Patent Application
20240156919
The invention relates to the field of molecular biology and recombinant nucleic acid technology. In particular, the invention relates to a method of treating a patient with Duchenne Muscular Dystrophy comprising the removal of at least one exon from the dystrophin gene using engineered nucleases
The instant application contains an electronic Sequence Listing which has been submitted in xml (ST.26) format via USPTO Patent Center, herein incorporated by reference in its entirety. Said xml copy, created on Dec. 20, 2023, is named “P89339 1330US.C5 Seq List.xml” and is 207,181 bytes in size.
Duchenne Muscular Dystrophy is a rare, X-linked muscle degenerative disorder that affects about 1 in every 3500 boys worldwide. The disease is caused by mutations in the dystrophin (DMD) gene, which is the largest known gene. DMD spans 2.2 Mb of the X chromosome and encodes predominantly a 14-kb transcript derived from 79 exons. The full-length dystrophin protein, as expressed in skeletal muscle, smooth muscle, and cardiomyocytes, is 3685 amino acids and has a molecular weight of 427 kD. The severe Duchenne phenotype is generally associated with the loss of full length dystrophin protein from skeletal and cardiac muscle, which leads to debilitating muscle degeneration and, ultimately, heart failure. A large number of different DMD mutations have been described, many of them resulting in either the severe Duchenne Muscular Dystrophy or the milder Becker Muscular Dystrophy. The Leiden University Medical Center maintains a database of mutations in the DMD gene (http://www.dmd.nl).
There are several therapeutic strategies being pursued for the treatment of Duchenne Muscular Dystrophy. First, “gene replacement” strategies are an active area of research (Oshima, et al. (2009) J of the Am. Soc. of Gene Ther. 17:73-80; Liu, et al. (2005) Mol. Ther. 11:245-256; Lai, et al. (2006) Hum Gene Ther. 17:1036-1042; Odom et al. (2008) Mol. Ther. 16:1539-1545). This approach involves delivering a functional copy of the DMD gene to patients using a viral delivery vector, typically adeno-associated virus (AAV). The large size of the DMD gene makes it incompatible with the limited carrying capacity of common viral vectors, however. This necessitates the use of a “micro-dystrophin” gene in which most of the repetitive central portion of the gene is removed to leave only the minimal functional protein. It is not clear, however, that expression of “micro-dystrophin” is sufficient for clinical benefit. In addition, this approach suffers from the possibility of random gene integration into the patient genome, which could lead to insertional mutagenesis, and the potential for immune reactions against the delivery vector.
A second approach to treating Duchenne Muscular Dystrophy involves the transplantation of healthy muscle precursor cells into patient muscle fibres (Peault et al. (2007) Mol. Ther. 15:867-877; Skuk, et al. (2007) Neuromuscul. Disord. 17:38-46). This approach suffers from inefficient migration of the transplanted myoblasts and the potential for immune rejection by the patient.
A third approach involves suppression of nonsense mutations using PTC124 (Welch, et al. (2007) Nature 447:87-91). This would require lifelong dosing of the drug, however, and the approach is yet to show any significant clinical benefit.
A fourth, and more promising, potential treatment for Duchenne Muscular Dystrophy is called “Exon Skipping” (Williams, et al. (2008) BMC Biotechnol. 8:35; Jearawiriyapaisarn et al. (2008) Mol Ther. 16:1624-1629; Yokota, et al. (2007) Acta Myol. 26:179-184; van Deutekom et al. (2001) Hum. Mol. Gen. 10:1547-1554; Benedetti et al. (2013) FEBS J. 280:4263-80; Rodino-Klapac (2013) Curr Neurol Neurosci Rep. 13:332; Verhaart and Aartsma-Rus (2012) Curr Opin Neurol. 25:588-96). In general, the N- and C-terminal portions of the dystrophin gene are essential for its role as a “scaffold” protein that maintains membrane integrity in muscle fibres whereas the central “rod domain”, which comprises 24 spectrin-like repeats, is at least partially dispensible. Indeed, the severe Duchenne phenotype is typically associated with mutations in the dystrophin gene that introduce frameshifts and/or premature termination codons, resulting in a truncated form of the dystrophin protein lacking the essential C-terminal domain. Mutations in the central rod domain, including large deletions of whole exons, typically result in the much milder Becker phenotype if they maintain the reading frame such that the C-terminal domain of the protein is intact.
