WO2019116306A1 - System for capture or immunoprecipitation of a protein-dna complex - Google Patents

Abstract

The present invention concerns a system for immunoprecipitation of a protein-DNA complex or for capture of a protein-DNA complex including: - a microfluidic device including a plurality of antibody units, wherein each antibody unit includes at least one microfluidic channel, wherein the at least one microfluidic channel includes at least one surface configured to receive a multi-layered assembly of biochemical species for capturing a target molecule to form a biofunctionalized surface; - at least one inlet for providing at least one sample containing DNA binding proteins to the microfluidic device; and - at least one outlet for providing DNA or eluted DNA for sequencing.

Description

SYSTEM FOR CAPTURE OR IMMUNOPRECIPITATION OF A PROTEIN-DNA COMPLEX

CROSS-REFERENCE TO RELATED APPLICATIONS

The present application claims priority to International Patent Application PCT/IB2017/057889 filed on December 13, 2017, the entire contents thereof being herewith incorporated by reference.

FIELD OF THE INVENTION

The present invention concerns a microfluidic device for rapid, multiplexed, bead-less chromatin immunoprecipitation, herein named FloChIP. The microfluidic device may additionally include on-chip DNA library preparation. More particularly, the present invention concerns a system or method for capture or immunoprecipitation of a protein-DNA complex, for example, chromatin immunoprecipitation of histone marks or transcription factors.

BACKGROUND

Chromatin immunoprecipitation followed by sequencing (ChIP-seq) is the most widely used method to study protein/DNA interactions in vivo in a genome-wide manner [1] Despite its popularity, the most common form ChIP-seq remains a long (>2 days), manually intensive, low-sensitivity and low-throughput approach. While a number of attempts have been made towards improving the experimental workflow, they have specifically tackled only a subset of these limitations, ignoring others.

ChIP-seq has been fruitfully performed on different organisms, including mouse and humans [2-4], and has been adopted by renowned international consortia like the ENCODE project [4] As of October 2017, ChIP-seq has been cited in more than 17Ό00 PubMed scientific articles.

Since its introduction in 2007, ChIP-seq has been modified in different ways [5-7]; however, these modifications have mainly addressed the pre-immunoprecipitation preparation of chromatin. On the other hand, solid-state functionalized beads, specific antibodies and DNA library preparation have remained a constant across all the different ChIP-seq implementations.

Despite being widely adopted, current ChIP-seq protocols present inherent limitations that hinder even wider adoption. For instance, ChIP-seq is an intensive manual protocol, requiring multiple steps, reagents, consumables and expensive antibodies [8] This weighs on laboratories budgets and personnel time, thus reducing the possible scope and scale of experiments. Currently, only one commercially available solution is offering an automated ChIP-seq pipeline: the SX-8G IP-Star^® Compact Automated System by Diagenode [9] However, despite relieving personnel from mundane pipetting, the IP-star is a bulky semi-automated robot which still requires a substantial amount of consumables and reagents.

Another important bottleneck for current ChIP-seq implementations is the quantity of cellular material needed in order to produce high quality sequencing data. Both manual and the above-mentioned automated ChIP-seq require millions of cells, thus limiting the application of the protocol to scarce but medically relevant samples like small tumors and tissue biopsies [10].

Recently, a novel microfluidic implementation pushed this lower boundary, demonstrating the feasibility of ChIP-seq on as low as 100 cells (it is to be noted though that these cells belonged to an immortal cell line and the feasibility of such an endeavor on medically relevant samples is yet to be demonstrated) [11, PI]. However, this device still requires magnetic beads, magnets to handle them, it is neither high- throughput nor scalable and involves hours of post-ChIP DNA library preparation.

The genome-wide distribution and dynamics of protein-DNA interactions constitute a fundamental aspect of gene regulation. Chromatin immunoprecipitation followed by next generation sequencing (ChIP-seq) has become the most widespread technique for mapping DNA-protein interactions genome-wide. ChIP- seq has been successfully applied to dozens of transcription factors (TFs), histone modifications, chromatin modifying complexes, and other chromatin-associated proteins in humans and other model organisms.

The ENCODE and modENCODE consortia have performed more than 8,000 ChIP-seq experiments, which have enhanced our collective understanding of how gene regulatory processes are orchestrated in humans as well as several model organisms.

In addition, ChIP-seq proved to be essential to acquire new insights into genomic organization and into the mechanisms underlying genomic variation-driven phenotypic diversity and disease susceptibility. More specifically, this assay proved crucial in determining the DNA binding properties of hundreds of TFs. Nevertheless, in comparison to other widespread NGS-based methods - e.g. RNA-seq, ATAC-seq, and Hi- C - ChIP-seq lags behind in some key metrics, i.e. throughput, sensitivity, modularity, and automation, which hinder its wider adoption and reproducibility.

For example, while RNA-seq can now be regularly performed on hundreds or thousands of single cells using readily available workflows, ChIP-seq has largely remained labor intensive and limited to few samples per run, each composed of millions of cells. Moreover, while a typical pre-amplification RNA-seq workflow consists of only three steps - i.e. cell lysis, RNA capturing and reverse transcription - ChIP-seq typically involves several pre-amplification steps (crosslinking, lysis, fragmentation, immunoprecipitation, end-repair and adapter ligation). Finally, any given RNA transcript is present in each cell in numerous copies, which increases the likelihood of its capture and detection, whereas, on the other hand, each locus-specific protein-DNA contact occurs a maximum of two times in a diploid cell. The combination of these idiosyncratic differences, together with the lack of enabling solutions, has thus far prevented the ChIP-seq technology, as opposed to other NGS-based methods, to reach its full potential in terms of adoption and overall utility.

In addition to the standard ChIP protocol, a modification of its workflow involving sequential chromatin immunoprecipitation (sequential-ChIP) has also been adopted to infer genomic co-occurrence of two distinct protein targets. In principle, sequential-ChIP consists in performing ChIP twice on the same input chromatin, which leads to a multiplication of the inefficiencies mentioned above. Therefore, not only does sequential-ChIP show the same limitations as regular ChIP-seq, but these also come in an augmented form due to its sequential nature. As a result, despite the multi-dimensional information provided, sequential- ChIP has also resisted wider adoption.

In recent years, several attempts have been made to alleviate some of the limitations of the ChIP-seq and sequential-ChIP approaches. Gasper et al.[14] and Aldridge et al. [15] addressed the issue of automation by implementing the manual steps of a conventional ChIP-seq workflow on robotic liquid handling systems. However, in these examples, automation came at the expense of sensitivity, which remains in the range of tens of millions of cells per experiment.

Van Galen et al.¹⁶ and Chabbert et al.[17] addressed the issue of throughput by barcoding and pooling chromatin samples before immunoprecipitation (IP). Although van Galen and colleagues also proved that their approach leads to higher sensitivity (500 cells per ChIP), both methods are not automated and were shown to work only for histone marks.

Ma et al.[18] and Rotem et al. [19] addressed the limits of sensitivity with two different microfluidic-based strategies. Ma et al. focused on improving the efficiency of the IP step by confining it within microfluidic channels. Although these researchers showed good IP efficiency down to as few as 30 cells, their approach requires impractical antibody-oligo conjugates, is not automated and was not shown to work for TFs. On the other hand, Rotem et al. achieved the remarkable feat of performing ChIP-seq in a single cell by integrating the concept of chromatin barcoding and pooling into a single droplet-based microfluidic chip.

However, even though the barcoding step has indeed single cell resolution, the most critical step - i.e. the IP step - is performed manually on 100 cells. As a result, their approach, which was also shown to work only for histone marks so far, yielded sparse single cell data and thousands of assays are needed to identify specific cell subpopulation signatures.

In a notable effort to simplify the sequential-ChIP workflow, Weiner et al. [20] complement the immunoprecipitation steps with sequential chromatin barcoding, thus achieving a high degree of multiplexing. However, their approach increases the number of experimental steps which makes it significantly more labor-intensive given that the workflow is not automated. In summary, previous valuable attempts at improving the technology selectively address specific limitations but typically at the expense of or ignoring others.

SUMMARY OF THE INVENTION

The present invention addresses the above-mentioned limitations by providing a system for immunoprecipitation of a protein-DNA complex or for capture of a protein-DNA complex according to claim 1 and a method for immunoprecipitation of a protein-DNA complex or for capture of a protein-DNA complex according to claim 40.

In order to address the above-mentioned drawbacks, the inventors designed, manufactured and optimized a microfluidic system, named FloChIP herein, which supports miniaturized, bead-less, automated, scalable, multiplexed and sensitive chromatin immunoprecipitation experiments. For the first time, it is shown that high-quality ChIP-seq data can be obtained by straightforward bio-functionalization of a microstructured polymer of the microfluidic device of the present invention.

