Figure 2B] (B) Sequence elements identified by LncLOOM as being conserved in at least 17 Cyrano sequences. Regions containing elements found in areas alignable by BLAST between human Cyrano sequences and zebrafish Cyrano sequences are circled. The numbers between elements indicate the range distance between the 18 elements. The circled numbers above each element indicate the element number used in the text and other panels. [Figure 2C] (C) Pairing of predicted binding elements in Cyrano with miR-25 / 92 miRNA and miR-7 miRNA. [Figure 2D] (D) Evidence of binding of PUM1 and PUM2 to the UGUAUAG motif (shaded region) in the human genome. ENCODE project CLIP data (top, K562 cells) and 22 (bottom, HCT116 cells). Shading is based on the strength of binding evidence as defined by the ENCODE project. [Figure 2E] (E) Binding and regulation of mouse Cyrano sequences by Pum1 / 2 and Rbfox1 / 2. Top: Pum1 / 2 CLIP and RNA-seq data from mouse brain and mESCs. Center: Rbfox1 CLIP from mouse brain and mESCs. Binding motifs of Pumilio and Rbfox are highlighted in yellow and blue, respectively. PhyloP sequence conservation scores are from the UCSC genome browser. Bottom: Binding of Ago2 in mouse brain to the miR-153 binding site region near the 3' end of Cyrano. [Figure 2F-1]CLIP data from (F). Top left: Alignment of the region surrounding the conserved AUGGCG motif near the 5' end of Cyrano. Top right and bottom: Composite Ribo-seq and RNA-seq data from multiple curated datasets. Chip-seq data of YY1 in the K562 cell line from the ENCODE project. Readout coverage and IDR peaks are shown. Sequences shown in the panel are marked as SEQ ID NOs. 33-42 and 53-67. [Figure 2F-2] Same as above. [Figure 3A] Figures 3A-E show the discovery of conserved elements in CHASERR lncRNA. (A) The human CHASERR gene structure is shown along with at least four conserved motifs, color-coded according to their degree of conservation. The final exon region is enlarged, and the motifs described in the text are highlighted. [Figure 3B] (B) An array logo of the two most preserved motifs adjacent to each other, with the shared AARAUGR motif shaded (the array shown on the panel is marked as sequence number 68). [Figure 3C] (C) Top: Mouse Chaserr locus highlighting the positions of primer pairs used in qRT-PCR and the regions targeted by GapmeR (the same as used for ASO) and ASO. Bottom: qRT-PCR using primers targeting Chaserr exons (shown at the top) or Chd2 exons in N2a cells treated with the indicated reagents. N=4 for ASO treatment, and N=5 for GapmeR. [Figure 3D] (D) Volcano plot for comparison of MS intensity between a pulldown with the WT sequence of the Chaserr final exon and the final exon with a mutated conserved element (Figure 8A). [Figure 3E] (E) qRT-PCR using primers targeting the indicated region after IP with the indicated antibody, n=4. Top right: Western blot using anti-DHX36 antibody on the indicated sample. The sequence shown in the figure is marked as SEQ ID NO: 68. [Figure 4-1] Figure 4 shows the identification of conserved elements in the 3'UTR of PUM1 and PUM2. Human sequences are shown, and motifs conserved in at least seven species are color-coded based on their conservation. Occurrences of the superconserved UGUACAUU (sequence number 14) motif are shown in a box. Sequences shown in the panel are marked as sequence numbers 69-70. [Figure 4-2] Same as above. [Figure 5A] Figures 5A-I show a global analysis of conserved motifs in the 3'UTR using LncLOOM. (A) Number of genes with varying numbers of orthologous sequences that did not have significant alignment with human sequences (black) or mouse, dog, and chicken sequences (gray). [Figure 5B] (B) Distribution of conserved unique k-mer combinations within the indicated number of sequences that were not aligned to human 3'UTR sequences. [Figure 5C] (C) Total number of intrinsic kmers (pink) identified by LncLOOM per species and quantification of all instances (dark red). The total number of widely conserved miRNA binding sites is shown in green, and the number of intrinsic kmers corresponding to these sites is shown in yellow. The number of genes containing any of the kmers is shown in gray, and the number of genes containing at least one kmer corresponding to a miRNA site is shown in black. [Figure 5D] (D) Top: Distribution of unique kmers identified in the first sequence that cannot be aligned to humans in multiple genes (gray). The number of kmers detected in invertebrate species in at least one gene is shown in black. Bottom: Unique kmers common to at least 50 genes and detected in invertebrate sequences. ARE-like kmers are colored red, PAS-like kmers are colored blue, and PRE-like kmers are colored green. [Figure 5E] (E) Comparison of genes containing widely conserved miRNA binding sites detected by LncLOOM and TargetScan within the human sequences of the analyzed genes. [Figure 5F](F) Number of widely conserved miRNA bindings detected by LncLOOM / number of sequences that cannot be aligned; percentage of genes with detected miRNA sites / number of layers that cannot be aligned (black); and number of unique kmers corresponding to the miRNA binding sites (yellow). [Figure 5G] (G) Top: Widely conserved miRNA binding sites in human sequences, as predicted by LncLOOM. Sites predicted by TargetScan and recovered by LncLOOM are shown in red, and new sites are shown in blue. Bottom: Conservation / species count of these sites. [Figure 5H] (H) Comparison of fractions of genes containing at least one miRNA site detected in the indicated species by TargetScan and LncLOOM. Only sites found in TargetScanHuman were used. [Figure 5I] (I) Percentage of genes containing miRNA sites detected by LncLOOM / number of sequences that cannot be aligned: (Red) miRNA sites previously predicted by TargetScan in human sequences and recovered by LncLOOM in additional sequences that are not part of the MSA used by TargetScan; (Blue) Novel miRNA sites predicted by LncLOOM but not previously predicted by TargetScan in human sequences. [Figure 6] Figure 6 shows elements conserved within libra lncRNA. Human sequences are shown, and motifs conserved in at least five species are color-coded based on their conservation. Pairs of vertical lines indicate the location of introns. Motifs matching miRNA seed sites are indicated by the miRNA family name above the motif. Regions that are part of the BLASTN alignment (E<0.001) between the human sequence and the spotted gar sequence are underlined. Sequences shown in the panel are marked as Sequence ID No. 71. [Figure 7]Figure 7 shows the gap in the genome assembly around the first exon of the Chaserr lncRNA locus. For each species, RNA-seq readout coverage is shown alongside the gap in the genome assembly (from UCSC browser). [Figure 8A] Figures 8A-D show the functional characterization of conserved elements in Chaserr lncRNA. (A) Sequence of the final exon of mouse Chaserr. Deeply conserved elements are shared. Conserved AUGG mutations in MS bait are shown in blue, and all other AUGG mutations are shown in green. Regions targeted by ASO are marked. [Figure 8B] (B) The ASO process shown is the same as in Figure 3C. [Figure 8C] (C) RNA-seq quantification of the expression of the indicated gene in HEK293 cells having the indicated genotype. [Figure 8D] (D) Data from RNA-seq quantification of the expression of the indicated gene in THP1 cells treated with non-targeted shRNA (shNT) or ZFR-targeted shRNA. The data from the sequence shown in 8A is marked as Sequence ID 72. [Figure 9-1] Figure 9 shows the identification of conserved elements in DICER 3'UTR. Human sequences are shown, and motifs conserved in at least eight vertebrate species are color-coded based on their conservation (9 species - conserved in lancelets; 10 species - conserved in lancelets and sea urchins). Regions of motifs that do not contain a motif of this length within 100 random sequences that maintain sequence identity are shaded in light yellow. Regions of motifs for which the exact motif is not found within the random sequences are shaded in light cyan. Sequences shown in the panel are marked as sequence number 73. [Figure 9-2] Same as above. [Figure 10A]Figures 10A–F show additional analysis of LncLOOM motifs identified within the 3'UTR. (A) Distribution of orthologous 3'UTR sequences. Top left: Frequencies of genes analyzed at various depths. Top right: Distribution of various combinations of non-amniotic sequences included in the 3'UTR sequence dataset. Bottom right: Total number of genes analyzed in the indicated species. [Figure 10B] (B) Distribution of the number of conserved unique k-mer combinations / unalignable sequences in the 3'UTR dataset. Alignment to humans, mice, dogs, and chickens was examined. [Figure 10C] (C) Distribution of unique kmers identified across amniotes and shared among multiple genes. Shows the number of kmers containing UUU (red line), AUAA (green line), or corresponding to widely conserved miRNA sites (yellow line). [Figure 10D] (D) Conservation of widely conserved miRNA sites detected by LncLOOM in genes in which TargetScan did not report any predictions. (Top) Number of genes / species containing detected miRNA sites (left), and number of sequences that cannot be aligned (right). (Bottom left) Number of genes / species containing detected miRNA sites. (Center) Number of new miRNA sites / species detected. (Right) Number of new miRNA sites / number of sequences that cannot be aligned. [Figure 10E] (E) Comparison of conserved miRNA sites detected per species by TargetScan and LncLOOM. Only sites previously identified by TargetScanHuman were compared. [Figure 10F] (F) Conservation of miRNA sites detected by LncLOOM in sequences that do not have alignment with human sequences. In human sequences, sites previously predicted by TargetScan are colored red, and new LncLOOM predictions are colored blue. [Figure 11A] Figures 11A-D show the constraints imposed on the LncLOOM graph. (A) Examples of LncLOOM graph scenarios and how they are represented in ILP. [Figure 11B](B) Conditional constraints on intersecting edges. An example of the near-optimal exclusion of repeated k-mers in complex paths during refinement in subsequent iterations, which can occur when all intersections are constrained. [Figure 11C] (C) Flowchart for defining conditional constraints on intersecting edges: A pair of intersecting edges is constrained only if there is at least one other edge from an intrinsic path that intersects either edge. [Figure 11D] (D) An example showing how a conditional constraint on crossing can relax the suboptimal exclusion of tandem repeating k-mers. The sequence shown in the panel is marked as sequence number 74. [Figure 12-1] Figure 12 shows the partitioning of the LncLOOM graph and the iterative refinement of selected repeating k-mers. Motif discovery is performed by an iterative process that starts from the deepest layer in the graph and searches for conserved motifs at progressively shallower depths. Here is an example of motif discovery starting with a 5-layer graph. The graph is solved, and the simple paths obtained in the solution (shown in green) are used to partition the graph into subgraphs that are individually solved in the next iteration, which is performed on the top 4 layers of the graph. Each simple path is immediately added to the final solution, while the complex paths (shown in blue and red) are refined during subsequent iterations of motif discovery. In this case, repeating k-mers that are removed during optimization are outlined in pink. [Figure 12-2] Same as above. [Figure 13A]Figures 13A-B illustrate the processing steps in the LncLOOM framework. (A) Construction of 5' and 3' graphs. LncLOOM predicts and extracts the 5' and 3' ends of individual sequences that have been stretched compared to other sequences in the graph, using the central positions of the first and final motifs identified in the primary ILP (where the full length of each sequence is considered). LncLOOM motif discovery is then performed on a subset of the extracted 5' and 3' regions. In this example, because a minimum depth of 3 is imposed, the AUUGCU (sequence number 15, blue) motif, which is conserved in only the top two sequences, is ignored, and CAUCCA (sequence number 16, dark red and underlined) is considered as the first node instead. [Figure 13B] (B) Illustration of motif neighborhoods. The reference sequence for each neighborhood is determined by combining all overlapping k-mers in the anchor sequence. Then, all k-mers stored at each depth in the graph and linked to one of the overlapping k-mers in the reference sequence are included in the neighborhood. The sequences shown in the panel are marked as sequence numbers 75-87. [Figure 14] Figure 14 is a flowchart illustrating a method suitable for analyzing a set of sequences, according to various exemplary embodiments of the present invention. [Figure 15] Figure 15 is a schematic diagram of a computing platform configured to analyze a set of sequences, according to various exemplary embodiments of the present invention. [Figure 16] Figure 16 is a graph showing the changes in gene expression of CHASERR, CHD2, and p21 (CDKN1A) after transfection with the indicated ASO (SEQ ID NOs. 128 and 134) compared to untransfected SH-SY5Y cells. [Figure 17] Figure 17 is a graph showing the changes in gene expression of CHASERR and CHD2 after transfection with the indicated ASOs (SEQ ID NOs. 128 and 134) compared to untransfected MCF7 and SH-SY5Y cells. [Modes for carrying out the invention]

[0041] In some embodiments, the present invention relates to compositions for use in the treatment of CHD2 haploinsufficiency, and methods for identifying the same.

[0042] Before describing in detail at least one embodiment of the present invention, it should be understood that the present invention is not necessarily limited in its application to the details described below or illustrated by the examples. Other embodiments of the present invention are possible, or it can be carried out or performed in various ways.

[0043] CHD2 haploinsufficiency is associated with neurodevelopmental delay, intellectual disability, epilepsy, and behavioral problems. Previous results have shown that CHD2 expression is tightly regulated by Chaserr, a conserved lncRNA located upstream of Chd2. Loss of Chaserr substantially increases the levels of Chd2 mRNA and protein, leading to alterations in gene expression, including transcriptional interference by inhibiting promoters found downstream of highly expressed genes.

[0044] During the conception of embodiments of the present invention, the inventors devised a novel algorithm for detecting conserved elements within a sequence that diverge beyond alignment possibilities and / or accumulate substantial lineage-specific sequences, such as transposable elements. Using this algorithm, or its embodiments referred to as "LncLOOM," the inventors identified and validated conserved regions of Chaserr that can be preferentially mutated / targeted to specifically inhibit the interaction of Chaserr with functionally relevant interactors and ultimately compensate for CHD2 haploinsufficiency.

[0045] Accordingly, according to one aspect of the present invention, a method is provided for increasing the amount of chromodomain helicase DNA-binding protein 2 (CHD2) in nerve cells, comprising introducing a nucleic acid agent into cells that downregulates the activity or expression of human Chaserr, wherein the nucleic acid agent is directed to the terminal exon of human Chaserr, thereby increasing the amount of CHD2 in nerve cells.

[0046] As used herein, “nucleic acid agent that downregulates the activity or expression of human Chaserr” refers to a nucleic acid molecule that inhibits the activity of human Chaserr or reduces the amount of human Chaserr.

[0047] According to some embodiments, the "nucleic acid agent for downregulating human Chaserr activity" includes one or more of the following: a nucleic acid agent that increases the expression of CHD2 (protein and optionally mRNA); a nucleic acid agent that increases the stability of CHD2 mRNA; a nucleic acid agent that induces the expression of CHD2 mRNA; and a nucleic acid agent that induces the translation of CHD2.

[0048] Accordingly, according to one aspect of the present invention, a nucleic acid agent is provided that activates or downregulates human Chaserr, and comprises a nucleic acid sequence that hybridizes to the final exon of human Chaserr (i.e., is complementary to the nucleotide sequence in the final exon of human Chaserr).

[0049] As used herein, “chromodomain helicase DNA-binding protein 2 (CHD2)” refers to the enzyme encoded by the CHD2 gene in humans. Examples of human CHD2 splice variants include NCBI reference sequences: NM_001271.4 and NM_001042572.

[0050] The splice mutant protein product is as described in the NCBI reference sequence: NP_001262.3 or NP_001036037.

[0051] As used herein, “haploinsufficiency” refers to a model of dominant gene action in diploid organisms, in which a single copy of the standard (so-called wild-type) allele at a locus, heterozygously paired with a mutant allele, is insufficient to produce a standard phenotype. Typically, only about half the amount of protein is produced compared to a healthy state where both alleles are wild-type.

[0052] As used herein, “increase the amount” means increasing the amount of the protein or RNA of interest by a statistically significant amount and by an amount that is useful for treating the haploinsufficiency of the protein or RNA of interest. In various embodiments, “increase the amount” of the protein or RNA of interest involves an increase of at least 10%, or in some embodiments, at least about 20%, at least 20%, 20–150%, 50–150%, for example, at least 50%, 60%, 70%, 80%, 90%, at least 1.2 times, 1.4 times, 1.5 times or more, for example, at least 2 times. According to a particular embodiment, the CHD2 level is restored to the amount found in normal cells (without haploinsufficiency) of the same type (i.e., neuronal) and developmental stage.

[0053] As used herein, “neuronal cells” refers to cells found in vivo or outside the body of a subject, such as tissue biopsy material, cell lines, and primary cultures.

[0054] Other cells, namely non-neuronal cells, are also considered.

[0055] Nerve cells may or may not be genetically modified; for example, they may be naive.

[0056] According to a particular embodiment, nerve cells are located in the central nervous system.

[0057] Methods for identifying cells in which CHD2 levels are modified or modified according to some embodiments of the present invention are well known in the art.

[0058] Contact between cells and the agent may be carried out under either in vivo or in vitro conditions, for example, by adding the agent to cells derived from the subject (e.g., primary cell cultures, cell lines) or to a biological sample containing it (e.g., a fluid, liquid containing cells), so that the agent comes into direct contact with the cells. According to some embodiments of the present invention, the cells of the subject are incubated with the agent. The conditions used to incubate the cells are selected for duration / cell concentration / concentration of the agent / cell-to-agent ratio, etc., which allow the drug to induce cellular changes such as an increase in CHD2 levels (amount) or related changes such as changes in the transcription and / or translation rate of specific genes, proliferation rate, differentiation, cell death, necrosis, apoptosis, etc.

[0059] CHD2 (mRNA and / or protein) levels may be analyzed before, simultaneously with, and / or subsequently to the introduction of the agent into cells. Additionally or alternatively, genomic DNA may be analyzed for modifications introduced by the agent, such as in the case of genome editing, as further described below herein.

[0060] Downregulation at the nucleic acid level (i.e., a reduction in the abundance of nucleic acids) is typically carried out using nucleic acid agents that have a nucleic acid backbone, DNA, RNA, their mimics, or combinations thereof. Nucleic acid agents can be encoded from DNA molecules or provided to the cell itself.

[0061] According to certain embodiments, the downregulation agent is a polynucleotide.

[0062] Nucleic acid agents are understood to be intended in this specification, to be coded from nucleic acid constructs, or as part of a pharmaceutical composition.

[0063] According to a particular embodiment, the downregulation agent is a polynucleotide or oligonucleotide that can hybridize to a gene or mRNA encoding CHD2.

[0064] According to certain embodiments, the downregulation agent directly interacts with the CHD2 gene or its RNA transcript.

[0065] According to a particular embodiment, the agent directly binds to the nucleic acid sequence within the final exon of Chaserr.

[0066] As used herein, "Chaserr" refers to CHD2 flanking repressive regulatory RNA. HGNC:48626 Entrez Gene:100507217

[0067] The exon structure of Chaserr is as follows: EXON1: nucleotides 1..344; EXON2: nucleotides 345..538; EXON3: nucleotides 539...608; EXON4: nucleotides 609...694; EXON5: nucleotides 695...763; EXON6: nucleotides 764...1787. Here, the final exon of Chaserr refers to nucleotide 764..1787 of sequence number 3 (NR_037601).

[0068] According to a particular embodiment, the nucleic acid agent hybridizes to a nucleic acid sequence element containing SEQ ID NO: 1 (AUG).

[0069] According to another embodiment, the nucleic acid agent hybridizes to a nucleic acid sequence element containing sequence number 2 (AUGG).

[0070] According to a particular embodiment, the nucleic acid agent hybridizes to a nucleic acid sequence element containing AAGAUGG (SEQ ID NO: 4), AAGAUG (SEQ ID NO: 5), or AAAUGGA (SEQ ID NO: 6).

[0071] According to another embodiment, the nucleic acid agent hybridizes to a nucleic acid sequence element containing sequence number 3 (aauaaa).

[0072] According to certain embodiments, the nucleic acid agent inhibits the binding of DHX36 to Chaserr.

[0073] As used herein, "DHX36" refers to the likely ATP-dependent RNA helicase DHX36, also known as DEAH box protein 36 (DHX36), MLE-like protein 1 (MLEL1), or G4 lysolbase 1 (G4R1), or the AU-rich element (RHAU)-associated RNA helicase, which is the enzyme encoded by the DHX36 gene in humans.

[0074] According to a particular embodiment, the nucleic acid agent comprises a nucleotide sequence complementary to UUUUUACCU (SEQ ID NO: 122).

