CONSTRUCT FOR EPIGENETIC MODIFICATION AND ITS USE IN THE SILENCING OF GENES

Abstract

The present invention concerns a construct for epigenomic modification of genes that includes the following components:

a) a Krppel-associated box zinc finger protein or homologous,
b) a DNA region capable of binding to the target gene or homologous,
c) a human DNA methyltransferase DNMT3A or homologous and
d) a murine DNA methyltransferase Dnmt3L or homologous
whereby components a), b), c) and d) are linked to each other either directly or via at least one linker. The construct is a designer epigenome modifier which can be used to silence genes coding for a protein in leukocytes which avoids the internalization of HI viruses in immune cells.

Claims

1. A construct for targeted epigenomic modification of genes comprising the following components: a) a Krppel-associated box zinc finger protein or homologous, b) a DNA region capable of binding to the target gene or homologous, c) a human DNA methyltransferase DNMT3A or homologous and d) a murine DNA methyltransferase Dnmt3L or homologous whereby components a), b), c) and d) are linked to each other either directly or via at least one linker.

2. The construct according to claim 1, characterized in that the single components are proteins and/or peptides.

3. The construct according to claim 1, characterized in that the construct is a nucleic acid based construct coding for the components a)-d) and at least a linker.

4. The construct according to claim 3, characterized in that the nucleic acid construct is contained within a vector.

5. The construct according to claim 4, characterized in that the vector is a vector derived from a lentivirus, an adenovirus or an adeno-associated virus.

6. The construct according to claim 3, characterized in that it is an mRNA molecule which may additionally comprise further components selected from the group consisting of a 7-methylguanosin cap (or an artificial cap analogue) at the 5-end, a non-coding region at the 5-end, a non-coding region at the 3-end and/or a polyA tail at the 3-end.

7. The construct according to claim 3, characterized in that the nucleic acid is a DNA.

8. A method for silencing a gene or genes of interest comprising the steps of: (a) introducing into target cells a construct for targeted epigenomic modification of genes according to claim 1; (b) silencing the gene or genes of interest in said target cells via epigenetic modification.

9. The method according to claim 8, characterized in that the target cells are primary cells.

10. The method according to claim 8, wherein step (a) is performed by introducing a polynucleotide RNA molecule coding for said construct for targeted epigenomic modification of genes according to claim 1 into said target cells.

11. The method according to claim 10, wherein step (a) is performed ex vivo.

12. The method according to claim 8, wherein the chromatin architecture of said target cells is dissected.

13. The method according to claim 8, wherein said method is applied to the treatment of a disease or disorder in a patient in need thereof.

14. The method according to claim 13, wherein said target cells are isolated lymphocytes, said genes of interest being silenced are genes required for the expression of a co-receptor required by an immune deficiency virus to enter into said cells and said disease or disorder being treated is an infection with said immune deficiency virus.

15. The method according to claim 13, wherein said target cells are isolated leukocytes, said genes of interest being silenced are genes required for the expression of a receptor required by human virus to enter into said cells and said disease or disorder being treated is an infection with said human virus.

16. The method according to claim 13, wherein said target cells are isolated human cells, said genes of interest being silenced are genes required for the multiplication of a human pathogen and said disease or disorder being treated is an infection with said human pathogen.

Description

BRIEF DESCRIPTION OF THE FIGURES

[0060] FIGS. 1A and 1B depict designer epigenome modifier (DEMs) and mechanism of action.

[0061] FIG. 1A shows a scheme of a designer epigenome modifier including the different components indicated on the right side in the legend. FIG. 1A shows four constructs, whereby, however, only the construct at the bottom of FIG. 1A (indicated as DEM) represents an embodiment of the present invention. The other constructs in the first three lines are embodiments already known and not covered by the present invention. In the first line the DNA-binding domain (TALE-DBD) is connected via a protein linker to KRAB to form a transient repressor (K). In the second and third line the DNA-binding site (TALE-DBD) is connected via linkers either to functional DNMT3A and Dnmt3L or to inactive dDNMT3a and Dnmt3L DNA methyltransferase to form a targeted methyltransferase (K-DEM) or an inactive effector used as negative control (K-dDEM).

[0062] FIG. 1B explains the principle of the invention. Upon delivery into the target cells, the DNA binding portion (TALE-DBD) will direct the DEM (construct) to a specific target site in the promoter (or any other regulatory element) of the target gene leading to the methylation of the neighboring CpG di-residues and the deacetylation of neighboring histone tails. As a consequence, the target gene of interest will be silenced. Since the modifications are inherited to the daughter cells, the silencing is permanent.