Duchenne Muscular Dystrophy is most frequently caused by the deletion of one or more whole exon(s), resulting in reading frame shift. For example, Exon 45 is frequently deleted in Duchenne patients. Because Exon 45 is 176 bp long, which is not divisible by three, deleting the exon shifts Exons 46-79 into the wrong reading frame. The same can be said of Exon 44, which is 148 bp in length. However, if Exons 44 and 45 are deleted, the total size of the deletion is 324 bp, which is divisible by three. Thus, the deletion of both exons does not result in a reading frame shift. Because these exons encode a portion of the non-essential rod domain of the dystrophin protein, deleting them from the protein is expected to result in a mild Becker-like phenotype. Thus, a patient with the Duchenne phenotype due to the deletion of one or more exon(s) can, potentially, be treated by eliminating one or more adjacent exons to restore the reading frame. This is the principle behind “Exon Skipping,” in which modified oligonucleotides are used to block splice acceptor sites in dystrophin pre-mRNA so that one or more specific exons are absent from the processed transcript. The approach has been used to restore dystrophin gene expression in the mdx mouse model by skipping Exon 23, which harbored a disease-inducing nonsense mutation (Mann, et al. (2001) Proc. Nat. Acad. Sci. USA 98:42-47). Oligonucleotide analogs which induce skipping of Exon 51 have also shown promise in early human clinical trials (Benedetti et al. (2013) FEBS J. 280:4263-80). The major limitations with this approach are: (1) the exon-skipping process is inefficient, resulting in relatively low levels of functional dystrophin expression; and (2) the exon-skipping oligonucleotide has a relatively short half-life so the affect is transient, necessitating repeated and life-long dosing. Thus, while Exon-Skipping approaches have shown some promise in clinical trials, the improvements in disease progression have been minimal and variable.
The present invention improves upon current Exon-Skipping approaches by correcting gene expression at the level of the genomic DNA rather than pre-mRNA. The invention is a permanent treatment for Duchenne Muscular Dystrophy that involves the excision of specific exons from the DMD coding sequence using a pair of engineered, site-specific endonucleases. By targeting a pair of such endonucleases to sites in the intronic regions flanking an exon, it is possible to permanently remove the intervening fragment containing the exon from the genome. The resulting cell, and its progeny, will express mutant dystrophin in which a portion of the non-essential spectrin repeat domain is removed but the essential N- and C-terminal domains are intact.
Methods for producing engineered, site-specific endonucleases are known in the art. For example, zinc-finger nucleases (ZFNs) can be engineered to recognize and cut pre-determined sites in a genome. ZFNs are chimeric proteins comprising a zinc finger DNA-binding domain fused to the nuclease domain of the FokI restriction enzyme. The zinc finger domain can be redesigned through rational or experimental means to produce a protein which binds to a pre-determined DNA sequence ˜18 basepairs in length. By fusing this engineered protein domain to the FokI nuclease, it is possible to target DNA breaks with genome-level specificity. ZFNs have been used extensively to target gene addition, removal, and substitution in a wide range of eukaryotic organisms (reviewed in S. Durai et al., Nucleic Acids Res 33, 5978 (2005)). Likewise, TAL-effector nucleases (TALENs) can be generated to cleave specific sites in genomic DNA. Like a ZFN, a TALEN comprises an engineered, site-specific DNA-binding domain fused to the FokI nuclease domain (reviewed in Mak, et al. (2013) Curr Opin Struct Biol. 23:93-9). In this case, however, the DNA binding domain comprises a tandem array of TAL-effector domains, each of which specifically recognizes a single DNA basepair. A limitation that ZFNs and TALENs have for the practice of the current invention is that they are heterodimeric, so that the production of a single functional nuclease in a cell requires co-expression of two protein monomers. Because the current invention requires two nucleases, one to cut on either side of the exon of interest, this would necessitate co-expressing four ZFN or TALEN monomers in the same cell. This presents significant challenges in gene delivery because traditional gene delivery vectors have limited carrying capacity. It also introduces the possibility of “mis-dimerization” in which the monomers associate inappropriately to make unintended dimeric endonuclease species that might recognize and cut off-target locations in the genome. This can, potentially, be minimized by generating orthogonal obligate heterodimers in which the FokI nuclease domains of the four monomers are differentially engineered to dimerize preferentially with the intended partner monomer.
Compact TALENs are an alternative endonuclease architecture that avoids the need for dimerization (Beurdeley, et al. (2013) Nat Commun. 4:1762). A Compact TALEN comprises an engineered, site-specific TAL-effector DNA-binding domain fused to the nuclease domain from the I-TevI homing endonuclease. Unlike FokI, I-TevI does not need to dimerize to produce a double-strand DNA break so a Compact TALEN is functional as a monomer. Thus, it is possible to co-express two Compact TALENs in the same cell to practice the present invention.