The major limitations of current ChIP-seq and sequential-ChIP solutions (throughput, sensitivity and automation) are tackled by the microfluidic multiprocessor device of the present invention, named FloChIP herein.

The device, system and method of the present invention shows that high quality one-day and parallelized ChIP-seq for histone marks (down to 500 cells and thus highly-sensitive) and transcription factors TFs (100Ό00 cells) can be achieved through a combination of microvalves, microstructures, flexible surface chemistry and on-chip chromatin tagmentation.

Moreover, the interconnected and modular device configuration of FloChIP enables straightforward re- immunoprecipitation of eluted chromatin, effectively enabling sequential-ChIP which allows to probe bivalent chromatin with unprecedented ease. As shown later, firstly the FloChIP has a wide dynamic input range (immunoprecipitation from 10^s to 500 cells) and it faithfully reproduces ENCODE data for all tested histone marks (immunoprecipitation on H3k27ac, H3k4me3, H3k4mel, H3k27me3 and H3k9me3).

The flexibility of the invention is demonstrated by performing H3k4me3 - H3k27me3 sequential ChIP-seq on the same chip, in the same day and with only 100Ό00 cells thus demonstrating the high sensitivity and efficiency of the present invention.

The high-throughput capabilities of the present invention are shown by performing ChIP-seq for a transcription factor (MEF-2A) on chromatin derived from 32 distinct lymphoblastoid cell lines.

Taken together, these results certify the present invention to be a flexible all-in-one automated ChIP-seq solution, which is a valuable enhancement to the existing known systems and methods.

The above and other objects, features and advantages of the present invention and the manner of realizing them will become more apparent, and the invention itself will best be understood from a study of the following description with reference to the attached drawings showing some preferred embodiments of the invention.

A BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

Figures 1A and IB show exemplary systems and devices of the present disclosure.

Figure 2 schematically shows an exemplary microfluidic device or chip of the present disclosure.

Figure 3 schematically shows exemplary elements of the microfluidic device or chip of the present disclosure.

Figure 4 is a schematic a cross-sectional view of an exemplary surface biofunctionalization of a surface in a channel of an antibody unit of the microfluidic device or chip of the present disclosure.

Figures 5a to 5f show data obtained using the system and method of the present disclosure performed on different chromatin-associated protein modifications namely, H3k27ac, H3k4me3, H3k4mel, H3k27me3 and H3k9me3 and this data is compared with publicly available ChIP-seq data generated by the ENCODE project.

Figures 6a and 6b presents data related to lower detection limit measurements of the system of the present disclosure.

Figure 6c presents a table containing data related to known ChIP-seq systems. Figure 7a to 7e generally show the microfluidic device architecture for miniaturized ChIP-seq. Figure 7a shows exemplary processing phases in descending chronological order of the system and method of the present disclosure. In the first "surface functionalization phase" (~S0 minutes), the inner walls are functionalised by sequentially introducing chemical species that firmly interact with both the previous and following layer of functionalization. Also in chronological order, these species can be, for example, biotin-BSA, neutravidin, biotin-Protein A/G and antibody. Following functionalization, IP takes place by flowing sonicated chromatin on-chip in a total time of 30-60 minutes, depending on the chromatin volume introduced. Subsequently, the antibody-bound chromatin is tagmented directly on-chip in order to introduce lllumina-compatible adapters. Finally, the tagmented chromatin is eluted off-chip by use of a SDS-containing buffer and high temperature.

Figure 7b shows a top-view of a portion containing numerous micro-pillars. Each portion is itself repeated several times along the length of one IP lane.

Figure 7c shows a top-view schematic of one IP lane. Each IP lane can be repeated n times across a FloChIP device and includes flow channels and control channels.

Figure 7d shows fluorescence micrographs showing biotin-BSA in the correct formation of the FloChIP's totem.

Figure 7e shows a top-view schematic of a high-throughput 64-unit FloChIP device, including flow channels and control channels.

Figures 8a to 8h generally concern the system's robust generation of chromatin landscapes for histone marks. Figure 8a shows a schematic depiction of FloChIP's mode 1: antibody multiplex. Each IP lane is functionalized separately by introducing different antibodies through the individual inlets. During IP, one sample is introduced through the common inlet and distributed equally across all IP lanes.

Figure 8b is a schematic depiction of FloChIP's mode 2: sample multiplex. One antibody solution is introduced through the common inlet and distributed equally across all IP lanes. During IP, each IP lane is loaded separately by introducing different samples through the individual inlets.

Figure 8c shows signal tracks for FI3k27ac, FI3k4mel and FI3k4me3 profiles obtained by the system of the present disclosure. For comparison, ENCODE data generated by conventional ChIP-seq are also shown. Figure 8d shows FI3k27ac profiles obtained by the system of the present disclosure with decreasing cell numbers. For comparison, ENCODE data generated by conventional ChIP-seq are also shown.

Figure 8e shows genome-wide correlation plots between FloChIP (x axis) and ENCODE (y axis) data for all targets tested, i.e. FI3k27ac, FI3k4mel, FI3k9me3, FI3k27me3 and FI3k4me3.

Figure 8f shows normalized read density profiles around transcription start sites for samples of decreasing cell numbers and ENCODE.

Figure 8g shows genome-wide correlation between pairs of samples with decreasing cell numbers and

ENCODE. Figure 8h shows a comparison in terms of fraction of reads in peaks (FRiP) between the system of the present disclosure and ENCODE for histone mark samples.

Figure 9a to 9d general relates to a "sequential IP" mode of the system of the present disclosure for the study of bivalent chromatin. Figure 9a shows exemplary sequential IP steps in descending chronological order for the case of FI3k4me3-FBk27me3. Chromatin coming from the first IP is collected into off-chip reservoirs connected to device. Following collection, the control channels are actuated in a way to isolate the first IP lane from the chromatin, while opening the path to the second IP lane. At this point, the chromatin flown into the second pre-functionalised IP lane. Finally, the bivalent chromatin is eluted again in off-chip reservoirs.

Figure 9b shows locus-specific signal tracks for the two individual IP libraries (H3k4me3 and H3k27me3) as well as the corresponding sequential IP samples (H3k27me3/H3k4me3 and FI3k4me3/ H3k27me3). Figure 9c shows a bivalency score distribution for HCP promoters. The shade-codes reflect the relative abundance of the two individual marks for each promoter.

Figure 9d shows Gene Ontology enrichment analysis for the first one thousand promoters with the highest bivalency score.

Figure 10a to g generally concern transcription factor IP in high-throughput mode performed by the system of the present disclosure. Figure 10a is a list of the 32 cell lines used in this study. Figure 10b shows qPCR enrichment for each library. The average across all libraries log 2 (fold change ) is 5.7. Figure 10c shows the percent of mapped reads for each library. The average mapping rate across all libraries is 46.6%. Figure lOd shows signal tracks reported for each library for three different genomic regions. Figure lOe shows the number of peaks called for each library (3374 peaks on average). Figure lOf shows the FRiP score for each library (6.9% on average). Figure lOg shows the MEF-2A motif enrichment for each library (a -log(Pvalue) of 9.2 on average).

Figure 11a also shows exemplary processing phases of the system and method of the present disclosure in descending chronological order in the case of chromatin/antibody pre-incubation. In the first "surface functionalization phase" (~80 minutes), the inner walls are functionalised by sequentially introducing chemical species that firmly interact with both the previous and following layer of functionalization. Also in chronological order, these species are for example biotin-BSA, neutravidin, biotin-Protein A/G. Following functionalization, the pre-incubated antibody/chromatin is flown on-chip in a total time of 30- 60 minutes, depending on the volume introduced. Subsequently, the antibody-bound chromatin is tagmented directly on-chip in order to introduce lllumina-compatible adapters. Finally, the tagmented chromatin is eluted off-chip by use of a SDS-containing buffer and high temperature.

Figure lib is a schematic of exemplary elements of the system of the present disclosure.

Figure 11c is an example of a COMSOL simulation used to optimise device architecture. Figure lid is a top-view schematic of the medium-throughput 8-unit microfiuidic device of the present disclosure including flow channels and control channels.