[0063] FIGS. 2A-2D depict the activity of CCR5-specific DEMs.

[0064] FIG. 2A shows schematically an eGFP (green fluorescent protein as indicator), an expression cassette driven by a minimal cytomegalovirus (minCMV) promoter and a fragment (400 bp in length) of the proximal CCR5 promoter which is randomly integrated in the genome of HEK293T cells using lentiviral vectors. The CCR5 promoter is relevant for the expression of an HIV co-receptor. When this gene is silenced in human T lymphocytes or macrophages, the co-receptor required by the virus can no longer be expressed and these leukocytes become resistant against HIV.

[0065] FIG. 2B shows the reporter cell line harboring the expression cassette shown in FIG. 2A which is transfected with mRNA coding for the indicated DEM using Lipofectamine. As a result of DEM activity, CpGs are methylated and eGFP is silenced. The decrease of GFP expressing cells is measured over time by flow cytometry. The effect of various constructs with regard to the expression of green fluorescent protein was measured. The following constructs were used:

TABLE-US-00011 K-dDEM #6 Control. Inactive DEM, lacking the KRAB domain and harboring the inactive dDNMT3A K #3 Control. Construct harboring only the KRAB without DNA methyltransferase targeting position #3 on the CCR5 promoter DEM #3 According to invention: Construct harboring the KRAB domain and the active DNMT3A with Dnmt3L targeting position #3 on the CCR5 promoter K-DEM #6 Control. Construct lacking the KRAB but harboring an active DNMT3A and Dnmt3L K-DEM #6 + K #3 Control. Split construct. One construct harboring only the KRAB binding to position 3 (K) and one construct lacking the KRAB but harboring the active DNMT3A and Dnmt3L binding to position 6. Position 3 and 6 are two sites on the CCR5 promoter. K #6 Control. Construct harboring only the KRAB without DNA methyltransferase targeting position #6 on the CCR5 promoter DEM #6 According to invention: Construct harboring the KRAB domain and the active DNMT3A with Dnmt3L targeting position #6 on the CCR5 promoter

[0066] FIG. 2C shows reporter cells transfected with DEM #6 from day 31 which were further manipulated with the aim of reactivating eGFP expression. Histogram shows the % of eGFP positive cells upon delivery of a transient transcriptional activator or a drug (5-AZA) that stably demethylates CpGs resulting in stable reactivation of eGFP expression.

[0067] The left part of FIG. 2C shows the expression of enhanced green fluorescent protein (eGFP) after transient transfection on day 2 (D2) and day 6 (D6). The right-hand part of FIG. 2C shows the effect of the drug 5-AZA at the same time points.

[0068] FIG. 2D shows the extent of CpG methylation induced by the construct of the present invention as is measured in the reporter cell line one month after the delivery of the indicated construct by sequencing of bisulfite converted genomic DNA.

[0069] FIGS. 3A-3C depict the activity of CXCR4-specific DEMs.

[0070] FIG. 3A shows schematically the CXCR4 gene including the promoter region. Upon DEM binding, the consequent CpG methylation and deacetylation of neighboring histone tails leads to CXCR4 silencing.

[0071] FIG. 3B shows the expression levels of the CXCR4 gene which were measured in HEK293T cell line via quantitative RT-PCR (TaqMan) at the indicated time points upon transfection of mRNA encoding the indicated CXCR4-specific DEMs. The histogram shows the extent of CXCR4 gene silencing as compared to the samples transfected with mRNA coding for the inactive K-dDEM. The gene expression levels are normalized to the housekeeping gene B2M.

[0072] FIG. 3C shows the CXCR4 protein levels measured by flow cytometry. The histogram shows the extent of CXCR4 positive cells as compared to the samples transfected with mRNA encoding for the inactive K-dDEM.

[0073] FIGS. 4A-4C depict the activity of DEMs in primary T cells.

[0074] FIG. 4A shows a time line of the experiment. Human CD4+ cells isolated from normal donors were thawed and activated with beads conjugated with antibodies recognizing CD28, CD3 and CD2 for three days. Subsequently, beads were removed and cells were transfected with mRNA encoding the indicated DEMs via nucleofection. Cells were re-activated every seven days and samples were collected for analysis (FACS and qPCR) on the indicated days ().

