BACKGROUND OF THE INVENTION

This invention relates to a single cell assay for determining the effect of chromosomal contact on the transcriptional activity of genes of interest in a cell and to methods of silencing gene expression in a cell by way of perturbing gene regulatory elements which are engaged in chromosomal contact.

Gene regulation begins with signal transduction cascades resulting in transcription in the nucleus. Gene expression is inherently stochastic with transcription occurring in bursts as gene switch from inactive to active states. Attempts have been made to decipher gene expression and relate it to biological noise and stochasticity, while simultaneously accounting for cell-to-ceil variation.

The highly conserved endogenous eukaryotic RNA interference (RNAi) pathway is the standard approach used to silence gene expression. This approach involves the transfection of exogenous small interfering RNAs (siRNA) and analysis of alterations in gene expression at a population level. Therefore, siRNA-approaches fail to reveal cetl-to-ce!! variability. Off-target effects represent another major challenge in the use of siRNA as gene knockdown tools.

The eukaryotic nucleus is an immensely crowded, yet paradoxically, highly organized environment. Its main constituent, DNA, is folded many times over reducing its one-dimensional length by over one million fold in space. Due to this tight compaction, regions of looped chromatin are permitted to interact or "kiss."

By identifying sites where DNA interacts, the population based chromosome conformation capture (3C) and derivative techniques (4C, 5C, HiC, ChlA-PET) allow the characterization of the global interactome (Dekker et a!., 2002, Fullwood et al., 2009). Analyses of Hi-C data reveal that chromatin is divided into smaller chromosomal interacting domains, referred to as topological^ associating domains (TADs), which range in size from several hundred Kb to -3 Mb (Lieberman-Aiden et al., 2009). By constraining the DNA topology, TAD structure may enhance the probability of specific loop-mediated interactions occurring thus playing a direct role in transcriptional activity.

The subdivision of the genome into subchromosomal domains, or TADs, represents a highly conserved feature of chromosomal organization (Dixon et al., 2012). This suggests that each cell across a population would possess a generally similar arrangement of its chromosomes. Paradoxically, traditional in situ studies reveal a large degree of heterogeneity across a population of genotypically identical ceils (Li et al., 2012). Since C-techniques report an ensemble of many nuclei, the dynamic nature of TAD structure at the single cell level and any consequent effects on transcription are obscured in the data. Clearly, interrogating chromosomal interactions at the level of a single cell is imperative to the interpretation of global interactome studies.

A characteristic feature of TADs is the enrichment of intra-domaSn chromatin contacts. FISH results confirm a spatial distinction between domains and contact arrangements within domains, as loci within a single domain are closer in nuclear space than those in different domains despite having similar genomic distances from one another. TADS therefore may provide a layer of structural regulation governing the principle of long-range chromatin contact.

Within TADS, and/or at the interface between TADS, chromosomal contact is strongly correlated with the transcriptional activity of interacting DNA elements (Li et al., 2012). Recently, we developed a novel single cell microscopy-based assay, to directly address the role of loop-mediated contact on the transcription of interacting genes. Using this assay, we demonstrated that chromosomal contact plays a central role in supporting transcription of co-regulated genes.

The invention described here relates to this single cell assay to silence gene expression through the discrete targeting (by cutting) of gene regulatory elements engaged in chromosomal contact. Major regulatory elements include; (I) enhancers, (li) sites within chromatin loops that engage in intra- or interchromosomal contact in multigene complexes and (III) regulatory sites within chromatin loops that determine loop structure.

Transcription of co-regulated genes occurs concurrently with long-range chromosomal contacts to form multigene complexes. Such contacts and transcription are lost in knockout studies of transcription factors and structural chromatin proteins. However, these approaches do not reveal the necessity of chromosomal contacts for cotranscription in multigene complexes. To interrogate in a discrete manner the role of loop-mediated contact on co-transcription, we devised a novel strategy using TALENs to cleave and disrupt gene loops in a well-characterized multigene complex. Monitoring this disruption using RNA FISH and immunofluorescence microscopy revealed that perturbing the site of contact has a direct effect on transcription of other interacting genes. Unexpectedly this effect on co-transcription was hierarchical, with dominant and subordinate members of the multigene complex engaged in both intra- and inter-chromosomal contact. This observation reveals the unprecedented level of influence of these chromosomal contacts on the transcription of co-regulated genes in a multigene complex. Transcription is replete with proximal and distal chromatin looping interactions whose formation represents the basic organizing principle of nuclear architecture and gene activity (Tan-Wong et al., 2012). Loop-mediated chromosomal contacts are usually identified on a genome-wide scale using population-based 'chromosome conformation capture' (3C) technologies (Dekker et al., 2002; Fullwood et a!., 2009; Lieberman-Aiden et al., 2009; Li et a!., 2012). Analyses of 3C-based data reveal a large heterogeneity in global chromatin interactions (Fullwood et al., 2009; Noordermeer et al., 201 1 ; Li et al., 2012). Therefore, interacting DNA elements identified by 3C-based technologies are verified at the single cell level using fluorescent in situ hybridization (FISH) assays (Papantonis et al., 2012). These highly sensitive assays can target either DNA or nascent mRNA, and have revealed the co-localization between FISH foci in a fraction of the population (Papantonis et al., 2010). This suggests that a subset of cells within the population may be enriched for specific chromosomal interactions. Chromosomes are large and constrained in their ability to roam the entire nuclear volume. Thus, it is reasonable to surmise that the topological arrangements after each cellular division shuffle chromosomal proximities such that their 3D arrangements are altered in 1 D space. This may lead to every cell in the population possessing unique spatial arrangements of its chromosomes.

Enhancer-promoter interactions utilize chromatin looping to trigger dynamic changes in transcription initiation (Deng et al., 2012). An example of this is the well- established model between the locus control region (LCR) and the promoter of the β- globin gene. In a tissue-specific manner, the LCR has been shown to physically contact the promoter of the β-globin gene, and initiate transcription (Deng et al., 2012). These LCR-mediated chromosomal interactions have been shown to result in variability in β-globin genes transcript levels, or variegated gene expression, across the population (Noordermeer et al., 201 1). In an otherwise identical population of cells, presumably through chromosomal interactions, such "jackpot" cells display higher levels of β-globin transcription (Noordermeer et al., 2011). Accordingly, the specific set of chromosomal interactions (and consequent gene expression that may depend on LCR-mediated interactions) will vary between cells across the population. This heterogeneity reveals the absolute requirement of single cell analysis in global interactome and gene loop studies.

Looping also brings distal genes into close proximity, enabling chromosomal contact in "multigene complexes" at a single focus of multiple RNA polymerases (Papantonis et al., 2012; Li et al., 2012). Numerous studies have demonstrated that the formation of loop-mediated contact coincides with alterations in gene expression (Fullwood et al., 2009). indeed, chromosomal contacts in multigene complexes appear to be the main modality of transcription in metazoan cells, as they are associated with over 95% of transcriptional activity (Li et al., 2012). in a comparable manner to enhancer-promoter interactions, specific chromatin interactions in multigene complexes are detected in a subset of cells within the population (Papantonis et al., 2010). Genome-wide chromatin interaction analysis with paired end tags (ChlA-PET), uncovered a multigene complex including the GREB1 locus and 3 other genes (Li et al., 2012). Of the 4 interacting genes, only GREB1 transcription is activated by the estrogen receptor-a (ERa) (Li et al., 2012). Intriguingly, despite the fact that this multigene complex may not assemble in every celt in the population, siRNAs targeting ERa disrupted all 4 interacting genes (Li et ai., 2012). Therefore, even though these chromosomal contacts may only occur in a fraction of the population they clearly play a significant roie in gene regulation. Moreover, this data supports a model of synergistic transcription, where chromosomal contact influences the transcription of the interacting genes. This would connote that the topological framework for transcriptional regulation is physical contact via chromosomal looping in multigene complexes.

Current siRNA and 3C-based experimental approaches cannot be applied to multigene complexes where all interacting genes are activated by the same transcription factor. Tumor necrosis factor alpha (TNFa) has been shown to induce the formation of such multigene complexes _> where ail interacting genes are activated by NF-KB (Papantonis et al., 2012). Ten minutes after TNFa stimulation, the promoters of genes located on the same chromosome (SAMD4A and TNFAIP) and on a different chromosome (SLC6A5) associate to form a NF-KB-dependent multigene complex (Papantonis et al., 2010). RNA FISH assays targeting the approximate sites of interaction identified by 3C, suggest an association between the formation of this NF-KB-regulated multigene compiex and the co-transcription of interacting genes (Papantonis et al., 2010). However, both 3C and FISH approaches fail to reveal the necessity of chromosomai contacts for co-transcription of these interacting genes. Therefore, to accurately interrogate a model of synergistic regulation, a discrete perturbation of a single site within a gene loop, whilst monitoring the transcriptional status of other members of the multigene compiex, is required. Importantly, owing to variegated gene expression (Noordermeer et al., 2011), this can only be achieved with a single cell approach.

Here we devise a single ceil strategy using TALE nucleases (TALENs) to discretely perturb sites within gene loops that are established to engage in chromosomal contact in the well-characterized NF-KB-regulated multigene complex (Papantonis et al., 2010). This enabled us to address the longstanding question of the requirement of loop-mediated contact for transcriptional co-regulation in a multigene compiex. Using RNA FISH and immunofluorescence (IF), we imaged the site of the disrupted loop simuitaneousiy with the transcriptional activity of other interacting genes in the NF-KB-regulated multigene complex. This unique single cell perspective revealed that perturbing loop-mediated contact between the NF-KB- reguiated genes, altered the transcriptional status of interacting genes. In addition, this effect on co-transcription was hierarchical, with dominant and subordinate members of the multigene complex engaged in intra-chromosomal contact at distances over 48mbp, as well as inter-chromosomal interactions. Furthermore, restoration of a disrupted gene loop re-established both chromosomal contacts and transcription of interacting genes in a sequenceindependent manner. The unexpected hierarchical organization within the TNFa-induced multigene complex reveals the unprecedented level of influence of these chromosomal contacts on the transcription of co-regulated genes in a multigene complex.

SUMMARY OF THE INVENTION

The present invention relates to a method for silencing gene expression and an assay for determining the effects of chromosomal contact on gene silencing.

in a first aspect of the invention there is provided for a method for silencing gene expression at a single cell level. The method includes a step of perturbing or interfering with at least one chromosomal contact in the cell. The point of chromosomal contact may include any region of chromatin and/or DNA. The method further includes a step of detecting the site of perturbation of the chromosomal contact; and further includes a step of detecting the effect of the perturbation of chromosomal contact on the transcriptional activity of a gene of interest. It will be appreciated, that a gene of interest may include a single gene and or a multigene complex.

In one embodiment, the chromosomal contact is perturbed by inducing a site- specific double stranded break in the chromatin or DNA.

in another embodiment, the transcriptional activity of the gene of interest may be abrogated or prevented by recruitment of proteins involved in a repair process to the site of the double stranded break, which when bound to the chromatin or DNA at the site of the double stranded break obstruct chromosomal contacts. Alternatively, the transcriptional activity of the gene of interest may be abrogated or prevented by means of enhancement of mobility of the region of chromatin or DNA containing the double stranded break, which results in a reduction in the capacity of a gene loop to engage in chromosomal contact, fn a further alternative, the transcriptional activity of the gene of interest may be abrogated or prevented through the loss of structural integrity of a gene loop, which results in the abrogation of chromosomal contact.

In yet another embodiment of the invention, the chromosomal contact may be an intergenic contact, an intragenic contact or both an intergenic and intragenic contact. In another embodiment of the invention, the region of chromatin or DNA that is perturbed or interfered with may comprise a gene or regulatory element selected from the group consisting of an enhancer and/or promoter, a site within a chromatin or DNA loop that engages in intra-chromosomal or inter-chromosomal contact, or a regulatory site within a chromatin or DNA loop which determines the loop structure. It will be appreciated that the chromosomal contact may be between chromatin and/or DNA located either inter-chromosomaily, intra-chromosomaliy or both inter- chromosomal!y and inter-chromosomally.

In yet another embodiment of the invention the double stranded break may be induced by a site-specific nuclease. The site-specific nuclease may be selected from the group consisting of a meganuclease, a zinc finger nuclease, a TALE nuclease, a BUD1 nuclease, and a CrispR nuclease. It will be appreciated that the site-specific nuclease may be delivered to the ceil by transfection of the cell with a vector encoding the site-specific nuclease and wherein the site-specific nuclease is endogenously expressed in the cell. Aiternativeiy, the site-specific nuclease may be exogenously expressed and the exogeounsty expressed site-specific nuclease may consequently delivered to the cell.

In yet another embodiment of the invention, the double stranded break is detected by the immunofluorescent staining of a protein involved in a cellular repair process in the cell, or by detecting the location of a recombinant protein expressing a fluorescent label which is involved in a cellular repair process in the celi.

in a further embodiment of the invention, the effect of the double stranded break on the transcriptional activity of the gene of interest is detected using a method selected from the group consisting of RNA fluorescent in situ hybridisation, live RNA fluorescent in situ hybridisation, immunogold labelling, molecular beacons and MS2 tagging.

In yet another embodiment of the invention the celi is a eukaryotic cell or prokaryotic cell.

A second aspect of the invention provides for a single cell assay for determining the effect of chromosomal contact on the transcriptional activity of at least one gene of interest in a cell. In one embodiment, a site-specific nuclease is used to induce or cause a double stranded break in a region of chromatin or DNA which is engaged in chromosomal contact in the celi. Consequently, an immunofluorescent probe or recombinant protein expressing a fluorescent label is used to detect the site of the double stranded break and one or more fluorescent oligonucleotide probes which are capable of hybridising to a target mRNA sequence produced by the transcription of at least one gene of interest are used to detect the presence or absence of transcription of the gene of interest by monitoring the fluorescence and the relative intensity of fluorescence of the fluorescent oligonucleotide probe hybridised to the target mRNA sequence. The fluorescence and relative intensity of fluorescence will be indicative of the effect of the chromosomal contact on the transcriptional activity of the gene of interest.

In one embodiment, of this aspect of the invention the transcriptional activity of the gene of interest may be abrogated or prevented by recruitment of proteins involved in a repair process to the site of the double stranded break, which when bound to the chromatin or DNA at the site of the double stranded break obstruct chromosomal contacts. Alternatively, the transcriptional activity of the gene of interest may be abrogated or prevented by means of enhancement of mobility of the region of chromatin or DNA containing the double stranded break, which results in a reduction In the capacity of a gene loop to engage in chromosomal contact. In a further alternative, the transcriptional activity of the gene of interest may be abrogated or prevented through the loss of structural integrity of a gene loop, which results in the abrogation of chromosomal contact.

In another embodiment, the chromosomal contact may be an intergenic contact, an intragenic contact or both an intergenic and intragenic contact.

In a further embodiment, the double stranded break perturbs or interfers with the chromatin or DNA engaged in chromosomal contact and consequently perturbs or interfere with the transcriptional activity of the gene of interest.

In another embodiment of the invention, the region of chromatin or DNA that is perturbed or interfered with may comprise a gene or regulatory element selected from the group consisting of an enhancer and/or promoter, a site within a chromatin or DNA loop that engages in intra-chromosomal or inter-chromosomal contact, or a regulatory site within a chromatin or DNA loop which determines the loop structure.

In yet another embodiment of the invention, the double stranded break may be induced by a site-specific nuclease. The site-specific nuclease may be selected from the group consisting of a meganuclease, a zinc finger nuclease, a TALE nuclease, a BUD1 nuclease, and a CrispR/Cas 9 nuclease. It will be appreciated by those skilled in the art that the site-specific nuclease may be delivered to the cell by transfection of the cell with a vector encoding the site-specific nuclease and wherein the site-specific nuclease is endogenously expressed in the cell. Alternatively, the site-specific nuclease may be exogenous!y expressed and the exogeounsly expressed site-specific nuclease may consequently delivered to the ceil.

In yet another embodiment of the invention the immunof!uorescent probe may be an antibody which binds to at least one of the proteins involved in the cellular repair process of a double strand break in the ceil.

In another embodiment, the location of the hybridized fluorescent oligonucleotide probes in the cell is observable by a technique selected from the group consisting of diffraction limited imaging techniques, sub-diffraction limit image resolution and other imaging techniques, such as imaging in three dimensions, wherein detection of the chromosomal location of gene expression for the target mRNA sequences and the diffraction limited or sub-diffractive location of the hybridised immunofluorescent probe is indicative of the location of the double stranded break.

In yet another embodiment, the cell is a eukaryotic cell or prokaryotic ceil. BRIEF DESCRIPTION OF THE FIGURES

Non-limiting embodiments of the invention will now be described by way of example only and with reference to the following figures:

Figure 1: Enhancer-promoter interactions. Enhancer-promoter interactions utilize looping to trigger dynamic changes in transcription. An example of this is the well-established model between the locus control region (LCR) and the promobter of the β-globin gene. These LCR-medsated chromosomal interactions have been shown to result in variability in β-globin genes transcript levels, or variegated gene expression, across the population (Noordmeer et al., (2011 )). Therefore, preventing enhancer-promoter interactions will have a significant impact on gene expression. To date, no assay exists to discretely perturb enhancer elements to silence gene expression.