Figure lie shows an exemplary microfiuidic 64-outlet multiplexer for pressure distribution into microfiuidic device of the present disclosure, and Figure Ilf shows an exemplary 16-outlet multiplexer. Figures 12a to 12f concern results for IP on histone marks. Figure 12a shows amplification cycle statistics for samples of decreasing ceil number, from 1 million to 500 ceils. Figure 12b shows fold enrichment statistics for samples of decreasing cell number, from 1 million to 500 ceils. Figure 12c shows a mapping rate for samples of decreasing ceil number, from IQO'OOQ to 500 ceils. Figure 12d shows a FRiP score for samples of decreasing cell number, from 1 million to 500 cells. Figure 12e shows normalized read density profiles around transcription start sites for H3k4me3, H3k27ac and H3k4mel. Figure 12f shows fold enrichment statistics for histone mark samples, namely H3k4me3, H3k27ac, H3k27me3, H3k9me3 and H3k4mel.

Figures 13a and 13b show signal tracks for individual and sequential IP libraries previously reported for the same loci.

Figure 13c shows correlation results between results provided by the system and method of the present disclosure and previously published Co-Chip data.

Figures 14a and 14b concern genome wide characterization of TF data by the system and method of the present disclosure. Figure 14a shows normalized read density profiles around transcription start sites for all sequenced libraries (thicker line being the average profile). Figure 14b shows correlation results between ail pairs of sequenced libraries.

Herein, identical reference numerals are used, where possible, to designate identical elements that are common to the Figures.

DETAILED DESCRIPTION OF SEVERAL EMBODIMENTS

Figures 1A and IB schematically show an exemplary system 1 of the present disclosure. The system 1 is, for example, a system for capture or immunoprecipitation of a protein-DNA complex; or a protein-DNA complex capture or immunoprecipitation system.

The system 1 is capable of supporting microfiuidic, bead-less, automated, scalable, multiplexed and sensitive immunoprecipitation (see, for example, Fig. 1A and IB) This system 1 is called FloChIP herein.

The system 1 is a bead-less system or a magnetic bead-less system.

The system 1 is, for example, an immunoprecipitation system configured to determine protein/DNA interactions in vivo. The system is, for example, configured to immunoprecipitate or capture a protein- DNA complex. The system 1 is, for example, a chromatin immunoprecipitation system.

The system 1 includes a microfluidic device or chip 108 comprising at least one or a plurality of antibody units 111, 201.

The antibody unit 111 (or each antibody unit 111) includes at least one microfluidic channel MFC or a plurality of microfluidic channels MFC.

The microfluidic channel MFC includes at least one or a plurality of surfaces configured to receive a multi layered assembly MLA of biochemical species (Figure 4). The multi-layered assembly MLA of biochemical species defines or forms a biofunctionalized surface. The biofunctionalized surface is configured to capture a target molecule.

The antibody unit 111 contains one or more side walls wl, w2 as well as an upper wall or ceiling CL and a lower wall or floor FL defining the microfluidic channel or channels MFC. These walls define a closed microfluidic channel MFC.

The microfluidic channel MFC may define a cavity having, for example, a substantially honeycomb or circular cross-sectional profile, or a substantially square or rectangular cross-sectional profile, as shown for example in Figure 4.

The microfluidic channel MFC has, for example, a height between 500nm and 500pm or between 5pm and 50pm (for example between the upper FL and lower CL walls FL), and/or a width between 500nm and 500pm or between 5pm and 50pm (for example, between side walls wl, w2), and/or a length between 5pm and 5cm or between 50pm and 2cm (for example, in the flow direction F (see Figure 4)).

Other separation walls may also be present to define a plurality of microfluidic channels in the unit 111 separated by these separation walls. The plurality of microfluidic channels may comprise independent microfluidic channels in which, for example, a different multi-layered assembly MLA of biochemical species is present. The plurality of microfluidic channels may alternatively or additionally be interconnected via a flow channel or lane fl (see for example Figure 3) that can be opened or closed to establish or not the interconnection using a controller or control channel.

The antibody unit 111 may alternatively comprise a plurality of elongated sections si containing solely one microfluidic channel extending therethrough and the plurality of sections may be interconnected using the previously mentioned flow channels fl (Figure 3). The walls include or define a surface configured to receive the multi-layered assembly MLA of biochemical species.

The walls may extend substantially parallel to each other. Alternatively or additionally, the walls can define a curved surface delimiting the microfluidic channel MFC. For example, the walls define a concave curvature 304c as shown in Figure 3.

The microfluidic channel MFC can include an array or plurality of elongated structures MP (see for example Figure 3). The elongated structure MP extends from a wall of the antibody unit 111 and inside the microfluidic channel MFC. The elongated structure or structures MP may extend from a first wall towards a second wall and may contact the second wall. For example, the elongated structure M P may extend from the lower wall or floor FL towards the upper wall or ceiling CL and may contact the upper wall or ceiling CL. Alternatively or additionally, the elongated structure or structures M P may extend from a first side wall wl towards the second side wall w2 and contact the second side wall w2.

The elongated structures extend from a surface of the microfluidic channel MFC and extend inside the microfluidic channel M FC.

The elongated structures may extend in a direction non-parallel or substantially perpendicular to a direction of fluid flow.

The elongated structures MP may also contact a wall and extend along the length of the wall. For example, contact and extend along the side wall wl and/or side wall w2 between the lower wall or floor FL towards the upper wall or ceiling CL.

The array of elongated structures is composed, for example, of a repeating pattern 304a, 304b of elongated structures. The pattern can, for example, comprise a plurality of elongated structures disposed as a matrix or grid array. The repeating pattern can extend partially or fully through the antibody unit 111.

The elongated structure MP is or defines, for example, a micropillar.

The or each elongated structure has, for example, a height between 5pm and 50pm (for example between the upper FL and lower CL walls FL), and/or a width between 5pm and 50pm (for example, in the X- direction (see Figure 3), and/or a length between 5pm and 50pm (for example, in the Y-direction (see Figure 3)). The topology or arrangement of the elongated structure MP in microfluidic channels MFC is determined in order to obtain a high surface area, to miniaturize the overall device footprint and ensure a flawless or extremely high distribution of chemical species that is without or with minimized dead volumes where undesired chromatin could accumulate.

The elongated structure can have or define for example a circular cross-section, a square or rectangular cross section 304b, or preferably a rhomboidal cross section 304a, for example, with the major axis aligned to the direction of the flow.

The plurality of elongated structures MP may all have the same cross-sections or different or a mix of different cross-sections.

The plurality of elongated structures MP in the antibody unit 111 or each antibody unit 111 defines, for example, a surface area to volume ratio of between 0.5 micrometer ¹ and 5 micrometer ¹. The plurality of elongated structures MP in the microfluidic channel MFC or each microfluidic channel MFC defines a surface area to volume ratio of between 0.5 micrometer ¹ and 5 micrometer¹.

Figure 3 shows the plurality of elongated structures MP in the microfluidic channel MFC of a subpart 303 of the antibody unit 111. The content of this subpart 303 is repeated extending through section si to define, for example, an immunoprecipitation (IP) lane (see also Figure 7a). The antibody unit 111 of Figure 3 includes a plurality of IP lanes, for example, fives interconnected IP lanes.

The microfluidic device or chip 108 can include a plurality of antibody units 111 in which the plurality or each antibody unit 111 includes at least one or a plurality of interconnected IP lanes, as for example shown in Figure lb or Figure 7a.

The presence of the plurality of elongated structures MP inside the microfluidic channel MFC defines or results in a plurality of interconnected microfluidic canals or passages MCA (see Figure 7) being present between adjacent elongated structures in the antibody unit 111 and through which a fluid can flow through.

The microfluidic canal or passage has, for example, a height between 5pm and 50pm (for example between the upper FL and lower CL walls FL), and/or a width between 5pm and 50pm (for example, in the X-direction (Figure 3) between two elongated structures MP), and/or a channel length between 5pm and 50pm (for example, in the Y-direction (Figure 3) between two elongated structures MP). In one exemplary embodiment, the microfluidic device 108 includes at most 7425 microfluidic canals or passages MCA, and at most 4400 elongated structures MP. Each antibody unit includes at most 135 microfluidic canals or passages MCA, and at most 80 elongated structures MP.

Each elongated structure MP also delimits or includes a surface configured to receive a multi-layered assembly MLA of biochemical species for capturing a target molecule. The multi-layered assembly MLA is provided on or attached to this surface of the elongated structure MP to provide a biofunctionalized surface for capturing a target molecule.

The surface of the walls defining the microfluidic channel MFC and/or the surface of the elongated structures MP comprise or consist of a hydrophobic surface or a hydrophilic surface.

The walls or the surface of the walls defining the microfluidic channel MFC and/or the elongated structures MP or the surface of the elongated structures MP include or consist solely of a polymer.