Engineered endonucleases based on the CRISPR/Cas9 system are also know in the art (Ran, et al. (2013) Nat Protoc. 8:2281-2308; Mali et al. (2013) Nat Methods. 10:957-63). A CRISPR endonuclease comprises two components: (1) a caspase effector nuclease, typically microbial Cas9; and (2) a short “guide RNA” comprising a ˜20 nucleotide targeting sequence that directs the nuclease to a location of interest in the genome. By expressing multiple guide RNAs in the same cell, each having a different targeting sequence, it is possible to target DNA breaks simultaneously to multiple sites in in the genome. Thus, CRISPR/Cas9 nucleases are suitable for the present invention. The primary drawback of the CRISPR/Cas9 system is its reported high frequency of off-target DNA breaks, which could limit the utility of the system for treating human patients (Fu, et al. (2013) Nat Biotechnol. 31:822-6).
In the preferred embodiment of the invention, the DNA break-inducing agent is an engineered homing endonuclease (also called a “meganuclease”). Homing endonucleases are a group of naturally-occurring nucleases which recognize 15-40 base-pair cleavage sites commonly found in the genomes of plants and fungi. They are frequently associated with parasitic DNA elements, such as group 1 self-splicing introns and inteins. They naturally promote homologous recombination or gene insertion at specific locations in the host genome by producing a double-stranded break in the chromosome, which recruits the cellular DNA-repair machinery (Stoddard (2006), Q. Rev. Biophys. 38: 49-95). Homing endonucleases are commonly grouped into four families: the LAGLIDADG family, the GIY-YIG family, the His-Cys box family and the HNH family. These families are characterized by structural motifs, which affect catalytic activity and recognition sequence. For instance, members of the LAGLIDADG family are characterized by having either one or two copies of the conserved LAGLIDADG motif (see Chevalier et al. (2001), Nucleic Acids Res. 29(18): 3757-3774). The LAGLIDADG homing endonucleases with a single copy of the LAGLIDADG motif form homodimers, whereas members with two copies of the LAGLIDADG motif are found as monomers.
I-CreI (SEQ ID NO: 1) is a member of the LAGLIDADG family of homing endonucleases which recognizes and cuts a 22 basepair recognition sequence in the chloroplast chromosome of the algae Chlamydomonas reinhardtii. Genetic selection techniques have been used to modify the wild-type I-CreI cleavage site preference (Sussman et al. (2004), J. Mol. Biol. 342: 31-41; Chames et al. (2005), Nucleic Acids Res. 33: e178; Seligman et al. (2002), Nucleic Acids Res. 30: 3870-9, Arnould et al. (2006), J. Mol. Biol. 355: 443-58). More recently, a method of rationally-designing mono-LAGLIDADG homing endonucleases was described which is capable of comprehensively redesigning I-CreI and other homing endonucleases to target widely-divergent DNA sites, including sites in mammalian, yeast, plant, bacterial, and viral genomes (WO 2007/047859).
As first described in WO 2009/059195, I-CreI and its engineered derivatives are normally dimeric but can be fused into a single polypeptide using a short peptide linker that joins the C-terminus of a first subunit to the N-terminus of a second subunit (Li, et al. (2009) Nucleic Acids Res. 37:1650-62; Grizot, et al. (2009) Nucleic Acids Res. 37:5405-19.) Thus, a functional “single-chain” meganuclease can be expressed from a single transcript. By delivering genes encoding two different single-chain meganucleases to the same cell, it is possible to simultaneously cut two different sites. This, coupled with the extremely low frequency of off-target cutting observed with engineered meganucleases makes them the preferred endonuclease for the present invention.
The use of engineered meganucleases for treatment of Duchenne Muscular Dystrophy was previously disclosed in WO 2011/141820 (the '820 application). In this patent application, the authors discuss the possibility of using engineered meganucleases to correct defects in the DMD gene via three different mechanisms (see WO 2011/141820
Finally, Ousterout et al. demonstrated that a DNA break can be targeted to the DMD coding sequence using a TALEN and that the break is frequently repaired via the mutagenic non-homologous end-joining pathway, resulting in the introduction of small insertions and/or deletions (“indels”) that can change the reading frame of the gene (Ousterout et al. (2013) Mol Ther. 21:1718-26). They demonstrated the possibility of restoring DMD gene expression in a portion of mutant cells by delivering a DNA break to the exon immediately following the mutation and relying on mutagenic DNA repair to restore the reading frame in some percentage of cells. Unlike the present invention, this approach involved a single nuclease and was targeted to the coding sequence of the gene.
The present invention is a method of treating Duchenne Muscular Dystrophy comprising delivering a pair of engineered nucleases, or genes encoding engineered nucleases, to the muscle cells of a patient such that the two nucleases excise one or more exons of the DMD gene to restore the normal reading frame. Cells so treated will express a shortened form of the dystrophin protein in which a portion of the central spectrin repeat domain is absent but the N- and C-terminal domains are intact. This will, in many cases, reduce the severity of the disease to the mild Becker phenotype.