[0075] According to certain embodiments, the nucleic acid agent inhibits the binding of CHD2 to Chaserr.

[0076] According to certain embodiments, the downregulation agent is an antisense, RNA silencing agent, or genome editing agent.

[0077] According to certain embodiments, the downregulation agent is antisense.

[0078] Antisense oligonucleotides are single-stranded oligonucleotides designed to hybridize to a target RNA, thereby inhibiting its function or level. Downregulation or inhibition of Chaserr RNA can be performed using antisense oligonucleotides that can specifically hybridize to Chaserr transcripts, for example, SEQ ID NOs: 1, 2, 4, or 6. Preferably, the hybridization of the antisense oligonucleotide prevents the binding of an effector element to Chaserr, but otherwise leaves the Chaserr RNA intact. According to certain embodiments, the nucleic acid agent does not recruit RNaseH.

[0079] In some embodiments, the antisense oligonucleotide does not recruit RNaseH. For example, the antisense oligonucleotide may substantially contain RNA nucleotides. In yet other embodiments, the antisense oligonucleotide recruits RNaseH and therefore contains at least one sequence of DNA nucleotides. For example, the antisense oligonucleotide may be a gapmer.

[0080] In certain embodiments, antisense sequences corresponding to the antisense oligonucleotides (ASOs) exemplified for mice in the Examples section below include, but are not limited to, CCATAGTAGACTGCCATCTT (SEQ ID NO: 7) targeting AAGATGGCAGTCTACTATGG (SEQ ID NO: 12), and ATCCACTGTCCATTTGTG (SEQ ID NO: 9) targeting CACAAATGGACAGTGGAT (SEQ ID NO: 10). For convenience, nucleotide sequences are represented here as complete DNA or RNA sequences, but it is understood that antisense oligonucleotides can be constructed as RNA nucleotides, DNA nucleotides, or mixtures thereof. That is, if an oligonucleotide exhibits a nucleotide thymine (T), it is understood that the nucleotide can be substituted with its RNA counterpart (uridine, i.e., U), and vice versa. Furthermore, it is understood that antisense oligonucleotides can be constructed using DNA nucleotide modifications and RNA nucleotide modifications, such as those well known in the art.

[0081] According to certain embodiments, the nucleic acid agent comprises a nucleotide sequence complementary to UUUUUACCU (SEQ ID NO: 122). As used herein, the term “complementary” refers to canonical (A / T, A / U, and G / C) base pairings.

[0082] According to certain embodiments, the nucleic acid agent inhibits the binding of CHD2 to Chaserr.

[0083] According to a particular embodiment, the antisense oligonucleotide has the nucleic acid base sequence described in SEQ ID NOs: 140-143 (corresponding to A40, 50, 51, and 52). In a modified version thereof, the antisense oligonucleotide is provided as SEQ ID NOs: 128, 131, 132, and 133.

[0084] The design of antisense molecules that can be used to efficiently inhibit or reduce the amount of Chaserr must be carried out while considering two aspects that are important to antisense methods. The first aspect is the delivery of an oligonucleotide to the nucleus of a suitable cell, and the second aspect is the design of an oligonucleotide that specifically binds to a designated RNA in the cell to inhibit the desired function.

[0085] Prior art has taught several delivery strategies that can be used to efficiently deliver oligonucleotides to a wide variety of cell types [e.g., Jaaskelainen et al. Cell Mol Biol Lett. (2002) 7(2):236-7; Gait, Cell Mol Life Sci. (2003) 60(5):844-53; Martino et al. J Biomed Biotechnol. (2009) 2009:410260; Grijalvo et al. Expert Opin Ther Pat. (2014) 24(7):801-19; Falzarano et al, Nucleic Acid Ther. (2014) 24(1):87-100; Shilakari et al. Biomed Res Int. (2014) 2014:526391; Prakash et al. Nucleic Acids See Res.(2014)42(13):8796-807 and Asseline et al. J Gene Med.(2014)16(7-8):157-65.

[0086] Furthermore, algorithms are available to identify sequences with the highest predicted binding affinity to target RNAs based on thermodynamic cycles that explain the energy mechanisms of structural changes in both target RNAs and oligonucleotides [see, e.g., Walton et al. Biotechnol Bioeng 65:1-9 (1999)]. Such algorithms have been successfully used to perform antisense techniques in cells.

[0087] Furthermore, several methods have been published for designing specific oligonucleotides using in vitro systems and predicting their efficiency (Matveeva et al., Nature Biotechnology 16:1374-1375 (1998)).

[0088] For example, suitable antisense oligonucleotides targeting Chaserr RNA are those of the sequences listed in Table 3 below (and considered to be an integral part of this specification), or any of the antisense oligonucleotides described in SEQ ID NOs. 140-143, or any of the antisense oligonucleotides having the modifications described in SEQ ID NOs. 128, 131, 132, or 133, corresponding to A40, 50, 51, and 52.

[0089] According to various embodiments, the antisense oligonucleotide may include a complete RNA nucleotide. Since such an antisense oligonucleotide does not recruit RNaseH, Chaserr should not be degraded by its antisense inhibition. In yet another embodiment, the antisense oligonucleotide includes a mixture of DNA nucleotides and RNA nucleotides (e.g., a gapmer) that can recruit RNaseH and degrade Chaserr RNA.

[0090] In some embodiments, the antisense oligonucleotide includes one or more nucleotides containing 2'-to-4' crosslinks, such as locked nucleotides (LNA) or restricted ethyl (cEt), and other crosslinked nucleotides as described herein.

[0091] In some embodiments, the antisense oligonucleotide comprises one or more (or all in some embodiments) nucleotides having a 2'-O modification, such as 2'-OMe or 2'-O-methoxyethyl (2'-O-MOE).

[0092] In some embodiments, the antisense oligonucleotide includes a modified skeleton such as a phosphorothioate or phosphorodithioate. In yet other embodiments, the antisense oligonucleotide includes a morpholino skeleton.

[0093] In some embodiments, the antisense oligonucleotide comprises one or more nucleotides having a modified base such as 5-methylcytosine.

[0094] Other possible nucleotide modifications are described elsewhere in this specification.

[0095] Alternatively, CHD2 downregulation can be achieved by RNA silencing. As used herein, the term “RNA silencing” refers to a group of RNA-mediated regulatory mechanisms that result in inhibition or “silencing” of RNA activity or availability [e.g., RNA interference (RNAi), transcriptional gene silencing (TGS), post-transcriptional gene silencing (PTGS), querying, and co-repression]. RNA silencing has been observed in many types of organisms, including plants, animals, and fungi.

[0096] As used herein, the term “RNA silencing agent” refers to RNA that can specifically inhibit or “silence” the expression of a target gene. In certain embodiments, RNA silencing agents can prevent the complete processing of mRNA molecules (e.g., complete translation and / or expression) via a post-transcriptional silencing mechanism. RNA silencing agents include non-coding RNA molecules, such as RNA double helix containing paired strands, and precursor RNAs that can produce such small non-coding RNAs. Exemplary RNA silencing agents include siRNA, miRNA, and dsRNA such as shRNA.

[0097] In one embodiment, the RNA silencing agent can induce RNA interference.

[0098] According to one embodiment of the present invention, the RNA silencing agent is specific to a target RNA and comprises the final exon of Chaserr (the following element: e.g., as described herein using SEQ ID NOs: 1, 2, 4, or 6), and is actually specific to a nucleic acid region that exhibits less than 99% overall homology to the target gene, as determined by PCR, Western blot, immunohistochemical analysis and / or flow cytometry, e.g., less than 98%, less than 97%, less than 96%, less than 95%, less than 94%, less than 93%, less than 92%, less than 91%, less than 90%, less than 89%, less than 88%, less than 87%, less than 86%, less than 85%, less than 84%, less than 83%, less than 82%, less than 81% overall homology to the target gene (or other exons within the same target).

[0099] RNA interference refers to the process of sequence-specific post-transcriptional gene silencing in animals, mediated by small interfering RNAs (siRNAs).

[0100] The following is a detailed description of RNA silencing agents that may be used according to specific embodiments of the present invention.

[0101] dsRNA, siRNA, and shRNA - The presence of long dsRNAs within cells stimulates the activity of a ribonuclease III enzyme called Dicer. Dicer is involved in the processing of dsRNA into shorter fragments known as small interfering RNAs (siRNAs). Small interfering RNAs derived from Dicer activity are typically about 21–23 nucleotides long and contain a double helix of about 19 base pairs. The RNAi response also features an endonuclease complex commonly called the RNA-induced silencing complex (RISC), which mediates the cleavage of single-stranded RNA having a sequence complementary to the antisense strand of the siRNA double helix. The cleavage of the target RNA occurs in the middle of the region complementary to the antisense strand of the siRNA double helix.

[0102] Therefore, some embodiments of the present invention intend to use dsRNA for downregulating protein expression from mRNA.

[0103] In one embodiment, dsRNAs longer than 30 bp are used. Various studies have demonstrated that long dsRNAs can be used to silence gene expression without inducing a stress response or causing significant off-target effects. See, for example, [Strat et al., Nucleic Acids Research, 2006, Vol.34, No.13 3803-3810; Bhargava A et al. Brain Res. Protoc. 2004; 13:115-125; Diallo M., et al., Oligonucleotides. 2003; 13:381-392; Paddison PJ, et al., Proc. Natl Acad. Sci. USA. 2002; 99:1443-1448; Tran N., et al., FEBS Lett. 2004; 573:127-134].

[0104] According to some embodiments of the present invention, dsRNA is delivered into cells in which the interferon pathway is not activated. See, for example, Billy et al., PNAS 2001, Vol 98, pages 14428-14433 and Diallo et al, Oligonucleotides, October 1, 2003, 13(5):381-392. doi:10.1089 / 154545703322617069.

[0105] According to one embodiment of the present invention, long dsRNAs are specifically designed not to induce the interferon and PKR pathways for downregulating gene expression. For example, Shinagwa and Ishii [Genes&Dev. 17(11):1340-1345, 2003] have developed a vector called pDECAP for expressing long double-stranded RNA from an RNA polymerase II (Pol II) promoter. Long dsRNAs from pDECAP do not induce an interferon response because transcripts from pDECAP lack both the 5'-cap structure and the 3'-poly(A) tail that facilitate the transport of ds-RNA into the cytoplasm.

[0106] Another method to circumvent the interferon and PKR pathways in mammalian systems is through the introduction of small inhibitory RNAs (siRNAs) via either transfection or endogenous expression.

[0107] The term "siRNA" refers to a small inhibitory RNA double helix (generally 18-30 base pairs) that induces the RNA interference (RNAi) pathway. Typically, siRNA is chemically synthesized as a 21-mer with a central 19-bp double helix region and symmetrical 2-base 3' overhangs at the ends. However, it has recently been reported that chemically synthesized RNA double helixes of 25-30 base pairs can exhibit up to a 100-fold increase in potency compared to 21-mers at the same position. The observed increase in potency obtained by using longer RNA during RNAi induction is attributed to providing the substrate (27-mer) instead of the product (21-mer) to the dicer, and this is suggested to improve the rate or efficiency of siRNA double helix entry into RISC.

[0108] The location of the 3'-overhang affects the potency of siRNA, and asymmetric double helix with the 3'-overhang on the antisense strand is generally found to be more potent than those with the 3'-overhang on the sense strand (Rose et al., 2005). This may be due to asymmetric strand loading on RISC, given that the opposite potency pattern is observed when targeting antisense transcripts.

[0109] The strands of double-stranded interfering RNA (e.g., siRNA) can be ligated to form a hairpin or stem-loop structure (e.g., shRNA). Thus, as mentioned, the RNA silencing agent in some embodiments of the present invention may also be short hairpin RNA (shRNA).

[0110] As used herein, the term “shRNA” refers to an RNA agent having a stem-loop structure containing first and second regions of complementary sequences, the degree of complementarity and orientation of the regions being sufficient for base pairing to occur between the regions, the first and second regions being linked by a loop region, the loop arising from the lack of base pairing between nucleotides (or nucleotide analogs) within the loop region. The number of nucleotides in the loop is 3–23, or 5–15, or 7–13, or 4–9, or 9–11. Some of the nucleotides in the loop may be involved in base pairing interactions with other nucleotides within the loop. Examples of oligonucleotide sequences that can be used to form loops are listed in International Patent Application Nos. WO2013126963 and WO2014107763. It will be recognized by those skilled in the art that the resulting single-stranded oligonucleotides form a stem-loop or hairpin structure containing a double-stranded region capable of interacting with the RNAi mechanism.

[0111] The synthesis of RNA silencing agents suitable for use with some embodiments of the present invention can be carried out as follows: First, the Chaserr mRNA sequence is scanned for AA dinucleotide sequences. The presence of each AA and the 3' adjacent 19 nucleotides is recorded as a potential siRNA target site.

[0112] Secondly, potential target sites are compared to appropriate genome databases (e.g., human, mouse, rat, etc.) using sequence alignment software such as BLAST software available from the NCBI server (www(dot)ncbi.nlm.nih(dot)gov / BLAST / ).

[0113] A suitable target sequence is selected as a template for siRNA synthesis. Preferred sequences are those with a low G / C content, as these have been shown to be more effective in mediating gene silencing than sequences with a G / C content of more than 55%. It is preferable that several target sites be selected along the length of the target gene for evaluation. For a better evaluation of the selected siRNA, it is preferable to use a negative control. The negative control siRNA preferably has the same nucleotide composition as the siRNA but lacks significant homology to the genome. Therefore, if it does not show any significant homology to any other gene, it is preferable to use a scrambled nucleotide sequence of the siRNA.

[0114] It will be understood that the RNA silencing agents of some embodiments of the present invention are not limited to molecules containing only RNA, as mentioned above herein, but further encompass chemically modified nucleotides and non-nucleotides.

[0115] miRNA and miRNA mimetics - According to another embodiment, the RNA silencing agent may be miRNA.

[0116] The terms "microRNA," "miRNA," and "miR" are synonymous and refer to a collection of non-coding single-stranded RNA molecules, approximately 19-28 nucleotides in length, that regulate gene expression. miRNAs are found in a wide range of organisms (viruses, humans, etc.) and have been shown to play some role in development, homeostasis, and disease pathogenesis.

[0117] The preparation of miRNA mimes can be carried out by any method known in the art, such as chemical synthesis or recombinant methods.

[0118] From the descriptions provided above in this specification, it will be understood that contacting cells with miRNA can be carried out, for example, by transfecting cells with mature double-stranded miRNA, pre-miRNA, or pri-miRNA.

[0119] In this specification, nucleic acid sequence modifications are also intended to improve bioavailability, affinity, stability, or a combination thereof.

[0120] According to one embodiment, the nucleic acid agent comprises at least one base (e.g., nucleic acid base) modification or substitution.

[0121] As used herein, “unmodified” or “natural” bases include the purine bases adenine (A) and guanine (G), and the pyrimidine bases thymine (T), cytosine (C), and uracil (U). “Modified” bases include, but are not limited to, other synthetic and natural bases, such as 5-methylcytosine (5-me-C); 5-hydroxymethylcytosine; xanthine; hypoxanthine; 2-aminoadenine; 6-methyl and other alkyl derivatives of adenine and guanine; 2-propyl and other alkyl derivatives of adenine and guanine; 2-thiouracil, 2-thiothymine, and 2-thiocytosine; 5-halouracil and cytosine; 5-propynyluracil and cytosine; 6-azouracil, cytosine This includes tosine and thymine; 5-uracil (pseudouracil); 4-thiouracil; 8-halo, 8-amino, 8-thiol, 8-thioalkyl, 8-hydroxyl and other 8-substituted adenines and guanines; 5-halo, especially 5-bromo, 5-trifluoromethyl and other 5-substituted uracils and cytosine; 7-methylguanine and 7-methyladenine; 8-azaguanine and 8-azaadenine; 7-deazaguanine and 7-deazaadenine; and 3-deazaguanine and 3-deazaadenine. Additional modified bases include those disclosed in U.S. Patent No. 3,687,808; Kroschwitz, J.I., ed. (1990), “The Concise Encyclopedia Of Polymer Science And Engineering,” pages 858-859; John Wiley & Sons; Englisch et al. (1991), “Angewandte Chemie,” International Edition, 30, 613; and Sanghvi, Y.S., “Antisense Research and Applications,” Chapter 15, pages 289-302, S. Crooke and B. Lebleu, eds., CRC Press, 1993. Such modified bases are particularly useful for increasing the binding affinity of the oligomeric compounds of the present invention.These include 5-substituted pyrimidines containing 2-aminopropyladenine, 5-propynyluracil, and 5-propynylcytosine, 6-azapyrimidines, and N-2, N-6, and O-6-substituted purines. 5-methylcytosine substitution has been shown to increase the stability of nucleic acid double helix by 0.6–1.2°C (Sanghvi, Y. Set al. (1993), “Antisense Research and Applications,” pages 276–278, CRC Press, Boca Raton), and is currently a preferred base substitution, and is even more particularly preferred when combined with 2'-O-methoxyethyl sugar modification. Additional base modifications are described in Deleavey and Damha, Chemistry and Biology (2012) 19:937–954, which are incorporated herein by reference.

[0122] According to one embodiment, the modification is in the backbone (i.e., in the internucleotide bonds and / or sugar portion).

[0123] The sugar modification of nucleic acid molecules is widely described in the art (see PCT International Publication Numbers WO92 / 07065, WO93 / 15187, WO98 / 13526 and WO97 / 26270, all of which are incorporated herein by reference; U.S. Patent No. 5,334,711; U.S. Patent No. 5,716,824; and U.S. Patent No. 5,627,053; Perrault et al., 1990; Pieken et al., 1991; Usman & Cedergren, 1992; Beigelman et al., 1995; Karpeisky et al., 1998; Earnshaw & Gait, 1998; Verma & Eckstein, 1998; Burlina et al., 1997). Such publications describe general methods and strategies for determining the site of incorporation of sugar, base, and / or phosphate modifications into nucleic acid molecules without modulating catalysis. Exemplary sugar modifications include, but are not limited to, 2'-modified nucleotides, such as 2'-deoxy, 2'-fluoro (2'-F), 2'-deoxy-2'-fluoro, 2'-O-methyl (2'-O-Me), 2'-O-methoxyethyl (2'-O-MOE), 2'-O-aminopropyl (2'-O-AP), 2'-O-dimethylaminoethyl (2'-O-DMAOE), 2'-O-dimethylaminopropyl (2'-O-DMAP), 2'-O-dimethylaminoethyloxyethyl (2'-O-DMAEOE), 2'-fluoroarabinooligonucleotide (2'-F-ANA), 2'-O-N-methylacetamide (2'-O-NMA), 2'-NH2, or locked nucleic acid (LNA). Additional sugar modifications are described in Deleavey and Damha, Chemistry and Biology (2012) 19:937-954, which are incorporated herein by reference.

[0124] Therefore, for example, oligonucleotides may be modified to enhance their stability and / or enhance their biological activity by modification with nuclease-resistant groups. For example, the nucleic acid agents of the present invention may contain 2'-O-methyl, 2'-fluorine, 2'-O-methoxyethyl, 2'-O-aminopropyl, 2'-amino, and / or phosphorothioate bonds. Including locked nucleic acids (LNAs), for example nucleic acid analogs in which the ribose ring is "locked" by a methylene bridge linking the 2'-O atom and the 4'-C atom, ethylene nucleic acids (ENAs), for example 2'-4'-ethylene-bridged nucleic acids, and specific nucleic acid base modifications, for example 2-amino-A, 2-thio (e.g., 2-thio-U), and G-clamp modifications can also increase binding affinity to targets. Including pyranose sugars in the oligonucleotide backbone can also reduce endonuclease cleavage. The binding arm may further contain peptide nucleic acids (PNAs) in which the deoxyribose (or ribose) phosphate backbone in the DNA is replaced by a polyamide backbone, or it may contain a polymer backbone, a cyclic backbone, or an acyclic backbone. The binding region may incorporate sugar mimes and may further contain protecting groups, particularly at its ends, to prevent undesirable degradation (as described below).