[0075] FIG. 4B shows the expression levels of CCR5 and CXCR4 genes which were measured in CD4+ T cells at the indicated time points upon nucleofection via quantitative RT-PCR (TaqMan). The histograms show CCR5 and CXCR4 gene expression levels as compared to the samples transfected with mRNA encoding the inactive K-dDEM #6 or K-dDEM #L1, respectively. The gene expression levels are normalized to the housekeeping gene B2M.

[0076] FIG. 4C shows CCR5 and CXCR4 protein levels measured by flow cytometry. The histograms show the extent of CCR5 and CXCR4 positive T cells as compared to the samples transfected with mRNA encoding for the inactive K-dDEM #6 or K-dDEM #L1, respectively.

[0077] FIGS. 5A-5C depict the extent of CpG methylation induced by DEM in primary human T cells.

[0078] FIG. 5A shows a scheme of the CCR5 locus (grey) with exons highlighted. The regions amplified from bisulfite converted genomic DNA extracted from primary human T cells and analyzed via next generation sequencing are depicted with number 1-12. Relative distance from the DEM #6 binding site is shown.

[0079] FIG. 5B shows the extent of CpG methylation which was assessed via next generation sequencing of the amplicons indicated in (A) from three independent experiments. The histogram shows the % of CpG methylation as ratio between total and methylated CpGs in the samples nucleofected either with the inactive K-dDEM #6 or the DEM #6, respectively.

[0080] FIG. 5C shows the extent of CpG methylation obtained from three independent experiments as heat map for a subset of the amplicons analyzed in B. The color ranges from light grey (basal methylation) to black (maximum methylation). The position of the CpG within the amplicon is indicated. (Student's t-test *p<0.05; **p<0.01).

[0081] FIGS. 6A and 6B depict the off-target activity of DEM measured via ATAC-seq in primary human T cells.

[0082] FIG. 6A is a circle plot illustrating the whole genome accessibility profile resulting from ATAC-seq data combined from two independent experiments in which primary human T cells were nucleofected with the inactive K-dDEM #6 (outer circle) or the DEM #6 (inner circle), respectively.

[0083] The DEM #6 binding site in CCR5 promoter is indicated with a line. Potential off-targets predicted via online tools (Table 1) or retrieved from ATAC-seq results are depicted.

[0084] FIG. 6B shows exemplary ATAC-seq data showing the reduced accessibility at the CCR5 locus as a lower ATAC-seq read counts (Upper panel). As a control, the lower panel shows the ATAC-seq read counts at the housekeeping gene locus B2M. The DEM #6 binding site is indicated with a vertical black line and the numbers in italics within the track indicate the size window of the area with reduced chromatin accessibility.

[0085] FIGS. 7A and 7B depict the off-target activity of DEM measured via next generation sequencing in primary human T cells.

[0086] FIG. 7A shows the extent of CpG methylation induced by the DEM at potential off-target (OT) sites identified with TAL Effector Nucleotide Targeter 2.0 and measured via next generation sequencing. A 300 bp amplicon centered on the potential binding site was amplified from bisulfite converted genomic DNA isolated from primary human T cells four days post nucleofection and sequenced via next generation sequencing. The histogram shows the % of CpG methylation as ratio between total and methylated CpGs in the samples nucleofected either with the inactive K-dDEM #6 or the DEM #6 respectively. PCR amplification at off-target 2 was not successful. (Student's t-test *p<0.05).

[0087] FIG. 7B shows ATAC-seq data showing low accessibility at the intergenic region centered on the OT5. The potential DEM #6 off-target binding site is indicated with a vertical black line.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0088] The following abbreviations were used:

DEM: designer epigenome modifier
KRAB: krppel-associated box
ZNF10: zinc finger protein 10
KOX1: zinc finger protein KOX1
DNMT3A: DNA methyltransferases member 3a (Human)
dDNMT3A: inactive or dead DNA methyltransferases member 3a (Human)
Dnmt3L: DNA methyltransferases member 3L (Murine)
TALE: transcription activator-like effector
DBD: DNA binding domain
CRISPR: clustered, regularly interspaced, short palindromic repeats
Cas9: Cas9 endonuclease
dCas9: inactive or dead Cas9 endonuclease
gRNA: guide RNA
bp: DNA base pairs
kbp: DNA kilo base pairs (1000 bp)
mRNA: messenger RNA
LV: lentivirus
IDLV: integrase-defective lentiviral vector
AAV: adeno-associated virus