Thus, in one embodiment, the invention provides a general method for treating Duchenne Muscular Dystrophy using a pair of nucleases. In another embodiment, the invention provides engineered meganucleases suitable for practicing the method. In a third embodiment, the invention provides engineered Compact TALENs suitable for practicing the method. In a fourth embodiment, the invention provides CRISPRs for practicing the method. In a fifth embodiment, the invention provides vectors and techniques for delivering engineered nucleases to patient cells.
The patent and scientific literature referred to herein establishes knowledge that is available to those of skill in the art. The issued U.S. patents, allowed applications, published foreign applications, and references, including GenBank database sequences, that are cited herein are hereby incorporated by reference to the same extent as if each was specifically and individually indicated to be incorporated by reference.
As used herein, the term “meganuclease” refers to an endonuclease that is derived from I-CreI. The term meganuclease, as used herein, refers to an engineered variant of I-CreI that has been modified relative to natural I-CreI with respect to, for example, DNA-binding specificity, DNA cleavage activity, DNA-binding affinity, or dimerization properties. Methods for producing such modified variants of I-CreI are known in the art (e.g. WO 2007/047859). A meganuclease may bind to double-stranded DNA as a homodimer, as is the case for wild-type I-CreI, or it may bind to DNA as a heterodimer. A meganuclease may also be a “single-chain meganuclease” in which a pair of DNA-binding domains derived from I-CreI are joined into a single polypeptide using a peptide linker.
As used herein, the term “single-chain meganuclease” refers to a polypeptide comprising a pair of meganuclease subunits joined by a linker. A single-chain meganuclease has the organization: N-terminal subunit—Linker—C-terminal subunit. The two meganuclease subunits, each of which is derived from I-CreI, will generally be non-identical in amino acid sequence and will recognize non-identical DNA sequences. Thus, single-chain meganucleases typically cleave pseudo-palindromic or non-palindromic recognition sequences. A single chain meganuclease may be referred to as a “single-chain heterodimer” or “single-chain heterodimeric meganuclease” although it is not, in fact, dimeric. For clarity, unless otherwise specified, the term “meganuclease” can refer to a dimeric or single-chain meganuclease.
As used herein, the term “Compact TALEN” refers to an endonuclease comprising a DNA-binding domain with 16-22 TAL domain repeats fused in any orientation to any portion of the I-TevI homing endonuclease.
As used herein, the term “CRISPR” refers to a caspase-based endonuclease comprising a caspase, such as Cas9, and a guide RNA that directs DNA cleavage of the caspase by hybridizing to a recognition site in the genomic DNA.
As used herein, with respect to a protein, the term “recombinant” means having an altered amino acid sequence as a result of the application of genetic engineering techniques to nucleic acids which encode the protein, and cells or organisms which express the protein. With respect to a nucleic acid, the term “recombinant” means having an altered nucleic acid sequence as a result of the application of genetic engineering techniques. Genetic engineering techniques include, but are not limited to, PCR and DNA cloning technologies; transfection, transformation and other gene transfer technologies; homologous recombination; site-directed mutagenesis; and gene fusion. In accordance with this definition, a protein having an amino acid sequence identical to a naturally-occurring protein, but produced by cloning and expression in a heterologous host, is not considered recombinant.
As used herein, the term “wild-type” refers to any naturally-occurring form of a meganuclease. The term “wild-type” is not intended to mean the most common allelic variant of the enzyme in nature but, rather, any allelic variant found in nature. Wild-type homing endonucleases are distinguished from recombinant or non-naturally-occurring meganucleases.
As used herein, the term “recognition sequence” refers to a DNA sequence that is bound and cleaved by an endonuclease. In the case of a meganuclease, a recognition sequence comprises a pair of inverted, 9 basepair “half sites” which are separated by four basepairs. In the case of a single-chain meganuclease, the N-terminal domain of the protein contacts a first half-site and the C-terminal domain of the protein contacts a second half-site. Cleavage by a meganuclease produces four basepair 3′ “overhangs”. “Overhangs”, or “sticky ends” are short, single-stranded DNA segments that can be produced by endonuclease cleavage of a double-stranded DNA sequence. In the case of meganucleases and single-chain meganucleases derived from I-CreI, the overhang comprises bases 10-13 of the 22 basepair recognition sequence. In the case of a Compact TALEN, the recognition sequence comprises a first CNNNGN sequence that is recognized by the I-TevI domain, followed by a non-specific spacer 4-16 basepairs in length, followed by a second sequence 16-22 bp in length that is recognized by the TAL-effector domain (this sequence typically has a 5′ T base). Cleavage by a Compact TALEN produces two basepair 3′ overhangs. In the case of a CRISPR, the recognition sequence is the sequence, typically 16-24 basepairs, to which the guide RNA binds to direct Cas9 cleavage. Cleavage by a CRISPR produced blunt ends.