Figure 2: Intra- or inter-chromosomal contact in muitigene complexes. Looping brings distal genes into close proximity, enabling chromosomal contact in muitigene complexes. In a comparable manner to enhancer-promoter interactions, specific chromatin interactions in muitigene complexes are detected in a subjet of cells within the population (Papantonis et al. (2010)). Loop-mediated contact between genes in a muitigene complex, influences their transcriptional status of interacting genes. Therefore, preventing loop-mediated contact will have a significant impact on the gene expression of interacting genes in a muttigene complex. To date, no assay exists to discretely perturb interacting genes in a multigene complex to silence expression.

Figure 3: Regulatory sites that determine loop structure. Many distinct regulatory elements have been identified that organize chromatin into loops, thereby altering gene expression. Knockout studies of these structural chromatin proteins reveal their role in regulating transcription. Hoewever, the interpretation of these studies is complicated by the global alterations in gene expression that inevitably occurs. Preventing chromating remodelling and structural proteins from bining to their target sites will have significant impact on gene expression of related genes. To date, no assay exists to discretely perturb sites where chromatin structural proteins bind, and monitor associated gene expression at a single ceil level.

Figure 4: The transcriptional response of co-regulated genes in a multigene complex is asymmetric. The promoters of genes located on the same chromosome (SAIV1D4A and TNFA!P) and on a different chromosome (SLC6A5) associate via RNA Polymerase tl (orange) to form a NF-KB-dependent multigene complex. B P4 and RCOR1 are non-responsive to TNFa, do not interact with SAMD4A (blue), TNFAIP2 (green) or SLC6A5 (red), and therefore serve as controls in this study.

Figure 5: The transcriptional response of co-regulated genes in a multigene complex is asymmetric. Chromosomal contact in a multigene complex is associated with cotranscription. Spectrally distinct RNA FISH intronic probes targeting the region of these genes involved in chromosomal contact reveal co- localization of nascent intronic RNA, as visualized by overlapping foci. Two-tailed Fisher exact test; * ^** P=4.5E-4, n, number of alleles.

Figure 6: RNA Pol ll-mediated SAMD4A, TNFAIP2 and SLC6A5 co- transcription may occur within TADs. Overlapping SAMD4A, TNFAIP2 and/or SLC6A5 RNA FISH foci consistently co-iocalize with the active, poised form of RNA Pol II (phosphoryiated at Serine 5). Bar, 10 pm.

Figure 7: RNA Pol ll-mediated SAMD4A, TNFAIP2 and SLC6A5 co- transcription may occur within TADs. HUVECs always display two spatially distinct DNA FISH foci, both before and after TNFa treatment. Ceils were counterstained with DAPI. Bar, 5 pm.

Figure 8: RNA Pol ll-mediated SAMD4A, TNFAIP2 and SLC6A5 co- transcription may occur within TADs. Analysis of overlapping DNA FISH foci revealed co-locaiized foci in both the unstimulated and the TNFa-treated HUVEC population. Cells were counterstained with DAPi. Bar, 5 pm.

Figure 9: The transcriptional response of co-regulated genes in a muitigene complex is asymmetric. The response to TNFa is asymmetric in diploid HUVECs. Allelic expression of each gene in the population of cells is shown. TNFAIP2 transcription occurs predominantly when SAMD4A is also transcribed (-86%) and SLC6A5 transcription occurs mainly when SAMD4A (-92%), or both SAMD4A and TNFA1P2 are transcribed (-62%). N, number of cells (N=166), n, number of alleles (n=332).

Figure 10: The transcriptional response of co-regulated genes in a muitigene complex is asymmetric. Data in (C) was replotted to show the combined allelic transcriptional status of SA D4A, TNFAIP2 and SLC6A5 in the total population (T/P). Colocalization frequencies relative to cells co-expressing SAMD4A, TNFAIP2 and/or SLC6A5 is also shown, n, number of alleles, Bar, 5 μιτι. See also Figures 6, 7 and 8.

Figure 11: Visualization of TALEN-mediated disruption of a gene loop. The SAMD4A TALEN targets the region in intron 1 involved in chromosomal contact 10mins post TNFa stimulation. TALE repeat domains are colored to indicate the identity of repeat variable diresidues (RVD); each RVD is related to the cognate targeted DNA base by the following code; NI=A, HD=C, NN=G/A, NG=T.

Figure 12: Detection of double strand breaks in HUVECs. Indirect immunofluorescence was performed using an antibody specific to Ser139 phosphorylation of histone variant, H2A.X. DSBs labeled by H2A.X were detected with a donkey-anti rabbit antibody conjugated to Atto488. Distinct nuclear staining was evident in HUVECs treated with 100μΜ Etoposide, a DSB-inducer, for 1hr. Cells were counterstained with DAPI. Bar, 5 pm.

Figure 13: Visualization of TALEN-mediated disruption of a gene loop. Successful detection of TALEN transfection at a single ceil levei. Discrete sites of H2A.X pSer139 were evident in -60% of HUVECs transfected with the SAMD4A TALEN. A higher portion of successfully transfected cells displayed single allelic DSBs than dual allelic DSBs.

Figure 14: A time course of SAMD4A TALEN nuclease activity. DSBs, labeled by an H2A.X pSer139 primary antibody, were detected with a donkey-anti rabbit secondary antibody conjugated to Atto488. Nuclease activity (a measure of transfection efficiency) was first evident after 6hr as evidenced by distinct sites of H2A.X pSer139 in -40% of HUVECs transfected with the SA D4A pcDNA TALENs. At this time point, the majority of ceil displayed no H2A.X foci and low levels of apoptotic cells, as evidenced by cells displaying multiple breaks (more than 2). Post 24hr transfection, -60% of ceils displayed DSBs, with slightly higher levels of apoptotic cells (-5%). After 4Shr post transfection, -70% of cells displayed DSBS and the level of apoptotic cells increased to -12%. At all time points, a higher portion of transfected cells displayed single allelic DSBs than dual allelic DSBs (at a ratio of -3:1).

Figure 15: Visualization of TALEN-mediated disruption of a gene loop. H2A.X immunofluorescence accurately labels the sites of the disrupted SA D4A gene loop. SA D4A DNA FISH foci consistently co-localize with SAMD4A TALEN- induced DSBs.

Figure 16: Visualization of TALEN-mediated disruption of a gene loop. Gel showing the surveyor nuclease result from the SAMD4A TALEN pair. NT, un- transfected control; GFP, GFP transfected cells; L, SA D4A left TALEN only; R, SAMD4A right TALEN only; L+R, cells transfected with pcDNA left and right TALENs. Red arrow, 246bp, blue arrow, 1 4bp. Cells were counterstained with DAPI. Bar, 5 m. See also Figures 12, 14, 21 , 32 and 41.

Figure 17: TALEN-mediated disruption of the SAMD4A gene loop abrogates SAMD4A RNA and protein expression. SAMD4A TALEN-induced DSB does not alter transcription up to the DSB. Nascent intronic SAMD4A RNA (detected by probe set i) transcribed 5' of the DSB was evident in 42% of HUVECs displaying DSB, as evidenced by H2A.X staining, n, number of alleles (n=194), NS, no significant difference.

Figure 18: TALEN-mediated disruption of the SAMD4A gene loop abrogates SAMD4A RNA and protein expression. The SAMD4A TALEN abrogates transcription downstream of the DSB. Nascent intronic SAMD4A RNA (detected by probe set ii) transcribed 3' of the DSB was never observed, n, number of alleles (n=84), Two-tailed Fisher exact test; ^*** P< 0.001.

Figure 19: TALEN-mediated disruption of the SA D4A gene loop abrogates SA D4A RNA and protein expression. Disrupting the SA D4A gene loop is sufficient to abrogate protein expression. TNFa induces SAMD4A protein expression 16hr post TNFa stimulation. Cells harboring dual allelic DSBs, as detected by H2A.X, displayed a significant reduction in protein expression, whilst cells harboring single allelic DSBs still express SA D4A protein. R.F.U., relative fluorescent units, meanlSD, ^* P<0.01 , Two-tailed unpaired Students t test, ceils were counterstained with DAP!, Bar, 5 μηι.

Figure 20: TALEN-mediated disruption of a single gene loop and the associated chromosomal contacts in a multigene complex alters the transcriptional status of other genes occupying the same complex. A DSB induced in the non-TNFa responsive gene, BMP4, did not alter transcription of SAMD4A, TNFAIP2 or SLC6A5 relative to the normal TNFa response (Figure 9).

Figure 21: Detection of TALEN efficiency. Gel showing the surveyor nuclease result from the BMP4 TALEN pair, NT, control from un-transfected cells; GFP, ceils transfected with GFP; L, BMP4 left TALEN only; R, BMP4 right TALEN only; L+R, cells transfected with the BMP4 pcDNA left and right TALENs. MW, O'GeneRuler 1kp Plus DNA ladder.

Figure 22: TALEN-mediated disruption of a single gene loop and the associated chromosomal contacts in a multigene complex alters the transcriptional status of other genes occupying the same complex. The disruption of the SA D4A gene loop abrogates TNFAIP2 and SLC6A5 transcription and co-localization. SAMD4A loop disruption detected by H2A.X was simultaneously monitored with transcription of TNFAIP2 and SLC6A5 by RNA F!SH.

Figure 23: Comparison between the phenotype in cells displaying a single allelic DSB to ceils displaying dual allelic DSBs. In cells where a single allele of SA D4A was targeted, virtually all transcription of TNFA I P2 and SLC6A5 was lost at the corresponding allele, with multigene transcription confined to the intact allele. Where both SAMD4A alleles were targeted, most transcription from TNFAIP2 and SLC6A5 was lost at both alleles. There was a significant reduction in TNFA1P2/SLC6A5 coiocalization frequencies in both cells displaying single and dual allelic DSBs. Two-tailed Fisher exact test; ^* P<0.05, ^*** P<0.001 , n, number of DSBs. Cells were counterstained with DAPI. Bar, 5 Mm.

Figure 24: TALEN-induced dual allelic DSBs abrogate protein expression of other members of the NF-kB multigene complex. Disrupting the SAMD4A gene loop is sufficient to abrogate protein expression of TNFAIP2 and SLC6A5. Dual indirect immunofluorescence was performed using a donkey-anti rabbit antibody conjugated to Atto488 to detect H2A.X and donkey-anti-goat secondary antibody conjugated to Atto565 to detect TNFAIP2 or SLC6A5. Cells displaying dual SAMD4A-induced DSBs, as detected by H2A.X, displayed a severe reduction in protein expression of TNFAIP2 and SLC6A5. Figure 25: The SAMD4A TALEN does affect the transcriptional status of RCOR1. RCOR1 , a gene with transcriptional activity comparable to GAPDH, located approximateiy 400kb 5' of TNFAIP2 was unaffected by the SA D4A TALEN. Two foci of RCOR1 are evident in each cell in the field of view. Cells were counterstained with DAPI. Bar, 10 pm

Figure 26: TALEN-mediated disruption of a single gene loop and the associated chromosomal contacts in a multigene complex alters the transcriptional status of other genes occupying the same complex. Disruption of the TNFAIP2 gene loop does not affect SAIV1D4A gene expression, but alters SLC6A5 transcription and SA D4A/SLC6A5 colocalization.

Figure 27: TALEN-mediated disruption of the TNFA1P2 gene loop abrogates TNFAIP2 gene expression. The TNFAIP2 TALEN targets the approximate region in intron 2 involved in chromosomal contact at 10mins post TNFa stimulation. Left and right TALENs were designed to contain 18 full monomer repeats and target a 20bp target region, where the first and last space specify the T at the N terminus and the 0.5 repeat respectively.

Figure 28: TALEN-mediated disruption of the TNFAIP2 gene loop abrogates TNFAIP2 gene expression. Post 24hr transfection, -60% of ceils displayed TNFAIP2 TALEN-induced DSBs and low levels of apoptotic ceils, as evidenced by cei!s displaying multiple breaks (more than 2). A higher portion of transfected cells displayed single allelic DSBs as opposed to dual allelic DSBs,

Figure 29: TALEN-mediated disruption of the TNFAIP2 gene loop abrogates TNFAIP2 gene expression, The TNFAIP2 TALEN abrogates transcription downstream of the DSB. Duai-transfected HUVECs were stimulated with TNFa for 10mins to allow for the first ~1.5kbp of the TNFAIP2 to be transcribed. Nascent intronic TNFAIP2 RNA (Atto565) transcribed downstream of the DSB (H2A.X - Atto488) was never observed. Two-tailed Fisher exact test; ^*** P<0.001. Cells were counterstained with DAPI, Bar, 5 pm.

Figure 30: TALEN-mediated disruption of the TNFAIP2 gene loop abrogates TNFAIP2 gene expression. TNFa induces TNFAIP2 protein expression 16hr post stimulation. Ceils were counterstained with DAPI. Bar, 5 pm.

Figure 31 : TALEN-mediated disruption of the TNFA1P2 gene loop abrogates TNFAIP2 gene expression. Disrupting the TNFAiP2 gene loop is sufficient to abrogate protein expression. Dual indirect immunofluorescence of TNFAIP2 and H2A.X (DSBs) was performed using a donkey-anti rabbit antibody conjugated to Atto488 and donkey-anti-goat conjugated to Atto565 respectively. DSBs, as detected by H2A.X, displayed a severe reduction in protein expression, R.F.U., relative fluorescent units, meantSD, ^** P<0.01 , Two-tailed unpaired Students t test, cells were counterstained with DAPI, Bar, 5 pm.

Figure 32: Detection of TALEN efficiency. Gel showing the surveyor nuclease result from the TNFAIP2 TALEN pair. NT, control from un-transfected cells; GFP, cells transfected with GFP; L, TNFAIP2 left TALEN only; R, TNFAIP2 right TALEN only; L+R, cells transfected with the pcDNA left and right TNFAIP2 TALENs.

Figure 33: Comparison between the phenotype in cells displaying a single allelic DSB to cells displaying dual allelic DSBs. In ceils displaying single allelic TNFAIP2 TALEN-induced DSBs, SAMD4A transcription remained unaffected, whilst SLC6A5 transcription was limited to the intact TNFAIP2 allele. SLC6A5 transcription was rarely observed in cells harboring dual allelic DSBs, whereas SAMD4A transcription remained unaffected. There was a reduction in SAMD4A SLC6A5 co- localization frequencies in both cells displaying single and dual allelic DSBs. Two- tailed Fisher exact test; *P<0.05, cells were counterstained with DAPI. Bar, 5 pm.

Figure 34: TALEN-induced dual allelic DSBs abrogate protein expression of other members of the NF-kB multigene complex. Disrupting the TNFAIP2 gene loop is sufficient to abrogate SLC6A5, but not SAMD4A protein expression. Dual indirect immunofluorescence of SAIV1D4A or SLC6A5 and DSBs was performed using a donkey-anti rabbit antibody conjugated to Atto488 and donkey-anti-goat conjugated to Atto565 respectively. Cells displaying dual TNFAIP2-induced DSBs, as detected by H2A.X, displayed a reduction in protein expression of SLC6A5, whereas SAMD4A protein expression did not appear to be affected.

Figure 35: TALEN-mediated disruption of a single gene loop and the associated chromosomal contacts in a multigene complex alters the transcriptional status of other genes occupying the same complex. Disruption of the SLC6A5 gene loop does not alter SA D4A TNFAIP2 transcription or co-localization. Two-tailed Fisher exact test; ^* P<0.05, ^** P<0.01, ** ^* P<0.001 , n, number of DSBs. Cells were counterstained with DAPI. Bar, 5 pm. See also Figures 23, 24, 27, 28, 29, 30, 31 , 33, 34, 36, 37, 38, 39, 40, 42 and .

Figure 36: TALEN-mediated disruption of the SLC6A5 gene loop abrogates SLC6A5 gene expression. The SLC6A5 TALEN targets the approximate region in intron 1 involved in chromosomal contact at 10min post TNFa stimulation. Left and right TALENs were designed to contain 18 full monomer repeats and target a 20bp target region, where the first and last space specify the T at the N terminus and the 0.5 repeat respectively.

Figure 37: TALEN-mediated disruption of the SLC6A5 gene loop abrogates SLC6A5 gene expression. Post 24hr transfection, -60% of cells displayed SLC6A5 TALEN-induced DSBs and low levels of apoptotic ceils, as evidenced by cells displaying multiple breaks {more than 2). Consistent with SA D4A and TNFAIP2 TALENs, a higher portion of transfected cells displayed single allelic DSBs as opposed to dual allelic DSBs.

Figure 38: TALEN-mediated disruption of the SLC6A5 gene loop abrogates SLC6A5 gene expression. The SLC6A5 TALEN abrogates transcription downstream of the DSB. Dual-transfecied HUVECs were stimulated with TNFa for 10min to allow for the first ~1.5kbp of SLC6A5 to be transcribed. Nascent intronic SLC6A5 (Atto647N) RNA transcribed downstream of the DSB (H2A.X - Atto488) was never observed. Two-tailed Fisher exact test; ^*** P<0.001. Cells were counterstained with DAPI. Bar, 5 pm.

Figure 39: TALEN-mediated disruption of the SLC6A5 gene loop abrogates SLC6A5 gene expression. TNFa induces SLC6A5 protein expression 16hr post stimulation. Ceils were counterstained with DAPI. Bar 5 pm.

Figure 40: TALEN-mediated disruption of the SLC6A5 gene loop abrogates SLC6A5 gene expression. Disrupting the SLC6A5 gene loop is sufficient to abrogate protein expression. Dual indirect immunofluorescence of SLC6A5 and H2A.X (DSBs) was performed using a donkey-anti rabbit antibody conjugated to Atto488 and donkey-anti-goat conjugated to Atto565 respectively. DSBs, as detected by H2A.X, displayed a significant reduction in protein expression. R.F.U., relative fluorescent units, mean±SD, ^** P<0.01, Two-tailed unpaired Students t test, cells were counterstained with DAPI, Bar, 5 pm.