The polymer is, for example, configured to passively adsorbs an attachment layer or first layer 401 of the multi-layered assembly MLA. The polymer can, for example, consist solely of or comprise polydimethylsiloxane (PDMS), or Polymethylmethacrylate (PMMA), or Cyclic-olefin copolymer (COC), or POLYSTYRENE (PS), or POLYCARBONATE (PC), or POLY-ETHYLENE GLYCOL DIACRYLATE (PEGDA), or POLYURETHANE (PU), or polyfluoropolyether diol methacrylate (PFPE-DMA), or TEFLON.

The surface of the walls defining the microfluidic channel MFC and/or the elongated structures MP is biofunctionalized and includes at least one antibody and/or at least one molecular capturing agent attached thereto. The antibody is, for example, specific to a DNA binding protein.

The microfluidic channel MFC and the elongated structures MP include, attached to or provided on the hydrophobic or hydrophilic surface, a first layer 401 assembled by deposition of a first protein species to the hydrophobic or hydrophilic surface. The first layer 401 is, for example, assembled by adsorption of a first protein species to the surface of the microfluidic channel MFC and/or the elongated structures MP. The first protein species comprises or consist of, for example, biotinylated bovine serum albumin (biotinylated-BSA).

The multi-layered assembly MLA of biochemical species includes a second layer 402 composed of a second protein species having (high) biophysical affinity to the first protein species. The second protein species comprises or consists of, for example, a protein of the avidin family or a glutaraldehyde. The multi-layered assembly MLA can further include a third layer 403 composed or consisting of a third protein species with (high) biophysical affinity to the second layer or assembly 402, or with an antibody or biotinylated antibody. The third protein species comprises or consists of a protein A or protein G or protein A/G, or a biotinylated protein A or biotinylated protein G or biotinylated protein A/G.

The multi-layered assembly MLA can further include a fourth layer 404 composed or consisting of an antibody or capturing agent.

The multi-layered assembly MLA can comprise or consist solely of the first layer 401, the second 402, the third 403, and the fourth 404 layer.

The first 401, second 402, third 403, and/or fourth 404 layers can be assembled by flowing a concentrated protein solution through the microfluidic channel or channels MFC of the antibody unit 111, 201.

Figure 4 schematically shows the multi-layered assembly MLA on one of the side walls w2. However, the multi-layered assembly MLA can also be preferably present on the other side wall wl and the upper CL and lower FL walls as wells on the surfaces defines by the elongated structures MP. The multi-layered assembly MLA is configured to target any molecule to which a specific antibody can bind. The multi layered assembly MLA is, for example, configured to target protein/DNA complexes, transcription factors, histone modifications or DNA methylation.

A method for fabricating a system 1 for immunoprecipitation of a protein-DNA complex or for capture of a protein-DNA complex of the present disclosure may include providing the system 1 wherein the microfluidic channel or channels MFC and/or the elongated structures MP are non-biofunctionalized; and flowing a concentrated protein solution through the microfluidic channel or channels MFC of the antibody unit 111, 201 to assembly a plurality of layers 401, 402, 403, 403 on at least one surface of the microfluidic channel or channels MFC and/or the elongated structures MP to biofunctionalize the microfluidic channel or channels (MFC) and/or the elongated structures MP.

The multi-layered assembly MLA defines a multilayered "totem" of molecular species. The totem is based on strong although non-covalent molecular interactions and culminates with the immobilization of an antibody of choice prior to immunoprecipitation.

The first layer can be, for example, obtained by flowing a concentrated solution of biotinylated-BSA into and through the microfluidic device or chip 108, which passively and thoroughly adsorbs to the

IB hydrophobic or hydrophilic walls and surfaces of the microfluidic device 108. This layer has both an insulating role, that is, it prevents non-specific adsorption of chromatin, and a docking role for the next layer, which can be obtained by flowing a solution containing, for example, neutravidin that strongly binds to the biotin groups of the first layer.

The third layer can be formed by flowing a solution of, for example, biotinylated-protein A/G, which gets firmly immobilized by the unsaturated binding sites of the previous layer. A recombinant protein having ability to strongly bind to a large number of different antibodies is thus preferably used. The multi-layered assembly MLA is configured to retain this ability and constitutes a general substrate for antibody pull down.

Each antibody unit 111 can, for example, be configured to capture a different target molecule or to carry out a different immunoprecipitation reaction.

The system 1 further includes at least one or a plurality of inlets 107 for providing a sample containing, for example, DNA binding proteins to the microfluidic device 108. The inlet 107 is configured for providing, for example, a cross-linked sheared or sonicated chromatin to the microfluidic device 108.

The inlets 107 are in fluid communication with the microfluidic device 108 and the microfluidic device 108 includes openings for fluid communication with the inlets 107. The inlets 107 are configured to be open or closed to allow or prevent fluid communication with the microfluidic device 108. This is done, for example, using control valves 104 actioned via a microcontroller 102 and computer/processor 101 using a system control program or software SW. This allows to select individual anti-body unit or units 111, or individual IP lane or lanes for sample input. Alternatively, all anti-body units 111, or IP lanes can be selected via use of a common inlet 204 (see for example Figure 2).

The system 1 also includes inlet collection tubes 106 in fluid communication with the inlets 107. The inlet collection tubes 106 are configured to provide input elements to the microfluidic device 108. The inlet collection tubes 106 are configured to hold and release samples to the microfluidic device 108 to be processed by the microfluidic device 108. The inlet collection tubes 106 are also used to form the multi layered assembly MLA in the microfluidic device 108. The contents of one inlet collection tubes 106 can, for example, be provided to all anti-body units 111 of the microfluidic device 108.

The system 1 also includes one or a plurality of outlets 110 for providing DNA or eluted DNA, from the microfluidic device 108, for sequencing. The outlets 110 are connected to the microfluidic device 108 in a same manner as that of the inlets 107 and are controlled in the same manner. The system 1 can further include outlet collection tubes 109 connected to the microfluidic device 108 outlets 110 for collecting processed samples from the microfluidic device 108.

The collecting processed samples from the microfluidic device 108 can also be re-inserted by additional inlets into the microfluidic device 108 for further processing, for example, by one or more different anti body units 111. This, for example, allows sequential ChIP-seq to be performed.

As shown, for example, in Figures la, lb and lib, the system 1 can further include a calculator or calculating means 101, a microcontroller unit 102 and a valve device comprising a plurality of electromechanical (EM) valves 104.

A temperature control device for controlling the temperature of system elements such as the microfluidic device 108 may also be optionally included. For example, a PCR device.

The calculator or calculating means 101 can be, for example, a general-purpose computing unit 101 used for experiment scripting. The system control program or software SW executed by the calculator or calculating means 101 allows operation of the system 1 to be carried out and the methods of the present disclosure to be performed.

The microcontroller unit 102 is connected to the computer 101 and to an arbitrary number or plurality of electromechanical (EM) valves 104. This unit 102 translates the computer instructions into voltage signals. It is connected to the EM valves by means of electric wires 103.

The arbitrary number or plurality of EM valves 104 are connected to the microcontroller unit 102 and to the set of collection tubes 106 which, for example, contain experimental samples or other solutions to be inputted to the microfluidic device or chip 108. EM valves 104 and collection tubes 106 are, for example, connected through airtight connections 105 in the form of plastic (or rubber) tubing (or wells). The EM valves 104 are configured to allow or block the passage of pressurized air to the microfluidic device or chip 108.

As previously mentioned, the microfluidic chip 108 contains an arbitrary number or plurality of antibody (or IP) units 111, 201. The microfluidic chip or device 108 is connected to the inlet collection tubes 106 and the outlet connection tubes 109 through airtight connections 110 in the form of, for example, plastic (or rubber) tubing (or wells). The valve device comprising the plurality of electromechanical (EM) valves 104 is configured to receive a pressurized fluid, for example, air and configured to distribute the pressurized air to multiple outlets of the valve device (see for example, Figure lib). The outlets each comprise an EM valve 104 that is controlled via the microcontroller unit 102. The air pressure value at the outlet can also be controlled and set to a desired value.

The plurality of EM valves 104 are connected, via the inlet collection tubes 106, to flow lines LF and control lines LC of the microfluidic chip or device 108 (see for example, Figure lib). The plurality of EM valves 104 are operated to allow or block the passage of a pressurized fluid (provided for example via the inlet collection tubes 106) to the microfluidic device 108. Valves 104 connected to flow lines LF permit to distribute a fluid through flow lines and compartments such as the microfluidic channels MFC to allow, for example, the immunoprecipitation reaction to be carried out inside the antibody units 111 of microfluidic chip 108. Valves 104 connected to a control line LC permit to select the direction or destination of the fluid flow in the microfluidic device 108. For example, these valves 104 connected to a control line LC permit to select which antibody units 111 fluid is flown to and are thus used in the microfluidic device 108, or the sequence of antibody units 111 a fluid is flow through. This allows different biofunctionalisation of the antibody units 111 and also to perform multiplexed measurements or processing of samples in the microfluidic device (108).