[0125] Exemplary nucleotide bond modifications include, but are not limited to, phosphorothioates, chiral phosphorothioates, phosphorodithioates, phosphotriesters, aminoalkyl phosphotriesters, methylphosphonates, alkylphosphonates (including 3'-alkylenephosphonates), chiral phosphonates, phosphinates, phosphoramidates (including 3'-aminophosphoramidates), aminoalkyl phosphoramidates, thionophosphoramidates, thionoalkylphosphonates, thionoalkylphosphotriesters, and boranophosphates (usually 3'- This includes those having a -5' bond, their 2'-5' bond analogues, and those with reverse polarity where adjacent nucleoside unit pairs are bonded from 3'-5' to 5'-3' or from 2'-5' to 5'-2', as well as boron phosphonates, phosphodiesters, phosphonoacetates (PACE), morpholino, amide carbamates, carboxymethyl, acetamide, polyamides, sulfonates, sulfonamides, sulfamates, formacetals, thioformacetals, alkylsilyls, substitutions, peptide nucleic acids (PNAs) and / or threose nucleic acids (TNAs). Various salt forms, mixed salt forms and free acid forms of the above modifications can also be used. The additional nucleotide bond modifications are described in Deleavey and Damha, Chemistry and Biology (2012) 19:937-954 and Hunziker & Leumann, 1995 and De Mesmaeker et al., 1994, which are incorporated herein by reference.

[0126] According to certain embodiments, the modification includes a modified nucleoside triphosphate (dNTP).

[0127] According to one embodiment, the modification includes an edge-blocker oligonucleotide.

[0128] According to certain embodiments, the edge-blocker oligonucleotide comprises a phosphate, an inverted dT, and an amino-C7.

[0129] According to one embodiment, the nucleic acid agent is modified to include one or more protecting groups, for example, a 5' and / or 3' cap structure.

[0130] As used herein, the term “cap structure” refers to a chemical modification incorporated into either end of an oligonucleotide (see, for example, U.S. Patent No. 5,998,203, incorporated herein by reference). These end modifications can protect nucleic acid molecules from exonuclease degradation and aid in intracellular delivery and / or localization. Cap modifications may be present at the 5' end (5'-cap) or the 3' end (3'-cap), or at both ends. In a non-limiting example, the 5'-cap may be an inverted abasic Selected from the group comprising: residue) (part); 4',5'-methylene nucleotide; 1-(beta-D-erythrofuranosyl) nucleotide, 4'-thionucleotide; carbocyclic nucleotide; 1,5-anhydrohexitol nucleotide; L-nucleotide; alpha-nucleotide; modified base nucleotide; phosphorodithioate bond; threopentofuranosyl nucleotide; acyclic 3',4'-seconucleotide; acyclic 3,4-dihydroxybutyl nucleotide; acyclic 3,5-dihydroxypentyl nucleotide, 3'-3'-inverted nucleotide part; 3'-3'-inverted debase part; 3'-2'-inverted nucleotide part; 3'-2'-inverted debase part; 1,4-butanediol phosphate; 3'-phosphoramide; hexyl phosphate; aminohexyl phosphate; 3'-phosphate; 3'-phosphorothioate; phosphorodithioate; or crosslinked or uncrosslinked methylphosphonate part.

[0131] In some embodiments, the 3'-cap is an inverted deoxynucleotide, e.g., inverted deoxythymidine, 4',5'-methylene nucleotide; 1-(beta-D-erythrofuranosyl) nucleotide; 4'-thionucleotide, carbocyclic nucleotide; 5'-amino-alkyl phosphate; 1,3-diamino-2-propyl phosphate; 3-aminopropyl phosphate; 6-aminohexyl phosphate; 1,2-aminododecyl phosphate; hydroxypropyl phosphate; 1,5-anhydrohexitol nucleotide; L-nucleotide; alpha-nucleotide; modified base nucleotide; phosphorodithioate; treop-nucleotide Selected from the group including antfuranosyl nucleotides; acyclic 3',4'-seconucleotides; 3,4-dihydroxybutyl nucleotides; 3,5-dihydroxypentyl nucleotides, 5'-5'-reverse nucleotide moieties; 5'-5'-reverse debase moieties; 5'-phosphoramides; 5'-phosphorothioates; 1,4-butanediol phosphates; 5'-aminos; crosslinked and / or uncrosslinked 5'-phosphoramides, phosphorothioates and / or phosphorodithioates, crosslinked or uncrosslinked methylphosphonates and 5'-mercapto moieties (generally refer to Beaucage & Iyer, 1993, incorporated herein by reference).

[0132] Nucleic acid agents can be further modified by incorporating a 3' cationic group or by reversing the orientation of the terminal nucleoside via a 3'-3' bond. Alternatively, the 3' end can be blocked by an aminoalkyl group, e.g., 3' C5-aminoalkyl dT. Other 3' conjugates can inhibit 3'-5' exonuclease cleavage. While not theoretically bound, 3' conjugates such as naproxen or ibuprofen may inhibit exonuclease cleavage by sterically blocking the exonuclease from binding to the 3' end of the oligonucleotide. Small alkyl chains, aryl groups, or heterocyclic conjugates or modified sugars (such as D-ribose, deoxyribose, and glucose) can also block 3'-5'-exonucleases.

[0133] According to one embodiment, the 5' end may be blocked by an aminoalkyl group, such as a 5'-O-alkylamino substituent. Other 5' conjugates can inhibit 5'-3' exonuclease cleavage. Although not limited to theory, 5' conjugates such as naproxen or ibuprofen may inhibit exonuclease cleavage by sterically blocking the exonuclease from binding to the 5' end of the oligonucleotide. Small alkyl chains, aryl groups, or heterocyclic conjugates or modified sugars (such as D-ribose, deoxyribose, and glucose) can also block 3'-5'-exonuclease.

[0134] According to certain embodiments, the modification includes locked nucleic acid (LNA) or other crosslinking nucleotides such as cEt, and / or 2'-O-(2-methoxyethyl) (abbreviated as 2'MOE) or 2'-OMe modifications, thereby modifying at least part or all of the sequence at the 2' position of each nucleotide. Examples, but not limited to, include A40, A50, A51, A35, A49, and A52.

[0135] Gapmers are also intended in this specification (see the Examples section below, see Table 5). Gapmers are chimeric antisense oligonucleotides containing a central block of deoxynucleotide monomers of sufficient length to induce RNase H cleavage.

[0136] Nucleic acid agents (and their modifications as described above) can also act at the DNA level, as summarized below.

[0137] Chaserr's downregulation can also be achieved by inactivating the gene (e.g., Chaserr) by introducing targeted mutations involving loss-of-function changes (e.g., point mutations, deletions, and insertions) within the gene structure.

[0138] As used herein, the term “loss-of-function alterations” refers to any mutation in the DNA sequence of a gene (e.g., in the final exon of Chaserr) that results in a downregulation of the expression level and / or activity of an expressed lncRNA product. Non-limiting examples of such loss-of-function alterations include: mutations in the promoter sequence, usually 5' to the transcription start site of a gene, that result in the downregulation of a particular gene product; regulatory mutations, i.e., mutations in an upstream or downstream region of a gene, or within a region of a gene, that affect the expression of a gene product; deletion mutations, i.e., mutations that delete any nucleic acid in the gene sequence; insertion mutations, i.e., mutations that insert a nucleic acid into the gene sequence, which may result in the insertion of a transcription termination sequence; inversion mutations, i.e., mutations that result in an inverted sequence; splice mutations, i.e., mutations that result in abnormal or incomplete splicing; and duplication mutations, i.e., mutations that result in a duplication sequence that may be within a frame or cause a frameshift.

[0139] According to certain embodiments, a loss-of-function alteration of a gene may include at least one allele of the gene.

[0140] As used herein, the term "allele" refers to one or more alternative forms of a gene locus, each allele of which is associated with a trait or characteristic. In diploid cells or organisms, the two alleles of a given gene occupy corresponding loci on homologous chromosome pairs.

[0141] According to other specific embodiments, the loss-of-function change in a gene involves both alleles of the gene. In such cases, for example, a mutation in the final exon of Chaserr can be homozygous or heterozygous.

[0142] Methods for introducing nucleic acid changes into genes of interest are well known in this field [e.g., Menke D. Genesis (2013) 51:-618; Capecchi, Science (1989) 244:1288-1292; Santiago et al. Proc Natl Acad Sci] See USA(2008)105:5809-5814; International Patent Application Nos. WO2014085593, WO2009071334 and WO2011146121; U.S. Patent Nos. 8771945, 8586526, 6774279 and U.S. Patent Publication Nos. 20030232410, 20050026157 and 20060014264, including genome editing by targeted homologous recombination, site-directed recombinases, PB transposases, and engineered nucleases. Agents for introducing nucleic acid changes into genes of interest can be designed from publicly available sources or commercially available from Transposagen, Addgene, and Sangamo Biosciences.

[0143] Examples include the use of genome editing agents such as CRISPR-Cas, meganucleases, zinc finger nucleases (ZFNs), TALENs, and transposons.

[0144] Genome editing using recombinant adeno-associated virus (rAAV) platforms – this genome editing platform is based on rAAV vectors that enable insertion, deletion, or substitution of DNA sequences within the genome of living mammalian cells. rAAV genomes are single-stranded deoxyribonucleic acid (ssDNA) molecules, either positive-sense or negative-sense, approximately 4.7 kb in length. These single-stranded DNA viral vectors have high transduction rates and possess the unique property of stimulating endogenous homologous recombination in the absence of double-stranded DNA breaks within the genome. Those skilled in the art can design rAAV vectors to target desired genomic loci and induce both whole and / or subtle endogenous genetic alterations within the cell. rAAV genome editing has the advantage of targeting a single allele and resulting in no off-target genomic alterations. rAAV genome editing technology is commercially available, for example, from Horizon® (Cambridge, UK) as the rAAV GENESIS® system.

[0145] Methods for identifying efficacy and detecting sequence changes are well known in the art and include, but are not limited to, DNA sequencing, electrophoresis, enzyme-based mismatch detection assays, and hybridization assays, such as PCR, RT-PCR, RNase protection, in situ hybridization, primer extension, Southern blotting, Northern blotting, and dot blotting analysis.

[0146] Sequence changes within specific genes can also be determined at the protein level using, for example, chromatography, electrophoresis, immunoassays such as ELISA, and Western blot analysis, as well as immunohistochemical tests.

[0147] Furthermore, those skilled in the art can easily design knock-in / knockout constructs containing positive and / or negative selection markers for efficiently selecting transformed cells that have undergone homologous recombination events with the construct. Positive selection provides a means for enriching clonal populations that have incorporated foreign DNA. Non-limiting examples of such positive markers include glutamine synthetase, dihydrofolate reductase (DHFR), neomycin, hygromycin, puromycin, and antibiotic resistance-constituting markers such as the blastocydin S resistance cassette. Negative selection markers are required to select against random incorporation and / or exclusion of marker sequences (e.g., positive markers). Non-limiting examples of such negative markers include herpes simplex thymidine kinase (HSV-TK), hypoxanthine phosphoribosyltransferase (HPRT), and adenine phosphoribosyltransferase (ARPT), which convert ganciclovir (GCV) to a cytotoxic nucleoside analog.

[0148] According to one embodiment, the technology relates to introducing RNA silencing molecules using transient DNA or a DNA-free method (such as RNA transfection).

[0149] According to one embodiment, the RNA silencing molecule (e.g., an antisense molecule) is delivered as a "naked" oligonucleotide, i.e., without the use of an additional delivery vehicle. According to one embodiment, the "naked" oligonucleotide includes chemical modifications to facilitate its tissue delivery (e.g., utilizing inverted nucleotides, phosphorothioate bonds, or locked nucleic acid incorporation, as described above).

[0150] Any method known in the art for the transfection of RNA or DNA, such as, but not limited to, microinjection, electroporation, or lipid-mediated transfection using liposomes or cationic molecules or nanomaterials, may be used in accordance with this instruction (as described below and further described in Roberts et al. Nature Reviews Drug Discovery (2020) 19:673-694, which is incorporated herein by reference).

[0151] According to one embodiment, as mentioned above, if the RNA silencing molecule (e.g., antisense) is unmodified, the RNA silencing molecule may be administered to target cells (e.g., senescent cells) as part of an expression construct. In this case, the RNA silencing molecule (e.g., antisense molecule) is ligated into a nucleic acid construct (also referred to herein as an “expression vector”) under the control of a cis-acting regulatory element (e.g., a promoter) that can constitutively or inductively lead to the expression of the RNA silencing molecule (e.g., antisense) within the target cells (e.g., nerve cells).

[0152] The expression constructs of the present invention may also include additional sequences (e.g., shuttle vectors) that make the expression constructs suitable for replication and integration in eukaryotes. Typical cloning vectors contain transcription start sequences and translation start sequences (e.g., promoters, enhancers) and transcription and translation terminators (e.g., polyadenylation signals). The expression constructs of the present invention may further include enhancers, which may be adjacent to or away from the promoter sequence and may function in upregulating transcription from the promoter sequence. Polyadenylation sequences may also be added to the expression constructs of the present invention to increase the efficiency of expression.

[0153] In addition to the embodiments already described, the expression constructs of the present invention may typically contain other specialized elements intended to increase the expression level of cloned nucleic acids or to facilitate the identification of cells having RNA silencing molecules (e.g., antisense). The expression constructs of the present invention may or may not contain eukaryotic replicons.

[0154] Nucleic acid constructs can be introduced into target cells of the present invention (e.g., nerve cells) using appropriate gene delivery vehicles / methods (e.g., transfection, transduction) and appropriate expression systems. Such methods are generally described in Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Springs Harbor Laboratory, New York (1989, 1992), Ausubel et al., Current Protocols in Molecular Biology, John Wiley and Sons, Baltimore, Md. (1989), Chang et al., Somatic Gene Therapy, CRC Press, Ann Arbor, Mich. (1995), Vega et al., Gene Targeting, CRC Press, Ann Arbor Mich. (1995), Vectors: A Survey of Molecular Cloning Vectors and Their Uses, Butterworths, Boston Mass. (1988), and Gilboa et al. [Biotechniques 4(6):504-512, 1986], and include, for example, stable or transient transfection, lipofection, electroporation, and infection with recombinant viral vectors. Furthermore, for information regarding positive-negative selection methods, please refer to U.S. Patent No. 5,464,764 and U.S. Patent No. 5,487,992.

[0155] Additionally or alternatively, lipid-based systems may be used to deliver the constructs or nucleic acid agents encoded thereby to the target cells of the present invention (e.g., senescent cells or cancer cells). Lipid-based systems include, for example, liposomes, lipoplexes, and lipid nanoparticles (LNPs). In some embodiments, antisense oligonucleotides or siRNAs include conjugated lipid or cholesteryl moieties.

[0156] The specificity of the method can be improved by using neuron-specific promoters. Examples of neuron-specific promoters include, but are not limited to, synapsin. Since synapsin is considered a neuron-specific protein (DeGennaro et al., 1983 Cold Spring Harb.Symp.Quant.Biol.1,337-345), its neuron-specific expression pattern can be leveraged to express the transgene in a neuron-specific manner. Adenovirus vectors and AAV vectors for local injection use minimal human synapsin promoters (Kugler et al. 2003 Human synapsin 1 gene promoter confers highly neuron-specific long-term transgene expression from an adenoviral vector in the adult rat brain depending on the transduced area. Gene Ther.10,337-347). AAV capsids that can reach the CNS after peripheral administration, such as AAV9 or other natural AAV serotypes, are advantageous for relatively non-invasive administration resulting in large-scale expression. Currently, several modified capsids exist that exhibit increased neuronal transduction efficiency. Lentiviruses with E / SYN promoters have been reported to show strong sustained expression in neurons (Hioki et al. Gene Therapy volume 14, pages 872-882 (2007)).

[0157] This instruction can be used in clinics to manage related diseases, syndromes, disorders, and conditions associated with CHD2 haploinsufficiency.

[0158] Accordingly, according to one aspect of the present invention, a method is provided for treating a disease or condition related to chromodomain helicase DNA-binding protein 2 (CHD2) haploinsufficiency in a subject requiring treatment, comprising administering a therapeutically effective amount of a nucleic acid agent that downregulates the activity or expression of human Chaserr to the subject, wherein the nucleic acid agent is directed to the terminal exon of human Chaserr, thereby providing a method for treating a disease or condition related to CHD2 haploinsufficiency.

[0159] In alternative or additional embodiments, a nucleic acid agent is provided that downregulates the activity or expression of human Chaserr, for use in treating diseases or conditions associated with chromodomain helicase DNA-binding protein 2 (CHD2) haploinsufficiency in subjects requiring treatment, and which is directed toward the terminal exon of human Chaserr.

[0160] As used herein, “chromodomain helicase DNA-binding protein 2 (CHD2) haploinsufficiency-related disease or condition” refers to a pathogenic condition characterized by reduced expression (protein and optionally mRNA) of CHD2, or in which the onset or progression is associated with reduced expression (protein and optionally mRNA) of CHD2.

[0161] According to certain embodiments, CHD2 haploinsufficiency-related disorders or conditions refer to CHD2-associated neurodevelopmental disorders typically characterized by early-onset epileptic encephalopathy (i.e., intractable seizures associated with frequent, ongoing epileptic-like activity and slowing or regression of cognition). Seizure onset is typically between 6 months and 4 years of age. Seizure types typically include falls, myoclonus, rapid onset of multiple seizure types associated with generalized spike waves on EEG, atonic-myoclonic-absence seizures, and clinical photosensitivity. Intellectual disability and / or autism spectrum disorder are common.

[0162] In a particular embodiment, the medical condition is selected from the group consisting of Lennox-Gastaut syndrome (LGS), myoclonus absence epilepsy (MAE), Dravet syndrome, intellectual disability with epilepsy, and autism spectrum disorder (ASD).

[0163] The diagnosis of CHD2-related neurodevelopmental disorders is established in subjects with heterozygous CHD2 single-nucleotide pathogenic variants, small indel (insertion / deletion) pathogenic variants, or partial or whole gene deletions detected by molecular genetic testing.

[0164] Mutations in the CHD2 gene can result from germline mutations or de novo somatic mutations.

[0165] The term "to treat" refers to inhibiting, preventing, or halting the onset of a pathological condition (disease, disorder, or state), and / or causing reduction, remission, or regression of the condition. Those skilled in the art will understand that various methods and assays can be used to assess the onset of a pathological condition, and similarly, various methods and assays can be used to assess reduction, remission, or regression of the condition.

[0166] As used herein, the term “prevent” means to prevent a disease, disorder, or condition from occurring in a person who is at risk of having the disease but has not yet been diagnosed with the disease.

[0167] As used herein, the term “Subject” includes mammals, preferably humans of any age suffering from the condition. Preferably, the term encompasses individuals at risk of developing the condition. It will be understood that the mammal may be an embryo or a fetus. Alternatively, the subject may be a child or adolescent up to 15 or 18 years of age.

[0168] For in vivo treatment, nucleic acid agents are administered to the subject either on their own or as part of a pharmaceutical composition.

[0169] As used herein, “pharmaceutical composition” refers to one or more preparations of the active ingredients described herein, including other chemical components such as physiologically suitable carriers and excipients. The purpose of the pharmaceutical composition is to facilitate the administration of the compound to a living organism.

[0170] In this specification, the term "active ingredient" refers to nucleic acid agents that can be the primary cause of biological effects.

[0171] Hereinafter, the terms "physiologically acceptable carrier" and "pharmaceutically acceptable carrier," which may be used interchangeably, refer to carriers or diluents that do not cause significant irritation to the organism and do not negate the biological activity and properties of the administered compound. Adjuvants are included under these terms.

[0172] In this specification, the term “excipient” refers to an inert substance added to a pharmaceutical composition to further facilitate the administration of the active ingredient. Examples of excipients include, but are not limited to, calcium carbonate, calcium phosphate, various sugars, and types of starch, cellulose derivatives, gelatin, vegetable oils, and polyethylene glycol.

[0173] Techniques for the formulation and administration of drugs can be found in “Remington's Pharmaceutical Sciences,” Mack Publishing Co., Easton, PA, latest edition, which is incorporated herein by reference.