Example 1

[0089] The ability to modulate the expression of target genes at will is a major need in the field of gene therapy. The present invention allows precise epigenome editing resulting in silencing of a target gene. The versatility of the platform concedes also that a silenced target gene can be transcriptionally reactivated by changing the combination of effector domains with transcription activator domains like herpes simplex virus-based transcriptional activator VP64 domain and DNA demethylase effectors from the TET family. The invention combines in a single molecule the highly specific DNA binding domain derived from transcription activator-like effectors (TALEs) identified in the plant pathogen Xanthomonas with the Krppel-associated box (KRAB) domain, the human derived DNA methyltransferase 3a and the murine DNA methyltransferase 3L (FIG. 1A). The resulting construct also referred to as designer epigenome modifier or DEM, thereby combines the KRAB-induced recruitment of the scaffold protein KRAB-associated protein 1 (KAP1) that is suggested to regulate transcription through the induction of histone modifications as deacetylation or trimethylation with the DNA methylation ability of DNA methyltransferase 3A capable of covalently adding one CH3 group to cytosines within CpG dinucleotides. Activity of DNMT3A is further enhanced by the Dnmt3L component. Both histone deacetylation and DNA methylation are commonly associated with closed and silenced chromatin structures, thereby the introduction of these repressive epigenetic marks in a promoter or enhancer region will lead to transcriptional silencing of the target gene (FIG. 1B).

[0090] The single experimental steps were in general performed in the examples as described below unless deviations from the general procedures are explicitly mentioned:

[0091] a) Generation of TALE-Based DEM Plasmids.

[0092] The TALE arrays targeted to the chosen CCR5 or CXCR4 sequences are generated using the platform described and can be cloned via Bpil restriction digestion using the different DEM plasmids depicted in FIG. 1A as Level 3 destination vectors. After ligation and transformation, colony PCR can be used to monitor for the successful cloning of the TALE-array. The positive colonies are expanded and plasmid DNA is extracted using a low-scale plasmid preparation kit. After sequence validation, the plasmids are further prepared using a large-scale plasmid purification kit.

[0093] b) In Vitro mRNA Transcription.

[0094] For the in vitro transcription of an mRNA encoding for the different DEMs depicted in FIG. 1A, 10 g of the corresponding expression plasmids are linearized with PspOMI restriction enzyme. The linearized plasmids are purified using the QIAquick PCR Purification Kit and 1 g of linearized plasmid is used for in vitro transcription using the mMessage mMachineT7 Ultra kit based on the T7 promoter. The T7 transcription reaction is performed for 2 h at 37 C. The mRNA encoding for the transgene of interest is recovered by lithium chloride precipitation as suggested by the provider and is subdivided in 2 g aliquots and stored at 80 C.

[0095] c) Generation of HEK293T-Based Reporter Cell Line.

[0096] The genomic region containing the DEM target site is amplified from human genomic DNA using a proof-reading polymerase. Upon purification, the amplicon is inserted in a lentiviral transfer plasmid via canonical molecular cloning. The lentiviral plasmid harboring the target site upstream of a minimal CMV promoter and the EGFP gene is subsequently used for the generation of a lentiviral vector as described. The vector is used to transduce HEK293T cells and three days after transduction, the cells are analyzed by flow cytometry to observe successful virus transduction and GFP expression. Sample showing low GFP expression (10% GFP positive cells) indicative of low integrated copy number of the lentiviral vector are expanded for single cell sorting. A selected HEK293T-GFP clone is used for the experiments reported.

[0097] d) Reporter Cell Transfection and Assessment of GFP Silencing.

[0098] HEK293T-GFP reporter cells are seeded 24 h before transfection and 2 g of the mRNA coding for the indicated DEM construct are used for transfection using Lipofectamine 2000 in optiMEM following the manufacturer instructions. The DEMs activity can be measured two days after transfection by analyzing the level of GFP expression via flow cytometry. Flow cytometry analysis is repeated at later time points to monitor the extent of stable silencing induced by DEMs.

[0099] e) Culturing, Activation and Nucleofection of Primary Human CD4+ T Cells.

[0100] CD4+ T cells are isolated from blood of normal individuals using magnetic beads. Activation beads containing antibodies recognizing CD2, CD3 and CD28 are prepared according to the manufacturer's instructions and mixed gently with the isolated CD4+ cells in a 2:1 (cells:beads) ratio. Three days later, the beads are removed using a magnet (DynaMag15, Invitrogen) and the cells are ready for nucleofection. To this end, according to manufacturer's protocol for stimulated human T cells (Lonza), CD4+ T cells are mixed with nucleofection solution and with the mRNA encoding for the indicated DEM and nucleofected using the program EO-115. Cells are harvested 4 days later for qPCR and flow cytometry analysis. Afterwards, cells are re-activated to allow for long term culturing and later time point analysis (up to day 18 post thawing).