The microfluidic device 108 can, for example, comprise a superposed control layer and flow layer in which are defined superposed control lines and flow lines separated by an intermediate blade or wall. Fluid pressure applied to the control line deforms the intermediate blade to allow opening and closing of the flow line to control fluid flow in the flow line.

The valves 104 are activated via the microcontroller unit 102 and the calculating means 101 with the desired processing in the microfluidic device (108) being determined via the experiment scripting or the system control program SW executed by the calculating means 101.

The actual immunoprecipitation reaction is carried out inside of the FloChIP microfluidic chip 108 (see for example, Fig. 2). As previously mentioned, the device 108 itself is composed of an arbitrary number or plurality of antibody units 201, one for each immunoprecipitation reaction. This array of units can be connected to the collection tubes 106 either in parallel through the common inlet 204 and common outlet 206, or individually through the individual inlets (202, one inlet per unit) and individual outlets (205, one outlet per unit). Inside of the microfluidic chip 108, important elements of the FloChIP approach are combined. These elements are: the antibody unit (Fig. 3) and an innovative surface biofunctionalisation (Fig- 4). Each antibody unit 111 can be composed of a large number of serially connected microfluidic microstructured units. Examples of these microstructured units 304a, 304b are reported in a close-up view of a subpart 303 of the antibody unit 111.

The microstructures MP have the role of increasing the surface area-to-volume ratio of the antibody unit 111. These microstructures MP can be implemented for example in the form of micropillars with different cross-sectional shapes. For example, the cross-section of the micropillars can be rhomboidal 304a, squared 304b, or circular. The micropillars are not essential for the basic functioning of the above described biofunctionalization chemistry and a unit without pillars 304c also functions. However, best results are obtained in the presence of micropillars. Given the importance of the device surface, the use of fluidic channels (or passages) and microstructures with microscopic dimensions is also important. Therefore, microfluidic channels (or passages) and microstructures are preferred to have heights ranging between 5 and 50 micrometers.

The surface biofunctionalisation is represented by the multilayered assemble MLA of biochemical species (Fig. 4, 7a, 11a). The function of this surface biofunctionalisation is accommodating a wide variety of antibodies and other molecular capturing agents in order.

The first layer 401 of FloChIP surface biofunctionalisation is assembled by passive adsorption of a protein species to the walls of the microfluidic channels. For this reason, any polymer that supports passive adsorption of protein species can be used for FloChIP. The polymer used is for example polydimethylsiloxane (PDMS) but any hydrophobic or hydrophilic polymer can be used as a viable substrate and the protein species used for this first layer can be, for example, biotinylated-BSA.

The second layer 402 is composed of a protein species with high biophysical affinity to the protein species of the first layer 401. In the case of FloChIP, this protein can be, for example, a protein of the avidin family.

The third layer 403 is also composed of a protein species with high biophysical affinity to the second layer 402. In the case of FloChIP this protein can be, for example, biotinylated protein A or protein G or protein A/G.

The fourth layer 404 is a layer composed of the antibody or capturing agent of choice.

All the layers mentioned above are assembled, for example, by flowing a concentrated protein solution through the microfluidic channels of the antibody unit 111. Alternative multilayered assemblies could be formed by using, for example, the following sequence of protein/chemical solutions:

Layer 1: biotinylated BSA; layer 2: avidin; layer 3: biotinylated antibody.

Layer 1: BSA; layer 2: glutaraldehyde; layer 3: protein A or G or A/G, layer 4: antibody.

Layer 1: BSA; layer 2: glutaraldehyde; layer 3: antibody.

Once assembled the multilayered surface biofunctionalisation is ready to be used to capture the target molecules. Considering that FloChIP allows in principle any antibody to be immobilized, the possible target molecules are any molecule to which a specific antibody can bind. Examples include, but are not limited to, transcription factors, histone modifications and DNA methylation.

The present disclosure also concerns a method for immunoprecipitation of a protein-DNA complex or for capture of a protein-DNA complex. The method includes the steps of providing the system 1 of the present disclosure, providing at least one sample containing DNA binding proteins to the microfluidic device 108, and recuperating, at the least one or a plurality of outlets of the microfluidic device 108, DNA or eluted DNA for sequencing.

The method is, for example, a chromatin immunoprecipitation method.

The providing step can include, for example, providing a cross-linked sheared or sonicated chromatin to the microfluidic device 108.

The method may further include the step of preparing the DNA for sequencing by providing a transposase loaded with NGS adapters in the microfluidic chip 108 for attachment to the DNA; and eluting the DNA in the microfluidic chip 108.

Eluting the DNA in the microfluidic chip 108 can be carried out by applying thermal energy and a salt buffer.

The method may further include the step of carrying out sequencing of the DNA. The sequencing can be carried out, for example, using next generation sequencing (NGS) processing.

The method may further include the step of carrying out on-chip library indexing to decrease downstream library preparation time. Carrying out on-chip library indexing to decrease downstream library preparation time can, for example, comprise flowing, in the microfluidic device 108, a transposase loaded with NGS adapters on top of the immunoprecipitated chromatin.

The provision of a sheared or sonicated chromatin to the microfluidic device 108 may include providing samples containing less than 1000, or less than 500 cells.

The present disclosure also concerns an apparatus configured to carry out the above described method.

In a preferred use of the technology, protein/DNA complexes are immunoprecipitated using FloChIP and subsequently the DNA is prepared for next generation sequencing (NGS) directly on chip.

This preparation is, for example, carried out by flowing in the microfluidic chip 108 a transposase loaded with NGS adapters on top of the immunoprecipitated chromatin. This is a variant of a patented technology called "ChIPmentation" [12, P2] The DNA is then eluted with high temperature and high salt buffer in order to be processed by next generation sequencing (NGS).

In order to demonstrate the feasibility of the system and method of the present diclosure, to the inventors performed FloChIP on different chromatin-associated protein modifications - namely, FI3k27ac, FI3k4me3, FI3k4mel, FI3k27me3 and FI3k9me3 - and compared their generated data with publicly available ChIP-seq data generated by the ENCODE project (Fig. 5). The cell line used was a standard lymphoblastoid cell line (GM12878) and the comparison is carried out in terms of dataset correlation based on the genomic co-localization of mapped reads. In other words, the higher the correlation, the more similar and thus consistent are the compared datasets. As can be seen in Fig. 5 from panel a) to e), FloChIP-derived data is very consistent with the benchmark data. Moreover, the proportion of mapped reads that fall within ChIP-seq peaks tend to be the same - if not higher - for FloChIP as compared to ENCODE (Fig. 5f), thus corroborating the efficiency of the system and method of the present disclosure. Finally, as observed in Fig. 5g, the superior inter-experiment correlation between two FI3k27ac replicates advocates for the high reproducibility of the system and method of the present disclosure. It is noteworthy that, ENCODE experiments were performed with an average of 10 million cells for each ChIP, whereas FloChIP was performed with only 100 thousand cells. This demonstrates that not only can FloChIP achieve the comparable genomic coverage as ENCODE, but it can do so with substantially less cellular input.

The lower limits of detection of the system and method of the present disclosure was also investigated. In order to this the inventors performed a multiplexed experiment in which was used the same antibody target (i.e. FI327ac in this case) for seven different samples of decreasing cellular input, from 400,000 cells to as few as 500 cells. As can be seen by the qPCR data replicates reported in Fig. 6a, the inventors obtained high enrichment for all input dilutions, therefore proving that the system and method of the present disclosure can selectively enrich the target sites for a very low number of cells. Subsequently, the inventors proceeded to sequence the two samples at the extremes of our dilutions (i.e. with 500 and 400,000 cells) and then evaluate the correlation. In Fig 6b, both samples show a good correlation with the respective ENCODE data. Most importantly, despite the extremely minute sample size, a high correlation is achieved between the data generate with only 500 cells, ENCODE and even more so with the parallel 400,000 cells sample. Taken together, these results illustrate the unparalleled sensitivity of FloChIP, which no other macroscale approach, manual or automated, can achieve.