Figure 41 : Detection of TALEN efficiency. Gel showing the surveyor nuclease result from the SLC6A5 TALEN pair. NT, control from un-transfected cells; GFP, cells transfected with GFP; L, SLC6A5 left TALEN only; R, SLC6A5 right TALEN only; L+R, cells transfected with the SLC6A5 pcDNA left and right TALENs.

Figure 42: Comparison between the phenotype in cells displaying a single allelic DSB to cells displaying dual allelic DSBs. Transcription, as well as co- localization of SAMD4A and TNFAIP2 was not affected in cells harboring both single allelic or dual allelic SLC6A5 TALENmediated DSBs. n, number of DSBs. Cells were counterstained with DAPI. Bar, 5 pm. Figure 43: TALEN-induced dual allelic DSBs abrogate protein expression of other members of the NF-kB multtgene complex. Disrupting the SLC6A5 gene loop does not alter protein expression of SA D4A or TNFAIP2. Dual indirect immunofluorescence of SAMD4A or TNFAIP2 and DSBs was performed using a donkey-anti rabbit antibody conjugated to Atto488 and donkey-anti-goat conjugated to Atto565 respectively. Cells displaying dual SLC6A5-induced DSBs, as detected by H2A.X, did not display any alteration in protein expressio of TNFAIP2 and SLC6A5. R.F.U., relative fluorescent units, meaniSD, Mann Whitney U test, *P<0.05, cells were counterstained with DAPI, Bar, 5 pm.

Figure 44: The SAMD4A gene loop was successfully repaired. Graphical representation of the repair strategy. The repair construct consists of an IRES and eGFP, flanked by splice sites.

Figure 45: The SAMD4A gene loop was successfully repaired. TNFa- induced activation of SAMD4A induces eGFP expression. eGFP signal was observed in -1% of HUVECs dual transfected with the pDT vector and PCR product for 72hr and stimulated with TNFa for 24hr. Cells were counterstained with DAPI. n, number of cells. Bar, 5 prrt. See also Figures 46 and 47.

Figure 46: The SAIV1D4A gene loop was successfully repaired with the IRES-GFP construct. HUVECs were transfected with a CMV-eGFP construct. eGFP positive or 'green' cells coincided with a large number of RNA FISH foci.

Figure 47: The SAMD4A gene loop was successfully repaired with the IRES-GFP construct. The SA D4A promoter does not constitutively express the IRES-eGFP mRNA. eGFP signal was observed in -1% of HUVECs dual transfected with the pDT vector and PCR product for 72hr and stimulated with TNFa for 24hr. -90% of cells displaying eGFP mRNA transcripts were not visualized as a 'green' cell. Cells were counterstained with DAPI. Bar 10 prn.

Figure 48: Repairing the disrupted SA D4A gene loop restores transcription of genes in a multigene complex. HUVECs were stimulated for 20min with TNFa, recapitulating transcription of the first -1.5kbp of SAMD4A as well as the IRES-eGFP. Distinct eGFP foci were evident in -10% of transfected cells, and these foci overlapped with SAMD4A mRNA. n, number of cells (n=1 18), Bar, 5 pm.

Figure 49: Repairing the disrupted SAMD4A gene loop restores transcription of genes in a multigene complex. There was no significant difference in the co-loca!ization frequencies between eGFP/SAMD4A, TNFAIP2 and SLC6A5 and mock-transfected cells. Ceils were countersiained with DAPI. N, number of cells (N=448), n, number of eGFP foci (n=45), NS, no significant difference, Bar, 5 Mm.

Figure 50: Hypothetical model of hierarchical transcription between coregulated genes in a multigene complex. TADs constrain genes to compartments in the nucleus that are admissible to long-range chromosomal interactions. Upon induction by TNFa, NF-kBresponsive genes engage in chromosomal interactions. As two of these genes reside on the same chromosome but almost 50 Mbp apart, and another resides on a different chromosome, these interactions most likely occur within a TAD. Single cell analysis 10mins post TNFa stimulation, reveals that the transcriptional response of interacting genes in a multigene complex is asymmetric (Figure 9). Allelic analysis revealed the following 4 categories were most prevalent; (a) no transcription of any gene, (b) SA D4A expression only, (c) SAMD4A, SLC6A5 and TNFAIP2 combined expression and (d) combined expression of SAMD4A and TNFAIP2. Despite the low frequency of these interactions (Figure 10), disrupting the sites within the gene loops that engage in chromosomal contact revealed a hierarchical organization between this TNFa-induced multigene complex.

Figure 51: Disrupting the !L8 enhancer abrogates transcription of the CXC proinflammatory genes. (A) The IL-8 enhancer is brough in close proximity to the IL8 by chromosomal looping. (B) Inducing a DSB in the IL8 promoter abrogates IL8 expression as well as the expression of a non coding RNA that is transcribed in the enhancer. Two-tailed Fisher exact test; ^* P<0.05, ^** P<0.01, ^*** P<0.001 , n, number of DSBs. Cells were countersiained with DAPI. Bar, 5 m.

Figure 52: TALEN-mediated disruption of a single gene loop and the associated chromosomal contacts in a multigene complex alters the transcriptional status of other genes occupying the same complex. (A) A DSB induced in the non- TNFa responsive gene, BMP4, did not alter transcription of SAMD4A, TNFAIP2 or SLC6A5 relative to the normal TNFa response. (B) The disruption of the SAMD4A gene loop abrogates TNFAIP2 and SLC6A5 transcription and co-localization. SAMD4A loop disruption detected by H2A.X was simultaneously monitored with transcription of TNFAIP2 and SLC6A5 by RNA FISH. (C) Disruption of the TNFAIP2 gene loop does not affect SA D4A gene expression, but alters SLC6A5 transcription and SAMD4A/SLC6A5 co-localization. (D) Disruption of the SLC6A5 gene loop does not alter SA D4A/TNFAIP2 transcription or co-localization. Two-tailed Fisher exact test; ^* P<0.05, ^** P<0.01 , ^** *P<0.001 , n, number of DSBs. Cells were countersiained with DAPI. Bar, 5 pm. Figure 53: Disrupting the CTCF site 3' of the TNFAIP2 gene abrogates transcription of TNFAIP2. (A) A CRISPR was designed to induce a DSB in the the consensus CTCF site located 3' of the TNFAIP2 gene. (B) The disruption of the 3' CTCF site was detected by H2A.X was simultaneously monitored with transcription of TNFAIP2 by RNA F!SH. Two-tailed Fisher exact test; ^* P<0.05, **P<0,01 , ^** *P<0.001 , n, number of DSBs. Cells were counterstained with DAP!. Bar, 5 μηπ.

DETAILED DESCRIPTION OF THE INVENTION

The present invention will now be described more fully hereinafter with reference to the accompanying drawings, in which some, but not all embodiments of the invention are shown.

The invention as described should not be limited to the specific embodiments disclosed and modifications and other embodiments are intended to be included within the scope of the invention.

Although specific terms are employed herein, they are used in a generic and descriptive sense only and not for purposes of limitation. As used throughout this specification and in the claims which foilow, the singular forms "a", "an" and "the" include the plural form, unless the context clearly indicates otherwise.

The terminology and phraseology used herein is for the purpose of description and should not be regarded as limiting. The use of the terms "comprising", "containing", "having" and "including" and variations thereof used herein, are meant to encompass the items listed thereafter and equivalents thereof as wefi as additional items.

The terms "nucleic acid" or "nucleic acid molecule" encompass both ribonuce!otides (RNA) and deoxyribonucieotides (DNA), genomic DNA, and synthetic DNA. The nucleic acid may be doubie-stranded or single-stranded. Where the nucleic acid is single-stranded, the nucleic acid may be the sense strand or the antisense strand. A nucleic acid molecule may be any chain of two or more covaiently bonded nucleotides, including naturally occurring or non-naturally occurring nucleotides, or nucleotide analogs or derivatives. By "RNA" is meant a sequence of two or more covaiently bonded, naturally occurring or modified ribonucleotides. The term "DNA" refers to a sequence of two or more covaiently bonded, naturally occurring or modified deoxyribonucieotides.

The term "chromatin" refers to the nucleoprotein structure comprising the cellular genome. Cellular chromatin comprises nucleic acid, primarily DNA, and protein, including histones and non-histone chromosomal proteins. The majority of eukaryotic cellular chromatin exists in the form of nucleosomes, wherein a nucleosome core comprises approximately 150 base pairs of DNA associated with an octamer comprising two each of histones H2A, H2B, H3 and H4; and linker DNA which extends between nucleosome cores. For the purposes of the present disclosure, the term "chromatin" is meant to encompass all types of cellular nucieoprotein, both prokaryotic and eukaryotic.

The term "gene of interest," refers to a nucleic acid sequence comprising a nucleotide sequence which includes a transcription unit, and which can be transcribed and translated into a protein. Using the methods and/or assay of the present invention the expression of a gene of interest may be interrupted or silenced as a result of perturbation of a chromosomal contact in the cell.

Chromosomal contact may be perturbed by means of by interfering with a region of chromatin or DNA in a cell, resulting in the abrogation of the transcriptional activity of a gene of interest. Transcriptional activity may be abrogated, stopped, prevented or decreased as a result of recruitment of proteins involved in DNA repair to a site of a double stranded break in the chromatin or DNA in a cell. As a result of the recruitment of these proteins chromosomal contact may be prevented. Alternatively, chromosomal contact may be perturbed by a protein which interferes with the transcriptional machinery of a cell as a result of being "parked" at a site of chromosomal contact and thus occluding chromosomal contact or blocking the assembly of transcription machinery at the site where the protein is parked, a non- limiting example of such a protein would include a deactivated Cas9 protein, which when bound to the chromatin or DNA obstructs chromosomal contact. Typically chromosomal contact is disrupted or prevent by the introduction of a double stranded break in the DNA.

Chromosomal contact may also be disrupted as a result of the enhancement of mobility of a region of chromatin or DNA which is in the vicinity of a double stranded break. Increased mobility of the region of chromatin or DNA results in a reduction in the capacity of a gene loop to engage in chromosomal contact, as a result the transcriptional activity of the gene of interest will be negatively affected.

Chromosomal contact may further be affected by the loss of structural integrity of a gene loop as a result of a double stranded DNA break, which results in the prevention of chromosomal contact, and consequently the disruption of transcriptional activity. The terms "protein," "peptide" or "polypeptide" refers to any chain of two or more amino acids, including naturally occurring or non-naturally occurring amino acids or amino acid analogues, irrespective of post-translational modification (e.g., glycosylation or phosphorylation).

The term "recombinant" means that something has been recombined. When used with reference to a nucleic acid construct the term refers to a molecule that comprises nucleic acid sequences that are joined together or produced by means of molecular biological techniques. The term "recombinant" when used with reference to a protein or polypeptide refers to a protein or polypeptide molecule which is expressed from a recombinant nucleic acid construct created by means of molecular biological techniques. Recombinant nucleic acid constructs may include a nucleotide sequence which is ligated to, or is manipulated to become ligated to, a nucleic acid sequence to which it is not ligated in nature, or to which it is ligated at a different location in nature. Accordingly, a recombinant nucleic acid construct indicates that the nucleic acid molecule has been manipulated using genetic engineering. Recombinant nucleic acid constructs may be introduced into a host cell by transformation. Such recombinant nucleic acid constructs may include sequences derived from the same host cell species or from different host cell species.

The present invention relates to a single cell assay to silence gene expression through the discrete targeting (by cutting) of gene regulatory elements engaged in chromosomal contact. It will be appreciated thai the cell used in the assay may be selected from a eukaryotic or prokaryotic cell. It will be further appreciated that a eukaryotic ceil may include, without limitation, a cell selected from a diploid or polyploid cell line, a diploid or polyploid oncogenic cell, a primary cell, stem cell, pluripotent stem cell or tissue sample or any other eukaryotic cell.

it wil! be appreciated by those of skill in the art that the fluorescent moiety used in the methods of the invention may include fluorophore selected from the group consisting of the Aiexafluor family of dyes, FA , TET or CAL FluorGold 540, HEX or JOE, V!CB, CAL Fluor Orange 560A; Cy3C or NEDB, Quasar 570A, Oyster 556D; TMR or CAL Fluor Red 590A; ROX or LC red 610E, CAL FLuor Red 61 OA; Texas red or LC red 610E, CAL Fluor Red 61 OA; LC red 640E or CAL Fluor Red 635A; Cy5C or LC red 670E, Quasar 670A, Oyster 645D; LC red 705E or Cy5.5C or 5-(2'-aminoethyl)aminonaphthalene- -sulfonic acid (EDANS), fluorescein, anthranilamide, coumarin, and terbium chelates. Major regulatory elements include; (I) enhancers, (II) sites within chromatin ioops that engage in intra- or interchromosoma! contact in multigene complexes and (Hi) regulatory sites within chromatin Ioops that determine loop structure. Each of the different types of regulatory elements is expanded below.

Enhancer-promoter interactions (Figure 1 )

Enhancers are DNA elements that regulate the activity of specific gene promoters. Enhancers may be separated from the genes they regulate by Iarge genomic distances. Enhancers are brought in close proximity to target genes by chromatin looping. Genome-wide chromatin interactome studies reveal that enhancer-promoter interactions are pervasive, and may occur in c/s even over Iarge genomic distances, or in trans across different chromosomes.

Enhancer-promoter interactions are the primary means in which cell-specific gene expression is achieved. Enhancers contain binding sites for transcription factors, chromatin remodelers and transcriptional co-activators. Enhancers with iarge numbers of regulatory protein binding sites (transcription factors, chromatin remodelers, co-activators), are termed "super-enhancers".

Super-enhancers have been shown to be more sensitive to small changes in factor concentration than those with small numbers of binding sites. Super-enhancer regulated genes may be identified by identifying genes located with the same TAD. Super-enhancer regulated genes have been shown to have important roles in embryonic stem ceil identity. Super-enhancer regulated genes have been shown to be more highly expressed than genes regulated by typical enhancers.

A remarkable recent observation is the transcription of a novel class of non- coding RNAs, including enhancer RNA (eRNA), activating RNAs and long non coding RNAs (IncRNA), arising from enhancer loci throughout the genome, including the chemokine locus. eRNAs are enriched for H3K4me1 and H3K27Ac chromatin marks, whilst IncRNAs possess, H3K4me3 and H3K36me3 chromatin marks. Through their interaction with promoters, cohesin and the mediator protein complex, eRNAs have been proposed to regulate chromatin looping.

Recent studies have used a combination of predictive polymer modelling followed by imaging and deletion of structural DNA regulatory elements such as CTCF/cohesin in cells, to ascertain variation in the internal structure of TADs between single cells. These studies have revealed the fluctuating structural environment within a TAD and the tremendous heterogeneity in TAD configurations ^

between single ceils that could be linked to transcriptional activity. Though DNA regulatory elements are identical in ali cells, eRNA activity is highly tissue specific. eRNAs are speculated to play key roles in organizing TAD structure, though this remains functionally uninterrogated.

A well-characterized example of enhancer-promoter interactions is the well- established mode! between the locus control region (LCR) and the promoter of the β- globin gene. These LCR-mediated chromosomal interactions have been shown to result in variability in β-giobin genes transcript levels, or variegated gene expression, across the population (Noordermeer et al., 2011 ). In cells unable to form this enhancer-promoter loop, forceful tethering of Ldb1 , a transcription factor thought to mediate the LCR-p-globin interaction, to the β-g!obin promoter showed that formation of the LCR- -globin loop underlies transcriptional activation (Deng et al., 2012). Therefore, preventing enhancer-promoter interactions will have a significant impact on gene expression. To date, no assay exists to discretely perturb enhancer elements to silence gene expression.

Regulatory sites within chromatin loops that determine loop structure (Figure 2)

CTCF, a sequence-specific DNA-binding protein, binds to consensus sites (CS) that frequently flank genes. At CS sites, the multiprotein cohesin 'ring-like' complex (including the Smc1-Smc3 heterodimer, Rad21 and Scc3/SA1/SA2) is loaded onto chromatin by Nipbl. The mediator complex (a multiprotein complex composed of over 30 proteins) may also be recruited to CTCF and/or cohesin occupied chromatin to stabilize loop topology, and regulate transcription initiation and elongation.

Studies to identify the architectural proteins shaping the three dimensional genome have revealed CCTC-binding factor (CTCF), Mediator and cohesin to have widespread and specific roles across the genome. Consistent with the above- mentioned role in enhancer-promoter interactions, Mediator and cohesin were found to specifically bridge short-range, cell-type-dtstinct interactions. On the other hand, long-range interactions were found to be bridged by CTCF and cohesin. Boundaries of TADs are enriched in binding sites for CTCF and cohesin complex, implicating their importance in maintaining domain integrity and loop-mediated transcription. Experiments in cells containing a deletion that spans the boundary between the Xist and Tsix TADs in the X-chromosome inactivation center, directly demonstrated that the loss of boundary had led to the partial fusion of the adjacent TADs and the formation of new and ectopic contacts, causing long-range transcriptional misregulation (Dixon et al., 2012). Furthermore, depletion of CTCF and cohesin has revealed that these factors contribute differentially to domain organization and transcriptional regulation. Particularly, disruption of cohesin reduces local chromatin interactions although TADs remain intact, whereas depletion of CTCF leads to a reduction in local intradomain interactions but also to an increase in interdomain interactions, in each case, different classes of genes are misreguiated, indicating that each factor has a distinct capacity in chromatin organization and gene regulation.