As used herein, the term “target site” or “target sequence” refers to a region of the chromosomal DNA of a cell comprising a recognition sequence for a meganuclease.
As used herein, the term “homologous recombination” or “HR” refers to the natural, cellular process in which a double-stranded DNA-break is repaired using a homologous DNA sequence as the repair template (see, e.g. Cahill et al. (2006), Front. Biosci. 11:1958-1976). The homologous DNA sequence may be an endogenous chromosomal sequence or an exogenous nucleic acid that was delivered to the cell.
As used herein, the term “non-homologous end-joining” or “NHEJ” refers to the natural, cellular process in which a double-stranded DNA-break is repaired by the direct joining of two non-homologous DNA segments (see, e.g. Cahill et al. (2006), Front. Biosci. 11:1958-1976). DNA repair by non-homologous end-joining is error-prone and frequently results in the untemplated addition or deletion of DNA sequences at the site of repair.
As used herein, the term “re-ligation” refers to a process in which two DNA ends produced by a pair of double-strand DNA breaks are covalently attached to one another with the loss of the intervening DNA sequence but without the gain or loss of any additional DNA sequence. In the case of a pair of DNA breaks that are produced with single-strand overhangs, re-ligation can proceed via annealing of complementary overhangs followed by covalent attachment of 5′ and 3′ ends by a DNA ligase. Re-ligation is distinguished from NHEJ in that it it does not result in the untemplated addition or removal of DNA from the site of repair.
As used herein, unless specifically indicated otherwise, the word “or” is used in the inclusive sense of “and/or” and not the exclusive sense of “either/or.”
The present invention is based, in part, on the hypothesis that certain deletions in the DMD gene that give rise to the Duchenne phenotype can be compensated for by deleting (an) additional exon(s) immediately up- or downstream of the mutation. The DMD-Leiden Database indicates that most of the mutations that cause Duchenne Muscular Dystrophy are deletions of one or more whole exons that cause a shift in reading frame. In many cases, the reading frame can be restored by eliminating the exon immediately before or after the mutation. As shown in Table 1, 29 different Duchenne-causing mutations, representing ˜65% of patients, can be compensated for by deleting a single exon adjacent to the mutation. For example, a patient with disease due to the deletion of DMD Exon 45, which occurs in approximately 7% of patients, can be treated with a therapeutic that deletes Exon 46. Notably, a therapeutic capable of deleting Exon 51 or Exon 45 could be used to treat 15% and 13% of patients, respectively.
It is known in the art that it is possible to use a site-specific nuclease to make a DNA break in the genome of a living cell and that such a DNA break can result in permanent modification of the genome via mutagenic NHEJ repair or via HR with a transgenic DNA sequence. The present invention, however, involves co-expression of a pair of nucleases in the same cell. Surprisingly, we found that a pair of nucleases targeted to DNA sites in close proximity to one another (less than 10,000 basepairs apart) can excise the intervening DNA fragment from the genome. Also surprisingly, we found that DNA excision using a pair of nucleases frequently proceeds via a mechanism involving the single-stranded DNA overhangs generated by the nucleases. In experiments involving a pair of meganucleases that generate complementary (i.e. identical) DNA overhangs, it was found that the overhang sequence was frequently conserved following fragment excision and repair of the resulting chromosome ends (see Examples 1 and 2). The mechanism of DNA repair, in this case, appears to direct re-ligation of the broken ends, which has not been observed in mammalian cells. Such precise deletion and re-ligation was not observed when using a pair of meganucleases that generated non-identical overhangs (see Example 3). Thus, in a preferred embodiment, the pair of nucleases used for DMD exon excision are selected to generate complementary overhangs.
To excise an exon efficiently, the pair of nuclease cut sites need to be relatively close together. In general, the closer the two sites are to one another, the more efficient the process will be. Thus, the preferred embodiment of the invention uses a pair of nucleases that cut sequences that are less than 10,000 basepairs or, more preferably, 5,000 basepairs or, still more preferably, less than 2,500 basepairs, or, most preferably, less than 1,500 basepairs apart.