The inventors next performed a thorough search of both published and commercially available solutions, which claimed to overcome some of the above-mentioned limitations of manual ChIP-seq (Table 1 of Figure 6c). Currently, Abeam and Active Motif offer standardized kits which are nevertheless still manually intensive. Active Motif's solution claims to allow for ChIP-seq on 1000 cells, although no peer reviewed publication supports this claim. As mentioned in the introduction, the state of the art for automated ChIP- seq the Diagenode's IP-star, whereas sensitivity is presently unsurpassed for the MowChIP-seq system. As drawbacks, the IP-star still requires a long ChIP duration, library preparation time and abundant input material. On the other hand, MowChIP-seq is not automated, requires long library preparation time and is neither scalable nor multiplexed. FloChIP presents itself as an all-in-one system, uniting the benefits of automation with the efficiency and sensitivity of a miniaturized system. On top of this, FloChIP also provides on-chip library indexing with consequent decrease of downstream library preparation time. Based on these observations and on the advantages summarized in Table 1 of Figure 6c, FloChIP is a preferable solution for user-friendly and reliable ChIP-seq experiments.

In one exemplary microfluidic device 108 fabrication, the microfluidic device comprises or consists, for example, of a flow and a control layer. Molds for each layer can be fabricated using standard lithography techniques on 4" silicon wafers. After exposure to 02 plasma the control layer mold is patterned with SU- 8 photoresist to a height of, for example, 10 pm. The mold for the fluidic layer was fabricated with AZ9260 photoresist to a height of, for example, 20 pm. Devices were cast in polydimethylsiloxane (PDMS), for example, using multilayer soft lithography. PDMS was prepared at a 20:1 ratio and spin-coated on the flow layer mold at 1700 rpm. PDMS at a 5:1 ratio was cast on the control layer mold to a thickness of about 4 mm. Both layers were baked at 80 °C for 30 min. The control layer was peeled off from its mold and manually aligned to the flow layer mold, followed by a baking step at 80 °C for 90 min.

In one exemplary method of device functioning and operation control lines on the microfluidic device 108 were primed with phosphate buffered saline (PBS) at 5 psi. When the control lines were fully primed the pressure was increased to 25 psi. Flow lines were operated at 3 psi. First, the surface area was derivatized by flowing, for example, a solution of biotinylated BSA, followed by a 1 min wash in PBS. Next a neutravidin solution in PBS was flown, followed by a 1 min wash in PBS. Next a biotinylated protein A/G in PBS was flown, followed by a 1 min wash in PBS. Next solutions of a ChIP-grade antibody of choice is either introduced in one single IP unit 111 or to all the IP units 111 of the array. Next, sonicated chromatin is introduced (again, individually or in parallel to all IP units). Next the chromatin-associated DNA is indexed by flowing a Tn5 transposase solution containing Nextera adapters. Next the indexed chromatin is eluted from the chip, amplified by 10-15 PCR cycles and sequenced using an lllumina NextSeq.

For chromatin preparation, GM12878 cells samples were centrifuged at l,600g for 5 min at room temperature in a swing bucket centrifuge with soft deceleration. Cells were then washed twice with PBS at room temperature by centrifugation and resuspension. Cells were cross-linked for 5 min with 1 ml of 1% freshly prepared formaldehyde. Cross-linking was terminated by adding 0.05 ml of 2.5 M glycine and shaking for 5 min at room temperature. Cross-linked cells were pelleted and washed with precooled PBS buffer and resuspended in sonication buffer. Cross-linked cells were sonicated with a Covaris E220 sonicator. The sonicated lysate was centrifuged at 14,000g for 10 min at 4 °C. Sonicated chromatin in the supernatant was transferred to a new 1.5-ml LoBind Eppendorf tube. From this stock chromatin preparation, samples equivalent to 400Ό00, 100Ό00 and 500 cells were divided into aliquots and diluted to give a final volume of 50 pi.

Further system 1 details and methods of the present disclosure are now presented.

The FloChIP system 1 is engineered for automated, bead-less and miniaturized ChIP-seq.

The two core elements of FloChIP's technology are the assembly of a multilayered "totem" of molecular species (Fig. 7a, Supp. Fig. 11a) and an engineered pattern of high surface-to-volume micropillars (Fig. 7b). The totem is based on strong although non-covalent molecular interactions and culminates with the immobilization of an antibody of choice prior to immunoprecipitation. The first layer is obtained, for example, by flowing on-chip a concentrated solution of biotinylated-BSA, which passively but thoroughly adsorbs to the hydrophobic walls of the microfluidic device 108. This layer has both an insulating role, i.e. it prevents non-specific adsorption of chromatin to the walls, and a docking role for the next layer, which is obtained for example by flowing on-chip a solution containing neutravidin that strongly binds to the biotin groups of the first layer. The third layer is formed, for example, by flowing a solution of biotinylated- protein A/G, which gets firmly immobilized by the unsaturated binding sites of the neutravidin layer. Protein A/G is a recombinant protein used in a variety of immunoassays due to its ability to strongly bind to a large number of different antibodies. This ability is retained by FloChIP's totem which thus constitutes a general substrate for antibody pull-down (Fig. 7a).

For the successful initiation of the antibody-capturing totem, the only substrate requirement is the hydrophobic or hydrophilic surface, for example, a hydrophobic surface of the polymer. Therefore, the inventors set out to optimize the topology of the microfluidic channels having three main goals in mind: obtaining as much surface area as possible, miniaturizing the overall device footprint and ensuring flawless distribution of chemical species - i.e. without dead volumes where undesired chromatin could accumulate. This optimization strategy led to a preferred design encompassing an array of micropillars of rhomboidal cross-section with the major axis aligned to the direction of the flow (Fig. 17b). To achieve a total estimated surface area that yields a sufficiently complex post-IP DNA library, the micropillar pattern is repeated multiple times across each IP-lane (Fig. 7c).

With the goal of visually confirming the outcome of the combination between totem assembly and the micropillar array and that every element is essential to this end, the inventors first sought IP chromatin derived from a FleLa FI2B-mCherry cell line using an anti-mCherry antibody. The resulting fluorescence micrographs confirmed the role of each layer of the totem for successful IP of cellular chromatin (Fig. 7d).

The IP-lane is an important element of the FloChIP architecture and it can itself be repeated n times, where n is the desired throughput of the device. For the low-throughput initial tests an exemplary 8-lane FloChIP device was used (Fig. lid), whereas for high-throughput an exemplary 64-unit device (Fig. 7e) was used. To gain accurate flow control, automation and multiplexing, a network of valves, for example, Quake-style microfluidic valves was used in the system 1. Moreover, by actuating distinct sets of valves, different multiplexing modes can be achieved with the same microfluidic architecture.

For instance, a "FloChIP mode 1" provides the option of multiplexing one sample into different IP units, hence equally distributing the same sample across multiple lanes, enabling multiple parallel IPs involving distinct antibodies (Fig. 8a). Alternatively, "FloChIP mode 2" provides the option of coating the whole device with one antibody, thus achieving sample multiplexing (Fig. 8b). it is noted that both multiplexing modes are compatible with the direct chromatin immunoprecipitation ChIP approach. Flowever, FloChIP is also fully compatible with the indirect ChIP strategy, in which the chromatin is pre-incubated with an antibody before flowing the sample-antibody mixture on-chip.

Interestingly, the inventors noticed that for low-affinity antibodies (e.g. certain TF antibodies), the indirect ChIP was preferable to the direct one. On the other hand, all tested histone marks antibodies showed high affinity for their epitopes and both direct and indirect ChIP yielded an equal data quality. Regardless of the chosen approach, it is important to emphasize that, due to its microfluidic nature, FloChIP's IP step is considerably shorter (30 to 60 minutes) than for the other macroscale alternatives.

The FloChIP system 1 reliably reproduces ENCODE data across a wide range of input cells. To benchmark the reliability of the system 1 of the present disclosure and its multiplexing features as well as its overall sensitivity, the inventors first set out to empirically estimate the overall binding capacity of each IP lane. To this end, the inventors performed FloChIP in multiplexing mode 2, i.e. "antibody multiplex", by functionalizing the whole chip with an anti-FI3K27ac antibody and immunoprecipitating different chromatin dilutions, from 1 million down to 500 cells. Since the amount of DNA that was typically recovered from the chip tended to be too small to be measured directly, the inventors used the number of amplification cycles needed to reach a given cycle threshold (Ct) as an indirect estimator of DNA amount. Following this metric, it was observed that for lower input amounts, a progressively greater number of amplification cycles is required to reach the same Ct value (Fig. 12a), indicating that below 100Ό00 cells, FloChIP functions in below-saturation conditions. The inventors therefore estimated that FloChIP's inner surface saturates at approximately 100Ό00 cells. Nevertheless, the inventors also obtained positive and stable fold enrichment results across the whole series of dilutions tested, suggesting that FloChIP can be carried out efficiently below and above its saturation point (Fig. 12b). To obtain a genome-wide perspective on its dynamic range, the inventors sequenced FloChIP's derived libraries for four dilutions, i.e. 100Ό00, 50Ό00, 5Ό00 and 500 cells. Although the rate of uniquely mapped reads remained high for all samples (Fig. 12c), the fraction of reads falling into peaks (FRiP score) slightly decreased with lowering input amounts - from over 60% for 100Ό00 cells, to just above 10% for 500 cells (Fig. 12d).