Knockout studies of these structural chromatin proteins (mediator, cohesin, CTCF) reveal their role in regulating transcription. However, the interpretation of these studies is complicated by the global alterations in gene expression that inevitably occurs. Preventing chromatin remodeling and structural proteins from binding to their target sites will have significant impact on gene expression of related genes. To date, no assay exists to discreiely perturb sites where chromatin structural proteins bind, and monitor associated gene expression at a single cell level.

Intra- or interchromosomal contact in muitigene complexes (Figure 3)

Within a TAD, or at the interface between TADs, chromosomal looping may also bring co-regulated genes into close proximity to permit chromosomal contact. These interactions are proposed to occur at discrete foci of active, hyper- phosphorylated RNA Pol II that have been referred to as muitigene complexes or transcription factories.

Sites within a chromatin or DNA loop which engage in intra-chromosomal or inter-chromosomal contact may include both coding and noncoding regions, including, but not limited to, introns, exons and 3' un-translated regions.

Numerous studies have demonstrated that loop-mediated contact between coregulated genes coincides with alterations in the transcription of interacting genes (Papantonis et al., 2010, Li et al., 2012). Over 95% of genes engaged in promoter- mediated chromosomal contact in muitigene complexes are associated with transcriptional activity. Therefore, chromosomal contact between coregulated genes in muitigene complexes appears to be a critical component of transcriptional regulation.

The present inventors developed a novel single cell microscopy-based assay, to directly address the role of loop-mediated contact on the transcription of interacting genes. Using this assay, they demonstrated that chromosomal contact piays a central role in supporting transcription of co-reguiated genes. Therefore, preventing loop-mediated contact will have a significant impact on the gene expression of interacting genes in a multigene complex. This is the first assay described to date, that enables the discrete perturbation of interacting genes in a multigene complex to siience expression.

The present invention provides a means to enhance the study of cellular processes through the discrete perturbation of chromatin regulatory elements that are engaged in chromosomai contact.

The invention provides a means for the discrete perturbation of chromatin regulatory elements resulting in the abrogation of transcription that can be applied in translational medicine. For example through the precise delivery of the nuclease to the correct celi type and location. Or through the manipulation ex vivo of patient cells with the nuclease that are then autologously transplanted into the patient with the nuclease modification having been conducted in the patient cells.

The perturbation is induced by site-specific nucleases (including meganucleases, zinc fingers (ZF), transcription activator-like effectors (TALE), the Type II CRISPR system (clustered regularly interspaced short palindromic repeats, BUD nucleases) which are designed to induce a double strand break (DSB) within the regulatory site (Li et al., 201 1 ).

Typically, these site-specific nucleases are used in 'cut and paste' experiments, whereby the cell's own repair responses, nonhomologous end-joining (NHEJ) or homologous recombination (HR), are exploited to repair the DSB. This is the first description of the use of these site-specific nucleases for the purpose of disrupting chromosomal contact.

Through the use of a catalytically inactive Cas9 coupled to a repressor domain, the CRISPRi system has been used as an efficient means to silence gene expression at a population level. However, the CRISPRi approach fails to reveal cell- to-cell variability. Further, as this approach only silences at the level of transcription, it is unable to perturb chromosomal interactions between regulatory elements.

Our results establish that the CRISPR system can be used as a modular and flexible DNA-binding platform for the recruitment of proteins to a target DNA sequence, revealing the potential of CRISPRi as a general tool for the precise regulation of gene expression in eukaryotic ceils. Site-specific nucleases are delivered into five celis by transfection (including microporation, electroporation, lipid transfection, calcium phosphate transfection) and endogenously expressed.

Cells for use for the assay encompass any type of eukaryotic cell, including but not limited to primary and induced pluripotent stem cell lines.

The DSB, or site of disruption, is detected at a single ceil level by immunofluorescent staining of a factor of the DSB repair process.

The disruption caused by the DSB may serve to do one of the following: (f) A large number of other proteins involved in the repair process are recruited to the DSB site. Therefore, one possibility could be that chromosomal contacts between genes or gene elements may still occur, but are "bridged" via the repair complex. This may occlude the ability of gene loops to "deliver" Pol I! to other interacting genes. (II) DSB induction has been shown to enhance the mobility of damaged chromatin. Therefore, due to increased movement, the probability that the disrupted gene loop will be able to interact with other loci is significantly reduced. (IN) Alternatively, it is unknown how the DSB, and associated repair factors, influence gene topology. Therefore, the DSB may prevent loop-mediated contact by destabilizing or collapsing the gene loop.

In parallel to the detection of the DSB by immunofluorescent staining of a factor of the DSB repair process, transcriptional activity of: (i) genes of interest, or (ii) interacting genes in the multigene complex, are detected using intronic single molecule RNA fluorescent in situ hybridization (smFISH). As introns are typically excised and degraded co-transcriptionaliy, intronic FISH foci represent the transcriptional start site (TSS).

The following examples are offered by way of illustration and not by way of limitation.

EXAMPLE 1

Cell Culture

Early passage HUVECs from pooled donors (Lonza) were grown to -80% confluence in Endothelial Basal edium-2 (EGM-2) with supplements (Lonza), serum-starved (18hr) in EGM-2 + 0.5% FBS, and treated with TNFa (10 ng/mi; Sigma) for up to 30mins. Prior to transfection cells were grown in antibiotic free EGM- 2. TALEN design

Software developed by the Bogdanove laboratory (https://talent. cac.cornelt.edu) was used to identify TALEN candidate binding sites (Doyle et al., 2012). Left and right TALENs were designed to contain 18 full monomer repeats, which target a 20bp sequence, where the first and last bases are specified by the thymine at the N terminus, and the 0.5 repeat, respectively. To facilitate Fokl dimerization, the left and right TALEN target sites were chosen with a spacer of 16- 19bp. For the pDT TALEN vector, the SAMD4A left and right arms were cloned into the pB!_CMV1 bidirectional promoter vector (C(ontech). The left TALEN was cloned into MCS1 of pBLCMVI {MM and tf/ndlH) and the right TALEN was cloned into CS2 of pBLCMVI (EcoRi and BglU).

TALEN binding sites and spacer regions

The SA D4A TALEN recognition sequences (for both pcDNA and pDT vectors) are: left TALEN 5'-TCC ACG TTT ATA AAT AGC TG-3' (SEQ ID NO:1) and right TALEN 5'-CAC TGG GGT GTG GAA GCA TA-3 ^! (SEQ ID NO:2), with a 16bp spacer. The TNFAIP2 TALEN recognition sequences are: left TALEN 5'-TTC GCG GCC CAC CTG GCC GC-3' (SEQ ID NO:3) and right TALEN 5 -CTG TGC GAG CGC GAC ACC TA-3' (SEQ ID NO:4), with a 16bp spacer. The SLC6A5 TALEN recognition sequences are: left TALEN 5'-TTG TCC CTT TAA AAC TTG AA-3' (SEQ ID NO:5) and right TALEN 5'-TTA TCA AAC TTG TAT TAT CA-3' (SEQ ID NO:6), with a 17bp spacer. The B P4 TALEN recognition sequences are: left TALEN 5'- TGC AGC GCC ACA GTC CCC GG-3' (SEQ ID NO:7) and right TALEN 5'-CAA CCG TTC AGA GGT CCC CAG-3' (SEQ ID NO:8), with a 19bp spacer.

TALEN synthesis

TALENs were generated using the protocol by Sanjana et al. 2012. Briefly, specific primers were used to amplify monomers from each of four plasmids to add the appropriate adapters (each plasmid corresponding to one RVD; NI=A, HD=C, NG=T and NN=A G). Monomers generated by PCR were then purified by gel extraction and DNA concentration was normalized to create a monomer library. In the first Golden Gate reaction step, monomers were simultaneously digested with the type IIS restriction enzyme BsmBI and ligated to generate circularized hexamers. Non-hexamers were removed by exonuclease treatment. Hexamers were then amplified by PCR, gel purified and the DNA concentration between samples was norma!ized using the Qubit High Sensitivity DNA quantification kit (Life Technologies), fn the second Golden Gate step, hexamers were simultaneously digested with Bsal, and ligated into the appropriate TALEN cloning backbone (corresponding to one of the four different bases targeted by the 0.5 repeat) to generate the final TALEN expression construct. Coiony PCR was used on E. coli transform ants to identify successful TALEN clones. HUVECs were then microporated with the respective TALENs using the Neon® Transfection System (Life Technologies) according to manufacturers instructions. Nuclease activity was assessed by the surveyor assay and the consistent co-tocaiization of DNA FISH with double strand breaks, as assessed by H2A.X immunofluorescence (Keogh et al., 2005).

Surveyor Assay

Genomic DNA of transfected cells was extracted using QuickExtract DNA extraction solution (Epicenter). The genomic region encompassing the nuclease target sites in human SAMD4A, TNFAIP2, SLC6A5 or BMP4 were PCR amplified and amplicons cleaned up with QIAquick PCR Purification Kit (Qiagen). For SAMD4A the primers were 5'-TGA GGG AGA TTC CAT TGA GC-3' (SEQ ID NO:9) and 5'- GGA AAA AGT GCT GCT CCA AC-3' (SEQ ID NO:10). For TNFAIP2 the primers were 5'-TGC AGG ACA GAC TCA GGA CA-3' (SEQ ID NO:1 1 ) and 5'-ATT TGG GTT GAG CAT TCC AC-3' (SEQ ID NO: 12). For SLC6A5 the primers were 5'-TGA TTT AAC CCC CTC CTT CC-3' (SEQ ID NO: 13) and 5'-CTT TAG GAG CCA CAG CCA AC-3' (SEQ ID NO: 14). For BMP4 the primers were 5'-CTA GTA CCT CCG CAC GTG GT-3' (SEQ ID NO: 15) and 5 -TCC AGC ACC ACT ATT GGA AA-3' (SEQ ID NO: 16). The DNA fragments were then subjected to digestion with mismatch- sensitive T7 endonuclease I (T7E1 ; New England BioLabs). For the T7E1 assay, DNA was denatured at 95°C for 5 min, slowly cooled down to room temperature to allow for formation of heteroduplex DNA, treated with 5 U of T7E1 for 1 hr min at 37°C, and then analyzed by 1 .2% agarose gel electrophoresis.

RNA FISH probes

RNA FISH was performed according to the protocol by Raj et al. 2008 using 48 20-mer probes (Biosearch) targeting the following genes: SAMD4A (probes set i = ~1.5kbp into intron 1 , probe set ii = -34 kbp into intron 1 , TNFAIP2 (intron2), SLC6A5 (intronl ), RCOR1 (intronl ) and 32 20-mer probes targeting eGFP (Table 1 ). Each 20-mer bares a 3'-amino-modifier C6-dT. The amino group was subsequently conjugated to the foflowing NHS-ester dyes: ATTO- 488, ATTO-565, ATTO-647 (ATTO-TEC) or Aiexa Fluor 647 (!nvitrogen). Briefly, oligonucleotide probes were ethanoi precipitated and resuspended in 0.1 M sodium tetraborate (Sigma).

Approximately 0.3 mg of the NHS-ester dye (ATTO-TEC) was dissolved in dimethy! sulphoxide (Sigma). The dye solution was added to the probe solution and incubated overnight in the dark at 37°C. Following conjugation reaction, the probes were ethano! precipitated overnight, and resuspended in 0.1 Triethyl ammonium (TEA, Sigma). Conjugated probes were separated and purified to enrich for dye-conjugated probes by reverse phase HPLC on a C18 column.

Table 1 : Sequences of RNA FISH probes

SAMD4A 1.5kbp into intron 1 SAMD4A 34kbp into intron 1

Sequence Sequence

Probe Sequence (5' - 3') Identity Probe Sequence (5" - 3') Identity

Number Number

ATT CAT G CT ACCGT AGCTAC SEQ ID NO-,25 CCCCACCCAATATGAATGAA SEQ ID NO:73

GGAAACTGCATGAGAGAAAA SEQ ID NO:26 CTGTTGAGACTCATTCTTGT SEQ ID NO:74

GACCTGACTTACTTATTTCC SEQ ID NO:27 GCTTTAAAATGTCGCTGGGT SEQ ID NO:75

CCAGAAAATCCAGACTCTAC SEQ ID NO:28 GTGATGGCCCTTTGATTTAC SEQ ID NO:76

GGAGGAAGGAAAAAACACAC SEQ ID NO:29 TAAT CCTGCCATTCCT AAG C SEQ ID NO:77

AGCCACGTTGCCCAAAAGAA SEQ ID NO:30 CCACTGTTTAACCG CAACAA SEQ ID NO: 78

CATGTGTTGCTGAAATCCAG SEQ ID NO:31 AATTCCAACATTGGAGCCTC SEQ ID NO:79

ATCCAAGCTTGGCTTCTGAA SEQ ID NO:32 CACACAGACACAAGATTCAA SEQ ID NO:80

TTCATGTACTCCTCACACAG SEQ ID NO:33 GGTTTCTGCCTAAAACAGCA SEQ ID NO:81

CCAGAATATCTGTGGGGAAA SEQ ID NO;34 ATTCTCTCCAAGTCATCAGC SEQ ID NO:82

TT CTG AAG ACG AAG CTCT AA SEQ ID NO:35 T AAGAGTGTCTG AG G G GTT A SEQ ID NO:83

TTACTAGTCTCTAGCGTCAC SEQ ID NO:36 GTCTAGCCAGACTGTAGTTT SEQ ID NO:84

GCAGTAAGCTTAACCGCATT SEQ ID NO:37 CTCACCCCAAAGTCTTTAAG SEQ ID NO:85

CAAGATCCGTATCAATATGG SEQ ID NO:38 TCTAAGGACTTAG CACC ATC SEQ ID NO:86

CCTTTCTC CAAG AC C CTTTT SEQ ID NO:39 G CCTTCG G AGTTTTCTTCTT SEQ ID NO:87

AAGT AAC C C ACTTCATGCCT SEQ ID NO:40 TTTCTCTCC AAC ATG GTTG C SEQ ID NO:88

GCAGGGTAATATGAAACGAT SEQ ID NO:41 GTTTCT AC ATTG CTTCCCTC SEQ ID NO:89

CATACTAGTTGAGGTGTCTG SEQ ID NO:42 TTTCTGTCCTGCTTTCTCCT SEQ ID NO:90

GCTGGACCTTTCGACTATAT SEQ ID NO:43 AAAGAAGTCTCTGGTACCAG SEQ ID NO:91

CCACGCTAGCAAATAGGAAA SEQ ID NO:44 GTCTAGCCCAAGTGAGAGAA SEQ ID NO:92

CCTACCCTCCAGGATATAAT SEQ ID NO:45 TAAACCAGAAGCCTGTCTCT SEQ ID NO:93

T C ACAACCAT C AG ACTTTCC SEQ ID N046 ATCTAATCCATGGCCAGCCA SEQ ID NO:94

GCAGCAGCATGAACTAAAGA SEQ ID NO:47 AAAGGGCATTGATTCCACAG SEQ ID NO:95

TAAACACTGGGGACTCTGTT SEQ ID N0.48 ACTTGGAAAGATGTCCCAAG SEQ ID NO:96

ACCCTGCATTCTTTTCTCTG SEQ ID NO:49 CCCAGTTAATCCTAAGGAAA SEQ ID NO:97

AACATGGAACAGCTGGAAGA SEQ ID NO:50 GCTTCTTCTTTGTACAG TG A SEQ ID NO:98

ACCTTGT C ATC AA ATG G CAG SEQ ID NO:51 CAGTCCTCTAGCAATGGAAA SEQ ID NO:99

ACTCACTTTAGTGTCTCCCA SEQ ID NO:52 GTGTGTGTGTGTGTGTGTAT SEQ ID NO:100

TCGCTTCTTGCTGCTCTGAA SEQ ID NQ:53 TTTCGTGTGTGTGTGTGTGT SEQ ID O:101

ATACTAGGGAGGAGGAATGA SEQ !D NO:54 AAAGCCACTGTCTGTGCTAT SEQ ID NO:102

TGTTTTCACCATCGTGCACA SEQ ID NO:55 AAAGGAAGTATCTTCATCCC SEQ ID NO:103 AT AT GG AAGC ATCC C ATTCT SEQ ID NO:56 CAGAAGAGACTAGAACAGAC SEQ ID NO 104