[0174] Preferred routes of administration may include, for example, systemic, oral, rectal, transmucosal, particularly transnasal, intestinal, or parenteral delivery, such as intramuscular, subcutaneous, and intrathecal injection, as well as intrathecal, direct intraventricular, intracardiac, intracardiac, such as intracavitary, intraright or left intraventricular, intracommon coronary artery, intravenous, intraperitoneal, intranasal, intratumoral, or intraocular injection.

[0175] According to a particular embodiment, the composition is for inhalation administration.

[0176] According to a particular embodiment, the composition is for intranasal administration.

[0177] According to a particular embodiment, the composition is for intraventricular administration.

[0178] According to a particular embodiment, the composition is for intrathecal administration.

[0179] According to a particular embodiment, the composition is for intratumoral administration.

[0180] According to a particular embodiment, the composition is for oral administration.

[0181] According to a particular embodiment, the composition is for local injection.

[0182] According to a particular embodiment, the composition is for systemic administration.

[0183] According to a particular embodiment, the composition is for intravenous administration.

[0184] Conventional methods for drug delivery to the central nervous system (CNS) include neurosurgical strategies (e.g., intracerebral injection or intraventricular infusion); molecular manipulation of agents to utilize one of the endogenous transport pathways of the blood-brain barrier (e.g., production of chimeric fusion proteins containing transport peptides with affinity for endothelial cell surface molecules in combination with agents that themselves cannot cross the BBB); pharmacological strategies designed to increase the lipid solubility of agents (e.g., conjugation of water-soluble agents with lipid or cholesterol carriers); and transient disruption of BBB integrity by hyperosmolar disruption (resulting from injection of mannitol solution into the carotid artery or the use of bioactive agents such as angiotensin peptides). However, each of these strategies has limitations, including inherent risks associated with invasive surgical procedures, size limitations imposed by inherent limitations of the endogenous transport system, potentially undesirable biological side effects associated with systemic administration of chimeric molecules composed of carrier motifs that may be active outside the CNS, and the potential risk of brain damage in the brain region where the BBB is disrupted, thus making each of these strategies a suboptimal delivery method.

[0185] Alternatively, the pharmaceutical composition may be administered locally rather than systemically, for example, by directly injecting it into a tissue area of ​​the patient.

[0186] Pharmaceutical compositions of some embodiments of the present invention may be produced by processes well known in the art, for example, by conventional mixing, dissolution, granulation, sugar-coated tablet production, wet grinding, emulsification, encapsulation, capture, or freeze-drying processes.

[0187] Accordingly, pharmaceutical compositions for use according to some embodiments of the present invention may be formulated conventionally using one or more physiologically acceptable carriers comprising excipients and adjuvants, thereby facilitating the processing of the active ingredient into pharmaceutically usable preparations. The appropriate formulation depends on the selected route of administration.

[0188] For injection, the active ingredient of the pharmaceutical composition may be formulated in an aqueous solution, preferably a physiologically compatible buffer, such as Hanks' solution, Ringer's solution, or physiological saline buffer. For transmucosal administration, a penetrating agent suitable for the barrier to be penetrated is used in the formulation. Such penetrating agents are generally known in the art.

[0189] For oral administration, pharmaceutical compositions can be readily formulated by combining the active compound with a pharmaceutically acceptable carrier known in the art. Such carriers enable the formulation of pharmaceutical compositions for oral intake by patients as tablets, pills, sugar-coated tablets, capsules, liquids, gels, syrups, slurries, suspensions, etc. Pharmacological preparations for oral use may be prepared by using a solid excipient, optionally adding suitable adjuvants to obtain a tablet or sugar-coated tablet core, and then optionally grinding the resulting mixture to process the granular mixture. Suitable excipients include fillers such as sugars containing lactose, sucrose, mannitol, or sorbitol; cellulose preparations such as corn starch, wheat starch, rice starch, potato starch, gelatin, tragacanth gum, methylcellulose, hydroxypropyl methylcellulose, and sodium carbomethylcellulose; and / or physiologically acceptable polymers such as polyvinylpyrrolidone (PVP). If desired, a disintegrant such as cross-linked polyvinylpyrrolidone, agar, or alginic acid or its salts, such as sodium alginate, may be added.

[0190] The sugar-coated tablet core is coated with a suitable coating. For this purpose, a concentrated sugar solution may be used which may optionally contain gum arabic, talc, polyvinylpyrrolidone, carbopole gel, polyethylene glycol, titanium dioxide, lacquer solution, and a suitable organic solvent or solvent mixture. Dyes or pigments may be added to the tablet or sugar-coated tablet coating for identification or to characterize various combinations of active compound dosages.

[0191] Pharmaceutical compositions for oral administration include gelatin-based suction capsules and soft-seal capsules made of gelatin and a plasticizer, such as glycerol or sorbitol. Suction capsules may contain the active ingredient mixed with a filler such as lactose, a binder such as starch, a lubricant such as talc or magnesium stearate, and optionally a stabilizer. In soft capsules, the active ingredient may be dissolved or suspended in a suitable liquid such as fatty oil, liquid paraffin, or liquid polyethylene glycol. Further stabilizers may be added. Any formulation for oral administration should be in a dosage appropriate to the selected route of administration.

[0192] When administered buccally, the composition may take the form of a conventionally formulated tablet or lozenge.

[0193] For administration by nasal inhalation, the active ingredient for use according to some embodiments of the present invention is conveniently delivered in the form of an aerosol spray from a pressurized pack or nebulizer using a suitable propellant, such as dichlorodifluoromethane, trichlorofluoromethane, dichlorotetrafluoroethane, or carbon dioxide. In the case of a pressurized aerosol, the dose unit may be determined by providing a valve for delivering the measured amount. For use in dispensers, for example, gelatin capsules and cartridges may be formulated containing a powder mixture of the compound and a suitable powder base such as lactose or starch.

[0194] The pharmaceutical compositions described herein may be formulated, for example, for parenteral administration by bolus injection or continuous infusion. The injectable formulations may be provided in unit dosage forms, for example, in ampoules, or optionally in multi-dose containers with preservatives. The compositions may be suspensions, solutions, or emulsions in an oily or aqueous vehicle and may contain compounding agents such as suspending agents, stabilizers, and / or dispersants.

[0195] Pharmaceutical compositions for parenteral administration include aqueous solutions of the active preparation in a water-soluble form. Furthermore, suspensions of the active ingredient may be prepared as suitable oily or aqueous injection suspensions. Suitable lipophilic solvents or vehicles include fatty oils such as sesame oil, or synthetic fatty acid esters such as ethyl oleate, triglycerides, or liposomes. Aqueous injection suspensions may contain substances that increase the viscosity of the suspension, such as sodium carboxymethylcellulose, sorbitol, or dextran. Optionally, the suspension may also contain suitable stabilizers or agents that increase the solubility of the active ingredient to enable the preparation of high-concentration solutions.

[0196] Alternatively, the active ingredient may be in powder form for preparation using a suitable vehicle before use, such as a sterile pyrogen-free aqueous solution.

[0197] The pharmaceutical compositions of some embodiments of the present invention may also be formulated into suppositories or rectal compositions such as retained enemas, using conventional suppository bases such as cocoa butter or other glycerides.

[0198] Pharmaceutical compositions suitable for use in connection with some embodiments of the present invention include compositions containing an active ingredient in an amount effective to achieve the intended purpose. More specifically, a therapeutically effective amount means an amount of an active ingredient (e.g., a nucleic acid agent) effective to prevent, alleviate or improve the symptoms of a disorder (e.g., associated with CHD2 haploinsufficiency) or to extend the survival time of the subject being treated.

[0199] Determining the effective dosage is well within the capabilities of those skilled in the art, particularly in light of the detailed disclosure provided herein.

[0200] In any preparation used in the method of the present invention, the therapeutically effective dose or effective amount can be initially estimated from in vitro assays and cell culture assays. For example, some dose may be formulated in an animal model to achieve a desired concentration or potency. Such information can be used to more accurately determine a useful dose for humans.

[0201] The toxicity and therapeutic efficacy of the active ingredients described herein can be determined in vitro by standard pharmaceutical procedures, or by standard pharmaceutical procedures in cell cultures or experimental animals. Data obtained from these in vitro assays, cell culture assays, and animal studies can be used to formulate various dosages for human use. Dosages may vary depending on the dosage form used and the route of administration utilized. The exact formulation, route of administration, and dosage may be selected by individual physicians in consideration of the patient's condition. (See, for example, Fingl, et al., 1975, in “The Pharmacological Basis of Therapeutics”, Ch.1 p.1).

[0202] Dosage and intervals can be individually adjusted to provide a sufficient level of the active ingredient (minimum effective concentration, MEC) to induce or inhibit the biological effect. While MEC varies from preparation to preparation, it can be estimated from in vitro data. The dose required to achieve MEC depends on individual characteristics and administration route. Plasma concentrations can be determined using detection assays.

[0203] Depending on the severity and responsiveness of the condition being treated, medication may be administered as a single or multiple doses, and the treatment series may continue for several days to several weeks, or until a cure is achieved or a reduction in the disease state is achieved.

[0204] The amount of the composition administered naturally depends on the patient being treated, the severity of the pain, the method of administration, and the judgment of the prescribing physician.

[0205] Compositions of some embodiments of the present invention may, if desired, be provided in packs or dispenser devices, such as FDA-approved kits, which may contain one or more unit dosage forms containing the active ingredient. Packs may include, for example, metal or plastic foil such as blister packs. Packs or dispenser devices may be accompanied by instructions for administration. Packs or dispensers may also be accompanied by a notice relating to a form of container prescribed by a government agency that regulates the manufacture, use or sale of a drug, which reflects the form of the composition or agency approval for human or veterinary administration. Such notices may be, for example, labels or approved product inserts approved by the U.S. Food and Drug Administration for prescription drugs. Compositions containing preparations of the present invention formulated in a suitable pharmaceutical carrier may also be prepared, placed in appropriate containers, and labeled for use in the indicated conditions, as further detailed above.

[0206] Treatment with nucleic acid agents according to the present invention may be enhanced using other control protocols known in the art, such as antiepileptic drugs (AEDs).

[0207] Figure 14 is a flowchart illustrating a suitable method for analyzing a set of sequences according to various exemplary embodiments of the present invention. Unless otherwise defined, the operations described below in this specification can be performed simultaneously or sequentially in many combinations or execution orders. Specifically, the order in the flowchart should not be considered limiting. For example, two or more operations appearing in a particular order in the following description or flowchart can be performed in a different order (e.g., in reverse order) or substantially simultaneously. Furthermore, some of the operations described below are optional and do not need to be performed.

[0208] At least some of the operations described herein can be performed by a data processing system, such as a dedicated circuit or a general-purpose computer, configured to receive data and perform the operations described below. At least some of the operations can be performed by a remote cloud computing facility.

[0209] Computer programs implementing the method of this embodiment may be generally provided to users via a communication network or, but not limited to, on a delivery medium such as a floppy disk, CD-ROM, flash memory device, or portable hard drive. Computer programs can be copied from the communication network or delivery medium to a hard disk or similar intermediate storage medium. The computer program may be executed by loading code instructions from either the delivery medium or the intermediate storage medium into the computer's executable memory and configuring the computer to operate according to the method of the present invention. During operation, the computer may store data structures or values ​​obtained by intermediate calculations in memory and retrieve these data structures or values ​​for use in subsequent operations. All of these operations are well known to those skilled in the art of computer systems.

[0210] The processing operations described herein may be performed by processor circuits such as DSPs, microcontrollers, FPGAs, ASICs, or any other conventional and / or dedicated computing system.

[0211] The method of this embodiment can be embodied in many forms. For example, the method of this embodiment can be embodied on a tangible medium such as a computer for performing method operations. The method of this embodiment can be embodied on a computer-readable medium equipped with computer-readable instructions for performing method operations. Furthermore, the method of this embodiment can be embodied in an electronic device having digital computer functions configured to execute a computer program on a tangible medium or to execute instructions on a computer-readable medium.

[0212] Referring here to Figure 14, the method begins at 10 and optionally, and preferably, is followed by 11 where a set of sequences is received. Typically, each sequence in the set describes a polynucleotide such as DNA or RNA, and the polynucleotides described by the various sequences in the set are determined to be homologous to one another, either manually or using bioinformatics tools known to those skilled in the art, such as Blastn, FASTA, as further described below and in the Examples section below. According to a particular embodiment, DNA is genomic DNA. According to another embodiment, DNA is cDNA or library DNA. According to a particular embodiment, DNA represents a gene locus. According to another embodiment, DNA is coding DNA or non-coding DNA. According to a particular embodiment, DNA includes exons, introns, or combinations thereof. According to a particular embodiment, sequence is an RNA sequence. According to a particular embodiment, RNA is coding RNA. According to another embodiment, RNA is non-coding RNA.

[0213] In some embodiments of the present invention, homologous polynucleotides are selected from the group consisting of 3'UTR, lncRNA, and enhancers.

[0214] The polynucleotides within a set can be complete sequences or partial sequences.

[0215] In some embodiments of the present invention, the method proceeds to 12, where the sequences in the set are aligned according to a predetermined order, for example, an order determined by evolution, to provide a multiple alignment having multiple alignment layers.

[0216] Alignments can be ordered as multiple alignments or using phylogenetic representations—dendrograms. Typically, in multiple alignments, the first alignment layer is a sequence describing the query polynucleotide. If the alignment is determined by evolution, the first layer is optionally, and preferably, a sequence describing the species of interest. For example, if one of the polynucleotides is a human polynucleotide, the first alignment layer may be a sequence of human polynucleotides.

[0217] Alignment may be performed by any technique known in the art. Typically, the alignment technique provides a score, and the order follows the score. For example, BLAST can be used to determine the order of sequences. If the alignment technique provides a score, the second alignment layer is preferably the sequence having the highest alignment score relative to the first alignment layer, the third alignment layer is preferably the sequence having the next highest alignment score relative to the first alignment layer, and so on. This provides an alignment in which the sequences in each layer have the best alignment score relative to the sequences in the previous layer. If the alignment technique does not provide a significant alignment for a particular alignment layer, the layers following that particular alignment layer will contain the next available sequences according to the order of the received set.

[0218] However, it should be understood that it is not necessary to perform operation 12. For example, the method can use the order at the time the set is received. Alternatively, the method can allow a user, for example, a user interface device, to select or input the order to be used by the method.

[0219] The method described above preferably follows step 13 in which the graph is constructed. The inventors have found it advantageous to transform the sequence analysis problem into a graph traversal problem in order to further define the constraints of the problem in the structured method described above. The graph is preferably a hierarchical and connected graph, where each edge of the graph connects nodes of consecutive layers. The layers of the graph preferably represent sequences, and the nodes within a layer represent k-mers within each sequence. Thus, for example, suppose the i-th layer of the graph represents a particular sequence of a set (e.g., the sequence of a canine organism). In this case, each node in the i-th layer represents a k-mer of that particular sequence. For example, the first node in the i-th layer may represent the first k-mer in that particular sequence (e.g., bases 1 to k of the sequence), the second node in the i-th layer may represent the second k-mer in that particular sequence (e.g., bases 2 to k+1 of the sequence), and so on. In various exemplary embodiments of the present invention, 6 ≤ k ≤ 12.

[0220] If operation 12 is not performed and the method does not receive user input regarding order, the method constructs layers of the graph according to the order of the arrays in the received set. Specifically, the first layer of the graph represents the first array in the received set, the second layer of the graph represents the second array in the received set, and so on. If the method receives user input regarding order, the method constructs layers of the graph according to the user input. Specifically, the first layer of the graph represents the array that is in the first order according to the user input, the second layer of the graph represents the array that is in the second order according to the user input, and so on. If operation 12 is performed, the method constructs layers of the graph according to alignment. Specifically, the first layer of the graph represents the array of the first alignment layer, the second layer of the graph represents the array of the second alignment layer, and so on.

[0221] In various exemplary embodiments of the present invention, the first layer of the graph represents a sequence describing a query polynucleotide.

[0222] The graph is optionally, and preferably, constructed such that each edge connects nodes representing identical or homologous k-mers. The advantage of this embodiment is that it allows for the identification of motifs that are conserved or substantially conserved across multiple polynucleotides.

[0223] According to some embodiments of the present invention, the homology between homologous k-mers linked by the edges of the graph is at least 60%, more preferably at least 70%, more preferably at least 80%, more preferably at least 90%, 95%, or more.

[0224] Representative examples of typical hierarchical graphs according to several embodiments of the present invention are shown in Figures 11B, 11D, and 12. In these figures, nodes are shown as strings corresponding to nucleotide bases that form kmers, edges are shown as straight solid lines, and layers are indicated by L1, L2, etc.

[0225] The method described above is followed by a search for continuous non-intersecting paths along the edges of the graph.14 The search can be performed using any known optimization technique, such as linear programming (e.g., integer linear programming), mixed linear programming, or any other method for finding the locally maximum solution, such as a greedy algorithm.

[0226] A path is non-intersecting in the sense that the edges connecting nodes representing a particular k-mer do not intersect with the edges connecting nodes representing k-mers that are not identical or homologous to that particular k-mer. However, it should be noted that if there are multiple edges connecting nodes representing a particular k-mer and belonging to two consecutive layers, these edges may intersect, but do not necessarily have to. For example, referring to the simplified graph at the bottom of Figure 11D, the graph contains two k-mers, namely, eight nodes representing the heptomer AGAAUCG and five nodes representing the hexamer CCGUAC. Edges connecting (identical or homologous) heptomers do not intersect with edges connecting (identical or homologous) hexamers. On the other hand, there are edges that connect heptomers and intersect each other (see, for example, the edge connecting the fourth node of layer L2 to the fourth node of layer L3, and the edge connecting the fifth node of layer L2 to the third node of layer L3). Nevertheless, some of the edges connecting the heptamers do not intersect with other edges (see, for example, the edge connecting the fourth node of layer L2 to the third node of layer L3 does not intersect with the edge connecting the fifth node of layer L2 to the fourth node of layer L3).

[0227] In some embodiments of the present invention, the search includes applying a criterion based on path depth as a constraint for the search, such that the search is prioritized over the search for deeper paths (i.e., paths that pass through layers with more graphs) over shallower paths (i.e., paths that pass through layers with fewer graphs).

[0228] The method described above proceeds from 14, optionally and preferably to 15, where the value of k decreases (preferably by 1), and then back to 13, where the graph is reconstructed according to the decreased value of k by including nodes in the graph that represent shorter k-quantities than those already represented by nodes already present in the graph. Preferably, the reconstruction involves adding nodes corresponding to shorter k-quantities while retaining at least some of the existing nodes, thus increasing the order (number of nodes) of the graph. Referring again to the simplified case in Figure 11D, the graph at the top of this figure has 8 nodes representing heptemers and does not include any nodes representing k-quantities for k<7. The middle graph in Figure 11D shows the reconstruction of the graph by adding 5 nodes representing hexamers so that the order of the graph increases from 8 to 8+5=13.

[0229] When nodes representing shorter k-mers are included in the graph, the method optionally, and preferably, updates the edges of the graph to connect identical or homologous k-mers in consecutive layers. This is illustrated in the central graph of Figure 11D, where edges have been added to the graph to connect newly added nodes representing hexamers. Layer L represents a specific k-mer. i If any node within the layer L represents the same specific k-mer, i+1 You can combine and add nodes so that they are connected to any node within it.

[0230] After each reconstruction of the graph, the method optionally, and preferably, re-executes operation 14 to provide continuous non-intersecting paths along the edges of the reconstructed graph. Such re-execution may result in the exclusion of previously acquired paths, for example, if it is found that a previously acquired path intersects with a newly added edge. This is illustrated in the upper part and graph of Figure 11D, where, for example, a path starting at the leftmost node of layer L1 and ending at the rightmost node of layer L3 is included in the upper graph of Figure 11D (before reconstruction) but is not included in the lower graph of Figure 11D (after reconstruction) because it was found to intersect with an edge connecting hexamers that was added during reconstruction.

[0231] The return from 14 to 13 via 15 is optionally and preferably continued iteratively. Preferably, in each iteration cycle, the method applies, as a constraint for the search, the path obtained in the previous iteration cycle. A representative example of such an application of the constraint is shown in FIG. 12 and further illustrated in the Example section below. The iteration is optionally and preferably repeated until there are no more k-mers to be added, or until there are no more new non-crossing paths found, or until some other predetermined stopping criterion is met.