Nevertheless, genome-wide analysis of the obtained libraries revealed expected accumulation of reads into regions in proximity of transcription start sites (TSS, Fig. 8f). Moreover, genome-wide correlations demonstrate the high accuracy of the system and method of the present disclosure by showing high correlation between all library pairs (between R²= 0.78 and R²= 0.98), including Encode-FloChIP pairs among which the highest correlation was obtained for the 100Ό00 cells samples, i.e. R²= 0.91.

After establishing 100' 000 cells as the optimal trade-off point between IP-lane saturation and data quality, the inventors set out to evaluate the reproducibility of the system and method of the present disclosure with other genomic targets. To this end, by using FloChIP's mode 1: "sample multiplex", the inventors ChIPed in parallel 5 histone marks (FI3K27ac, FI3K4me3, FI3K27me3, FI3K4mel and FI3K9me3), going from chromatin to sequencing-ready libraries, in just one day. By visual inspection of locus specific genomic regions, the inventors found that the obtained signal tracks closely resemble those of Encode (Fig. 2c). In addition, to evaluate FloChIP's performance more precisely, the inventors determined the extent of genome-wide correlation between FloChIP and Encode datasets. Comparison of signal intensities between the respective datasets confirmed an overall high genome-wide correlation (FI3K4me3: R²= 0.82, H3K27ac: R²= 0.88, H3K4mel: R²= 0.91, H3K27me3: R²= 0.56, H3K9me3: R²= 0.86 ; Fig. 8e).

Moreover, comparison in terms of the FRiP score showed that, despite the ChIP input for Encode being two orders of magnitude greater than that of FloChIP, the system and method of the present disclosure consistently yields highly enriched libraries, with FRiP scores between 1.07x and 4.12x higher for FloChIP compared to Encode (expect for FI3k27me3). These data show that FloChIP can be used to robustly generate chromatin landscapes for histone marks with a wide input dynamic range.

The system and method of the present disclosure advantageously allows immunoprecipitation from 10^s cells to as low as 500 cells as well as immunoprecipitation of different histone marks. The inventors obtain data with quality comparable to the benchmark ENCODE data with less cells and, in shorter times, in an automated and parallelized way. Thus, a system of high sensitivity, efficiency and multiplexing is assured.

The FloChIP "sequential-IP" mode or method of the system 1 of the present disclosure additionally provides genome-wide information on bivalent promoters. Conventional ChIP-seq provides information on the genome-wide localization of one specific protein or histone modification at a time. Flowever, DNA regulatory elements generally harbor the interaction of several transcription factors and histone modifications in order to regulate gene expression. For instance, it has been shown that promoters showing both repressive (FI3k27me3) and activating (FI3k4me3) marks are a characteristic feature in embryonic stem (ES) cells. This class of promoters have been originally named "bivalent" and are strongly associated to key developmental genes. In order to obtain direct information on the genomic location of bivalent promoters, a variant of the standard ChIP protocol called sequential-ChIP was developed.

Sequential-ChIP relies on the consecutive IP of two different antigens and, as opposed to simply intersecting two ChIP-seq datasets, provides unbiased information on bivalent regions. Despite the advantage of sequential ChIP over standard ChIP in discerning true bivalency, its manual involvement and impracticality have thus far prevented widespread usage. Moreover, due to the inefficiency of the method, few studies have so far performed sequential ChIP followed by next generation sequencing (sequential-ChIP-seq), since most of them relied on qPCR to validate putative bivalent regions (sequential ChIP-qPCR). To address the technical limitations of the current sequential-ChIP workflow, the inventors exploited FloChIP's intrinsic modularity, highly efficient IP and multiplexing features to derive the example of an automated and miniaturized sequential-ChIP solution (Fig. 9a). The inventors validated their method by focusing on bivalent chromatin given its well studied role in embryonic development. Specifically, the inventors acquired genome-wide direct co-occupancy profiles for H3K27me3 and H3K4me3 in mouse embryonic stem cells (mESCs) in both IP directions - i.e. H3K27me3 first followed by H3K4me3 (H3K27me3/H3K4me3) and vice versa.

As mentioned above, H3K4me3 and H3K27me3 bivalency has been originally attributed to promoters of developmental genes, leading to the hypothesis that a bivalent state maintains genes in a poised state. In a previous study, it has been suggested that, based on the promoter read coverage comparison of ChlP- seq and sequential-ChIP-seq data, promoters show three distinct patterns of bivalency, i.e. pseudo bivalency, partial bivalency and full bivalency. However, the inventors consider these classes an artificial construct that does reflect the more fine-grained distribution of bivalency levels. Therefore, instead of assigning promoters to specific classes, the inventors compute for each TSS a "bivalency score" (bvScore, Fig. 9c, detailed further below).

To evaluate the performance of sequential FloChIP, the inventors focused on three distinct regions which have been previously used as proof-of-concept models by Bernstein and colleagues using ChIP-qPCR and sequential ChIP-qPCR to illustrate the methylation status difference among 4me3-only (Tc/4 TSS), 27me3- only (upstream of Hoxa3) and bivalent (Irx2 TSS) regions. FloChIP-based genomic profiles (Fig. 9b, Fig. 13a) and bvScore distributions (Fig. 9c) validated these previous findings as the inventors observed that the Tc/4 promoter shows high H3K4me3 but low H3K27me3 enrichment, and thus very low bivalency (bvScore=0.83).

In contrast, Hoxa3 was mainly marked by H3K27me3, with low H3K4me3 and bivalency signals (bvScore=0.44). Finally, the TSS of Irx2 showed true bivalency (bvScore=3.34), with all four genomic tracks showing high coverage. In addition to considering specific loci, the inventors also validated their data on a genome-wide scale by achieving high correlation with the results obtained by Weiner et. Al²⁰ using their Co-ChIP system (Fig. 13b). The inventors' results are also consistent with a study by Mikkelsen and colleagues[21], who analyzed the genome-wide co-occurrence of H3K4me3 and H3K27me3 in mESCs by conventional ChIP-seq. Their findings suggested that, at the embryonic stage, most "high-CpG" (HPC) promoters, are associated with intervals of H3K4me3 enrichment, while the remaining ~22% appear to be bivalent. While the inventors analysis confirmed that the majority (80%) of HPC promoters is marked by H3K4me3, as reflected in Fig. 9c, the inventors found that the remaining 20% of promoters is bivalent (~15%) but also marked by mainly H3K27me3 (~5%) especially for very low bivalency scores.

Finally, as an independent validation of this analysis, the inventors performed gene ontology enrichment on the first one thousand promoters with the highest bivalency score. As expected, the inventors, found that these promoters are highly enriched in genes involved in a number of developmental processes, from anatomical structure development to neurogenesis (Fig. 9d).

Taken together, data indicates that FloChIP's "sequential IP" mode or method provides the first example of an automated, low-input (100' 000 cells) and rapid (between 5-6 hours) sequential-ChIP-seq workflow for the study of bivalent promoters.

The system and method of the present disclosure advantageously allows sequential ChIP-seq on histone marks. The obtained data shows a quality comparable to benchmark previously published data with less cells and, shorter times, in an automated and parallelized way and a system of high sensitivity, efficiency and multiplexing is provided.

The FloChIP system 1 is capable of ChIPing TFs in a "high-throughput" mode. As mentioned earlier, previous attempts at improving the sensitivity and multiplexing ability of ChIP-seq experiments were shown to perform well only in the context of histone modifications. The reason for this is that performing TF ChIP-seq poses additional challenges as compared to histone marks (FIM) including the fact that i) TF/DNA interactions are less abundant and less robust than HM/DNA interactions and ii) antibodies for TFs normally show lower affinity for their epitopes as compared to HM antibodies.