[0101] f) DNA Methylation Analysis Via Bisulfite Sequencing.

[0102] Genomic DNA is isolated from HEK293T-GFP or primary human T cells transfected with the indicated DEMs at the chosen time points via the QIAamp DNA Blood Mini Kit. Bisulfite conversion is performed on 500 ng of the purified genomic DNA with the EZ DNA Methylation Gold Kit following the instructions. To interrogate the methylation status of the CpG dinucleotides in a certain genomic location (on-targets or off-targets), the region of interest is amplified via PCR using the bisulfite converted DNA as template. Amplification is performed with the PyroMark PCR Kit. PCR amplicon are cloned via the CloneJET PCR Cloning Kit into pJET plamid. To know the extent of CpG methylation, the pJET plasmids are sequenced via Sanger sequencing since sodium bisulfite conversion of the genomic DNA changes non methylated cytosines to uraciles leaving unaltered methylated cytosines. Thereby, Sanger sequencing can be used to distinguish the two different nucleotides that will appear as thymine if originally not methylated or as cytosine if originally methylated. In particular, CG to TG mutations represent non methylated cytosines whereas the presence of a CG is indicative of methylated cytosines. For a more quantitative analysis, PCR amplicons are also sequenced via next generation sequencing (NGS).

[0103] g) Analysis of Chromatin Accessibility Via ATAC-Seq.

[0104] ATAC-Seq was performed as described. Library fragments were amplified with NEBNext Ultra II Q5 Master Mix and purified and size-selected with AMPure XP beads (Beckman Coulter). Libraries were sequenced on a HiSeq 2000 system as single reads. Reads were aligned to hg19 with Bowtie software (Johns Hopkins University). Two biological replicates were performed. Circos software package was used for visualization and the replicates were combined. The top ten potential off-targets were determined as 3-fold tag difference between the control and modified cells in both replicates, a minimum of 10 normalized tags in control cells and within 10-kb of TSS of active gene.

Example 2

[0105] In order to prove the efficacy of the invention CCR5 was used as exemplary target gene which is known to be essential for infection of CD4+ T cells with the human immunodeficiency virus (HIV). Individuals that harbor natural mutations inactivating the CCR5 gene are largely resistant to HIV infection. DEMs targeted to the CCR5 promoter were generated and to test their activity, a fluorescent reporter harboring an enhanced green fluorescent protein (eGFP) gene driven by a minimal CMV (minCMV) promoter fused to a fragment of the CCR5 promoter containing the DEM target sites was prepared. The reporter was integrated in the genome of HEK293T cell line via lentiviral transduction resulting in a cell line expressing GFP (green fluorescent proteins) under the control of the CCR5-minCMV fusion promoter (FIG. 2A). Delivery of effectors having only the KRAB domain or only the DNA methyltransferase domains resulted in a transient silencing of the eGFP gene expression which came back to normal after two weeks in culture. However, upon delivery of mRNA encoding for DEMs, the eGFP expression was silenced in about 80% of cells with a very fast kinetics (i.e. after only 6 days) and remained stable over time up to two months post mRNA delivery (FIG. 2B). Importantly, although delivering the different effectors in a separate fashion, meaning the KRAB in one molecule and the DNA methyltransferases DNMT3A and Dnmt3L in a second molecule binding to two adjacent sites, resulted in eGFP silencing, the extent of eGFP silencing was lower than when a single construct, i.e. a DEM, was used (FIG. 2B, see sample indicated with KDEM #6+K #3).

[0106] This experiment clearly shows the advantage of using a single molecule DEM over multiple molecules targeting adjacent sites. Importantly, the eGFP silencing induced by DEM could be reverted by using DNA demethylating agents, as 5-AZA (FIG. 2C), but not using transient activators (i.e. a TALE-DBD fused to a VP64 activator domain) highlighting that silencing is indeed due to specific methylation of the eGFP promoter. Bisulfite sequencing was used to determine the extent of induced CpG methylation and to define the window in which methylation occurred. To this end, genomic DNA was extracted from the reporter cell line one month after DEM delivery and was converted using bisulfite reaction. This leads to conversion of all non-methylated cytosines into thymine while methylated cytosines remain unaltered. Upon PCR amplification of regions at variable distance from the DEM binding site and subsequent sequencing, we assessed that up to 80% of the CpG resulted methylated as compared to control and the levels of methylation were reduced to 40-60% at 2-kb distance from the DEM binding site (FIG. 2D). This experiment unequivocally shows that silencing mediated by DEM is indeed due to target DNA methylation and that the long term effect can be achieved by a single administration of an mRNA molecule encoding for the DEM.