CACTTCCACCCTATGATTCT SEG ID NO:57 ACTAGTTTCTGTTCCACGCA SEQ ID NO 105

AACTGTGAAGATTTCCAGCG SEQ ID NO:58 CTTCATCTCCTTAAGCTAGC SEQ ID NO 106

AG TCCTGTCTGGTCAG GAAA SEQ ID NO:59 CATGTCATTCCCACAAGAAC SEQ ID NO 107

TTGGCCATGCAGGATCTTTC SEQ ID NO:60 TGAGAAGGTTTCTGTCCATG SEQ ID NO 108

ATGGAATCTCCCTCAATGTG SEQ ID NO:61 AAAGCTTCCCAATTCTAGCC SEQ ID NO 109

AAACGTGGAGCTCAACCAAA SEQ ID NO:62 CCGAAGGTGAATGTCTTAAG SEQ ID NO 1 10

AATTATGCTTCCACACCCCA SEQ !D NO:63 ATCTCCTAATGCTATCCCTC SEQ ID NO 1 11

AGAGGGTGGATCATCAGTTA SEQ !D NO:64 CTGCACCCATTAACTCATCA SEQ ID NO 1 12

CCGCTGAAATTAAGGAAGGA SEQ ID NO:65 TATACATGTGCCGTGTTGGT SEQ ID NO 1 13

AA AACCG AG C AG GT AAACC A SEQ ID NO:66 CAACGTGCAGGTTTGTTACA SEQ ID NO 114

CCTTGTTTGTCAAACCTGAG SEQ ID NO:67 GAGAGAATGTAAGAAGGCGA SEQ ID NO 1 15

GCGTAAATCCTCTGCAATCT SEQ ID NO:68 TTTAGCCTGACAGGCATGAA SEQ ID NO 116

CCAGCACTGAAATCAAGCAT SEQ ID NO:69 GTG CACTTG AAGT AACA AG G SEQ ID NO 1 17

AGATGAGTTTGAACAGTCCC SEQ ID NO:70 TGATGTCCTTGATACTGGTC SEQ ID NO 1 18

GT G CTGCTC C AACATTTG AA SEQ ID NO:71 GCTGCATTTCCAAGAAGATC SEQ ID NO 1 19

CTCCATTACTCAAAGGGAAA SEQ ID NO:72 ATTCAGTAGGTCTGCAGTGT SEQ ID NO 120

TNFAIP2 intron 2 SLC6A5 intron 1

Sequence Sequence

Probe Sequence (5 ^* - 3') Identity Probe Sequence (5' - 3') Identity

Number Number

TACCTGCCCTATCACCCCTC SEQ ID NO:121 CCCTTTCCTCTTGAAAGAAC SEQ ID NO:169

ACCTTCCTCTCACCTTCTCC SEQ !D NO:122 AACGAATCTGCTTTCCCTGT SEQ ID NO:170

AGAGGACATGCTCAGGAGCA SEQ ID NO:123 GCACAGACATCTCCAAAAGA SEQ ID NO:171

TAGCTCTGCTCAGCTCCAGA SEQ ID NO:124 AGAATTGTTCTCCTCGTCGC SEQ !D O:172

ATCACTTGGAACTCTCCTGC SEQ ID NO:125 AGTACAAATACCTAGGGCTG SEQ ID NO:173

TAACAAGACACGTGGCCCTA SEQ ID NO: 126 ACAGCCTAGGAAACCTCTTT SEQ ID NO:174

ACAGGGTAACCCTACACAAT SEQ !D NO:127 TCTTTAGGAGCCACAGCCAA SEQ ID NO: 175

TGTACTCAGCAGGGGTACCA SEQ ID NO:128 TTTTCAGCACCGAGGACAGT SEQ ID NO: 176

ATCT ATG G AGTCAAG GG AC G SEQ ID NO:129 AGAGCAACTCGCCTCTGTAC SEQ ID NO:177

ATGCAGACCGTGCGAATGAC SEQ ID NO: 130 TTTGTCTCAGTAGCCTCTAG SEQ ID NO: 78

ACACTG CAGTAAGCCTCTCT SEQ ID NO:131 AAATCTAG GT ACCTCG GCTT SEQ ID NO:179

AC AATAAC AG CCCCTGG ACT SEQ ID NO:132 GCAGAATAGAACTCCTCGAT SEQ ID NO:180

TTCACACTAGGATCTGGGGC SEQ ID NO.-133 TCTAACCCCCTTTCACAAAC SEQ ID NO:181

CTCCAGGCTTCCTTCCTAAA SEQ ID NO:134 TCCTATTATCTTCGCCTCCA SEQ ID NO:182

AC ACTG CCC ACACTG T CTAT SEQ !D NO:135 GCTTCCTAAAC ATT AG C ACC SEQ ID NO:183

TTCAACACGAACTGAGCCCA SEQ ID O:136 TCCCCTCGTGTTATCTTTCA SEQ ID NO:184

AT CCTTCTG C C AG AGGCTG A SEQ ID NO:137 CGGTGGTTTTTCTATCCACT SEQ ID NO:185

CTCTCCACTGGACCCTTTCT SEQ ID NO:138 CGCTTTTTCCCAATTCACTT SEQ ID NO:186

TATGCCCTGCTGACTCCTCT SEQ ID NO:139 GAGTACCAAAACTCATGACT SEQ ID NO:187

TTATCTCTCCCGTGGCCCCT SEQ ID NO:140 CTTCACAACTCTAAGCTTGC SEQ ID NO:188

AGCAAGCGGGAACTGGCTCT SEQ ID NO:141 CGTAACTATTCACAGGAGTC SEQ ID NO:189

CATTTCCTGGTGAGTCAGGA SEQ ID NO:142 CTTCTTCTTCCTCTTCTTCT SEQ ID NO:190

TCACTGGGACAGAAAGACTG SEQ ID NO:143 TCTTCTTCTTCTTCTTCTTC SEQ ID NO:191

TTGCCTGTTGCTAACCCCAG SEQ ID NO: 144 CTTCTTCTTCTTCTTCTTCT SEQ SD O:192

TCCATGCCCAGCTTGGCTTT SEQ ID NO:145 GGTCTTCTTCTTCTTCTTCT SEQ ID NO:193

ACACAG GCTTC AACG ATGCC SEQ ID NO:146 CCCAGACATGGAAAAGAATA SEQ ID NO:194

CTCCGTT C CCC AC ATTTCTG SEQ ID NO: 147 CAACTCGTATTCACTTCCAG SEQ ID NO:195

CAATCCAGATGCAGCTGTGT SEQ !D NO:148 G ACACAG AGTAGTAACT AG A SEQ ID NO:196

AAATGACCTTCTCTCTGCCC SEQ ID NO:149 CCACAAAAAAACCTACAACG SEQ ID NO:197 CTTTGTCTGCCTAGAGGTTT SEQ ID NO:150 CAG GT AG C AA AG AC AAG GTT SEQ ID NO:198

AGCCCTGGCACTTCCTAAGG SEQ ID NO:151 TGAGGCCAAATTGGTTGAGT SEQ ID NO:199

ATCTTGGCTCACTGTCCTGG SEQ ID NO: 152 ACAAAGCTGATGCCCTGGCA SEQ ID NO:200

CCCCAAACTGCTCCACAGAC SEQ ID NO:153 TTTTCCCCCATAGACAAGCT SEQ ID NQ:201

TAGGCCAGAATGGGCAGGAA SEQ ID NO:154 CATTCAACAATGCACACTGC SEQ ID NO:202

AG GTCTTG GG C ATCTC ACCA SEQ ID NO:155 TTTTCCGCCCCATTTCCTTT SEQ ID NO:203

ACTCACAGCAGCCCTGGAAT SEQ ID NO:156 CC ATCACC ACTGTAG AA AG A SEQ ID NO:204

TGTCACCCAGATAAAACCCT SEG 1D 0:157 TTGG ATTC C AC ATC ACAG CT SEQ ID NO:205

TTCTTCCTGCTCCAAGACTA SEQ ID NO:158 AAATAGCCCTGGAGTCATGA SEQ ID NO:206

GAGCCACCTTTAAGATCTGA SEQ ID NO:159 TGTAAGGGTAAAGGGGGAAT SEQ !D NO:207

AATT CCTTCTTC ACC C AG CA SEQ ID NO:160 G GATG AACT AAG AA AAG AGG SEQ ID NO:208

TGCTCCTTCCCTCACCTCCA SEQ ID NO:161 CTAATATTCCCTCCCTAGTG SEQ ID NO:209

TCACTGCAAAAGCCTCCTCC SEQ ID NO:162 CCCGTTT C ATC AAT AACC AG SEQ ID NO;210

AAATCAGCAGCCTCATGCCA SEQ ID NO:163 GGCTGCAATCAGTATTTCTC SEQ ID NO:21 1

TAG GT CCT G CTCC AAAAT AG SEQ ID NO:164 GGGAAGCAAATTGGAATCTG SEQ ID NO-.212

ATCATGCCTCCTCTGACTCT SEQ ID NO:165 TCAGGTCCAGACACAATATC SEQ ID NO:213

TT GG AAC AG A AACT CTGAAG SEQ ID NO:166 TAGATATGGAGACGTTCGCA SEQ ID NO;214

TCCTCCATCTACTTAGTTTG SEQ 1D 0:167 C A AAACAAAGTC ACAAC AG C SEQ ID NO:215

GGTCCCCATTTCAGATGATG SEQ ID NO:168 CAAACACAATGTCAAGTTCG SEQ ID O:216

RC0R1 intronl eGFP

Sequence Sequence

Probe Sequence (5' - 3') identity Probe Sequence (5' - 3') identity

Number Number

CAAAAAGAAAGTTGTCGGCG SEQ ID NO:217 TCCTCGCCCTTGCTCACCAT SEQ ID NO:265

GTTTGCAGAACACTCGTGTG SEQ ID NO:218 ATGGGCACCACCCCGGTGAA SEQ ID NO:266

GCTCACCTGGACGCAGGGAC SEQ ID NO:219 TTACGTCGCCGTCCAGCTCG SEQ ID NO:267

GCAGGACACTAACTCTCCGG SEQ ID NO:220 GACACGCTGAACTTGTGGCC SEQ ID NO:268

GGACCTCCTAAGTCCGGGGC SEQ ID NO:221 GGCATCGCCCTCGCCCTCGC SEQ ID NO:269

GTTCCGAGAAAGGGGTCTCT SEQ ID NO-.222 TCAGGGTCAGCTTGCCGTAG SEQ ID NO:270

CAGTCAGGAGCCAGCAAGAG SEQ ID NO:223 TTGCCGGTGGTGCAGATGAA SEQ ID NO:271

GTTCTTCTACTCGCCTTCTG SEQ ID NO:224 GGTGGGCCAGGGCACGGGCA SEQ ID NO:272

AG GTCCT AAAATC C AACTG C SEQ ID NO:225 CGTAGGTGAAGGTGGTCACG SEQ ID NO:273

GGCTAAAGCTTCTGATGCTT SEQ ID NO:226 TAGCGGCTGAAGCACTGCAC SEQ ID NO:274

CTAGGGTGTAGGGCTTTCTT SEQ ID NO:227 GTGCTGCTTCATGTGGTCGG SEQ ID NO:275

CCAGGGAACTAAGTGTTCCA SEQ ID NO:228 G C ATGG CG G ACTTG AAG AAG SEQ ID NO:276

TTGCTTCTACACAGAATAGC SEQ ID NO:229 CGCTCCTGGACGTAGCCTTC SEQ ID NO:277

TGACGAATTCCTTTGTCTGA SEQ ID NO:230 GTCGTCCTTGAAGAAGATGG SEQ ID NO:278

TG GCTAG AATCTCAT G CATT SEQ ID NO:231 CGGCGCGGGTCTTGTAGTTG SEQ ID NO:279

AAAGACAGTCTCATACAGAC SEQ ID NO;232 GTGTCGCCCTCGAACTTCAC SEQ ID NO:280

CTCCAGCTTTACAAGATGCT SEQ ID NO:233 TTCAGCTCGATGCGGTTCAC SEQ ID NO:281

TCCAACCATGGAGGATTGTT SEQ ID NO:234 CGTCCTCCTTGAAGTCGATG SEQ ID NO:282

TGGTATAACCCACATACAGA SEQ ID NO:235 AGCTTGTGCCCCAGGATGTT SEQ ID NO:283

TTTCCAAGGAAAAAACATGT SEQ ID NO:236 GTGGCTGTTGTAGTTGTACT SEQ ID NO:284

ATTGCAATTTTCTGCAGTCT SEQ ID NO:237 TGTCGGCCATGATATAGACG SEQ ID NO:285

CGCCTTTTTCAAGGACTAGA SEQ ID NO:238 ACCTTGATGCCGTTCTTCTG SEQ ID NO:286

TTAATAACAGCTTGAGAGCA SEQ ID NO:239 TGTTGTGGCGGATCTTGAAG SEQ ID NO:287

CTTC AAAGTG CAG TTT AGT A SEQ ID NO:240 AGCTGCACGCTGCCGTCCTC SEQ ID NO:288

GTCCAAGAGCTCACCTAGGA SEQ ID NO:241 TGTTCTGCTGGTAGTGGTCG SEQ ID NO:289

AG G A AGTC ATTT ATTTCC AG SEQ ID NO:242 CACGGGGCCGTCGCCGATGG SEQ ID NO:290

TCAA AATTT AG AAG GCCA AT SEQ ID NO:243 CTCAGGTAGTGGTTGTCGGG SEQ ID NO:291 AAG AA ACACAAAC G CG AGT A SEQ ID NO:244 CTTTGCTCAGGGCGGACTGG SEQ ID NO:292

AGAGGAAACTATCAAAAACA SEQ ID NO:245 TGATCGCGCTTCTCGTTGGG SEQ ID NO:293

CCTGTCACTCATAAACGAGA SEQ ID NO:246 CACGAACTCCAGCAGGACCA SEQ ID NO:294

ATAGCCTGGCCACCTAGTTC SEQ ID NO:247 CGAGAGTGATCCCGGCGGCG SEQ ID NO:295

AGGAGCCAAGCATACAAGAG SEQ ID N0.248 CACTTGTACAGCTCGTCCAT SEQ ID NO-.296

CTGGTGGCCAACCAAGGAAC SEQ !D NO:249

GTAACCAAAAAACAGTATGC SEQ ID NO:250

CAAAATGAAGACAGGGTCAA SEQ ID NO:251

AAGCAAAGATAGACTTGACC SEQ ID NO:252

ACAAGCTTTATGTTCTGTGT SEQ ID NO:253

CTCAAAGATTTGTGGAAGTC SEQ ID NO;254

TGTTGTTG AC AAAT ATG C CC SEQ ID NO:255

TCGATGAATGTCAGAAATCT SEQ ID NO:256

AATGCAGAGCTGATGTCAAT SEQ ID NO:257

GTATAACTCCAAAGAAGTCA SEQ ID NO:258

AACAAAAAGCCCAAAGACAG SEQ ID NO:259

GCTGGTCACTAGCAAGTTTA SEQ ID NO:260

G ACT ATG C AGC AAAAC AG TT SEQ ID NO:261

AT ACTC AACCAATTG G CAG C SEQ ID NO:262

GCTTAGGTTGGCAAAACAAG SEQ ID NO:263

GTAAAGTCTCAATTTTCTGC SEQ ID NO:264

Immuno-RNA RNA FISH

For each experiment, early passage HUVECs on coversiips were grown to ~80% confluence, treated with TNFa, fixed in 3.7% formaldehyde for 10 min at room temperature, then washed three times in PBS. Cells were permeabilized in ice-cold 90% methanol for ten minutes, then washed twice with PBS and incubated in blocking buffer (1 % BSA/PBS) for 30 minutes at room temperature on an orbital shaker. Cells were then incubated in antibody solution (diluted in 1 % BSA PBS) for 1 hr. Double strand breaks were detected with rabbit polyclonal anti-phospho- histone H2A/X (Ser139) (Sigma). Goat polyclonal anti-SAMD4A C-15 (SantaCruz), mouse monoclonal anti-TNFAIP2 F-6 (SantaCruz) and goat polyclonal anti- SLC6A5/GLYT2 N-20 (SantaCruz) were used to detect SAMD4A, TNFAIP2 and SLC6A5 proteins respectively. Coversiips were then washed 5 times with wash buffer (0.05% Tween-20/PBS), following incubation with secondary antibodies conjugated to either Atto-488 or Atto-565 for 1 hr. Coversiips were then washed 5 times with wash buffer (0.05% Tween-20/PBS), and post-fixed with 3.7% formaldehyde/PBS for 10 minutes at room temperature, followed by further permeabilization in 70% ethanol overnight. For RNA FISH detection, coversiips were washed twice in PBS and incubated in wash buffer (10% formamide, 2X SCC - 1X SCC is 0.15 M NaCI plus 0.015 sodium citrate) for 5min. Cells were then hybridized overnight in a humidified chamber at 37°C in 50μΙ of Hyb buffer (10% dextran sulfate, 1 pg/μΙ E. coli tRNA, 2mM Vanadyl ribonucleoside complex, 0.02% RNAse-free BSA, 10% formamide) combined with 50ng of single molecule FISH probes. Coversiips were then washed 3X (30 min each on the orbital shaker) in wash buffer (10% formamide, 2X SCC). Cells were then incubated in equilibration buffer (0.4% glucose, 2X SCC) for 5 min and counter stained with 1 pg/ml DAPI (4',6- diamidino-2-pheny!indole; Life Technologies). Coversiips were mounted in glox buffer (3.7 g/yil glucose oxidase, U catalase) and imaged. image acquisition and processing

Cells were imaged on a custom built Nikon Ti Eclipse widefieid TIRF microscope using a 100X N.A. 1.49 Nikon Apochrornat TIRF oil immersion objective. Imaging was done using mercury lamp illumination through the appropriate filter sets at low camera gain in each of the fluorescent channels using an Andor iXion897 EMCCD camera cooled to -80°C. The microscope was controlled using pmanager open source microscope management software (NIH and UCSF, USA). A 20 ms exposure time was used for DAPI, Exposure times ranged from 200 to 500 ms for other dyes. Each field of view was captured as a series of images acquired on multiple focal planes through the samples, across a range of 2-10pm in the axial plane. A 0.2μηι piezo step-size was used for these z-stacks. Chromatic aberration was verified before image capture by alignment of Focal Check Fluorescent Microspheres (Molecular Probes). Signal intensities were measured using Fiji. The contrast of pictures shown was adjusted to fit a 16 bit grey scale. To facilitate the comparison between different fields of view on the same coverslip, IDV values were normalized relative to the intensity of fluorescent beads.