[0232] At 16, an output is generated. The output preferably identifies, as a nucleic acid sequence of interest functionally, a k-mer corresponding to at least one of the paths. The output can be displayed graphically or textually on a display device or stored on a computer-readable storage medium for subsequent use.

[0233] The method ends at 17.

[0234] Figure 15 is a schematic diagram of a client computer 130 having a hardware processor 132 that typically includes input / output (I / O) circuits 134, a hardware central processing unit (CPU) 136 (e.g., a hardware microprocessor), and hardware memory 138 that typically includes both volatile and non-volatile memory. The CPU 136 communicates with the I / O circuits 134 and the memory 138. The client computer 130 preferably includes a graphical user interface (GUI) 142 that communicates with the processor 132. The I / O circuits 134 preferably communicate information with the GUI 142 in a well-structured manner. A server computer 150 is also shown, which may similarly include a hardware processor 152, I / O circuits 154, a hardware CPU 156, and hardware memory 158. The I / O circuits 134 and 154 of the client computer 130 and the server computer 150 can operate as transceivers that communicate information with each other via wired or wireless communication. For example, the client computer 130 and the server computer 150 can communicate with each other via a network 140 such as a local area network (LAN), a wide area network (WAN), or the internet. In some embodiments, the server computer 150 may be part of a cloud computing resource of a cloud computing facility that communicates with the client computer 130 via the network 140.

[0235] The GUI 142 and the processor 132 may be integrated within the same housing, or they may be separate units that communicate with each other.

[0236] GUI142 may optionally, and preferably, be part of a system including a dedicated CPU and I / O circuitry (not shown) that enables GUI142 to communicate with processor 132. Processor 132 sends graphic and text outputs generated by CPU 136 to GUI142. Processor 132 also receives signals from GUI142 relating to control instructions generated by GUI142 in response to user input. GUI142 may be any type known in the art, for example, a keyboard and display, a touchscreen, etc. In a preferred embodiment, GUI142 is the GUI of a mobile device such as a smartphone, tablet, or smartwatch. If GUI142 is the GUI of a mobile device or processor 132, the CPU circuitry of the mobile device can function as processor 132 and execute the code instructions described herein.

[0237] The client computer 130 and the server computer 150 may each further comprise one or more computer-readable storage media 144, 164. The media 144 and 164 are preferably non-temporary storage media that store computer code instructions for performing the methods further detailed herein, and the processors 132 and 152 execute these code instructions. The code instructions may be executed by loading each code instruction into the respective execution memories 138 and 158 of the respective processors 132 and 152.

[0238] Each of the storage media 144 and 164 can store program instructions that, when read by their respective processors, cause the processors to perform the methods described herein. In some embodiments of the present invention, a set of sequences describing a plurality of homologous polynucleotides is received by processor 132 via I / O circuit 134. Processor 132 constructs a graph, searches the graph for continuous non-crossing paths, and generates output that functionally identifies kmers corresponding to at least one path as nucleic acid sequences of interest, as further detailed above herein. Alternatively, processor 132 can send the set of sequences to server computer 150 via network 140. Computer 150 receives the set of sequences, constructs a graph, searches the graph for continuous non-crossing paths, and functionally identifies kmers corresponding to at least one path as nucleic acid sequences of interest, as further detailed above herein. Computer 150 sends back the functionally interested nucleic acid sequences to computer 130 via network 140. Computer 130 receives the nucleic acid sequences and displays them on GUI 142.

[0239] Once a motif is identified, it can be validated using molecular biological techniques, for example, by cloning it into an expression vector that typically contains a reporter sequence.

[0240] As used herein, the term "approximately" refers to ±10%.

[0241] The terms "comprises," "comprising," "includes," "including," "having," and their cognates all mean "to include, but not limited to."

[0242] The term "consisting of" means "including and limited to."

[0243] The term "consisting essentially of" means that the composition, method, or structure may include additional components, steps, and / or parts, but only if the additional components, steps, and / or parts do not substantially alter the basic and novel features of the claimed composition, method, or structure.

[0244] As used herein, the singular forms "a," "an," and "the" include plural references unless otherwise explicitly indicated by the context. For example, the term "compound" or "at least one compound" may include multiple compounds, including mixtures thereof.

[0245] Throughout this application, various embodiments of the present invention may be presented in range form. It should be understood that range form descriptions are merely for convenience and brevity and should not be interpreted as inflexible limitations on the scope of the invention. Therefore, range descriptions should be considered to specifically disclose all possible subranges and the individual numbers within those ranges. For example, a range description such as 1-6 should be considered to specifically disclose subranges such as 1-3, 1-4, 1-5, 2-4, 2-6, 3-6, and the individual numbers within those ranges, e.g., 1, 2, 3, 4, 5, and 6. This applies regardless of the width of the range.

[0246] Whenever a numerical range is indicated in this specification, it means that any of the cited numbers (fractions or integers) within that range are included. The phrases “ranging / ranges between” between the first and second indicators and “ranging / ranges from” the first indicator to the second indicator are used interchangeably in this specification and mean that the ranges include the first and second indicators and all fractions and integers between them.

[0247] As used herein, the term “method” means, but is not limited to, a method of accomplishing a given task, including methods, means, techniques and procedures that are known to those skilled in the art in the fields of chemistry, pharmacology, biology, biochemistry and medicine, or that can be readily developed from methods, means, techniques and procedures known to those skilled in the art.

[0248] It is understood that RNA antisense sequences may be provided herein as DNA sequences in which U is replaced by T.

[0249] When referring to a specific sequence listing, such reference should be understood to also include sequences substantially corresponding to its complementary sequence, including minor sequence mutations resulting from, for example, sequencing errors, cloning errors, or other changes that result in base substitutions, base deletions, or base additions, provided that the frequency of such mutations is less than one per 50 nucleotides, or less than one per 100 nucleotides, or less than one per 200 nucleotides, or less than one per 500 nucleotides, or less than one per 1000 nucleotides, or less than one per 5000 nucleotides, or less than one per 10000 nucleotides.

[0250] For clarity, certain features of the Invention described in the context of separate embodiments may be provided in combination in a single embodiment. Conversely, various features of the Invention described in the context of a single embodiment for brevity may also be provided separately, in any preferred partial combination, or as suitable for any other described embodiment of the Invention. Certain features described in the context of various embodiments should not be considered essential features of those embodiments unless the embodiments are unable to operate without those elements.

[0251] The various embodiments and aspects of the present invention described above and claimed in the following claims section are supported by experiment in the following examples. [Examples]

[0252] Here, refer to the following examples which non - limitly show some embodiments of the present invention together with the above description.

[0253] Materials and Methods Input to LncLOOM LncLOOM functions on a set of sequences obtained from various species. Typically, each sequence corresponds to a putative homolog of sequences obtained from various species. Currently, the inventors are studying using only one sequence isoform per species, but if there are multiple sequences per species, for example, it is adaptable to alternative splicing products. Input sequences are typically constructed by RNA - seq and EST data as well as manual inspection of existing annotations. Some of the input sequences may be incomplete, and it should be noted that according to some embodiments of the present invention, this framework includes specific steps to handle such scenarios. Before graph construction, the set is filtered to remove identical sequences. This can be further adjusted by the user to remove sequences with a percentage identity exceeding a threshold, in which case LncLOOM calculates the percentage identity between each sequence pair using MAFFT MSA and retains the sequence that first appears in the input dataset.

[0254] Ordering of Sequences The LncLOOM framework is constructed around an ordered set of sequences that ideally should be from species (in this manuscript, human in all examples) having a monotonically increasing evolutionary distance with respect to an anchor sequence. The order of the sequences can be provided by the user or determined by using BLAST. When BLAST is used, the anchor sequence is defined as the first sequence in the dataset. The second sequence is the one with the highest alignment score to the anchor sequence. Each subsequent sequence is the one with the best alignment score to the previous sequence among the sequences that have not yet been ordered. If no significant alignment is found, the next available sequence in the original input is selected.

[0255] Overview of the LncLOOM method Once the sequence order is established, LncLOOM identifies a set of short, conserved k-mer combinations for various values ​​of k by reducing each nucleotide sequence to a k-mer sequence, each represented by a node in the graph. Identical k-mers in adjacent sequences are linked in the graph using additional constraints (Figures 11A-D) and integer linear programming (ILP) to find a set of long, non-crossing paths within these graphs. The set of paths identified within each graph is used in subsequent iterations to define constraints on the graph and to partition the graph (an example of graph partitioning is shown in Figure 12). Starting with the largest k and decreasing iteratively, LncLOOM constructs an initial main graph for every k-mer length within a given range. The main graph is constructed over all ordered sequences in the dataset and then pruned layer by layer into a set of subgraphs (until only the top two sequences remain), with the ILP problem for each subgraph solved independently. At any given depth, the subgraphs may be partitioned into an additional set of even smaller subgraphs based on the paths found in previous iterations. In practice, this technique allows for prioritizing the identification of longer, more deeply conserved motifs over shorter, less conserved ones, keeping the size of the ILP program below 1,000 edges so that it can be solved quickly, and keeping the overall execution time of LncLOOM to a few minutes even when applied to dozens of long sequences.

[0256] Graph Construction Given a dataset of lncRNA sequences derived from species D and the kmer length k (k-mer length k) (6-15nt), LncLOOM constructs a directed graph G=(V,E), where V is the set of all nodes in the graph and E is the set of edges. The graph consists of D layers, where D is the number of sequences in the dataset. Each sequence is located in a layer (L1, L2...L D Modeled as, layer L i(corresponding to an array of length N(i)) consists of nodes (v1, v2... v N(i)-k+1 ), where each node v n represents the k-mer at position n within the i-th array (Figure 1B). All pairs of nodes that represent the same k-mer and are found in consecutive layers (for j = i + 1, L i and L j ) are connected by an edge x uv = (u, v), where u ∈ L i■ and v ∈ L j■ . Since each substring typically appears multiple times within one array, the number of edges can potentially far exceed the number of nodes within the graph. The deeply conserved ordered combinations of k-mers do not cross (i.e.,

[0257]

Number

[0258] ), and correspond to long paths within G that have nodes in L1. Thus, the goal is to find a set S within E such that each edge is reachable from L1 via an edge within S, and no two edges within S cross. Ideally, it is desirable to find the largest S that is subject to potential additional constraints. For example, short paths may not be desirable, and thus this requires that any edge within S is found on a path that reaches a specific layer.

[0259] Identification of Long Non-Crossing Paths Using ILP In the ILP problem, each edge within G is represented by a variable x uv to which a value of 1 is assigned if (u, v) is within S. The objective function is defined to maximize |S| as follows.

[0260]

Number

[0261] The additional constraints imposed on this model are derived from several considerations. Firstly, LncLOOM aims to identify short, conserved kmers that appear in the same order within an LncRNA sequence. However, it is unlikely that a kmer will appear only once within each sequence. Therefore, the constraints applied to the ILP model should allow for complex pathways containing multiple repetitions of a single kmer in one or more layers, as long as they do not intersect with pathways of mismatched kmers that do not have the same depth (Figures 1B and 11A). To ensure the selection of non-crossing pathways, the following constraints are imposed on any pair of edge crossings between two consecutive layers:

[0262]

number

[0263] The above constraint considers only the starting position of each node, thus excluding intersecting edges that connect identical k-mers repeated within two consecutive layers. When a k-mer is repeated within both consecutive layers, a network of edges is constructed from each iteration-to-iteration connection (Figure 11B). This network of edges may invalidate the selection of other paths that are equally conserved but connect fewer k-mers. Therefore, it is important to impose this constraint on edges connecting identical k-mers because it facilitates the complex partitioning of paths into multiple non-intersecting paths scattered by paths of intrinsically present k-mers. However, if a network of edges connecting identical iterations is constrained by each other only when no other paths exist, the ILP solver can select any possible solution of edges from multiple iteration-to-iteration connections. This can lead to suboptimal exclusion of repeated k-mers during subsequent iterations of graph refinement (scenario shown in Figure 13B). To avoid this scenario, if there is at least one other path of equal depth that intersects the repeating k-mer network, the intersection constraint is imposed only on edges that connect identical k-mers.

[0264]

number

[0265] Here, Z and P represent the respective subsets of all direct successor and successor nodes of node v, y is the minimum depth requirement, and M is a sufficiently large constant (in practice, 100 was used). Under this constraint, from L1 to at least L y Only paths with continuous connectivity up to a certain point are selected. At the same time, this constraint allows for the selection of complex connected paths containing k-mers that are repeated tandem within one or more layers (Figure 1B).

[0266] In graph G, each layer L i A node (v) can start at any consecutive position in the sequence and have a length of k bases. 1、 v2...v N(i)-k+1 ) consists of. This is from set S to set S union However, this follows that it can be formed by merging edges that connect mutually overlapping adjacent nodes. When the ILP is solved, these overlapping nodes are combined into a single, longer kmer. This process may encounter scenarios where a set of adjacent kmers represents a region of sequence containing a single repeating string of bases (see Figure 1B for an example). The resulting merged kmer may then contain layer-specific insertions. To overcome this, L is used such that the start and length of the overlapping region are equal between two adjacent nodes in each layer. i or L j The following constraints are imposed on any pair of edge elements that connect adjacent k-mers that overlap in any of the following ways:

[0267]

number

[0268] ILP is a well-known NP-hard problem that poses a significant challenge to the scalability of LncLOOM for very long sequences or large datasets. To overcome this limitation, the framework includes several steps to reduce the complexity of ILP for each graph and to prioritize the selection of deeply conserved k-mers. These include graph pruning, graph partitioning based on simple paths, additional constraints on edge construction, and iterative refinement of complex non-intersecting paths.

[0269] Graph pruning The LncLOOM framework uses two pruning steps. The first step involves removing nodes corresponding to k-mers that are excessively repeated within one or more layers. The number of repetitions allowed per layer can be adjusted by the user, and if a smaller k (e.g., 6) is used, the edge density in longer sequences can be significantly reduced. For a given k-mer length, this step is performed during the construction of the initial graph for all sequences in the dataset, and then the removed nodes are excluded from any resulting subgraphs. The second pruning step is performed at each iteration of subgraph construction at a given level, removing any nodes that do not have a connected path from L1 to the current depth.

[0270] Divide the graph to reduce computational complexity. The constraints imposed on the ILP problem allow for the selection of simple or complex paths, where a simple path is defined as one containing only one node per layer. Simple paths should not intersect with shallower paths, and therefore consist of deterministically selected edges that present boundaries into which the graph can be divided into even smaller subgraphs that can be solved independently (Figure 12). Currently, these graphs are being solved sequentially, but in the future, there is room to use parallel computing to process larger datasets, provided that at least one simple path is found. The division is based on simple paths of the current k-quant length found at each level in each layer-by-layer iteration. Each subgraph consists of two simple paths τ with depth = y aand τ b It is constructed by selecting a subset of nodes located between and , where the boundary is from each layer L1 to L y-1 Regarding this, W={v} is defined as the end and start positions of the nodes within each path: n |q+k-1 <n,n+k-1<r,v q ∈τ a ,v r ∈τ b (The final layer is removed for the next iteration.) If k-mers of adjacent simple paths overlap, the k-mers are first combined, and the boundary is defined by the start and end positions of the longer combined k-mer.

[0271] Refinement of complex non-crossing paths In contrast to simple paths, complex paths, especially those selected in early iterations when the graph is unconstrained, can include branches that connect repeating k-mers. In an unconstrained graph, it is impossible to decipher which iterations will randomly appear in each layer. Therefore, complex paths are not used in subsequent iterations to constrain edge selection in the graph. Instead, the set S found in each iteration is divided into 1) a subset of simple paths used for partitioning and defining edge constraints, and 2) a subset of complex paths that are stored separately in subsequent iterations and sequentially refined. During refinement, complex paths are optimized to remove branches that intersect with newly discovered paths (Figure 12). Refinement of complex paths is performed in two stages during layer-by-layer exclusion. Firstly, before solving the subgraph spanning y layers, the individual graphs of only complex paths are used to define a subset LC of longer k-mers with depth = y. d=y And a subset C from the path of the k-mer length at that point in time that has a minimum depth y+1 (the complex path selected in the previous iteration at the k-mer length at that point in time). d>y It is constructed from the above. Then, according to the ILP problem, a subset C of the refined complex paths is constructed. refined It can be found. However, LC d=y C across any of the shallower paths within d>y To ensure the selection of any complex path within the system, the following additional constraints are imposed:

[0272]

number

[0273] Under this constraint, C d>y For each path τ within the set, at least one repeating k-mer is selected from L1. When this constraint is imposed along with the above constraint, the solution includes refined paths spanning at least y layers. Set C refined Once found, subgraphs of all k-mers of that length and depth are constructed. Then C refined Every path within is added to the subgraph at that point, C refined The ILP problem is solved by imposing additional constraints to favor the selection of each path τ within the graph. This solution is then split into a set of simple paths and a set of complex paths for subsequent iterations. LncLOOM also includes an option to memorize and refine the simple paths so that simple paths are preferred over longer, shallower k-mers with greater depth. However, when this option is applied, the graph is not split, and no constraints are imposed on edge construction in subsequent iterations. Therefore, this option is computationally expensive and can only be used to analyze small datasets with short sequences.

[0274] Using BLAST High-Scoring Pairs (HSPs) to Reduce Graph Complexity BLAST can be used as an optional step in the LncLOOM graph construction process. BLAST HSPs are local non-gap alignments between segments of sequences found in consecutive layers that have significant similarity. We use these HSPs to constrain edge construction so that any pair of nodes not contained within the same HSP between two consecutive layers are not connected. HSPs found by BLAST are redundant in that HSPs may overlap with each other and any segment may match multiple segments in the target sequence. With respect to any set of overlapping HSPs, only the most significant pair is included in the HSPs used for graph construction. Similarly, if one segment aligns with multiple segments in the target sequence, only the highest scoring alignment is included. These constraints derived from BLAST analysis can effectively reduce the number of possible paths in the graph and promote the correct placement of edges between layers where parts of the sequence are incomplete (Figure 1A).

[0275] Graph size limitations While the process includes steps to reduce the complexity of the ILP problem, in some scenarios the graph becomes too large to solve within a reasonable time. To address this bottleneck, the total number of edges in the graph is limited. By default, the maximum number of edges allowed in the ILP problem is 1200, but this can be set to any number greater than 50. If, during any iteration, the number of edges in the graph G exceeds the limit, the graph is split into a series of subclusters, each in which the ILP problem is solved individually. Starting with the path with the fewest edges (the fewest repeating k-mers), each individual graph is divided into each path τ in G and the C that intersects it. refined It is constructed using only the paths within. Then, the acceptable edges within this subcluster of G are optimized using ILP, and then C refined The map is updated to include these edges, and path τ is removed from G. This process ensures that all paths are C refinedThe process is either optimized individually for each path, or repeated for each remaining path in G until the number of edges in G reaches a maximum limit, at which point all remaining paths in G are optimized for each other in a single ILP problem. If the number of edges in the graph constructed from individual subclusters of intersecting paths exceeds a maximum limit, the ILP does not proceed, and C refined Only paths from this point are retained in the solution.

[0276] Discovery of motifs within the extended 5' and 3' regions of the sequence The input to LncLOOM may occasionally contain sequences with incomplete 5' or 3' ends. Since the dataset is ordered by homology, not completeness, these sequences may be found in any layer of the graph, potentially disrupting layer-wise connectivity of nodes within these regions. In this scenario, to reduce the possibility of losing preserved motifs, motif discovery is performed in three stages. In the first stage, LncLOOM identifies motifs from a primary graph constructed over all sequences in the dataset (sum of Dsequences). LncLOOM then determines which sequences potentially have elongated 5' or 3' ends by considering the positions of the first and final motifs within each sequence relative to the central position across all sequences (Figure 13A). Based on this, LncLOOM constructs and solves individual graphs of the elongated 5' and 3' regions of the more complete sequences in the dataset. To construct the 5' elongation graph, LncLOOM considers each layer L1~L D The first node inside

[0277]

number

[0278] The starting position is the center position M q First, calculate q. i >t·M q In this case, node W = {v n |n+k-1 i A subset of} is in each layer L i ​Extracted from, where t is some tolerance defined by the user. Nodes in the extended 3' graph are extracted based on the end position of the final motif relative to the length of each sequence. Specifically, LncLOOM is extracted for each layer L1~L D The last node within

[0279]

number

[0280] The central relative position M of the end position Re Calculate, here,

[0281]

number

[0282] That is. Next, Re i <M Re ·In the case of (1+t), node W={v n |n <r i A subset of {+k-1} is L in each layer i The motifs are extracted from the following. By default, t=0.5 is used for the extraction of both 5' and 3' graphs, but a tolerance can be defined independently for each graph. This step of motif discovery proceeds only if nodes from the extended region of the anchor array are included in the graph. For example, to avoid a scenario where shallowly preserved motifs prevent the identification of 5' or 3' cuts in deeper layers, such as when a motif found near the 5' end is preserved only in the first two layers, the "minimum depth" parameter can be applied to select the positions of the first and last motifs in each array from a subset of motifs preserved to a specified depth. When the minimum depth parameter is applied, any motifs that do not meet the specified depth requirement are also removed from the solution.