Example 3

[0107] To prove that the invention is easily scalable to different genes, four TALE-based DNA binding domains were constructed targeting the CXCR4 gene (FIG. 3A) with the aim to silence also the second co-receptor used by the HIV to infect the target CD4+ T cells. The CXCR4-specific TALE-DBDs were incorporated in the different construct depicted in FIG. 1A and delivered in form of mRNA into HEK293T cells. Three days after lipofection of the mRNA, the levels of CXCR4 transcripts were measured via qPCR. Interestingly, all four CXCR4-specific DEMs reduced 4- to 5-folds the CXCR4 transcripts levels with an average reduction of about 70% in gene expression (FIG. 3B). Similarly, protein levels measured via flow cytometry were also reduced to the same extent (FIG. 3C).

Example 4

[0108] To test the therapeutic relevance of the invention, its potency was assessed in normal human CD4+ T cells isolated from blood of healthy donor. After isolation, cells were activated for three days prior nucleofection with mRNA encoding CCR5- or CXCR4-specific DEMs (FIG. 4A). To increase the potency, combinations of multiple DEMs targeting the same locus at a neighboring target sequence were tested. Four days after nucleofection, part of the cells were harvested for analyzing CCR5 or CXCR4 mRNA levels via qPCR and their corresponding protein levels via flow cytometry. CD4+ T cells express very high levels of CXCR4 as compared to CCR5. Indeed, four days after nucleofection, effect of DEMs was particularly evident at the CCR5 locus (both mRNA and protein) with a peak of 50% reduction in expression levels as compared to control by using a combination of two DEMs (FIGS. 4B and 4C, upper left graphs), while CXCR4 levels were only mildly affected (FIGS. 4B and 4C, upper right graphs). Delivering the effectors on two separate molecules had no effect in primary CD4+ T cells (FIGS. 4B and 4C, columns identified with AK-DEM#6+K#3). Cells were further activated and cultivated for two additional weeks to prove that DEMs induce long term epigenetic silencing of the target locus. Surprisingly, 18 days post nucleofection, the levels of CXCR4 were dramatically reduced of about 50% as compared to control (FIGS. 4B and 4C, lower right graphs). In contrast, CCR5 was re-activated in most of the cells (FIGS. 4B and 4C, lower left graphs) which is somewhat expected because the in vitro T cell activation procedure is necessary to keep the cells in culture and it does induce CCR5 expression, which may lead to selection of cells expressing CCR5.

Example 5

[0109] To prove that the silencing effect observed at four days post nucleofection was indeed due to increased methylation of the CCR5 promoter, the status of the CpG in a region of 9-Kb surrounding the CCR5-specific DEM target site was analyzed via bisulfite high throughput sequencing (FIG. 5A). Strikingly, we could observe increased CpG methylation in a 1-kb window surrounding the DEM binding site which was more pronounced in proximity of the target site (FIGS. 5B and 5C).

Example 6

[0110] To assess the propensity of DEMs to introduce off-targeted methylation, an assay for transposase-accessible chromatin using sequencing (ATAC-seq) was performed. This is a genome wide method to assess chromatin accessibility through the introduction of small sequence tags that are used for next generation sequence as readout for genomic accessibility (FIG. 6A).

[0111] It was hypothesized that off-target DNA methylation can reduce chromatin accessibility at potential off-target sites and thereby reduce the introduction of tags at those sites. Four days after nucleofection of the mRNA encoding for the CCR5-specific DEM, cells were harvested and ATAC-seq revealed reduced accessibility at the CCR5 on-target site in a window of about 2-kb from the DEM binding site (FIG. 6B; upper panel). Interestingly, no differences were detected at an unrelated housekeeping gene (B2M, FIG. 6B; lower panel). The top 10 potential off-target sites in silico using the TAL Effector Nucleotide Targeter 2.0 online tool (Table 1) were predicted and the potential off-target methylation via high throughput bisulfite sequencing was analyzed.