DNA FISH probes

DNA FISH probes were constructed using either one of two methods, in the first method, they were constructed from BAC clones RP11-299D5 (SAMD4A), RP1 1-1102D6 (TNFAIP2) and RP11-207A15 (SLC6A5) (Empire Genomics). Using the FISH Tag DNA Multicolor Kit (Molecular Probes), these clones were nick- translated to incorporate aminoallyl-dUTP, followed by dye-labeling of the amine- modified DNA. Alexa Fluor 488, 555 or 647 dye was incorporated. Alternatively, a PCR-based protocol was followed to construct "high definition" FISH probes (Bienko et al., 2012). Each locus was ampiified from genomic DNA using twenty pairs of PCR primers obtained from the van Oudenaarden FiSH probe database (Table 2; www.hdfish.eu). Ampiified DNA was pooled, purified and labeled with a fluorescent dye (FISHBright 495, 550 or 647) using the FISHBright Nucleic Acid Labeling Kit (Kreatech).

Table 2: Sequences of PCR primers used to construct "high definition" FISH probes

SAMD4A HD DNA FISH probe primers

Primer Name Primer Sequence (5' - 3') Sequence

Identity Number

SAMD4A5'1 F CTCC AAAAACCTG ACTTCG C SEQ ID NO:297

SA D4A5'1 R TGTCCTAATTGGCCTGATCC SEQ ID NO:298

SAMD4A5'2F CCACCCAACATGCTCTCTTT SEQ ID NO:299

SAMD4A5'2R CCATCATTTACCAGCCCTGT SEQ ID NO:300

SAMD4A5'3F GCTCATTGCCCTGCAGATAG SEQ ID NO:301

SA D4A5'3R TGATCCTCCAGTAACCGCAT SEQ ID NO:302

SA D4A5'4F AGACCTGCATTTCCTTCCCT SEQ ID NO:303

SA D4A54R ACTGTGGCTCTGTTGCCTTT SEQ ID NO:304

SAMD4A5'5F CCCTCAGACCCACTTCACAT SEQ ID NO:305

SAMD4A5'5R CCCTTCTGCACGTCTCTTCT SEQ ID NO:306

SA D4A5'6F CAGTGTCTATTCAGAAACCACAGAA SEQ ID NO:307

SAMD4A5'6R CGTGAGTTCAGGCCTACTGTC SEQ ID NO:308

SAMD4A57F C ATG CAG AG AAG ATC ACC ATT SEQ ID NO:309

SAMD4A57R GGGCCAAGTCCTCAAGATAA SEQ ID NO:310

SA D4A5'8F TGGCAATTATCCATTGTCATTT SEQ ID NO:311

SA D4A5'8R TGTTTAAGAGGATGCCTGGG SEQ ID NO:312

SAMD4A5'9F TTAGTAATCTTTGGAACATCTGAACA SEQ ID NO:313

SAMD4A5'9R GGCACTGAATACACGATGGG SEQ ID NO:314

SAMD4A5'10F TTATTGCCTAACCCAGTGCC SEQ ID O:315

SA D4A5'10R TCCT ATGTC AACCTG G AC CC SEQ ID NO:316

SAMD4A5'1 1 F TCAAGCCTTTCTGCCTCTACA SEQ ID NO:317

SA D4A5'1 1 R G AC ACGTG ATAG C ACCTAAC CA SEQ ID NO:318

SAMD4A5'12F ACATATGCCGGAGTTGGAAA SEQ ID NO:319

SAMD4A5'12R TCTGGGCTATCTCCGTGATT SEQ ID NO:320

SAMD4A5'13F GCACTCTCCAGATCCAGGTG SEQ ID NO:321

SAMD4A5'13R CTGGCTGAGTCCCACTCTTC SEQ ID NO:322

SAMD4A5'14F AACCACATTCCCAAGGACAA SEQ ID NO:323

SA D4A5"14R GGTCGGGTAGAAGGAACCTC SEQ ID NO:324

SAMD4A5'15F GTCCCCACTTAGGAATCCGT SEQ ID NO:325

SAMD4A5'15R GCTGCTGCCTCCAGTGTC SEQ ID NO:326

SA D4A5'16F ACTCGAAGCACCGCACTC SEQ ID NO:327

SAMD4A5'16R GAGGGGACCCCCAGTGAC SEQ ID NO:328

SA D4A5'17F AGCAGGCTTCTCGCTGTTAC SEQ ID NO:329

SAMD4A5'17R GGAGAGAAAGAGCTGCAGTGA SEQ ID NO:330

SA D4A5'18F ACCACCGGAACGTAACTGAA SEQ ID NO:331

SAMD4A5'18R CGAAACATCATGGTTAGGGG SEQ !D N0.332

SAMD4A5'19F GGATTTTGCCCACTGTGAAC SEQ ID NO:333

SAMD4A5'19R CGGGCATCTATCATGTGTTG SEQ ID NO:334

SA D4A5'20F TTCCTATTTGCTAGCGTGGG SEQ ID NO:335

SAMD4A5'20R ATGGAACAGCTGGAAGAGGA SEQ ID NO:336 TNFAIP2 HD DNA FiSH probe primers

Sequence

Primer Name Primer Sequence {5' - 3')

identity Number

TNFAIP23'1 F GGGCAAAGCAATCAATAACC SEQ ID NO:337

TNFAIP23'1 R TTCTCCCCTACCTGCCCTAT SEQ !D NO:338

TNFA!P23'2F TACCCCTGCTGAGTACAGCC SEQ ID NO:339

TNFAIP23'2R ACCTCCAGGCTTCCTTCCTA SEQ ID NO:340

TNFAIP23'3F GAGGAGTCAGCAGGGCATAG SEQ ID NO:341

TNFAIP23'3R AGCTTGGCTTTAGTTGCCTG SEQ ID NO:342

TNFAIP23'4F AGGGGCAGAGAGAAGGTCAT SEQ ID NO:343

TNFAIP23'4R AG GTCTTG G GC ATCTC ACC SEQ ID NO:344

TNFAIP23'5F TTTGGAGCAGGACCTAATGG SEQ ID NO:345

TNFA!P23'5R ACCACTCAAGCTAGAGCCCA SEQ ID NO:346

TNFAiP23'6F CTGCCTGCAGTGACATCATC SEQ ID NO:347

TNFAtP23'6R CTGCCCATGTCCTGTCTGT SEQ ID NO:348

TNFAIP237F GCCAATGTGAGGGAGTTGAT SEQ ID NO:349

TNFAIP237R CATGGGTATGCACACAGGAC SEQ ID NO:350

TNFAIP23'8F CTGGACTTGGGCTCACAGAT SEQ ID NO:351

TNFAIP23'8R GCTGAGTGGGAAACAACTCC SEQ ID NO:352

TNFAIP23'9F TCTGACCTCCACCAGGATTC SEQ ID NO:353

TNFAIP23'9R TAGGAGTGAGCGTGTGGTTG SEQ ID NO:354

TNFAIP23'10F AGATGGGCTGGTACCCTCTT SEQ ID NO:355

TNFAIP23'10R GTAGCTGGCGAAACCAGAAG SEQ ID NO:356

TNFAIP23'1 1 F TTCTGGAGAGAGGCAAGCAG SEQ ID NO:357

TNFAIP23'1 1 R TCCTCACACATGTTGCTGGT SEQ ID NO:358

TNFAIP23'12F CAAGCACACAGGCAGATGTT SEQ ID NO:359

TNFAIP23'12R CAATGGCTCCCCATTCTCTA SEQ ID NO:360

TNFAIP23'13F GGATGTCCATGGAGCAGAAT SEQ ID NO:361

TNFAtP23'13R CTTT GCTTCTGCC ACTCCT C SEQ ID NO:362

TNFA1P23'14F ATTCAGACACAGCCCAGTCC SEQ ID NO:363

TNFAIP23'14R CCCCACTTTACACCCTGCTA SEQ ID NO:364

TNFAIP23'15F GGGTGACCTTGGATAAGGGT SEQ ID NO:365

TNFAIP23'15R GAGACAGGAGAGGCAGGATG SEQ ID NO:366

TNFAtP23'16F ATCATCCAACTCAGCAAGGG SEQ ID NO:367

TNFAIP23'16R GGGCTTAGAGAGGCACCAG SEQ ID NO:368

TNFAIP23'1 7F ATTGAGGTGGCCACTTATGC SEQ ID NO:369

TNFAIP23'17R TCGTAGTGCTGTGGTGAAGG SEQ ID NO:370

TNFA1P23'18F TGGCCCACACTCTTAGCTTT SEQ ID NO:371

TNFAIP23"I 8R AACATGTCAGAGGACCCAGC SEQ ID NO:372

TNFAiP23'19F AAGGGAAGCCAGGTCTCAGT SEQ ID NO:373

TNFA!P23'19R CTTCCCACCCCTTAGGTCTC SEQ ID NO:374

TNFAIP23'20F AGTAGGGGTGTGGGTGACAG SEQ ID NO:375

I TNFAIP23'20R GCCACAGGAAAAGCTAACCA SEQ ID NO:376

SLC6A5 HD DNA FiSH probe primers

Sequence

Primer Name Primer Sequence (5' - 3')

identity Number

SLC3'1 F GTTGGCTGTGGCTCCTAAAG SEQ ID NO:377

SLC3'1 R CGCACTCCTCTCGTCTAACC SEQ ID NO:378

SLC3'2F CCTTGTCTTTGCTACCTGGC SEQ ID NO:379

SLC3'2R TCCCCATCACCACTGTAGAA SEQ ID NO:380

SLC3'3F CCATTCCCCCTTTACCCTTA SEQ ID NO:381

SLC3'3R GCAACTCAG GTCC AG AC AC A SEQ ID NO:382 SLC3'4F ATAAACTGCCAGCCAACAGC SEQ ID NO:383

SLC3'4R GAGTTTGCAAGACCCCACTC SEQ ID NO:384

SLC3'5F CG CTGCACTGTAAG ATCCCT SEQ ID NO:385

SLC3'5R AATGCAACCTGCTTACCTGG SEQ ID NO:386

SLC3'6F ACGTATGCTGATGGGGAAAC SEQ ID NO:387

SLC3'6R CAACTCTTTCGTGGATCTGGA SEQ ID NO:388

SLC37F AGGTGGGTTTACCTGAGGCT SEQ ID NO:389

SLC3'7R GTGGGTTTTTCACGCCTCTA SEQ ID NO:390

SLC3'8F ATCTCCCACTTGCCTTTCCT SEQ ID NO:391

SLC3'8R CCATTCCTAATCCCACCCTT SEQ ID NO:392

SLC3'9F TGTCCTGCACACCCTGTAGA SEQ ID NO:393

SLC3'9R GGTCCTGGAATGCCTCACTA SEQ ID NO:394

SLC3'10F GCATCTTCGTGTGAGCTTGA SEQ ID NO:395

SLC3'10R TGGG AAG AG ACAAG CCATCT SEQ ID NO:396

SLC3'1 1 F GAAAGTAAGGCAGGGGCTCT SEQ ID NO:397

SLC3'11 R CCCAACCCCTTACCAAGAAT SEQ ID NO:398

SLC3'12F TACAGGTTCAAGGTTTGGGG SEQ ID NO:399

SLC3 12R ATGCATCAGCAAGCAGGAC SEQ ID NO:400

SLC3'13F GCCTGCTTGTGGACCTACTC SEQ ID NO:401

SLC3'13R AGAGCGGAAAAGCCATACCT SEQ ID NO:402

SLC3'14F ACAGCCTCCTTAGGCTCCAG SEQ ID NO:403

SLC3'14R AACACTTGACCGCTAGCACC SEQ ID NO404

SLC3'15F GTTGAGGCTGTGCTGTCAAA SEQ ID NO:405

SLC3'15R AGCCGATGCAAGCCTAACTA SEQ ID NO406

SLC3'16F CACTGCTTCACCCCATAGGT SEQ ID NO:407

SLC3'16R ATGGCAAATGCTCTGAGGTC SEQ ID NO:408

SLC3'17F ACACGAATAGAGGCACCCAG SEQ ID NO:409

SLC3'17R AACCCTGAGATCTGCCCC SEQ ID NO:410

SLC3'18F TTTTCAGGAGCAGAGGAGGA SEQ ID N0:411

SLC3'18R AGCGGAGGGAAAACTGATCT SEQ ID NO:412

SLC3'19F CCTCAAAGAATGCTGAAGGG SEQ ID NO:413

SLC3'19R CCTTCTGAGTCCTCCCACAG SEQ ID NO:414

SLC3'20F TGACCTTGAAGGATGGAAGG SEQ ID NO:415

SLC3'20R CCCAGGACCCAAAATCTTCT SEQ ID N0416

Immuno-DNA FISH

Immunofluorescence was performed as above and ceils were post-fixed with 3.7% formaldehyde/PBS for 20 min at room temperature. DNA FISH hybridization was performed as follows. Cells were washed twice in PBS for 5 min each and permeabilized in 0.5% Triton X-100 for 10 minutes. Ceils were treated with 10 U of RNAseA in 2X SCC for ihr at 37°C. Cells were then washed twice with 2 X SSC, and dehydrated in 70%, 85% and 100% ethanol for 2 min each. After air-drying, ceils were denatured in 70% formamide, 2X SCC for 3 min at 73°C, and dehydrated in ice- cold 70%, 85% and 100% ethanol for 2 min each. Cells were then hybridized overnight in a humidified chamber at 37°C in 10μ! of Hyb buffer (10% dextran sulfate, 50% formamide, 4X SSC) combined with 30ng of DNA FISH probes that were freshly denatured for 5 min at 72°C and cooled on ice. Coverslips were then washed 3X (5 min each on the orbital shaker) in each of three wash buffers - 1 ) 50% formamide, 2X SCC 2) 1X SSC and 3) 4X SSC, 0.01 % Tween 20. Cells were counter stained with DAPl, mounted in glox buffer (3.7 Mg/ Ι glucose oxidase, 1 U catalase) and imaged.

Repair construct

The repair construct pCR-SAMD4A-IRES-GFP includes two segments, SAMD4A 5' Intron 1 and SA D4A 3' Intron 1 of 564 bp and 563 bp respectively which are homologous to regions either side of the SAMD4A intron 1 TALEN cleavage site. Segment SA D4A 5' Intron 1 includes 3' splice acceptor site and 52 bp of 3' introntc sequence derived from a modified chimeric intron used in the mammalian expression vector pCI-neo (Promega, Wl, USA). Similarly, segment SAMD4A 3 ^* Intron 1 includes a 5 ^! splice donor site and 52 bp of 5' intronic sequence derived from a modified chimeric intron used in the mammalian expression vector pCI-neo (Promega, Wl, USA). The resulting artificial exon, which includes an IRES- GFP cassette, allowed for independent in situ GFP expression off the SAMD4A promoter following repair (homologous recombination) and splicing. The primers used to amplify the 564 bp 5' SA D4A fragment were as follows: SAMD4A 5' Intron F: 5'-TGC TGC TGC AGG AGG GTG-3' (SEQ ID NO: 17); SA D4A 5' Intron R: 5'- GAT CGC TAG CAC CTG TGG AGA GAA AGG CAA AGT GGA TGT CAG TAA GAC CAA TAG GTG CCT ATC ATG GCC CTC CAG CTA TTT ATA AAC GTG-3' (SEQ ID NO:18) (Bold = splice acceptor site; Underlined = intron extension from a modified chimeric intron derived from pCI-neo [Promega]). The primers used to amplify the 563 bp 3' SAMD4A fragment were as follows: SAMD4A 3' Intron F: 5'- GAT CGC TAG CAG GTA AGT ATC AAG GTT ACA AGA CAG GTT TAA GGA GAC CAA TAG AAA CTG GGG CCT CCT TCA CTG GGG TGT G-3' (SEQ ID NO: 19) (Bold = splice donor site; Underlined = intron extension from a modified chimeric intron derived from pCI-neo [Promega]); SAMD4A 3' Intron R: 5'-CTT TTG TAT ATC TAC ATC ATT TAG CAG CAT G-3' (SEQ ID NO:20). Both amplicons were digested with Nhei (NEB) and ligated. Full length ligated products were selected by PCR with primers SAMD4A 5' intron F and SA D4A 3' intron R and cloned into the TA vector pCR.2.1 (Invitrogen) to produce pCR-SAMD4A- 5/3. The (HCV) IRES-GFP sequence was amplified from pHiV7-IRES-GFP (a gift from John Rossi) and cloned into the Nhe\ site of pCR-SAMD A-5/3 to produce pCR-SAMD4A-IRES-GFP. The primers used for amplification include: IRES F: 5'-GAT CGC TAG CCC CCC TAA CGT TAC TGG CCG-3' (SEQ ID NO:21 ) and GFP R: 5'-GAT CGC TAG CGG ATC CTC ACT TGT ACA GCT CGT CCA TGC C-3 ^* (SEQ ID NO:22). To generate double stranded PCR product harboring the 20 and 18bp homologous arms, the following primers were used: S'-CCA CAC CCC AGT GAA GGA G-3' (SEQ ID NO:23) and 5 ^: -TAA ATA GCT GGA GGG CCA TG-3' (SEQ ID NO:24). The PCR product was purified by QIAquick PCR Purification Kit (Qiagen) prior to transfection.