[0283] Motif module and neighborhood calculation Once the ILP problem is solved for every subgraph in the framework, each set of non-intersecting paths selected from the linear graph, 5'-extended graph, and 3'-extended graph is processed into motif modules and neighborhoods. A motif module is defined as an ordered combination of at least two unique motifs stored in a set of arrays, where each motif can have any number of tandem iterations. By default, modules are L1 through L i By extracting paths that span all layers up to L, we can access every layer of the graph, L i The calculation is performed for |2≦i≦D|. If a minimum depth d is specified in the parameters, the module is used for any layer L. i |d≦i≦D| is calculated. As described above, motif discovery is performed by an iterative process of exclusion layer by layer. This selects longer identity regions as the set of sequences is successively reduced to include more closely related sequences. As a result, shorter motifs that are more deeply preserved are often embedded in longer motifs that are preserved only between the top layers (Figure 13B). We define these regions in the graph as motif neighborhoods, where each neighborhood includes all nodes in the graph, connected to a single region of overlapping nodes in L1, along with the adjacent regions of each node in each layer. To compute motif neighborhoods, LncLOOM first combines all overlapping nodes in L1 to form a set of reference k-mers representing each neighborhood. For each reference k-mer, all paths connected to each shorter k-mer embedded within the reference k-mer are included in its neighborhood. For each motif in each layer, the length of the adjacent region is calculated for the position of the motif within the reference k-mer (Figure 13B). The motif modules and neighborhoods from the linear graph, 5' elongated graph, and 3' elongated graph are presented in HTML and plain text file formats.

[0284] Calculation of motif significance Motif significance is inferred by calculating the empirical p-value for each motif within two genres of a random dataset. Firstly, L iFor a motif of length k stored in the input sequence, the inventors have found that L is a set of random sequences having the same percentage identity among consecutive layers observed in the input sequence. i Within the dataset, the empirical probability of finding the exact motif found in any combination of the actual dataset and any number of motifs of the same length or more is determined at least once. This is achieved by using MAFFT to generate an MSA of the input sequence, and then running multiple iterations (100 for the analysis described in this manuscript) of LncLOOM iterations in which the columns of the MSA are randomly shuffled. Secondly, the inventors determine the L of a set of random sequences generated such that each layer has the same length and the same dinucleotide composition as the corresponding layer in the input sequence. i We determined the empirical probability of finding at least once within a layer an exact motif and any combination of any of the motifs of the same length and number (without preserving % identity between layers). In the analysis described in this manuscript, only the p-value of the former was used. Multiprocessing was performed to run the iterations in parallel.

[0285] Functional annotation of motifs LncLOOM has annotation functions that exist for two purposes. First, discovered motifs can be mapped to miRNA binding sites by identifying complete base pairings with seed regions of conserved (mammalian-conserved) and widely conserved (typically found throughout vertebrates) miRNAs from TargetScan. For each motif, the type of pairing (hexamer, heptamer, heptamer-A1, heptamer-M8, or octamer) is determined in each sequence by considering the motif together with the directly adjacent bases from both sides of the motif. A match is found only when the complete seed region (hexamer) directly matches the motif. Second, motifs found in genes expressed in HepG2 or K562 cell lines can also be mapped to RBP binding sites identified by eCLIP in the ENCODE project. To determine the chromosomal coordinates of each motif in the selected query sequence, LncLOOM aligns the sequence to the genome using BLAT (Kent, 2002) and then calculates the overlap with the coordinates of the RBP binding sites extracted from the ENCODE bigBed file using the pyBigWig package. Alternatively, the user can upload a bed file specifying the chromosomal coordinates and lengths of each exon in the query sequence. The extracted eCLIP data is filtered to exclude any peaks with enrichment < 2 beyond the mock input. RBPs that bind to a large portion of the anchor sequence are marked because the overlap between their binding peaks and any conserved motif is unlikely to be functionally relevant to that particular motif.

[0286] Implementation and Availability of LncLOOM Graph construction is performed using the networkx package. The integer programming problem is modeled using PuLP and solved by either the open-source COIN-OR Branch- and -Cut solver (CBC) (www(dot)coin-or(dot)org / ) or the commercial Gurobi solver (www(dot)gurobi(dot)com / ). LncLOOM utilizes the following alignment programs, namely BLAST, BLAT, and MAFFT, during graph construction, motif annotation, and empirical evaluation of motif significance. The multiprocessing python package is used to compute statistical iterations in parallel.

[0287] Calculation of motif enrichment To evaluate the enrichment of specific motifs within a sequence, the inventors generated 1,000 sets of random sequences matching the dinucleotide composition of the input sequence, counted the occurrences of the motifs, and calculated the predicted number of motifs and the empirical p-value.

[0288] LncLOOM analysis of lncRNA and 3'UTR Using LncLOOM, we analyzed Cyrano sequences from 18 species, libra sequences (Nrep in mammals) from 8 species, Chaserr sequences from 16 species, DICER1 sequences from 12 species, and PUM1 and PUM2 sequences from 16 species. For all genes, LncLOOM parameters were set to search for kmers of 15–16 nucleotide lengths, and the sequences were reordered by BLAST using human sequences defined as anchor sequences in each case. No HSP constraints were imposed. Motif significance was calculated over 100 repeats. Table 1 shows the sequence order of each gene as a representation in the LncLOOM framework.

[0289] Additionally, LncLOOM was used to analyze 2,439 3'UTR genes. TargetScan7.2 miRNA target site prediction suite 10A dataset was constructed from 3'UTR MSAs generated by [tool name], and the dataset contained human, mouse, dog, and chicken sequences ranging from 300 to 3,000 nts. Depending on availability and length (>200 nucleotides), sequences from frogs, sharks, zebrafish, gars and lampreys, cioan, and flies were obtained from Ensembl and added to their respective gene datasets. For each dataset, BLASTN was used with a cutoff E value of 0.05 to classify which sequences from each species did not have a detectable alignment to their human orthologues, as well as sequences that did not align to mouse, dog, and chicken. kmers identified by LncLOOM were fitted to seeds of the widely conserved miRNA family reported by TargetScanHuman as hsa-miRNA. To evaluate the sensitivity of LncLOOM, we compared the widely conserved miRNA binding sites identified by LncLOOM with the predictions reported by TargetScan (www(dot)targetscan(dot)org / cgi-bin / targetscan / data_download.vert72.cgi). Specifically, we compared only miRNA sites from genes for which TargetScan reported the same representative human transcript sites used in this LncLOOM dataset. In total, this represented 2,359 out of 2,439 genes.

[0290] tissue culture 10% fetal bovine serum and 100 U penicillin / 0.1 mg ml -1 Neuro2a cells (ATCC) were routinely cultured in DMEM containing streptomycin in a humidified incubator at 37°C and 5% CO2. The cells were routinely tested for mycoplasma contamination, but authentication was not performed.

[0291] Mass spectrometry sample preparation As previously mentioned 47The samples were subjected to trypsin digestion in solution using suspension trapping (S-trap). In summary, pull-down proteins were eluted from the beads using 5% SDS in 50 mM Tris-HCl. The eluted proteins were reduced with 5 mM dithiothreitol and alkylated in the dark with 10 mM iodoacetamide. Each sample was loaded onto an S-Trap microcolumn (Protifi, USA) according to the manufacturer's instructions. After loading, the samples were washed with 90:10% methanol / 50 mM ammonium bicarbonate. The samples were then digested with trypsin at 47°C for 1.5 hours. The digested peptides were eluted using 50 mM ammonium bicarbonate. Trypsin was added to this fraction and incubated overnight at 37°C. Two further elutions were performed using 0.2% formic acid and 0.2% formic acid in 50% acetonitrile. The three eluates were pooled together, centrifuged under vacuum, and dried. The sample was kept at -80°C until further analysis was performed.

[0292] Liquid chromatography ULC / MS grade solvents were used for the entire chromatography process. Dried digested samples were dissolved in 97% H2O / acetonitrile + 0.1% formic acid. Each sample was loaded using a splitless nano-Ultra Performance Liquid Chromatography (10 kpsi nanoAcquity; Waters, Milford, MA, USA). The mobile phases were A) H2O + 0.1% formic acid and B) acetonitrile + 0.1% formic acid. Sample desalting was performed online using a reversed-phase Symmetry C18 trapping column (180 μm inner diameter, 20 mm length, 5 μm particle size; Waters). Peptides were then separated using a T3 HSS nanocolumn (75 μm inner diameter, 250 mm length, 1.8 μm particle size; Waters) at 0.35 μL / min. The peptide was eluted from the column to the mass spectrometer using the following gradient: 4% to 30% B at 55 minutes, 30% to 90% B at 5 minutes, maintained at 90% for 5 minutes, and then returned to the initial conditions.

[0293] mass spectrometry Using a FlexIon nanospray system (Proxeon), a nanoUPLC was coupled online to a quadrupole orbital mass spectrometer (Q Exactive HF, Thermo Scientific) via a nanoESI emitter (10 μm tip; New Objective; Woburn, MA, USA).

[0294] Data was acquired in data-dependent acquisition (DDA) mode using the Top10 method. MS1 resolution was set to 120,000 (200 m / z), mass range to 375-1650 m / z, AGC to 3e6, and maximum injection time to 60 milliseconds. MS2 resolution was set to 15,000, quadrupole isolation to 1.7 m / z, AGC to 1e5, dynamic exclusion to 20 seconds, and maximum injection time to 60 milliseconds.

[0295] Processing and analysis of mass spectrometry data Raw data was processed using MaxQuant v1.6.6.0. Data was searched using the Andromeda search engine against the mouse (Mus musculus) protein database downloaded from Uniprot (www(dot)uniprot(dot)com), and common experimental protein contaminants were appended. Enzyme specificity was set to trypsin, allowing for a maximum of two cleavage errors. Immobilization modification was set to cysteine ​​carbamide methylation, and variable modification was set to methionine oxidation and protein N-terminal acetylation. Peptide precursor ions were searched with a maximum mass deviation of 4.5 ppm, and fragment ions with a maximum mass deviation of 20 ppm. A decoy database strategy (MaxQuant's "Revert" module) was used to filter peptide and protein identifications with a 1% FDR. The minimum peptide length was 7 amino acids, and the minimum Andromeda score for modified peptides was 40. The run-to-run agreement option was checked to ensure peptide identification was reflected across the entire sample. The unlabeled quantification option was selected for the search. Quantitative comparisons were calculated using Perseus v1.6.0.7. Decoy hits were excluded. Student's t-test was used after logarithmic transformation to identify significant differences between experimental groups across biological replicas. Scale changes were calculated based on the ratio of geometric means of different experimental groups.

[0296] RNA pull-down assay A template for in vitro transcription was generated by amplifying a synthetic oligo (Twist Bioscience) and adding the T7 promoter to the 5' end of the sense sequence and to the 3' end of the antisense control sequence (see Table 2 for the complete sequence). Biotinylated transcripts were generated using the MEGAscript T7 in vitro transcription reaction kit (Ambion) and a biotin RNA labeling mixture (Roche). Template DNA was removed by treatment with DNase I (Quanta). Neuro2a cells (ATCC) were lysed on ice for 15 minutes using RIPA supplemented with a protease inhibitor cocktail (Sigma-Aldrich, #P8340) + 100 U / ml RNase inhibitor (#E4210-01) and 1 mM DTT. The lysates were clarified by centrifugation at 21130 × g for 20 minutes at 4°C. Streptoavidin magnetic beads (NEB #S1420S) were washed twice with buffer A (0.1M NaOH and 0.05M NaCl), once with buffer B (0.05M NaCl), and then resuspended in two binding / washing tubes (1M NaCl, 5mM Tris-HCl pH 7.5, and 0.5mM EDTA supplement containing PI + 100U / ml RNase inhibitor and 1mM DTT). One tube of beads was washed three times in RIPA supplemented with PI and 1mM DTT, then cell lysates were added and the tube was pre-clarified by rotating it upwards at 4°C for 30 minutes. The second tube was equally divided into individual tubes for each RNA probe. Then, 2–10 pmol of biotinylated transcript was added to each tube and rotated upwards at 4°C for 30 minutes. Next, the beads were washed three times with binding / wash buffer, and then equal volumes of pre-clarified cell lysates were added to each bead and RNA probe sample. The samples were then rotated upwards at 4°C for 30 minutes. After rotation, the beads were washed three times with high-salt CEB (10 mM HEPES pH 7.5, 3 mM MgCl2, 250 mM NaCl, 1 mM DTT, and 10% glycerol). Proteins were then eluted from the beads in 5% SDS in 50 mM Tris pH 7.4 over 10 minutes at room temperature.

[0297] Antisense oligonucleotide and LNA GapmeR transfection An ASO (Integrated DNA Technologies) was designed to target the conserved ATGG site identified by LncLOOM in the final exon of mouse Chaserr (Figure 8A). The entire ASO was modified with a 2'-O-methoxyethyl base. An LNA gapmer (Qiagen) targeting the Chaserr intron was used for Chaserr knockdown (see Table 3 for the complete oligo sequence). Transfection: 2 × 10⁶ 5 Neuro2A cells were seeded in 6-well plates and transfected with lipofectamine 3000 (Life Technologies, L3000-008) to a final concentration of 25 nM using a mixture of LNA1-4, or ASO1, ASO2, and ASO3, or a mixture of ASO1 and ASO3, or a mixture of ASO1-3, according to the manufacturer's protocol. The endpoint of all experiments was 48 hours after transfection, after which cells were harvested using TRIZOL for RNA extraction and evaluation by RT-qPCR analysis.

[0298] RNA immunoprecipitation (RIP) Neuro2a cells (ATCC) were harvested, centrifuged at 94×g for 5 minutes at 4°C, and washed twice with ice-cold phosphate-buffered saline (PBS) supplemented with a ribonuclease inhibitor (100 U / mL, #E4210-01) and a protease inhibitor cocktail (Sigma-Aldrich, #P8340). Next, the cells were lysed on ice for 10 minutes in 1 mL of lysis buffer (5 mM PIPES, 200 mM KCl, 1 mM CaCl2, 1.5 mM MgCl2, 5% sucrose, 0.5% NP-40, supplemented with protease inhibitor cocktail + 100 U / mL RNase inhibitor and 1 mM DTT). The lysates were sonicated three times at 30% amplitude with 1 second ON, 30 seconds OFF (Vibra-cell VCX-130), followed by centrifugation at 21130×g for 10 minutes at 4°C. Next, the supernatant was transferred to a new 2 mL tube and supplemented with 1 mL of IP-binding / washing buffer (150 mM KCl, 25 mM Tris (pH 7.5), 5 mM EDTA, 0.5% NP-40, supplemented with a protease inhibitor cocktail + 100 U / ml RNase inhibitor and 0.25 mM DTT). Then, the sample was rotated at 4°C for 2–4 hours with 5 μg of antibody per reaction cycle. 50 μl of GenScript A / G beads (#L00277) per reaction cycle were washed three times with IP-binding / washing buffer, and then added to the lysate and incubated overnight with rotation. After incubation, the beads were washed three times with IP-binding / washing buffer. 10% of each sample was collected and boiled at 95°C for 5 minutes for subsequent analysis by Western blotting. For RNA extraction and evaluation by RT-qPCR analysis, the remaining beads were resuspended in 0.5 mL of TRIZOL, and the immunoprecipitation material was normalized relative to the entire cell lysate.

[0299] Western blot Protein samples collected from RIP were separated on 8-10% SDS-PAGE gels and transferred to polyvinylidene fluoride (PVDF) membranes. After blocking with 5% skim milk in PBS containing 0.1% Tween-20 (PBST), the membranes were incubated with a primary antibody, followed by a secondary antibody conjugated with horseradish peroxidase. Blots were quantified using Image Lab software. The primary antibody used was anti-Dhx36 (Bethyl, #A300-525A, 1:1,000 dilution), and the secondary antibody was anti-rabbit (JIR#111-035, 1:10,000 dilution).

[0300] qRT-PCR Total RNA was extracted from transfected N2a cells using TRIREAGENT (MRC) according to the manufacturer's protocol. cDNA was synthesized using the qScript Flex cDNA synthesis kit (95049, Quanta) with random primers. Fast SYBR Green master mix (4385614) was used for qPCR. Gene expression levels were normalized for the housekeeping genes Actin and Gapdh.

[0301] [Table 1]

[0302] [Table 2]

[0303] [Table 3]

[0304] [Table 4]

[0305] Example 1 LncLOOM framework LncLOOM receives a collection of sequences that are putatively homologous to the genome sequence of interest. In one embodiment, the focus is on lncRNA and 3'UTR, but other elements such as enhancers can be used just as readily. In the case of lncRNA, only the exon sequence is used for motif identification, but LncLOOM visualizes the location of exon-exon junctions. The input sequences are provided in a specific order that can be automatically set based on sequence similarity, ideally coinciding with the evolutionary distance between species (Figure 1A). The precise definition of the data structure and algorithm used in LncLOOM appears in "Materials and Methods," and an overview of the framework is shown in Figures 1A-B. LncLOOM represents each RNA sequence as a "layer" of nodes in a network graph (Figure 1B), where each node represents a short k-mer (e.g., k from 6 to 15). The order of the layers reflects the evolutionary distance of the input sequence from the query sequence placed in the first layer of the graph (human in the analysis described herein), and sequences from other species are placed in additional contiguous layers of the graph. Edges in the graph connect nodes that have the same k-quanta in consecutive layers. It is understood that it is also possible to connect "similar" k-quanta. Under these definitions, the objective is to identify combinations of long "paths" in the graph that do not intersect each other, and thus connect short motifs that maintain the same order in various sequences. Since the objects of interest are typically motifs that reside in the top layer, it is a necessary condition that the paths begin within them. For k=1, the problem of identifying the maximum set of such paths is computationally difficult, as it is equivalent to the longest common subsequence problem. However, the results so far show that this can be transformed into a problem of solving integer linear programming (ILP), and while finding an optimal solution for integer linear programming (ILP) is computationally difficult, efficient solvers are available (Figure 1B and "Methods").

[0306] Once the graph is constructed, the process begins by identifying the pathways with the largest k values, and then using these pathways to constrain the possible locations of pathways with smaller k values ​​(if found). This technique allows for the preference of longer conserved elements, but also allows for the identification of significantly conserved shorter k-mers. Once all k values ​​have been tested, the resulting graphs are merged to obtain combinations of motifs and the depths in which they are conserved. To calculate the statistical significance of motif conservation, an MSA of the input sequence is generated, and the alignment column is shuffled to derive random sequences with similar internal similarity structures to the input sequence. The full LncLOOM pipeline is then applied to these sequences, and for each motif found in the original input sequence that is conserved in layer D, the empirical probability of identifying a combination of either the exact same motif or the same number of motifs of its length that is conserved in layer D is applied. Additional P-values ​​are calculated for less stringent controls, where random sequences with the same dinucleotide composition are generated, but the intersequence similarity structures are not conserved.

[0307] A rich HTML-based suite is used to visualize these motifs in various ways, for example, by color-coding them based on their depth of conservation and highlighting the motifs in both query sequences and other sequences (see Figures 3A-E and 4 for examples of LncLOOM output). The LncLOOM output also includes a custom color-coded track of motifs identified within the query sequence, which can be viewed in the UCSC Genome Browser. Motifs are annotated using a set of conserved microRNA seed sites (from TargetScan) and RBP binding sites found in eCLIP data from the ENCODE project.