As shown in
In alternative embodiments, as diagrammed in
In the preferred embodiment, as diagrammed in
Treating Duchenne Muscular Dystrophy using the invention requires that a pair of nucleases be expressed in a muscle cell. The nucleases can be delivered as purified protein or as RNA or DNA encoding the nucleases. In one embodiment, the nuclease proteins or mRNA or vector encoding the nucleases are supplied to muscle cells via intramuscular injection (Maltzahn, et al. (2012) Proc Natl Acad Sci USA. 109:20614-9) or hydrodynamic injection (Taniyama et al. (2012) Curr Top Med Chem. 12:1630-7; Hegge, et al. (2010) Hum Gene Ther. 21:829-42). To facilitate cellular uptake, the proteins or nucleic acid(s) can be coupled to a cell penetrating peptide to facilitate uptake by muscle cells. Examples of cell pentrating peptides known in the art include poly-arginine (Jearawiriyapaisarn, et al. (2008) Mol Ther. 16:1624-9), TAT peptide from the HIV virus (Hudecz et al. (2005), Med. Res. Rev. 25: 679-736), MPG (Simeon, et al. (2003) Nucleic Acids Res. 31:2717-2724), Pep-1 (Deshayes et al. (2004) Biochemistry 43: 7698-7706, and HSV-1 VP-22 (Deshayes et al. (2005) Cell Mol Life Sci. 62:1839-49. Alternatively, cell penetration can be facilitated by liposome encapsulation (see, e.g., Lipofectamine™, Life Technologies Corp., Carlsbad, CA). The liposome formulation can be used to facilitate lipid bilayer fusion with a target cell, thereby allowing the contents of the liposome or proteins associated with its surface to be brought into the cell.
In some embodiments, the genes encoding a pair of nucleases are delivered using a viral vector. Such vectors are known in the art and include lentiviral vectors, adenoviral vectors, and adeno-associated virus (AAV) vectors (reviewed in Vannucci, et al. (2013 New Microbiol. 36:1-22). In some embodiments, the viral vectors are injected directly into muscle tissue. In alternative embodiments, the viral vectors are delivered systemically. Example 3 describes a preferred embodiment in which the muscle is injected with a recombinant AAV virus encoding a pair of single-chain meganucleases. It is known in the art that different AAV vectors tend to localize to different tissues. Muscle-tropic AAV serotypes include AAV1, AAV9, and AAV2.5 (Bowles, et al. (2012) Mol Ther. 20:443-55). Thuse, these serotypes are preferred for the delivery of nucleases to muscle tissue.
If the nuclease genes are delivered in DNA form (e.g. plasmid) and/or via a viral vector (e.g. AAV) they must be operably linked to a promoter. In some embodiments, this can be a viral promoter such as endogenous promoters from the viral vector (e.g. the LTR of a lentiviral vector) or the well-known cytomegalovirus- or SV40 virus-early promoters. In a preferred embodiment, the nuclease genes are operably linked to a promoter that drives gene expression preferentially in muscle cells. Examples of muscle-specific promoters include C5-12 (Liu, et al. (2004) Hum Gene Ther. 15:783-92), the muscle-specific creatine kinase (MCK) promoter (Yuasa, et al. (2002) Gene Ther. 9:1576-88), or the smooth muscle 22 (SM22) promoter (Haase, et al. (2013) BMC Biotechnol. 13:49-54). In some embodiments, the nuclease genes are under the control of two separate promoters. In alternative embodiments, the genes are under the control of a single promoter and are separated by an internal-ribosome entry site (IRES) or a 2A peptide sequence (Szymczak and Vignali (2005) Expert Opin Biol Ther. 5:627-38).
It is envisioned that a single treatment will permanently delete exons from a percentage of patient cells. In preferred embodiments, these cells will be myoblasts or other muscle precurser cells that are capable of replicating and giving rise to whole muscle fibers that express functional (or semi-functional) dystrophin. If the frequency of exon deletion is low, however, it may be necessary to perform multiple treatments on each patient such as multiple rounds of intramuscular injections.
This invention is further illustrated by the following examples, which should not be construed as limiting. Those skilled in the art will recognize, or be able to ascertain, using no more than routine experimentation, numerous equivalents to the specific substances and procedures described herein. Such equivalents are intended to be encompassed in the scope of the claims that follow the examples below.
1. Meganucleases that Recognize SEQ ID NO: 19 and SEQ ID NO: 42
An engineered meganuclease (SEQ ID NO: 135) was produced which recognizes and cleaves SEQ ID NO: 19. This meganuclease is called “DYS-1/2”. A second engineered meganuclease (SEQ ID NO: 136) was produced which recognizes and cleaves SEQ ID NO: 42. This meganuclease is called “DYS-3/4” (
Human embryonic kidney (HEK-293) cells were co-transfected with mRNA encoding DYS-1/2 and DYS-3/4. mRNA was prepared by first producing a PCR template for an in vitro transcription reaction (SEQ ID NO: 139 and SEQ ID NO: 140. Each PCR product included a T7 promoter and 609 bp of vector sequence downstream of the meganuclease gene. The PCR product was gel purified to ensure a single template. Capped (m7G) RNA was generated using the RiboMAX T7 kit (Promega) according to the manufacturer's instructions and. Ribo m7G cap analog (Promega) was included in the reaction and 0.5 ug of the purified meganuclease PCR product served as the DNA template. Capped RNA was purified using the SV Total RNA Isolation System (Promega) according to the manufacturer's instructions.