[0112] No increase in CpG methylation at these sites except for the OT5 was observed (FIG. 7A). However, the already high level of CpG methylation in the control sample at this site and the poor accessibility measured via ATAC-Seq (FIG. 7B) suggest that this effect is not relevant. We expanded our off-target analysis by predicting in silico all the potential off-target sites harboring one or two mismatched nucleotides as compared to the on-target site using COSMID. We identified a total of 53 potential off-target sites (Table 2). Importantly, at all these sites we observed no decrease in DNA accessibility via ATAC-Seq with most of them showing only poor chromatin accessibility suggesting that the nearest genes are not expressed in CD4+ T cells. This highlights the high safety profile associated with the use of DEMs.

TABLE-US-00012 TABLE1 Listofpotentialoff-targetsitesidentified withTALEffectorNucleotideTargeter2.0 Distance Mis- Start fromTSS COSMID Cosmid ID Chr Strand Gene matches Score Position TargetSequence (bp) score mismatch 0 3 Plus CCR5 0 4.98 46411596 TGACCATATACTTATGTCA 19 0 (SEQIDNO:15) 1 7 Plus LOC101927668 2 6.24 20121800 TAACCATATACTTATCTCA 42968 ID#396 ID#35 (SEQIDNO:16) 2 4 Plus Intergenic 1 7.31 165401212 TGAACATATACTTATGTCA n/a ID#1 ID#1 (SEQIDNO:17) 3 18 Plus YES1 2 8.04 779684 TGACCATATACCTATCTCA 32626 ID#567 ID#40 (SEQIDNO:18) 4 17 Plus Intergenic 3 8.24 8563008 TCACCATATACATATATCA n/a ID#573 ID#580 (SEQIDNO:19) 5 20 Plus Intergenic 3 8.24 12726816 TCACCATATACATATATCA n/a ID#575 ID#582 (SEQIDNO:20) 6 5 Plus Intergenic 3 8.24 97609784 TCACCATATACATATATCA n/a ID#574 ID#581 (SEQIDNO:21) 7 12 Plus TEAD4 2 8.36 3076665 TGAACATATACTTATCTCA 8186 ID#405 ID#36 (SEQIDNO:22) 8 5 Plus LOC101927421 3 8.63 124565833 TAACCATATATTTATATCA 193308 ID#536 ID#545 (SEQIDNO:23) 9 X Minus Intergenic 3 8.63 112150169 TAACCATATATTTATATCA n/a ID#535 ID#544 (SEQIDNO:24) 10 2 Minus NYAP2 3 8.9 226408801 TAGCCATATACTTATATCA 143199 ID#427 ID#440 (SEQIDNO:25)