[0308] Example 2 LncLOOM identifies deeply conserved elements within Cyrano lncRNAs. Cyrano lncRNAs are widely and highly expressed lncRNAs. 12、13Despite being conserved across vertebrates, Cyrano exhibits approximately 5-fold variation in overall exon length (from 2,340 nt in medaka to 10,155 nt in opossums, Figure 2A). The previously identified 67 nt highly constrained element of Cyrano is the only region for which BLAST reports significant similarity when comparing zebrafish and human sequences. Furthermore, the entire Cyrano locus is not alignable between mammals and fish in 100-way whole genome alignment (UCSC Genome Browser). The highly conserved element contains a very broadly complementary range of miR-7 binding sites required for miR-7 degradation by Cyrano.

[0309] To identify additional conserved elements, Cyrano sequences were curated from 18 species where available RNA-seq data could be located, including 8 mammal species, chickens, *X. tropicalis*, 7 vertebrate fish species, and the elephant shark (not shown). LncLOOM identified 7 elements conserved across all species, 9 elements conserved across all species except sharks (Figure 2B), and 37 motifs conserved across all mammals. The following study focuses on the 9 elements conserved across all species except sharks (numbered 1-9 in Figure 2B).

[0310] AUGGCG (Sequence ID 17) UGUGCAAUA (Sequence ID 18) ACAAGU (Sequence ID 19) CAACAAAAU (Sequence ID 20) GUCUUCCAUU (Sequence ID 21) UGUAUAG (Sequence ID 22) UGCAUGA (Sequence ID 23) CUAUGCA (Sequence ID 24) GCAAUAAA (Sequence ID 25)

[0311] Seven of these were found to be statistically significant by both LncLOOM tests (P<0.01) (see Materials and Methods). Only elements 3-6 fall within a 67nt conserved region identifiable by BLAST, including two corresponding to pairing with the 5' and 3' of miR-7 (Figure 2C), as well as another UGUAUAG (SEQ ID NO: 22) similar to the Pumilio Recognition Element (PRE, element #6). This element actually binds to PUM1 and PUM2 in CLIP data from humans and mice (Figures 2D-E), and in the brains of neonatal mice with relatively high Cyrano levels, depletion of Pum1 and Pum2 is consistent with the function of these proteins in RNA decay. 15 This leads to increased Cyrano expression (adjusted P-value 3.49 × 10⁻⁶). -3 , 14 Data from Figure 2E). This suppression is likely due to a combined effect of this highly conserved PRE with others. Eighteen Cyrano sequences from different species had an average of 3.2 consensus PREs (P<0.001, compared to an average of 1.3 in 1,000 randomly shuffled sequences, including 2 in mouse sequences; see "Methods").

[0312] Several additional conserved elements within the Cyrano sequence, identified by LncLOOM, can be assigned putative biological functions. The necaper UGUGCAAUA (element #2 in Figure 2B, SEQ ID NO: 35), conserved across all 18 input species, is found approximately 60 nt upstream of the miR-7 binding site, outside the region alignable by BLAST. This element corresponds to a seed match for the miR-25 / 92 family (Figure 2C), and has recently been shown to be bound and regulated by members of the miR-25 / 92 family in mouse embryonic hearts. 16At the 3' end of Cyrano, one conserved element (SEQ ID NO: 25, GCAAUAAA) corresponds to the Cyrano polyadenylation signal (PAS) and the miR-137 site. Another sequence, CUAUGCA (SEQ ID NO: 24), found approximately 100 nt upstream of the PAS, corresponds to a seed match for miR-153, and this region is bound by Ago2 in the mouse brain (Figure 2E). Interestingly, Cyrano levels in HeLa cells decrease by 41% and 11% after transfection with miR-137 and miR-153, respectively. 17 Therefore, beyond the reported interactions with miR-7 and miR-25 / 92, Cyrano is under highly conserved regulation by additional microRNAs.

[0313] Approximately 55 nt downstream of the conserved Pumilio binding site is a conserved WGCAUGA motif (W=A / U, SEQ ID NO: 27) that coincides with the consensus binding motif of Rbfox RBP. This motif, as well as an additional region containing an example of WGCAUGA in the 3' half of Cyrano, is bound by Rbfox1 / 2 in mice (Figure 2E). Indeed, analysis of 18 Cyrano species showed significant enrichment of WGCAUGA (9.8 examples predicted by chance vs. 4.5 examples, P<0.001, see "Methods"). In contrast to the miRNA binding site and Pumilio binding site, examination of various RNA-seq datasets of Rbfox1 / 2 loss of function did not identify any effect on Cyrano levels (not shown), suggesting that widespread and conserved binding by Rbfox1 / 2 may affect Cyrano functionality rather than expression.

[0314] Another highly conserved hexamer, AUGGCG (SEQ ID NO: 17), is found precisely at the 5' end of Cyrano. Examination of Cyrano sequences and Ribo-seq data from human, mouse, and zebrafish revealed that this hexamer corresponds to the first two codons of a conserved short 2-3aa ORF (Figure 2F). A clear ribosome association is observed at the 5' end of Cyrano in this ORF, and a very limited number of ribosome-protective fragments are observed downstream of this element in both humans and zebrafish (Figure 2F), suggesting efficient translation and ribosome release in this short ORF. The situation of the AUG start codon within the ORF perfectly matches the 12 bases of the TISU motif, a regulatory element that affects both transcription and translation. TISU, located at the 5' end of the transcript, acts as a YY1 binding site that can determine the transcription start site, and as a highly efficient and precise cap-dependent translation initiator element for translation that operates without scanning. 18、19 The genomic region of this motif exhibits strong YY1 binding to DNA (Figure 2F). It is suggested that this motif may have a dual function: as a YY1 element regulating Cyrano expression, as suggested for other lncRNAs, and as the initiation of a short ORF that may contribute to Cyrano function. 20 Overall, the presumed biological functions can be assumed for eight of the nine conserved elements in Cyrano: four as miRNA binding sites, two as RBP binding sites, one as a conserved short ORF, and one as a PAS. These elements are separated by long stretches of non-conserved sequences (Figure 2B), which highlights the ability of LncLOOM to be combined with annotation and orthogonal data to elucidate lncRNA biology.

[0315] Example 3 LncLOOM identifies deeply conserved elements within libra lncRNA. As another example of LncLOOM's ability to find conserved elements in transcripts, which is known to be relevant to miRNA biology, we applied this to eight homologs of libra lncRNA in zebrafish and to the mammalian Nrep protein. This is one of the few examples of a gene that has transformed from a possible ancestral lncRNA into a protein-coding gene while retaining substantial sequence homology within its 3' region. 12、21 Libra causes the degradation of miR-29b in zebrafish and mice via highly conserved and highly complementary sites. 21 When comparing zebrafish (libra) sequences with human and mouse sequences using BLASTN, approximately 250 nt of alignment was recovered from a human sequence of about 2.2 kb, and even shorter significant alignments were found in spotted gar (E value < 0.001). LncLOOM identified 17 elements conserved across all species and >25 elements conserved across all species except zebrafish (Figure 6). These included miR-29 sites and conserved binding sites for eight additional miRNAs, three of which were found outside the alignment region between mammalian and fish species by BLAST (Figure 6). Thus, Cyrano and libra, two lncRNAs that have been shown to effectively induce target-directed miRNA degradation (TDMD), possess several additional highly conserved miRNA binding sites, but these appear to be “regular” seed sites that can affect lncRNA levels rather than miRNA levels, in contrast to TDMD-mediated sites.

[0316] Example 4 LncLOOM identifies conserved motifs within CHASERR lncRNAs. To test LncLOOM's ability to identify conserved modules within sequences, which are unsuitable for BLAST comparison, the inventors focused on CHASERR, an lncRNA recently characterized as essential for mouse survival. 27CHASERR homologs can be easily identified in various species based on their proximity (<2kb) to the transcription start site of CHD2 and their characteristic 5-exon gene structure. 27 The inventors manually curated CHASERR sequences from 16 vertebrate species, ranging in length from 579 to 1313 nt, four of which are likely to be 5' incomplete, due to the extremely G / C-rich promoter of CHASERR and gaps in some of the genomic assemblies around the first exon. 27 (Figure 7). BLASTN found significant alignment (E-value < 0.01) between human CHASERR and nine sequences derived from amniotes, but not in any of the other six vertebrate species. Conversely, when zebrafish sequences were used as queries, BLAST found homology only in other fish species and opossums. When CHASERR sequences are supplied to ClustalO MSA... 28 Only three identical positions are found. Therefore, the limited preservation of CHASERR poses a challenge to analysis using tools commonly used for comparative genomics.

[0317] LncLOOM identified two kmers conserved throughout the entire layer: AAUAAA (sequence number 3) at the 3' end corresponding to PAS, and AAGAUG (sequence number 2) (motif 1 in Figure 3A) found once or twice in the final exon of the entire CHASERR sequence. The AAUAAA (SEQ ID NO: 1) motif was found near the 3' end of CHASERR and was most likely to correspond to a polyadenylation signal (PAS), so it was not further tested. Examination of the CHASERR sequence revealed that the AAGAUG motif (SEQ ID NO: 5) was substantially overpresented. CHASERR homologs had an average of 2.1 motifs compared to the only 0.45 predicted by chance (P<0.01). The motif saturation was also typically similar across these 34 instances, with motifs typically followed by purines (Figure 3B). The clearly related motif AUGG (Motif 2 in Figure 3A) (SEQ ID NO: 2) was conserved in 11 sequences. Including adjacent sequences, motif 2 shares an ARAUGR core with motif 1 (Figure 3B). These sequences did not appear to be consistent with any known binding preferences of RBPs, and examination of eCLIP data did not reveal any obvious binder candidates. Therefore, the functionality of these sequences was further explored experimentally.

[0318] To test the functional significance of the conserved elements, we designed antisense oligonucleotides (ASOs) complementary to three examples of conserved motifs in mouse Chaserr (Figure 8A) and transfected mouse Neuro2a (N2a) cells. It has been previously shown that Chaserr depletion results in increased levels of Chd2 RNA and protein. 27 The human sequences corresponding to these ASOs are CCATAGTAGACTGCCATCTT (sequence number 7), which targets AAGATGGCAGTCTACTATGG (sequence number 12), and ATCCACTGTCCATTTGTG (sequence number 9), which targets CACAAATGGACAGTGGAT (sequence number 10).

[0319] Transfection with ASO1 and ASO3 individually or in combination resulted in a significant increase in Chd2 levels comparable to that caused by Chaserr knockdown (Figure 3C). Interestingly, ASO treatment resulted in an increase in Chaserr levels, as assessed by RT-PCR primer pairs found either upstream or downstream of the ASO targeting region (Figure 3C).

[0320] To identify proteins that potentially bind to conserved regions, we used in vitro transcription to produce biotinylated RNAs containing the WT sequence of the final exon of Chaserr, the same sequence with AUGG→UACC mutations in four conserved motifs, and a second mutant in which all seven AUGG sites in the final exon were mutated to UACC (Figure 8A). These sequences were incubated with lysates derived from N2a cells along with their antisense controls, and proteins associated with various RNA mutants were isolated and identified using mass spectrometry. As is typical in these experiments, a large number of proteins, 938 (not shown), were identified as being associated with the WT sequence, of which 74 were enriched more than three times with the antisense sequence, but only 9 of these had more than twice the recovery of both mutants when using the WT sequence (Figure 3D). We then investigated publicly available RNA-seq datasets to look for evidence of changes in Chd2 and / or Chaserr levels when these proteins were disrupted. Such evidence was available for DHX36 and ZFR (Figure 8B-C). A significant association between Chaserr and DHX36 (the protein showing the highest enrichment compared to the mutant sequence) was examined using RNA immunoprecipitation (RIP) and specific antibodies (Figure 3D). Interestingly, DHX36 is known to bind to G-quadrivalent sequences. 29、30 The conserved elements do indeed contain GG pairs, but they are quite far apart from each other, and a typical G-quadruplicate contains at least 3G runs. QGRS Mapper 31While it predicts one G quadruple chain within Chaserr's final exon (Figure 8A), G4RNA scanners incorporating different scoring systems 32 Other tools, including this one, did not find any high-scoring G quadruples within Chaserr's final exon. Non-canonical G quadruple formation may occur within this sequence, or it may have a different recognition mode by DHX36.

[0321] Therefore, LncLOOM can identify functionally relevant elements within lncRNAs, which can serve as a basis for designing targeted reagents to disrupt lncRNA function, and enables the use of proteomic methods to identify specific and functionally relevant lncRNA interaction partners.

[0322] Example 5 Deeply conserved elements within the 3'UTR of DICER1 mRNA and Pumilio mRNA The inventors then sought to evaluate the applicability of LncLOOM for comparing sequences over longer evolutionary distances beyond lncRNA. The 3'UTR can determine the RNA stability and translation efficiency of mRNA and evolves much more rapidly and typically than other mRNA regions. 34Orthology between 3'UTRs is fairly easy to define based on their adjacent coding sequences, which are often readily comparable over very long evolutionary distances. However, there are few known examples of long-range conservation of functional elements within the 3'UTR between vertebrates and invertebrates. To test 3'UTR conservation using LncLOOM, we first focused on genes that act during post-transcriptional regulation, as they typically undergo particularly complex post-transcriptional regulation. Using available RNA-seq and expression sequence tag (EST) data, we compiled a collection of 3'UTR sequences of DICER1 encoding key components of the miRNA pathway from 12 species, including eight vertebrates, lancelets, lampreys, sea urchins, C. intestinalis, and two DICERs in Drosophila. BLASTN allowed us to align human DICER1 to 3'UTRs derived from vertebrate species, but not beyond. LncLOOM identified 15 elements conserved across all vertebrate sequences, six of which were of lengths not found in random sequences (P<0.01, Figure 9). Eight of the conserved motifs were conserved beyond vertebrates (and could not be evaluated by MSA or BLAST), and one corresponding to a conserved miR-219 binding site was found in all species, including the Dicer2 3'UTR of the fly.

[0323] Next, the inventors focused on the 3'UTR of PUM1 mRNA and PUM2 mRNA, which encode Pumilio proteins that repress gene expression after transcription. Pumilio proteins are deeply conserved, with two Pumilio proteins, PUM1 and PUM2, existing in vertebrates, and a single ortholog in other chordates and flies. 3'UTR sequences from 12 vertebrates and 4 invertebrates (lamprey, lancelet, C. intestinalis, and Drosophila) were curated. The human and zebrafish 3'UTRs were readily alignable by BLASTN, and significant homology was even found between the human PUM1 3'UTR and the lamprey and lancelet Pumilio mRNA 3'UTRs, but not between those of flies and C. intestinalis. LncLOOM identified eight elements conserved across the entire vertebrate PUM1 3'UTR, one of which, UGUACAUU (SEQ ID NO: 14), was conserved in all 16 3'UTRs analyzed, including the fly pum 3'UTR (Figure 4, top). In PUM2, three elements were found conserved across vertebrates, including UGUACAUU, which was found in the entire sequence (Figure 4, bottom). Interestingly, this UGUACAUU motif partially matches the pre-consensus UGUANAUA (SEQ ID NO: 28) and is linked by both PUM1 and PUM2 in human ENCODE data, suggesting that this ancient element is part of a self-regulatory program known to exist in Pumilio mRNA. 15 Therefore, LncLOOM can identify deeply conserved elements within 3'UTR sequences, including those separated by more than 500 million years, when other available tools fail to detect significant sequence conservation.

[0324] Example 6 A systematic analysis of the preserved motifs within 3'UTR reveals deeply preserved elements. To broadly evaluate the predictive power of LncLOOM, a comprehensive analysis of 3'UTR sequences was performed. The inventors focused on clearly defined 3'UTRs based on highly conserved coding sequences adjacent to the 3'UTR, which allowed them to construct a highly reliable input dataset spanning hundreds of millions of years of evolution, from which thousands of elements could be systematically tested using LncLOOM. The dataset is based on 2,439 genes with 3'UTR MSAs generated as part of the TargetScan7.2 miRNA target site prediction suite. 10 For each gene, in each of the four species (human, mouse, dog, and chicken), a dataset of 3'UTR sequences was generated for LncLOOM analysis, including aligned sequences from TargetScan MSA, only if they were between 300 and 3,000 nt in length. For genes with several 3'UTR isoforms, the inventors selected the longest 3'UTR. Subsequently, the inventors added Ensembl-annotated 3'UTR sequences from additional species to the dataset, if available, if they were longer than 200 base pairs. These included sequences from five non-amniotic vertebrate species (frog, shark, zebrafish, gar, and lamprey) and two invertebrates (sea squirt and fly). The primary objective was to evaluate LncLOOM's ability to identify deeply conserved elements; therefore, only genes with at least one suitable sequence from a non-amniotic species were used. Figure 10A shows the number of sequences that can be analyzed at various depths. Of the 2,439 3'UTR datasets, 2,117 contained at least one sequence for which BLASTN did not report any significant alignment to human sequences (E value < 0.05), and 2,031 datasets contained at least one sequence that did not have a significant alignment to any of the four species (Figure 5A). Therefore, it was possible to analyze a large number of sequences whose entire conservation depth could not potentially be examined by MSA-based methods.

[0325] Using LncLOOM, we searched for conserved motifs with a minimum length of 6 nucleotides and P<0.05 across all LncLOOM studies. LncLOOM detected over 150,000 significant motifs in human sequences, of which 27,826 (18.3%) corresponded to seed sites of widely conserved miRNA families (as defined by TargetScan). 11,725 ​​kmers were conserved across amniotes, with 3,897 of these detected in at least one non-alignable sequence (Figures 5A-I and 10). LncLOOM detected at least one unique kmer in 1,640 of the 2,117 genes containing sequences that did not align to their respective human orthologues, and combinations of at least three unique kmers were found in 1,088 genes (Figure 5B). Considering only sequences that did not align to any of the four amniote species, at least one unique k-mer was detected within the first unalignable sequences in the 1,529 dataset (Figures 10A–F). Conservation was found across vertebrates in 114 genes, and from humans to Drosophila in 97 genes. A total of 170 unique k-mers (265 instances) were found in fly genes, of which only two coincided with widely conserved miRNA binding sites (Figure 5C).

[0326] The inventors then examined specific conserved kmers shared across the 3'UTRs of multiple genes. Of the kmers detected within non-alignable sequences, 42 were common to at least 50 genes, of which only 2 corresponded to widely conserved miRNA binding sites, and 30 were conserved within invertebrate sequences (Figure 5D). Of these 30, 18 kmers contained UUU sequences in an A / U-rich context similar to AU-rich elements (AREs), and 5 kmers contained AUAA similar to PAS. The other kmers contained a UGUA core similar to PRE. Thus, these three groups of miRNA-unrelated elements are also often very deeply conserved within the 3'UTR, and their conserved occurrences can be detected by LncLOOM.

[0327] To evaluate the sensitivity of LncLOOM, we compared the widely conserved miRNA binding sites identified by LncLOOM with TargetScan predictions for each of 2,439 genes, of which TargetScan predicted binding sites within human sequences for 2,121. LncLOOM predicted binding sites within 2,330 genes, including 217 where TargetScan alignment did not identify widely conserved sites (Figure 5E). An overview of all miRNA sites predicted by LncLOOM can be found at github(dot)com / LncLOOM / LncLOOM. In a significant number of cases (29% of 2,117 genes), LncLOOM found miRNA binding sites that were significantly conserved in species where the 3'UTR could not be aligned to human sequences in MSA (Figure 5F). To more accurately compare lncLOOM predictions with TargetScan predictions, we focused on 2,359 genes for which TargetScan predicted binding sites in the same human transcripts used for lncLOOM analysis (Figure 5E). Of these, lncLOOM recovered 90.24% of all widely conserved sites predicted by TargetScan within human sequences (Figure 5G). Of the 217 genes, 42 had sites conserved beyond mammals, and some genes showed conservation in fish and fruit fly species (Figures 10A-F). In addition to the recovered miRNA sites, lncLOOM identified an additional 21,615 widely conserved sites that had not been previously predicted. Comparing the depth of conservation, lncLOOM detected more sites recovered by TargetScan in more distal species (Figures 5G and 10A-F). Importantly, in 24% and 13% of the genes, respectively, 831 recovered predictions and 331 new predictions were detected within sequences that could not be aligned.