1.5×106 HEK-293 cells were nucleofected with 1.5×1012 copies of DYS-1/2 mRNA and 1.5×1012 copies of DYS-3/4 mRNA (2×106 copies/cell) using an Amaxa Nucleofector II device (Lonza) according to the manufacturer's instructions. 48 hours post-transfection, genomic DNA was isolated from the cells using a FlexiGene kit (Qiagen) according to the manufacturer's instructions. The genomic DNA was then subjected to PCR using primers flanking the DYS-1/2 and DYS-3/4 cut sites (SEQ ID NO: 141 and SEQ ID NO:142). When PCR products were resolved by agarose gel electrophoresis, it was apparent that cells co-expressing DYS-1/2 and DYS-3/4 yielded two PCR products with apparent lengths of 1079 basepairs and 233 basepairs whereas genomic DNA from untransfected HEK-293 cells yielded only the larger product (
The smaller PCR product was isolated from the gel and cloned into a bacterial plasmid (pUC-19) for sequence analysis. Three plasmid clones were sequenced, all of which were found to have Exon 44 deleted (
We have demonstrated that it is possible to use a pair of engineered single-chain meganucleases to excise a fragment from the human genome in a cultured cell line. The DNA removal and repair process appears to have proceeded via a mechanism that involves the 3′ overhangs produced by the nucleases, suggesting that the process is more efficient when the overhangs are complementary and able to anneal to one another.
1. Meganucleases that Recognize SEQ ID NO: 62 and SEQ ID NO: 74
An engineered meganuclease (SEQ ID NO: 137) was produced which recognizes and cleaves SEQ ID NO: 62. This meganuclease is called “DYS-5/6”. A second engineered meganuclease (SEQ ID NO: 138) was produced which recognizes and cleaves SEQ ID NO: 74. This meganuclease is called “DYS-7/8” (
2. Deletion of DMD Exon 45 in HEK-293 cells
Human embryonic kidney (HEK-293) cells were co-transfected with mRNA encoding DYS-5/6 and DYS-7/8. mRNA was prepared by first producing a PCR template for an in vitro transcription reaction (SEQ ID NO: 143(20) and SEQ ID NO: 144(21). Each PCR product included a T7 promoter and 609 bp of vector sequence downstream of the meganuclease gene. The PCR product was gel purified to ensure a single template. Capped (m7G) RNA was generated using the RiboMAX T7 kit (Promega) according to the manufacturer's instructions and. Ribo m7G cap analog (Promega) was included in the reaction and 0.5 ug of the purified meganuclease PCR product served as the DNA template. Capped RNA was purified using the SV Total RNA Isolation System (Promega) according to the manufacturer's instructions.
1.5×106 HEK-293 cells were nucleofected with 1.5×1012 copies of DYS-5/6 mRNA and 1.5×1012 copies of DYS-7/8 mRNA (2×106 copies/cell) using an Amaxa Nucleofector II device (Lonza) according to the manufacturer's instructions. 48 hours post-transfection, genomic DNA was isolated from the cells using a FlexiGene kit (Qiagen) according to the manufacturer's instructions. The genomic DNA was then subjected to PCR using primers flanking the DYS-5/6 and DYS-7/8 cut sites (SEQ ID NO: 145 and SEQ ID NO:146). When PCR products were resolved by agarose gel electrophoresis, it was apparent that cells co-expressing DYS-5/6 and DYS-7/8 yielded two PCR products with apparent lengths of 1384 basepairs and 161 basepairs whereas genomic DNA from untransfected HEK-293 cells yielded only the larger product (
The smaller PCR product was isolated from the gel and cloned into a bacterial plasmid (pUC-19) for sequence analysis. 16 plasmid clones were sequenced, all of which were found to have Exon 45 deleted (
We have demonstrated that it is possible to use a pair of engineered single-chain meganucleases to excise a fragment from the human genome in a cultured cell line. The DNA removal and repair process appears to have proceeded via a mechanism that involves the 3′ overhangs produced by the nucleases, suggesting that the process is more efficient when the overhangs are complementary and able to anneal to one another.