The transcriptional response of co-regulated genes in a muitigene complex is asymmetric.

TNFa has been shown to rapidly and synchronously shape the transcriptional response in early passage human umbilical vein endothelial cells (HUVECs) by systematically inducing the formation of a large variety of different muitigene complexes (Papantonis et al., 2012). Interacting genes in TNFa- induced muitigene complexes are activated by NF-κΒ (Papantonis et al., 2010). The SAMD4A, TNFAIP2 and SLC6A5 genes associate in a NF-xBdependent muitigene complex (Figure 4), which has been extensively characterized by 3C, 4C (3C capture-on-chip), tiling microarray and FISH, and thus is an ideal mode! to interrogate the role chromosomal contacts play in cotranscription at a single eel! level (Papantonis et al., 2010; Papantonis et al., 2012). HUVECs were arrested in the GO phase of the cell cycle by serum deprivation. Importantly, the transcription of SAMD4A, TNFAIP2 and SLC6A5 is rapidly induced 10mins post TNFa stimulation (Figure 5; Papantonis et al., 2010). Concurrent to their expression at 10mins, prior 3C data including our own (data not shown) (Papantonis et al., 2010), indicate that chromosomal contacts between these genes occur at sites ~1.5kbp downstream of the TSS (Figure 5). To interrogate whether the formation of these contacts are associated with cotranscription, we designed RNA FISH probes to target the intronic sites where these contacts occur 10mins post TNFa treatment (Figure 5). As these introns are typically spliced and degraded co-transcriptionally, these probes label the transcriptional start site (TSS). Each set of gene-specific RNA FISH intronic probes was conjugated to spectrally distinct fluorophores. By simultaneously performing RNA FISH on the 3 NF- KBreguiated genes, we were able to investigate the frequency of these interactions across a population of cells. Consistent with previous studies (Papantonis et a(., 2010, Papantonis et al, 2012), analysis of overlapping RNA FISH foci revealed co- localized foci in only a fraction of all alleles in the total population (-5%) (Figure 5). Chromosomal interactions between transcribed genes have been shown to occur at discrete foci of active RNA Pol II, as well as at nuclear speckles. Overlapping SA D4A, TNFAIP2 and/or SLC6A5 RNA FISH foci consistently co-localize with the active, poised form of RNA Pol i! (phosphorylated at Ser5) (Figure 6). Collectively, these data suggest that the co-transcription of these genes at RNA pol II foci may only occur in a small fraction of the HUVEC population. Chromosomes in primary cells occupy distinct territories in the nucleus. As HUVECs are primary cells, they always display two spatially distinct DNA FISH foci, both before and after TNFa treatment (Figure 7). Within territories, chromosomes are segregated further into topotogicaliy associating domains (TADs). As a consequence of this compartmentaiization, interacting genes in multigene complexes may be confined within much smaller genomic neighborhoods. By simultaneously performing DNA FISH on these 3 co-regulated genes, we were able to investigate whether the DNA of these 3 genes were in close vicinity prior to activation by TNFa. Interestingly, analysis of overlapping DNA FISH foci revealed proximal foci in both the unstimulated and the TNFa- treated HUVEC population (Figure 8). This data suggest that in a fraction of the HUVEC population, the DNA of the SAMD4A, TNFAIP2 and SLC6A5 genes may be constrained within a TAD, prior to TNFa induction.

Through an enrichment of chromosomal interactions in TADs, and higher levels of transcription, "jackpot" cells may contribute to variable, or stochastic effects, in gene expression (Noordermeer et al., 201 1 ). As these NF-KBregulated genes respond stochastically to TNFa (Papantonis et al., 2010), we assessed the monoallelic and biallelic expression of the three genes. Nine of the 33 possible phenotypes were observed in the majority of the population (-84%) (Figure 9). With the exception of ceils displaying no foci (-29%), all cells in this category displayed either a single or dual SAMD4A foci. With respect to the individual expression of each of the 3 genes, not all cells respond similarly to TNFa, with approximately half of the alleles expressing SAMD4A, a lower proportion expressing TNFAIP2 and approximately one fifth expressing SLC6A5 (Figure 9). As chromosomal contact occurs between these 3 genes (Figure 5; Papantonis et al., 2010), the asymmetric transcriptional response of these genes to TNFa suggests a hierarchical regulation in the assembly of this multigene complex. Supporting this hypothesis, TNFAIP2 transcription occurs predominantly when SAMD4A is also transcribed (-86%) and SLC6A5 transcription occurs mainly when SAMD4A (-92%), or both SAMD4A and TNFAIP2 are transcribed (-62%) (Figure 9).

As SAMD4A, TNFAIP2 and SLC6A5 are not always co-expressed (Figure 9), we also represented the co-localization frequencies relative to cells coexpresstng SAMD4A, TNFAIP2 and/or SLC6A5 (Figure 10). Allelic coexpression analyses between the 3 genes revealed that co-expression of SAMD4A and TNFA1P2, as well as triple expression of the 3 genes was most prevalent across the population (-12% and -14% respectively) (Figure 10). SLC6A5 was rarely expressed (-4%) in the absence of TNFA I P2 expression at the corresponding allele (Figure 10). Moreover, co-transcription of TNFAIP2 and SLC6A5 in the absence of SAMD4A transcription was extremely rare (-1 %) (Figure 10). This provides further evidence for a relationship between the transcriptional status of SAMD4A and the transcriptional activation of TNFAIP2 and SLC6A5. With respect to coexpressed alleles, SAMD4A and TNFAIP2 were co-localized at -26% of coexpressed alleles (3% in the total population (T/P)), whilst SAMD4A, TNFA1P2 and SLC6A5 were co-localized at -40% of all co-expressed alleles (5% in the T/P) (Figure 10). Collectively, these data suggest a hierarchical mode of regulation between these genes, in which chromosomal contact favors cotranscriptional activation.

Visualization of TALEN-mediated gene loop disruption

Hierarchical regulation dependent on chromosomal contact was revealed in the GREB1 muitigene complex (Li et al., 2012). qPCR analysis revealed that siRNAs targeting ERa disrupted not only GREB1 transcription, but were sufficient to cause more than a 2 to 4-fold reduction in the transcription of interacting members. siRNA approaches cannot be applied to the TNFa- induced muitigene complex, as these 3 genes are transcriptionally activated by NF-κΒ (Papantonis et al., 2010). Toward this end, we developed a single cell assay allowing the discrete disruption of individual gene loops at the sites of chromosomal contacts (Figure 5). in parallel, to decrypt the role of a single individual gene loop on co-transcription in the same muitigene complex, we visualized transcriptional activity of other genes in the muitigene complex using highly sensitive RNA FISH (Figure 5). We constructed our microscopy-based assay upon TALENs, the orthogonal and robust, wellestablished genome editing tool derived from TAL effectors of Xanthomonas sp. (Christian et al., 2010, Li et al., 201 1 ). Within the TALE structure, a central repeat domain mediates highly specific DNA recognition (Boch et a!., 2009; Christian et at., 2010). Fusion of this domain to Fokl endonuclease enables TALENs to induce site-specific double strand breaks (DSB) (Christian et al., 2010, Li et al., 201 1 ). As the TALENs used in this study induce a DSB at the approximate site of chromosomal contact that occurs 10mins post TNFa treatment (Figure 11), we reasoned that should loop-mediated contact be required for co-transcription, the DSB would serve to rupture chromosomal contact between these genes. To ascertain whether the TALEN had successfully disrupted a gene loop, we stained for induction of a DSB using histone variant H2A.X (Figure 12 and 13). Rapidly phosphorylated at Ser139 within minutes following DNA rupture, this modification persists throughout the entire DNA repair process. To initially test our TALEN system and its ability to disrupt a single gene loop and concurrent assembly of the multigene complex, we introduced into HUVECs, by high efficiency microporation, our TALEN vectors that were able to disrupt the gene loop of the longest gene in the multigene complex, SA D4A. We used H2A.X staining of DSBs to carefully establish a time course of TALEN activity (also a measure of transfection efficiency), noting that nuclease activity was first evident after ~6hrs, and was sustained until ~48hrs (Figure 14). We chose to assay for nuclease activity after 24hrs, as there were high levels of DSBs and low levels of cytotoxicity at this time point. Discrete sites of H2A.X phosphorylation were evident in -60% of cells 24hrs post transfection, with a higher proportion of these cells displaying single allelic DSBs, with fewer cells displaying dual allelic DSBs (Figure 3). The high specificity of the SAMD4A TALEN was demonstrated by the consistent co-localization of SAMD4A DNA with the DSB, visualized by DNA FISH (Figure 15). TALEN cleavage efficiency was further supported by the results of "surveyor assays" (Figures 16, 21, 32 and 42). Thus, we were able to assay uniquely for the disruption of the SAMD4A gene loop in its native environment at a single cell level.

TALEN-mediated disruption of a gene loop abrogates RNA and protein expression

Upon stimulation by TNFa, RNA pol II engages the SAMD4A promoter, triggering a wave of transcription that propagates down the gene. RNA FISH tiling array analysis revealed that the transcriptional cycle takes approximately 85min (Papantonis et al., 2010). Accordingly, RNA transcribed ~1.5kbp downstream of the TSS appears within 10mins (probe set i, Figure 17), and ~34kbp into intron 1 after 30mins post TNFa stimulation (probe set ii, Figure 18). Importantly, the DSB induced by the SA D4A TALEN occurs at a site in between these two regions where different sets of intronic RNA FISH probes bind. HUVECs were dual transfected for 24hrs with the SAMD4A TALENs and exposed to TNFa for 10mins. This recapitulated SA D4A transcriptional activity, allowing for the first ~1.5kb of SAMD4A to be transcribed. By using RNA FISH to monitor intronic RNA transcription 10mins post TNFa stimulation, SA D4A transcription was frequently detected (-42%), either overlapping or in close proximity to the DSB (Figure 17). Thus, these data indicate that despite disrupting the SAMD4A gene loop, the DSB did not appear to influence the ability of RNA Poi II to access the SAMD4A promoter, permitting transcriptional initiation and elongation up to the DSB. The half-life of SA D4A intronic RNA transcripts transcribed at 10mins is between 3 and 6mins. Consequently, RNA transcripts transcribed downstream of the TSS at 10mins, have degraded by the SOmins time point. In a separate experiment, dual-transfected HUVECs were stimulated with TNFa for 30mins to allow for the first ~34kbp of SAMD4A to be transcribed (Figure 18). Notably, the TALEN-induced perturbation was able to abrogate transcription of SAMD4A in this region, as no transcription of the intronic region of SAMD4A was ever evident beyond the DSB (Figure 18)

It has been published that DNA lesions alter RNA Polymerase il occupancy distal to DSBs, resulting in transcriptional inhibition (Shanbhag et a!., 2010). This effect is mediated by ATM kinase activity, which results in the abrogation of Pol II- dependent elongation beyond the DSB. Importantly, ATM-mediated local inhibition of Pot II occurs in ci$ to DSBs (Shanbhag et a!., 2010). Hence, in this study, the local inhibition of Pol II that occurs distal to the DSB, should not affect the intact allele in cells displaying single allelic DSBs.) IF in untransfected cells revealed that TNFa induction resulted not only in the robust transcription of members of the multigene complex, but also in an increase in protein expression (Figure 19). Consistent with the ability of the TALEN to abrogate transcription, there was a significant reduction in SA D4A protein expression in cells harboring dual allelic DSBs as detected by IF, whilst cells harboring single allelic DSBs were sti!l able to express SAMD4A protein presumably from the intact allele (Figure 19). Overall this robust assay offered an unprecedented perspective into the dynamics of TALEN activity, transcriptional status of the targeted gene and related protein levels at single cell resolution. Importantly, the observed disruption in SAMD4A RNA (Figure 18) and protein expression (Figure 19) strengthened our assay, as it validates that the DSB occurs at the intended site of TALEN activity where H2A.X staining occurred.

TALEN-mediated disruption of a gene loop abrogates expression of interacting members

Satisfied that our TALEN assay was able to discretely disrupt SAMD4A at the site that engages in chromosomal contact, we repeated the prior experiment with an important modification: we used RNA FiSH to visualize transcription of two other interacting genes that engage in intra- and inter-chromosomal with the region of SAMD4A that had been disrupted. By simultaneously monitoring SAMD4A gene loop disruption, as well as transcription of additional members of this multigene complex, we were able to interrogate discretely the effect of the disrupted SA D4A gene loop on the transcriptional status of two other members of the complex.

Firstly, to exclude the possibility that the DSB was capable of inducing ceil cycle arrest, thereby altering global transcription, we designed a TALEN to rupture B P4, a non-TNFa responsive gene, located ~600kb 5' of SAMD4A on chromosome 14 (Figure 20). We observed no change in the transcription of any of the three members of the multigene complex relative to the normal TNFa-induced transcriptional response (Figure 9). Therefore, inducing a DSB 5 ¹ of SAMD4A has no effect on transcription of members of the multigene complex.

We then monitored the effects of perturbation of the chromatin loop of SAMD4A at its first intron. We observed a significant reduction of transcription of TNFAIP2 and SLC6A5 at the SAMD4A TALEN-induced DSB (Figure 22). In cells where a single allele of SAMD4A was targeted, virtually all transcription of TNFAIP2 and SLC6A5, as assessed by RNA FISH, was lost at the corresponding allele or DSB, with multigene transcription confined to the intact allele (Figure 23). Importantly, as the SAMD4A protein is still observed in cells harboring a single allelic DSB (Figure 8), this observation excludes the possibility that the SA D4A protein is required for the transcription of TNFAIP2 and SLC6A5.

Initially, a possible explanation for the transcriptional response observed in Figures 20, 22, 26 and 35 could be that the SAMD4A protein is required for the expression of TNFAIP2 and SLC6A5. However, microarray analysis demonstrates that it takes ~85mtns for RNA Pol II to transcribe the SAMD4A gene. The observed reduction in TNFAIP2 and SLC6A5 expression by the SA D4A TALEN-induced DSB is observed 10mins post TNFa stimulation (Figures 22 and 26). No transcription, or chromosomal contact, is observed between these co-reguiated genes in unstimulated HUVECs (Papantonis et al., 2010). Therefore, these effects cannot be attributed to the entire SAMD4A transcript, or protein, as neither is present at the I Qmins time point. HUVECs are primary cells with defined chromosomal territories, and therefore always display two spatially distinct DNA FISH foci (Figure 7). The comparison between cells harboring single allelic DSBs and dual allelic DSBs (Figure 19) offers a unique perspective on the influence of protein expression in the assembly of the multigene complex, in cells displaying a single allelic break (disrupted SA D4A transcription at one allele) the following is observed; reduced transcription of two other genes that interact with SAMD4A near to the TALEN- induced DSB (Figure 23), and translation of SAMD4A protein, presumably from the intact allele (Figure 19). The reduction of TNFAIP2 and SLC6A5 transcription was also observed in cells harboring dual allelic DBSs for SA D4A (Figure 23). Cells displaying dual allelic SA D4A TALENinduced DSBs also have reduced protein expression (Figure 19). Collectively, these data clearly show that the protein product of the SA D4A gene is not required for the expression of TNFAIP2 and SLC6A5. Currently, no other biochemical technique offers the same insight.) Supporting the observed reduction in transcription at single allelic DSBs, in cells where both SAMD4A alleles were targeted, virtually all transcription from TNFA1P2 and SLC6A5 was lost at both aileies (Figures 23, 33 and 42). Furthermore, protein levels of TNFAIP2 and SLC6A5 were also severely reduced, as assessed by IF, in cells where both SAMD4A alleles were successfully targeted (Figures 24, 34 and 43). Hence, in a manner analogous to GREB1 (Li et a!., 2012), SAMD4A appeared to be influencing the transcription of other co-regulated and interacting genes within the same multigene complex.

We sought to determine whether the loss of transcription of TNFA1P2 was due to the DSB abrogating the transcription of genes located between SAMD4A and TNFAIP2, on chromosome 14. RCOR1 is a gene located 400kb 5' of TNFAIP2 and displays transcriptional activity comparable to GAPDH (Papantonts et al., 2012). We repeated the TALEN-medsated disruption of SAMD4A gene loop whilst monitoring transcription of RCOR1 (Figure 25). Transcription remained unaffected at the RCOR1 locus (Figure 25). This observation concurred with the hypothesis that loop- mediated chromosomal contact of SA D4A specifically impacts the transcription of other members of the multigene complex but has no effect on genes outside of the complex. This extended to genes on the same chromosome that were interspersed between two genes in the multigene complex.