[0328] Thus, LncLOOM is also a powerful tool for analyzing 3'UTR sequences, revealing a greater depth of conservation of miRNA or other functional binding sites than is possible with MSA-based methods, while making only limited compromises in sensitivity.

[0329] Example 7 Targeting CHASERR induces CHD2 upregulation in neuroblasts. The array is provided below.

[0330] [ka]

[0331] (ASO targeting CHASERR:) A35: The same ASO used for the mouse. This ASO is complementary to the mouse sequence. A40: In mice, it targets the same region as ASO1, but is a completely complementary ASO to the human sequence. A49: Similar to A35 and A40, but with the potential to base pair with both human and mouse sequences using GU pairing. A50 is identical to A40, but has a 2'MO modification instead of a 2'MOE, and is cleaved by only two bases at the 3' end. A51 is identical to A40, but has a 2'MO modification instead of a 2'MOE, and is cleaved by only two bases at the 5' end. A52: Same as A40, but includes LNA modification.

[0332] result The effects on CHD2 mRNA and protein levels were compared with those of untargeted ASO A27 and A28. A28 was compared with A27 because it induces p21 upregulation and stress response in SH-SY5Y cells (Figure 16).

[0333] Cells 2.5 × 10 5 Cells were plated at a density of / 35 mm. Cells were transfected with 25 nM ASO using DharmaFECT4 transfection reagent (T-2004-03, Horizon). RNA was extracted 48 hours after transfection.

[0334] ASO A40, A50, A51, and A52 were the most potent in upregulating CHD2 compared to untransfected cells or cells transfected with a control ASO (Figure 16).

[0335] Example 8 Targeting CHASERR induces CHD2 upregulation in MCF7 cells and SH-SY5Y cells. Antisense oligonucleotide and LNA GapmeR transfection 10% fetal bovine serum and 100 U penicillin / 0.1 mg ml -1 MCF7 cell line (obtained from ATCC) was cultured in DMEM containing streptomycin. 10% fetal bovine serum and 100U penicillin / 0.1mg ml were added. -1 SH-SY5Y cell lines (obtained from ATCC) were cultured in DMEM / Nutrient Mixture F-12 Ham (Sigma: D6421) containing streptomycin and 2 mM GlutaMAX (Thermofisher: 35050061). All cells were cultured at 37°C in a humidified incubator with 5% CO2 and routinely tested for mycoplasma contamination. The first set of ASOs, ASO1 (A40, SEQ ID NO: 128) and ASO3 (A41, SEQ ID NO: 134), were modified with 2'-O-methoxyethyl base. An LNA gapmer targeting the second intron of human Chaserr was used for Chaserr knockdown. Transfection: 2 × 10⁶ 5Individual MCF7 or SH-SY5Y cells were seeded in 6-well plates and transfected with a mixture of ASO1 (ASO40) and ASO3 (ASO41), or Chaserr gapmeR (Table 5), using Dharmafect4 (Dharmacon) transfection reagent according to the manufacturer's protocol, to a final concentration of 50 nM. The endpoint of all experiments was 48 hours after transfection, after which cells were harvested using TRIZOL for RNA extraction and evaluation by RT-qPCR analysis. The effects on Chasser and CHD2 expression are shown in Figure 17.

[0336] [Table 5]

[0337] While the present invention has been described in conjunction with its specific embodiments, it is evident that many alternative, modified, and variant forms are apparent to those skilled in the art. Therefore, it is intended to encompass all such alternative, modified, and variant forms that fall within the spirit and broad scope of the appended claims.

[0338] All publications, patents, and patent applications referenced herein are incorporated herein by reference in their entirety, as if each individual publication, patent, or patent application were specifically and individually referred to when it is mentioned that they are incorporated herein by reference. Furthermore, any citation or specification of any reference in this application should not be construed as an acknowledgment that such reference is available as prior art of the present invention. Section headings should not necessarily be construed as restrictive insofar as they are used. Furthermore, all priority documents of this application are incorporated herein by reference in their entirety.

[0339] References (The text includes other references) 1. Ulitsky, I. & Bartel, D. P. lincRNAs: genomics, evolution, and mechanisms. Cell 154, 26-46 (2013). 2. Iyer, M. K. et al. The landscape of long noncoding RNAs in the human transcriptome. Nat. Genet. 47, 199-208 (2015). 3. Ulitsky, I. Evolution to the rescue: using comparative genomics to understand long non-coding RNAs. Nat. Rev. Genet. (2016) doi:10.1038 / nrg.2016.85. 4. Hezroni, H. et al. Principles of Long Noncoding RNA Evolution Derived from Direct Comparison of Transcriptomes in 17 Species. Cell Rep. (2015) doi:10.1016 / j.celrep.2015.04.023. 5. Wang, A. X., Ruzzo, W. L. & Tompa, M. How accurately is ncRNA aligned within whole-genome multiple alignments? BMC Bioinformatics 8, 417 (2007). 6. Bartel, D. P. Metazoan MicroRNAs. Cell 173, 20-51 (2018). 7. Dominguez, D. et al. Sequence, Structure, and Context Preferences of Human RNA Binding Proteins. Mol. Cell 70, 854-867.e9 (2018). 8. Maier, D. The Complexity of Some Problems on Subsequences and Supersequences. (1978). 9. Atamturk, A. & Savelsbergh, M. W. P. Integer-Programming Software Systems. Ann. Oper. Res. 140, 67-124 (2005). 10. Agarwal, V., Bell, G. W., Nam, J.-W. & Bartel, D. P. Predicting effective microRNA target sites in mammalian mRNAs. Elife 4, e05005 (2015). 11. Van Nostrand, E. L. et al. A Large-Scale Binding and Functional Map of Human RNA Binding Proteins. bioRxiv 179648 (2017) doi:10.1101 / 179648. 12. Ulitsky, I., Shkumatava, A., Jan, C. H., Sive, H. & Bartel, D. P. Conserved function of lincRNAs in vertebrate embryonic development despite rapid sequence evolution. Cell 147, 1537-1550 (2011). 13. Kleaveland, B., Shi, C. Y., Stefano, J. & Bartel, D. P. A Network of Noncoding Regulatory RNAs Acts in the Mammalian Brain. bioRxiv (2018). 14. Zhang, M. et al. Post-transcriptional regulation of mouse neurogenesis by Pumilio proteins. Genes Dev. 31, 1354-1369 (2017). 15. Goldstrohm, A. C., Hall, T. M. T. & McKenney, K. M. Post-transcriptional Regulatory Functions of Mammalian Pumilio Proteins. Trends Genet. 34, 972-990 (2018). 16. Li, X., Pritykin, Y., Concepcion, C. P., Lu, Y. & La Rocca, G. High-resolution in vivo identification of miRNA targets by Halo-Enhanced Ago2 Pulldown. bioRxiv (2019). 17. McGeary, S. E., Lin, K. S., Shi, C. Y., Bisaria, N. & Bartel, D. P. The biochemical basis of microRNA targeting efficacy. doi:10.1101 / 414763. 18. Elfakess, R. & Dikstein, R. A translation initiation element specific to mRNAs with very short 5’UTR that also regulates transcription. PLoS One 3, e3094 (2008). 19. Elfakess, R. et al. Unique translation initiation of mRNAs-containing TISU element. Nucleic Acids Res. 39, 7598-7609 (2011). 20. Housman, G. & Ulitsky, I. Methods for distinguishing between protein-coding and long noncoding RNAs and the elusive biological purpose of translation of long noncoding RNAs. Biochim. Biophys. Acta (2015) doi:10.1016 / j.bbagrm.2015.07.017. 21. Bitetti, A. et al. MicroRNA degradation by a conserved target RNA regulates animal behavior. Nat. Struct. Mol. Biol. 25, 244-251 (2018). 22. Munschauer, M. et al. The NORAD lncRNA assembles a topoisomerase complex critical for genome stability. Nature 561, 132-136 (2018). 23. Lovci, M. T. et al. Rbfox proteins regulate alternative mRNA splicing through evolutionarily conserved RNA bridges. Nat. Struct. Mol. Biol. 20, 1434-1442 (2013).

[0340] 24. Jangi, M., Boutz, P. L., Paul, P. & Sharp, P. A. Rbfox2 controls autoregulation in RNA-binding protein networks. Genes Dev. 28, 637-651 (2014). 25. Chi, S. W., Zang, J. B., Mele, A. & Darnell, R. B. Argonaute HITS-CLIP decodes microRNA-mRNA interaction maps. Nature 460, 479-486 (2009). 26. Michel, A. M. et al. GWIPS-viz: development of a ribo-seq genome browser. Nucleic Acids Res. 42, D859-64 (2014). 27. Rom, A. et al. Regulation of CHD2 expression by the Chaserr long noncoding RNA gene is essential for viability. Nat. Commun. 10, 5092 (2019). 28. Sievers, F. et al. Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol. Syst. Biol. 7, (2011). 29. Chen, M. C. et al. Structural basis of G-quadruplex unfolding by the DEAH / RHA helicase DHX36. Nature 558, 465-469 (2018). 30. Sauer, M. et al. DHX36 prevents the accumulation of translationally inactive mRNAs with G4-structures in untranslated regions. Nat. Commun. 10, 2421 (2019). 31. Kikin, O., D’Antonio, L. & Bagga, P. S. QGRS Mapper: a web-based server for predicting G-quadruplexes in nucleotide sequences. Nucleic Acids Res. 34, W676-82 (2006). 32. Garant, J.-M., Perreault, J.-P. & Scott, M. S. G4RNA screener web server: User focused interface for RNA G-quadruplex prediction. Biochimie vol. 151 115-118 (2018). 33. Haque, N., Ouda, R., Chen, C., Ozato, K. & Hogg, J. R. ZFR coordinates crosstalk between RNA decay and transcription in innate immunity. Nat. Commun. 9, 1145 (2018). 34. Shabalina, S. A., Ogurtsov, A. Y., Rogozin, I. B., Koonin, E. V. & Lipman, D. J. Comparative analysis of orthologous eukaryotic mRNAs: potential hidden functional signals. Nucleic Acids Res. 32, 1774-1782 (2004). 35. Kirk, J. M. et al. Functional classification of long non-coding RNAs by k-mer content. Nat. Genet. 50, 1474-1482 (2018). 36. Quinn, J. J. et al. Rapid evolutionary turnover underlies conserved lncRNA-genome interactions. Genes Dev. 30, 191-207 (2016). 37. Tycowski, K. T., Shu, M. D., Borah, S., Shi, M. & Steitz, J. A. Conservation of a triple-helix-forming RNA stability element in noncoding and genomic RNAs of diverse viruses. Cell Rep. 2, 26-32 (2012). 38. Deveson, I. W. et al. Universal Alternative Splicing of Noncoding Exons. Cell Syst 6, 245-255.e5 (2018). 39. Katoh, K., Misawa, K., Kuma, K.-I. & Miyata, T. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res. 30, 3059-3066 (2002). 40. Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. Basic local alignment search tool. J. Mol. Biol. 215, 403-410 (1990). 41. Karp, R. M. Reducibility among Combinatorial Problems. in Complexity of Computer Computations: Proceedings of a symposium on the Complexity of Computer Computations, held March 20-22, 1972, at the IBM Thomas J. Watson Research Center, Yorktown Heights, New York, and sponsored by the Office of Naval Research, Mathematics Program, IBM World Trade Corporation, and the IBM Research Mathematical Sciences Department (eds. Miller, R. E., Thatcher, J. W. & Bohlinger, J. D.) 85-103 (Springer US, 1972). 42. Hagberg, A., Swart, P. & S Chult, D. Exploring network structure, dynamics, and function using NetworkX. www(dot)osti(dot)gov / biblio / 960616 (2008). 43. Mitchell, S., OSullivan, M. & Dunning, I. PuLP: a linear programming toolkit for python. The University of Auckland, Auckland, New Zealand (2011). 44. Kent, W. J. BLAT-The BLAST-Like Alignment Tool. Genome Res. 12, 656-664 (2002). 45. Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15-21 (2013). 46. Li, B. & Dewey, C. N. RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics 12, 323 (2011).

[0341] 47. Elinger, D., Gabashvili, A. & Levin, Y. Suspension Trapping (S-Trap) Is Compatible with Typical Protein Extraction Buffers and Detergents for Bottom-Up Proteomics. J. Proteome Res. 18, 1441-1445 (2019). 48. Cox, J. & Mann, M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat. Biotechnol. 26, 1367-1372 (2008). Exemplary embodiments of the present invention are described below. <1> A method for increasing the amount of chromodomain helicase DNA-binding protein 2 (CHD2) in nerve cells, comprising introducing a nucleic acid agent into the cells that downregulates the activity or expression of human Chaserr, wherein the nucleic acid agent is directed to the terminal exon of human Chaserr, thereby increasing the amount of CHD2 in the nerve cells. <2> A method for treating a disease or condition associated with chromodomain helicase DNA-binding protein 2 (CHD2) haploinsufficiency in a subject requiring treatment, comprising administering to the subject a nucleic acid agent that downregulates the activity or expression of human Chaserr in a therapeutically effective amount, wherein the nucleic acid agent is directed to the terminal exon of human Chaserr, thereby treating the disease or condition associated with CHD2 haploinsufficiency. <3> A nucleic acid agent for downregulating the activity or expression of human Chaserr, for use in the treatment of diseases or conditions related to chromodomain helicase DNA-binding protein 2 (CHD2) haploinsufficiency in subjects requiring treatment of said diseases or conditions, wherein the nucleic acid agent is directed towards the terminal exon of human Chaserr. <4> A nucleic acid agent containing a nucleic acid sequence that hybridizes to the final exon of human Chaserr, and a nucleic acid agent for the activity or expression of human Chaserr. <5> The human Chaserr is an alternatively spliced ​​variant selected from the group consisting of SEQ ID NOs: 11 (NR_037600), 12 (NR_037601), and 13 (NR_037602). Includes a variant <1> ~ <4> The method described in any one of the items or a nucleic acid agent. <6> The nucleic acid agent contains a sequence complementary to sequence number 2 (AUGG), <1> ~ <5> The method described in any one of the items or a nucleic acid agent. <7> The nucleic acid agent contains a sequence complementary to AAGAUG (SEQ ID NO: 5) or AAAUGGA (SEQ ID NO: 6), <1> ~ <5> The method described in any one of the items or a nucleic acid agent. <8> The nucleic acid agent contains a sequence complementary to UUUUUACCU (Sequence ID 122), <1> ~ <5> The method described in any one of the items or a nucleic acid agent. <9> The nucleic acid agent inhibits the binding of DHX36 to Chaserr. <1> ~ <8> The method described in any one of the items or a nucleic acid agent. <10> The nucleic acid agent inhibits the binding of CHD2 to Chaserr. <1> ~ <8> The method described in any one of the items or a nucleic acid agent. <11> The nucleic acid agent is an antisense oligonucleotide. <1> ~ <9> The method described in any one of the items or a nucleic acid agent. <12> The nucleic acid agent is one or more nucleotides containing a 2'-to-4' crosslink, or one or more nucleotides having a 2'-O modification. <1> ~ <11> The method described in any one of the items or a nucleic acid agent. <13> The antisense oligonucleotides are as shown in Sequence IDs 92-99. <9> Methods or nucleic acid agents. <14> The antisense oligonucleotides are as shown in SEQ ID NOs: 128, 131, 132, 133, 140, 141, 142, or 143. <10> or <12> Methods or nucleic acid agents. <15> The antisense oligonucleotide comprises two or more antisense oligonucleotides. <11> , <12> , and <13> The method described in any one of the items or a nucleic acid agent. <16> The two or more antisense oligonucleotides include ASO40 of SEQ ID NO: 140 or SEQ ID NO: 128, and ASO41 of SEQ ID NO: 144 or SEQ ID NO: 134. <15> Methods or nucleic acid agents. <17> The nucleic acid agent is an RNA silencing agent. <1> ~ <10> The method described in any one of the items or a nucleic acid agent. <18> The nucleic acid agent is a genome editing agent. <1> ~ <10> The method described in any one of the items or a nucleic acid agent. <19> The nucleic acid agent is inducibly active. <1> ~ <18> The method described in any one of the items or a nucleic acid agent. <20> The nucleic acid agent is active in a tissue-specific or cell-specific manner. <1> ~ <10> The method described in any one of the items or nucleic acid agents 。 <21> The diseases or conditions associated with the aforementioned chromodomain helicase DNA-binding protein 2 (CHD2) haploinsufficiency are selected from the group consisting of intellectual disability, autism, epilepsy, and Lennox-Gastaut syndrome (LGS). <2> ~ <20> The method described in any one of the items or a nucleic acid agent. <22> A method for analyzing a set of sequences describing multiple homologous polynucleotides, The method involves constructing a graph having multiple nodes arranged in layers and multiple edges connecting nodes in consecutive layers, where the first layer represents a sequence describing a query polynucleotide, each node represents a k-mer within that sequence, and each edge is the same The nodes representing homologous k-mers are connected, and each layer represents a sequence of the set such that k is between 6 and 12; Searching the graph by finding continuous non-crossing paths along the edges of the graph; and To generate an output that identifies kmers corresponding to at least one pathway as nucleic acid sequences of interest in function. A method that includes this. <23> Before generating the output, the process includes iteratively repeating the construction and the search, each time making it shorter. <22> Methods used. <24> The method includes applying the path obtained in the previous iteration as a constraint for the search in each iteration cycle. <23> Methods used. <25> The search includes applying a criterion based on path depth as a constraint for the search, such that the search is prioritized over the search for deeper paths. <22> ~ <24> The method described in any one of the items. <26> The aforementioned search includes applying an integer linear programming (ILP) method to the graph. <22> ~ <25> The method described in any one of the items. <27> The homologous polynucleotides are DNA sequences. <22> ~ <25> The method described in any one of the items. <28> The homologous polynucleotide is an RNA sequence. <22> ~ <25> The method described in any one of the items. <29> The method includes aligning the sequences in the set in a predetermined order to provide a multiple alignment having multiple alignment layers, wherein the first layer is the query polynucleotide from the plurality of homologous polynucleotides, and the plurality of alignment layers correspond to each layer of the graph. <22> ~ <28> The method described in any one of the items. <30> The predetermined order is evolution-dictated, and the query is the most evolved among the homologous polynucleotides. <29> Methods used. <31> The homology between the aforementioned homologous k-mers is 70% or more. <22> ~ <30> The method described in any one of the items. <32> The homologous polynucleotides include a partial sequence, <22> ~ <31> The method described in any one of the items. <33> The homologous polynucleotide is selected from the group consisting of 3'UTR, lncRNA, and enhancer. <22> ~ <32> The method described in any one of the items.

Claims

1. A nucleic acid agent for use in the treatment of a disease or condition related to chromodomain helicase DNA-binding protein 2 (CHD2) haploinsufficiency in subjects requiring treatment of said disease or condition, wherein the nucleic acid agent is directed to the final exon of human Chaser and contains a sequence complementary to AAGAUGG (SEQ ID NO: 4) or AAAUGGA (SEQ ID NO: 6), and does not recruit RNaseH.

2. The nucleic acid agent according to claim 1, comprising a sequence complementary to CACAAAATGGACAGTGGAT (SEQ ID NO: 10) or AAGATGGCAGTCTACTATGG (SEQ ID NO: 8).

3. The nucleic acid agent according to claim 1, comprising the sequence CCATAGTAGAACTGCCATCTT (SEQ ID NO: 7) or ATCCACTGTTCCATTTTG (SEQ ID NO: 9), wherein the thymine nucleic acid base may be substituted with a uracil nucleic acid base, and the cytosine nucleic acid base may be substituted with 5-methylcytosine.

4. A nucleic acid agent according to any one of claims 1 to 3, having a phosphorothioate skeleton.

5. A nucleic acid agent according to any one of claims 1 to 4, comprising one or more nucleotides having a 2'-to-4' crosslink, and / or one or more nucleotides having a 2'-O modification selected from 2'-O-methoxyethyl (2'-O-MOE) and 2'-OMe.

Images