The standard mouse model of DMD is the mdx mouse, which has a point mutation in Exon 23 that introduces a premature stop codon (Sicinski et al. (1989) Science. 244:1578-80). In the mouse, DMD Exon 23 is 213 basepairs, equivalent to 71 amino acids. Thus, we reasoned that it should be possible to delete Exon 23 in its entirety and thereby remove the stop codon while maintaining the reading frame of the DMD gene. To this end, we developed a pair of single-chain meganucleases called “MDX-1/2” (SEQ ID NO: 147) and “MDX-13/14” (SEQ ID NO: 148). The former recognizes a DNA sequence upstream of mouse DMD Exon 23 (SEQ ID NO: 149) while the latter recognizes a DNA sequence downstream of mouse DMD Exon 23 (SEQ ID NO: 150). The nucleases were tested, initially, using a reporter assay called “iGFFP” in CHO cells as shown in
A mouse myoblast cell line (C2C12) was co-transfected with in vitro transcribed mRNA encoding the MDX-1/2 and MDX-13/14 nucleases. mRNA was produced using the RiboMAX T7 kit from Promega. 1e6 C2C12 cells were Nucleofected with a total of 2e6 copies/cell of mRNA encoding each MDX enzyme pairs (1e6 copies of each mRNA) using an Amaxa 2b device and the B-032 program. After 96 hours, cells were cloned by limiting dilution in 96-well plates. After approximately 2 weeks growth, cells were harvested and genomic DNA was isolated using a FlexiGene kit from Qiagen. A PCR product was then generated for each clone using a forward primer in DMD Intron 22 (SEQ ID NO: 151) and a reverse primer in Intron 23 (SEQ ID NO: 152). 60 of the PCR products were then cloned and sequenced. 20 of the sequences had deletions consistent with meganuclease-induced cleavage of the DMD gene followed by mutagenic DNA repair (
To produce AAV vectors for simultaneous delivery of MDX-1/2 and MDX-13/14 genes, we first produced a “packaging” plasmid called “pAAV-MDX” (
Recombinant AAV1 virus particles carrying the MDX-1/2 and MDX-13/14 genes were produced as described above. Three hindlimb TA muscles from a pair of mdx mice were injected with virus as described in Xiao, et al (Xiao, et al. (1998) J. Virology 72:2224-2232). One muscle from one mouse was not injected as a negative control. Muscles from the two mice were harvested at 4 days or 7 days post-injection and genomic DNA was isolated from the muscle tissue. The genomic region surrounding DMD Exon 23 was amplified by PCR using a first primer pair (SEQ ID NO:151 and SEQ ID NO: 152). This reaction was then used to template a second PCR reaction using a “nested” primer pair (SEQ ID NO:174 and SEQ ID NO: 175) to eliminate non-specific PCR products. PCR products were then visualized on an agarose gel and it was found that genomic DNA from the three AAV1 injected muscles, but not the un-injected control muscle, yielded smaller PCR products that were consistent in size with the product expected following deletion of DMD Exon 23 by the MDX-1/2 and MDX-13/14 meganucleases. The smaller PCR products were then cloned and sequenced. Three unique sequences were obtained, each of which comprised a portion of the mouse DMD gene including part of Intron 22 and Intron 23 but lacking Exon 23 and all of the sequence intervening the cut sites for the MDX-1/2 and MDX-13/14 meganucleases (SEQ ID NO: 176-178). Thus, we have demonstrated that a pair of meganucleases delivered by AAV can be used to delete a portion of the DMD gene in vivo from mouse muscle.
We have demonstrated that the genes encoding a pair of engineered single-chain meganucleases can be delivered to cells and organisms using recombinant AAV vectors and that meganucleases so delivered are able to cleave genomic DNA in the cell and delete fragments of DNA from the genome. We have further demonstrated that a pair of meganuclease-induced DNA breaks that do not generate compatible overhangs will not re-ligate to yield a defined sequence outcome following removal of the intervening sequence. Thus, for therapeutic applications in which a defined sequence outcome is desirable, it is preferable to use a pair of nucleases that generate identical overhangs.
This application is a continuation of U.S. application Ser. No. 17/033,331, filed Sep. 25, 2020, which is a continuation of U.S. application Ser. No. 16/503,396, filed Jul. 3, 2019, now abandoned, which is a continuation of U.S. application Ser. No. 16/284,733, filed Feb. 25, 2019, now abandoned, which is a continuation of U.S. application Ser. No. 15/124,984, filed Sep. 9, 2016, now abandoned, which is a National Stage Entry of PCT/US2015/020205, filed Mar. 12, 2015, which claims the benefit of priority to U.S. Provisional Patent Application No. 61/951,648, filed Mar. 12, 2014, all of which are hereby incorporated by reference.
| Number | Date | Country | |
|---|---|---|---|
| 61951648 | Mar 2014 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 17033331 | Sep 2020 | US |
| Child | 18490484 | US | |
| Parent | 16503396 | Jul 2019 | US |
| Child | 17033331 | US | |
| Parent | 16284733 | Feb 2019 | US |
| Child | 16503396 | US | |
| Parent | 15124984 | Sep 2016 | US |
| Child | 16284733 | US |