We sought to determine whether the presence of other gene loops in the multigene complex were equally required for co-transcription, since all genes in the multigene complex were bound by the NF- Β transcription factor (Papantonis et a!., 2010). We hypothesized that if gene loops were equally required for co-transcription, then the disruption of any other gene loop wouid have a similar effect to that observed for SA D4A. Alternatively, an asymmetrical relationship between gene loops infers that the disruption of one gene loop may have no bearing on the transcriptional status of other genes in the complex. To differentiate between these two hypotheses we designed TALENs targeted to TNFAIP2 (Figures 27, 28, 29, 30 and 31 ) and SLC6A5 (Figures 36, 37, 38, 39 and 40). The sites for these TALENs were identified by 3C (Papantonis et al., 2010) indicating where the chromosomal contacts occurred between these genes. Using an identical approach to our initial assay, we delivered a TALEN to TNFA1P2 whilst monitoring transcription of SA D4A and SLC6A5. The TNFAIP2 TALEN targets the site of chromosomal contact (~1.5kbp downstream of the TSS), whilst our RNA FISH probes interrogated a region downstream of this site (Figure 27). Remarkabiy, we observed that the disruption of the TNFAIP2 loop had no significant effect on SAMD4A transcription, yet transcription of both TNFAIP2 (Figure 29) and SLC6A5 was reduced (Figure 26). Transcriptional loss was observed at either a singie or both targeted alleles (Figure 33). When we repeated the experiment with a TALEN targeting the SLC6A5 gene loop on chromosome 11 (Figure 36), the hierarchical effect was more pronounced with transcription unaffected in both SA D4A and TNFAIP2 (Figure 35 and Figure 42). However at all successfully targeted single and dual alleles of SLC6A5, transcription was lost (Figure 38). This result lends credence to the hypothesis that the disrupted site of chromosomal contact precludes the ability of the gene loop to access critical transcriptional machinery due to a requirement for loop-mediated chromosomal contacts. This unique single cell perspective reveals a highly unexpected hierarchical organization within the TNFa-induced multigene complex, providing the first direct evidence that chromosomal contacts p!ay a central role in supporting transcription and determining the hierarchy.

Disrupted SA D4A gene loops can be successfully repaired

If the integrity of the gene loop topology and chromosomal contacts were essential for co-transcription, then we hypothesized that restoring the chromosomal contacts would restore transcription of interacting genes in the multigene complex. DSBs are generally repaired in most cells by two highly conserved mechanisms: rapid but error-prone non-homologous end joining (NHEJ), or the slower, but highly precise, homology-directed repair (HDR). We generated a repair construct designed to span the SAMD4A TALEN target site, to exploit HDR to restore SAMD4A gene loop integrity by inserting an exonic eGFP sequence into intron 1 of SAMD4A (Figure 44). We included an internal ribosome entry sequence (IRES), as well as splice sites flanking the IRES-eGFP, to facilitate the co-transcriptional splicing and independent translation of the repair construct (Figure 44). Once repaired, through the activation of SA D4A, TNFa stimulation would then be able to induce SA D4A/IRES-eGFP transcription. We used the identification of eGFP positive cells to establish a time course for TNFa treatment. Due to HDR being rare, and as the IRES-eGFP is under the control of the endogenous SAMD4A promoter, we sought to enhance the efficiency of the repair. We constructed a novel bidirectional TALEN system abie to combine both targeting arms of the TALEN into a single bi-directional promoter plasmid (pDT; Figure 45), to ensure that equimoiar quantities of both left and right TALENs were synthesized in the same cell throughout the DSB-induction process. HUVECs dual transfected with the pDT and the repair plasmids demonstrated poor repair efficiency in the transfected cells as evidenced by the lack of eGFP signal (data not shown). To modify the repair construct, we amplified the IRES-eGFP sequence with associated flanking splice sites and included short 5' 20bp and 3' 18bp homologous extensions on either side of the SAMD4A DSB to induce efficient HDR (Figure 45). Notably, eGFP protein signal was observed in -1% of HUVECs dual transfected with the pDT piasmid and PCR product for 72hr and stimulated with TNFa for 24hr (Figure 45). However, eGFP mRNA transcripts, which can only arise following successful HDR, were observed in -10% of cells in the population (Figures 46 and 47). Notably, eGFP positive or 'green' cells coincided with a large number of RNA FISH foci (Figures 46 and 47). As the SAMD4A promoter does not constitutively express the IRES-eGFP mRNA, it is reasonable that very few cells in the population would transcribe sufficient IRES-eGFP mRNA to enable the detection of a 'green ¹ cell. Therefore, detection of IRES-eGFP mRNA was the more sensitive approach to identifying successful HDR and was used to interrogate the effects of a repaired gene loop.

Restoration of the SAMD4A gene loop restores transcription of interacting members in a sequence-independent manner

Satisfied that the repair experiment was fully functional, we sought to investigate whether the re-estabiishment of contact was sufficient to restore transcription of interacting genes. We stimulated HUVECs for 20mins with TNFa, recapitulating transcription of the first ~1.5kbp of SA D4A and IRESeGFP. We detected transcription of eGFP using RNA FISH probes binding to its RNA (Figures 46 and 47), and related the position of eGFP transcription to members of the multigene complex. Distinct eGFP foci were evident in -10% of transfected cells and these foci overlapped with SAIV1D4A intronic mRNA (Figure 48). In the instances where we had successful HDR mediated repair, we observed no change in the transcription of any of TNFAIP2 and SLC6A5 relative to the normal TNFa-induced transcriptional response (Figure 49). Moreover, there was no significant difference in the co-localization frequencies between eGFP/SAMD4A, TNFAIP2 and SLC6A5 in the repaired and mock-transfected cells (Figure 49). This result indicates that reestablishment of an intact SAMD4A loop, in a sequence-independent manner, restores chromosomal contacts as we!l as transcription of TNFA1P2 and SLC6A5 in this multigene complex.

EXAMPLE 2

Disrupting the IL8 enhancer abroaqtes 1L8 expression

Recently, Hi-C studies revealed that the pro-inflammatory chemokines CC Chr.17 (CCL2, CCL7, CCL1 1 ) and CXC Chr.4 (IL8, CXCL1, CXCL3, CXCL2) are organized into TADs and engage in chromosomal contact.

Using intronic smFISH, we were able to show that the CXC chemokines are only induced following TNF induction. Furthermore, the smFISH foci of co-expressed CXC genes always co-localize.

Deeper bioinformatic analysis of both Hi-C and ChlA-PET data in the CC and CXC TADs identified a large cluster of transcriptional enhancers. At the 5' end of the CXC TAD, we identified a putative 'super-enhancer' region, spanning ~80kb and forming extensive chromosomal contacts with the proinflammatory genes (unpublished data). Typically, such regions are densely occupied by chromatin regulators over tens of kb. Accordingly, this region is highly enriched for eRNA chromatin marks, H3K4me1 and H3K27Ac. Using the recently published 'eRNA encyclopedia', we identified several eRNAs transcribed from this region in a tissue- and cell-specific manner. Using smFISH, we were able to detect some of these eRNAs across different primary eel! types and observed different expression patterns (unpublished data).

Through their interaction with the mediator complex, eRNAs have been proposed to be the architects of chromatin organization and therefore putatively TADs. We found that depleting Mediator 12 (Med 12) abrogates the expression of all the CXC cytokines. We designed TALENs to induce a DSB at various sites within the superenhancer region that was established by ChlA-PET to engage in contact. In one instance the DSB was induced at the site of an eRNA. The DSB, or site of disruption, was detected by immunofluorescent staining of H2A.X pSer139, a factor of the DSB repair process. In parallel to the detection of the DSB, transcriptional activity of interacting genes in the multigene complex was determined using intronic RNA FISH. We observed that perturbing the superenhancer in the CXC TAD silenced all of the pro-inflammatory CXC genes (IL8, CXCL2, CXCL1 and CXCL3).

EXAM PLE 3

Disrupting intra- or interchromosomal contact in multigene complexes abrogates the transcription of interacting genes (see Figure 52)

TNFa, a major proinflammatory cytokine, is a stimulus that induces the coordinated assembly of coreguiated genes in NF- B-dependent multigene complexes (Papantonis et aL, 2012). SA D4A, a -221 Kb gene on chromosome 14, is rapidly switched on by TNFa in primary human umbilical vein endothelial cells (HUVEC). 4C analysis reveals that prior to stimulation with TNFa, SAMD4A seldom interacts with other genes.

After activation by TNFa, SAMD4A interacts with multiple coreguiated genes to form a multigene complex. TNFAIP2, a gene located on the same chromosome but ~50 b downstream, and SLC6A5 on chromosome 1 , are two well-characterized interacting partners of SA D4A. siRNA approaches cannot be utilized to interrogate loop-mediated dynamics in the SAMD4A TNFAIP2/SLC6A5 multigene complex, as all three genes are activated by the same transcription factor. This exposes the necessity to develop functional assays able to discretely alter gene loop topology, without the global ablation of transcriptional regulators.

We used TALENs to discretely perturb sites within gene loops that are established to engage in chromosomal contact in the weil-characterized NF-KB- regulated multigene complex (Papantonis et al., 2010).

Using the TALEN single cell assay, we observed that perturbing the SAMD4A gene loop did not alter transcription 5' of the DSB. However, consistent with other studies, we observed silencing of SAMD4A transcription 3' of the DSB. Intriguingly, despite occupying distal genomic locations, the transcription of TNFAIP2 and SLC6A5 were also significantly reduced (Figure 52(i)). Further, the effect on co- transcription was hierarchical, with the disruption of TNFAIP2 altering SLC6A5 expression, but having no influence on SAMD4A (Figure 52(H)), whereas perturbing SLC6A5 did not influence SAMD4A or TNFAIP2 transcription (Figure 52(HI)). This suggests a hierarchical assembly between these three genes, whereby TNFAIP2 'collects' NA Po! II from SA D4A, which are then able to recruit SLC6A5 to the multigene complex.

EXAMPLE 4

Perturbing CTCF sites flanking TNFAIP2 abrogates transcription

CTCF, binds to consensus sites that frequently flank genes. TNFAIP2, a -11 Kb gene on chromosome 14, is flanked by CTCF consensus binding sites. We used CRISPRs to discretely perturb sites within CTCF consensus binding site (Figure 53). We observed that perturbing the 3' CTCF binding site with a DSB in the silenced the expression of TNFAIP2.

Discussion

Here we show that disrupting sites within gene loops that engage in chromosomal contact significantly impacts the transcription of interacting genes in a multigene complex. We revealed this level of gene regulation by implementing a novel single cell strategy that allows discrete perturbation within chromatin loops. Initial evidence for this regulation was our single cell based observation of the hierarchical transcriptional response of these 3 genes to TNFa induction (Figure 9). Despite the clear asymmetric transcriptional response, co-localization of RNA FISH foci was only observed in a subset of the population (Figure 5). The functional relationship between chromosomal contact and co-transcription of interacting members of a multigene complex remains opaque. However, if loop-mediated contact is indeed a prerequisite for co-transcription, the RNA FISH co-localization data suggests that only certain cells in the population may possess the correct spatial arrangements of their chromosomes to permit such regulation (Figure 5). Such variegated gene expression is suggested to occur in TADs, highly conserved compartments within mammalian chromosome territories (Dixon et al., 2012). We speculate that TADs may ensure robust SAMD4A, TNFAIP2 and SLC6A5 gene regulation by constraining these genes to areas permissible to long-range interactions (Figure 50). Indeed, analysis of DNA in unstimulated HUVECs revealed that despite occupying distal genomic locations, in a fraction of the HUVEC population, the SAMD4A, TNFAIP2 and SLC6A5 DNA are in close proximity prior to TNFa induction (Figure 8). The enrichment of chromosomal interactions in these 'jackpot' cells, may contribute to hierarchical SAMD4A, TNFAIP2 and SLC6A5 gene expression. However, current biochemical and imaging technologies lack the spatiotemporat resolution to interrogate whether loop-mediated co-transcription between these genes can only occur in these 'jackpot' ceils. Alternatively, given more time, whether these chromosomal interactions may occur in most cells across the population.

Chromosomal translocations are natural perturbations in chromosome structure, which alter the spatial positioning of DNA within chromosome territories. Interestingly, these discrete perturbations in chromosome structure do not only influence genes located near to the breakpoint, but are sufficient to modify gene expression in cis and trans. Therefore, through the repositioning of chromosomes and relocation of DNA into different TADs, translocations may alter transcription by disrupting intra- and interchromosomal interactions. We were able to reveal that perturbing the SAMD4A gene loop has a direct effect on the transcriptional status of TNFAIP2 and SLC6A5 (Figure 22). In addition, corresponding to the transcriptional response (Figure 9), the effect on co-transcription was hierarchical, with the perturbation of TNFAIP2 altering SLC6A5 expression, but having no influence on SAMD4A (Figure 26). Furthermore, perturbing SLC6A5 did not impact either SAMD4A or TNFAIP2 transcription (Figure 35). This observation raises the question: if all the factors necessary for transcription are present, why does transcription of these interacting genes not occur? A possible explanation is that the recruitment of the multitude of repair proteins to the DSB acts as an obstruction to the normal chromatin contacts between these genes loops. Alternatively, contacts between interacting genes may still be present but are 'bridged' by the repair machinery. Although this "bridging" maintains the contact between gene loops, it is still inadequate for transcriptional activity to occur. It is important to note that the DNA of these 3 loci appear to be in close proximity in unstimulated HUVECs, despite the absence of 3C contact (Papantonis et al., 2010). Thus, it would be very difficult to definitively show that the disruption of a gene loop abrogates contact between other interacting genes, at least as measured by diffraction limited co-localized DNA FISH foci. Another possible explanation could be that the DNA repair machinery is occluding the entry of RNA Pol II to the other interacting genes. However, the observed unidirectional loss of transcription between these three genes (Figures 20, 22, 26 and 35) excludes this possibility. Thus, taken together, the most likely conclusion is that the DSB serves to disrupt chromosomal contact between interacting genes. The repair experiment reveals that restoration of an intact SAMD4A gene loop, in a sequence-independent manner, is sufficient to restore co- transcription, as ef! as co-localization, of TNFA!P2 and SLC6A5 (Figure 49). Therefore, we speculate that by disrupting chromosomal interactions, the topological framework (comprised of gene loops and RNA Pol !l) is unable to assemble, thus, disrupting transcription. Collectively, these data provide strong evidence that intact chromatin is a requirement for loop-mediated co-transcription.

One way enhancer-promoter interactions are proposed to enhance transcription, is by bringing protein complexes to the promoter (Deng et al., 2012). In an analogous manner, through interchromosomai interactions, NF-KB has been shown to be delivered to the promoter of inducible genes. Similarly, we propose SAMD4A to be the dominant member of the NF-κΒ multigene complex that "organizes" transcription through loop-mediated contact. Recent published data in live cells reveals that RNA Pol II is mobile and clustering precedes transcriptional elongation and is linked to transcriptional initiation (Cisse et al., 2013). Therefore, we speculate that the SAMD4A gene loop provides a topological platform that serves as a scaffold on which a focus of many RNA Pol II molecules can cluster and engage in transcription of subordtnate members of the multigene complex. When TNFAIP2 or SLC6A5 cannot engage in chromosomal contact, their ability to access the focus of Pol II is limited. The strict hierarchical relationship between the interacting members further suggests a hand off or "collector" process, where dynamic chromosomal contacts with the dominant gene loop are formed (Figure 50). This could occur by a mechanism where TNFAIP2 via chromosomal contact "collects" its Pol II from an intact SA D4A gene loop and in turn SLC6A5 from an intact TNFAIP2 loop. Therefore, these data argue neither in favor of nor against putative "transcription factories," as they are currently defined, but suggest that such factories may be dynamically assembled rather than immobile structures. A recent study revealed that sequences within the promoter might drive the co-locaiization between these NFkB- regulated genes (Papantonis et al., 2013). Therefore, an alternative model might involve sequence specific elements within the promoters that facilitate the accumulation of different thresholds of general transcription factors for the 3 genes. Therefore, only upon the creation of a nuclear subcompartment of sufficient activity by SAMD4A can TNFAIP2 and then SLC6A5 be activated.

Our observation shifts the general paradigm of how transcriptional regulation in three dimensions occurs. Although these chromosomal interactions are rare and stochastic (Noordermeer et al., 2011), our single cell view strongly suggests that chromosomal contact between genes engaged in multigene complexes have a significant impact on co-transcription of interacting genes. Such long-range interactions of co-transcribed genes could serve to organize transcription in nuclear space, using hierarchical relationships between gene loops. Without gene loops of dominant members, subordinate members of the multigene complex cannot engage in long-range chromosomal contacts, nor can they participate in transcription. Finally, as looping enables stochastic chromosomal contact between genes in multigene complexes, this study provides supporting evidence to prior work showing that gene looping is a fundamental requirement of transcriptional activity (Tan-Wong et al., 2012). Importantly, such chromatin looping within multigene complexes (Li et al., 2012) may be governed by similar hierarchical regulation. Perturbation of members of multigene complexes, through knockouts, gene deletions, chromosomal translocations or silencing of transcription factors, all of which disrupt loop-mediated contact, may inadvertently result in unintended consequences to transcription of other members of a given multigene complex.

REFERENCES

Boch, J., et al. (2009) Science 326, 1509-1512.

Christian, M., et al. (2010) Genetics 186, 757-761.

Cisse, 1.1., et al. (2013) Science 341, 664-7.

Dekker, J., et al. (2002) Science 295, 306-13 1.

Deng, W., et al. (2012) Cell 149, 1233-1244.

Dixon, J.R., et al. (2012) Nature 485, 376-380.

Fanucchi et al. (20 3) Cell 155, 606-20.

Fullwood, .J., et al. (2009) Nature 462, 58-64.

Li, T., et al. (201 1 ) Nuc!. Acids Res. 39, 359-372.

Li, G„ et al. (2012) Celt 148, 84-98.

Lieberman-Aiden, E., et al. (2009) Science 326, 289-293.

Noordermeer, D., et al. (2011 ) Nat Cell Biol 13, 944-951.

Papantonis, A., et al. (2010) PLoS Biol 8: e1000419.

Tan-Wong, S.M., et al. (2012) Science 338, 671- 675.

Bienko, M., et al. (2013) Nat Methods. 10, 122-124.

Keogh, M.C., et al. (2005) Nature 439, 497-501.

Shanbhag, N.M., et al. (2010) Cell 141, 970-981.