TABLE-US-00013 TABLE2 Listofpotentialoff-targetsitesidentified withCOSMID(Orderedbynumberofmismatches) Distance Mis- Start from ID Chr Strand Gene matches Position TargetSequence TSS(bp) 0 3 plus CCR5 0 46411595 TGACCATATACTTATGTCA(SEQIDNO:26) 19 1 4 minus Intergenic 1 165401211 TGAACATATACTTATGTCA(SEQIDNO:27) n/a 2 12 minus Intergenic 2 126201878 AGACAATATACTTATGTCA(SEQIDNO:28) n/a 3 4 minus Intergenic 2 34747592 TCACTATATACTTATGTCA(SEQIDNO:29) n/a 4 18 plus MIB1 2 19371552 TTACAATATACTTATGTCA(SEQIDNO:30) 86634 5 12 plus Intergenic 2 66048068 TGAGAATATACTTATGTCA(SEQIDNO:31) n/a 6 8 minus Intergenic 2 111862001 TGATTATATACTTATGTCA(SEQIDNO:32) n/a 7 13 plus SERP2 2 44968765 TGAGCATCTACTTATGTCA(SEQIDNO:33) 20787 8 17 plus Intergenic 2 14261495 TGACCCTCTACTTATGTCA(SEQIDNO:34) n/a 9 4 minus Intergenic 2 106456883 TGATCATATCCTTATGTCA(SEQIDNO:35) n/a 10 18 minus ATP9B 2 76945488 TTACCATATAGTTATGTCA(SEQIDNO:36) 116094 11 3 plus KCNMB2 2 178293606 TGTCCATATATTTATGTCA(SEQIDNO:37) 16988 12 13 plus DCLK1 2 36353811 TGACCAAATTCTTATGTCA(SEQIDNO:38) 76169 13 8 plus Intergenic 2 33669235 TGATCATATAGTTATGTCA(SEQIDNO:39) n/a 14 6 plus C6orf203 2 107362597 TGACTATATATTTATGTCA(SEQIDNO:40) 13190 15 5 plus Intergenic 2 160357986 TGACCATTTGCTTATGTCA(SEQIDNO:41) n/a 16 7 plus Intergenic 2 84533770 TGACCAAATATTTATGTCA(SEQIDNO:42) n/a 17 11 minus Intergenic 2 127541818 TGACCATCTATTTATGTCA(SEQIDNO:43) n/a 18 8 plus Intergenic 2 31056581 TTACCATATACATATGTCA(SEQIDNO:44) n/a 19 X plus Intergenic 2 99230790 TGTCCATATACATATGTCA(SEQIDNO:45) n/a 20 3 minus KIAA1257 2 128677399 TGTCCATATACATATGTCA(SEQIDNO:46) 35499 21 6 plus RNGTT 2 89528614 TGACCATACATTTATGTCA(SEQIDNO:47) 144716 22 3 minus ST6GAL1 2 186686208 TGACCGTATACATATGTCA(SEQIDNO:48) 37896 23 10 minus Intergenic 2 109301891 TGAGCATATACTGATGTCA(SEQIDNO:49) n/a 24 1 minus Intergenic 2 115995413 TGACCATATGTTTATGTCA(SEQIDNO:50) n/a 25 8 minus TRPS1 2 116567819 TGACCACATACTGATGTCA(SEQIDNO:51) 113388 26 5 minus Intergenic 2 13083831 TGACCATACACTGATGTCA(SEQIDNO:52) n/a 27 16 minus Intergenic 2 61077413 TTACCATATACTTTTGTCA(SEQIDNO:53) n/a 28 12 plus SCN8A 2 52015830 TGATCATATACTTCTGTCA(SEQIDNO:54) 30810 29 X minus Intergenic 2 98322767 TGATCATATACTTTTGTCA(SEQIDNO:55) n/a 30 13 minus Intergenic 2 48720381 TGACCTTATACTTCTGTCA(SEQIDNO:56) n/a 31 8 minus PXDNL 2 52364466 TGACCATCTACTTGTGTCA(SEQIDNO:57) 357518 32 10 plus JMJD1C 2 65030745 TGGCCATATACTTAAGTCA(SEQIDNO:58) 194959 33 1 plus Intergenic 2 81197612 TGACCAAATACTTAGGTCA(SEQIDNO:59) n/a 34 X plus REPS2 2 17017334 TGACCATATACATGTGTCA(SEQIDNO:60) 52520 35 7 plus LOC101927668 2 20121799 TAACCATATACTTATCTCA(SEQIDNO:61) 58232 36 12 plus TEAD4 2 3076664 TGAACATATACTTATCTCA(SEQIDNO:62) 8186 37 4 minus METTL14 2 119610622 TGACCATAAACTTATTTCA(SEQIDNO:63) 4051 38 12 minus DDX47 2 12967242 TGACCATATCCTTATTTCA(SEQIDNO:64) 1104 39 10 minus TCTN3 2 97447765 TGACCATATTCTTATCTCA(SEQIDNO:65) 6114 40 18 plus YES1 2 779683 TGACCATATACCTATCTCA(SEQIDNO:66) 32626 41 2 minus Intergenic 2 224060353 TGACAATATACTTATGACA(SEQIDNO:67) n/a 42 1 minus NEGR1 2 72084753 TGACCATATCCTTATGGCA(SEQIDNO:68) 481840 43 Y minus Intergenic 2 17815850 TGACCATATAATTATGCCA(SEQIDNO:69) n/a 44 9 minus LINGO2 2 29004177 TGAGCATATACTTATGTAA(SEQIDNO:70) 208800 45 10 plus Intergenic 2 47982402 TGACCATGTACTTATGTAA(SEQIDNO:71) n/a 46 10 plus Intergenic 2 51927697 TGACCATGTACTTATGTAA(SEQIDNO:72) n/a 47 10 minus Intergenic 2 52532086 TGACCATGTACTTATGTAA(SEQIDNO:73) n/a 48 5 plus PRLR 2 35132581 TGACCATATTCTTATGTAA(SEQIDNO:74) 98092 49 20 minus Intergenic 2 12544854 TGACCATATAGTTATGTAA(SEQIDNO:75) n/a 50 2 plus Intergenic 2 47956567 TGACCATATACGTATGTAA(SEQIDNO:76) n/a 51 11 plus BDNF 2 27732515 TGACTATATACTTATGTCT(SEQIDNO:77) 8761 52 7 plus Intergenic 2 134102729 TGACCATATTCTTATGTCT(SEQIDNO:78) n/a 53 X minus Intergenic 2 119842107 TGACCATATACTTTTGTGA(SEQIDNO:79) n/a