These challenges, whose severity varies on a case by case basis, translate into the need for greater sample inputs and longer incubation times. Indeed, the inventors also experienced these challenges and for most of the TF antibodies that were tested, FloChIP's indirect method - i.e. with 2-4 hours antibody/chromatin pre-incubation in tubes - appeared to be the only way to obtain high quality results. Nevertheless, by slowly and intermittently flowing the pre-incubated antibody/chromatin mixture on-chip, the inventors succeeded in performing TF immunoprecipitation on only 100Ό00 cells, proving for the first time the feasibility of miniaturized and automated TF ChIP-seq (Fig. 10).

After establishing a working protocol for TF ChIP-seq, the inventors set out to concurrently demonstrate the high-throughput capabilities of the system 1. To this end, by using half of the 64 IP lanes of the FloChIP device, the inventors performed MEF2-A ChIP-seq on chromatin derived from 32 different lymphoblastoid cell lines (LCLs) (Fig. 10a, A map of human genome variation from population-scale sequencing.)

Before sequencing, the inventors verified the immunoprecitation quality of each library by qPCR (Fig. 10b). Amplification results indicate consistently good fold enrichment across the 32 IP lanes (log2(Fold enrichment), mean=5.7, stdev=1.5). Another positive aspect of FloChIP is its apparent internal normalization effect on the immunoprecipitated DNA. To note, even without normalizing the input chromatin across the 32 cell lines, the amount of recovered DNA after FloChIP was extremely similar. In fact, all post-ChIP DNA libraries were amplified the same number of cycles (17 PCR cycles), adding to the convenience of the system and method of the present disclosure. The inventors attribute this particular feature to the fact the FloChIP provides very efficient IP and reaches saturation levels even with lowly abundant samples. When saturated, the microfluidic device 108 cannot IP any more chromatin and, given that the geometrical structure of each IP lane is the same, this leads to uniform amounts of DNA recovered even from very different samples.

Subsequently, the inventors sequenced at low coverage all 32 samples and observed variation in the percentage of uniquely mapped reads (Fig. 10c, mean=42%, stdev=22%), which in turn translated into variable genomic coverage of the libraries and variable number of peaks called (Fig. lOd and Fig. lOe). Despite this variable coverage, the inventors were able to assess the quality of the generated libraries through different criteria. By visual inspection, both genomic profiles and genome wide TSS annotation suggested that the obtained reads accumulated as expected near transcription start sites and that the respective regions and were clearly visible in the genome browser (Fig. lOd and Fig. 14a).

Moreover, in order to analyze the genome-wide agreement of the 32 datasets in an unbiased way, the inventors considered a set of peaks obtained by merging all the 32 alignment files together and counted the number of reads mapped within each peak of each library. This allowed to observe high pairwise correlation of the 32 libraries (Fig. 14b) as well as FRiP scores (Fig. lOf, mean=6.9%, stdev=2.7%) similar to the ones obtained for Encode with its own data (i.e 4.7%).

Finally, the inventors examined the enrichment of the MEF2-A motif for each set of peaks, generate individually from each sample. Despite the low number of peaks of some libraries, the expected motif is found in all libraries with a p-value lower 0.001 (Fig. lOf, -log(Pvalue), mean=9.1, stdev=6.9).

The interaction between DNA and proteins constitutes a fundamental aspect of gene regulation. ChIP-seq allows to probe DNA-protein interactions on a genome-wide scale, thus achieving high-throughput in terms of DNA sequence space coverage. On the other hand, for what concerns throughput in terms of proteins and biological samples, ChIP-seq remains at the lowest level possible, with only one protein species and one sample tested per experiment. Aggravating this aspect, the long and manually intensive protocol prevents straightforward development towards higher throughput. Community-led efforts like ENCODE have therefore been put in place in order to perform ChIP-seq for a large number of proteins and cell types. However, despite the valuable data generated, ENCODE still sampled only a small portion of a much larger combination space. In addition to limited throughput and manual involvement, standard ChIP-seq is also restricted by the input requirements for biological material. The requirement for at least one million cells, has precluded ChIP-seq from performing reliably on smaller but possibly biologically relevant samples. Understanding the impact of these limitations, several groups attempted to improve the original protocol. However, these attempts have addressed specifically certain issues while overlooking others. The system and method of the present disclosure address all major ChIP-seq limitations by introducing a new technology, FloChIP, that allows for rapid, high-throughput, automated and sensitive chromatin immunoprecipitation. The two core technological aspects of FloChIP are its surface chemistry and its microfluidic architecture. The former confers FloChIP the ability to perform solid-state bead-less IP with most off-the-shelf antibodies, while the latter provides the structural substrate for miniaturized IP, rapid washing, multiplexing and straightforward automation.

Following IP, another distinctive advantageous feature of FloChIP is the direct on-chip tagmentation of captured chromatin (Fig. 7a). On bead-bound chromatin, direct solid-state tagmentation reduces time, cost and input requirements of ChIP experiments. To the inventors knowledge, this is the first time solid- state tagmentation is shown to function efficiently and reproducibly on the walls on a microfluidic chip and, in general, on a substrate other than microbeads.

In order to demonstrate its reliability and applicability, the inventors performed FloChIP for a variety of targets and samples. Initially, the inventors aimed to empirically gain insights into FloChIP's dynamic input range. By obtaining high H3k27ac qPCR fold enrichment and high correlation with the respective ENCODE data for inputs ranging between 10^s and 500 cells, the inventors show that FloChIP can be used across a wide range of inputs.

Next, as a more comprehensive benchmark, the inventors performed FloChIP for four more histone marks, namely H3k27me3, H3k4me3, H3k4mel and H3k9me3. Despite the much lower input used for FloChIP, the results showed high correlation with ENCODE data and superior FRiP scores, thus advocating the robustness and efficiency of the system and method of the present disclosure.

Next, by designing pair-wise interconnect IP lanes, the inventors show that the chromatin eluted after the first IP step can be re-directed into a second IP lane, therefore achieving straightforward sequential immunoprecipitation. The inventors validated FloChIP's sequential IP by recapitulating previously published qPCR and sequencing data on bivalent promoters in mouse embryonic stem cells. To the best of the inventors knowledge, this is the fastest (1/2 day) and most sensitive (100Ό00 cells) example of sequential ChIP-seq. Moreover, this is the first automated, microfluidic and bead-less example of sequential ChIP-seq.

Finally, the inventors sought to simultaneously demonstrate FloChIP's applicability on TFs and throughput by ChIPping MEF-2A from 32 different lymphoblastoid cell lines.

Overall, it is demonstrated that the FloChIP system and method of the present disclosure are a robust, sensitive and high-throughput all-in-one ChIP-seq solution. Given its advantages and wide applicability, the FloChIP system 1 can be a widely adopted tool for the study of genome-wide protein-DNA interactions.

Further details of the exemplary methods and materials used to obtain the above results are now presented.

Chromatin preparation:

Cell fixation

GM 12878 cells (5-10 millions) were harvested, washed once with PBS and resuspended in 1ml crosslinking buffer (1% PFA in PBS) for 10 minutes shaking. Crosslinking was stopped by adding 50mI of 2.5M glycine and shaking for other 5 minutes. Fixed cells were then washed twice with ice-cold PBS, pelleted, deprived of the supernatant, snap frozen and stored and -80°C.

Lysis and sonication

The frozen cell pellet was resuspended in ice-cold PBS at 4°C agitating for 30 minutes, spun at lOOOg for 5 minutes, resuspended in lysis buffer (50 mM Flepes pH 7.8, 140 mM NaCI, ImM EDTA, 0.5% NP40, 10% glycerol, 0.25% Triton and freshly added protease inhibitor), incubated with mild agitation for 10 minutes, spun for 5 minutes at lOOOg, resuspended in nuclei wash buffer (20 mM T ris-HCI pH 8.0, 200 mM NaCI, 1 mM EDTA, 0.5 mM EGTA and freshly added protease inhibitor), incubated with mild agitation for 10 minutes, spun for 5 minutes at lOOOg and resuspended in sonication buffer (20 mM T ris-HCI pH 8.0, 200 mM NaCI, 1 mM EDTA, 0.5 mM EGTA, 0.5% Na-Deoxycholate, 0.5% N-laurosylsarcosine and freshly added protease inhibitor). Nuclei were sonicated on a covaris E220 machine with the following settings: MOW intensity, 5% duty factor and 200 bursts/cycle. Chromatin was then aliquoted ("Ί00Ό00 cells/aliquot) in PCR tubes and snap frozen until ChIP.

In an exemplary fabrication of the microfluidic device 180 of the system 1, microfluidic device designs were generated using Tanner L-Edit and fabricated using multilayer standard soft lithography (Thorsen, T., Maerkl, S. J. & Quake, S. R. Microfluidic large-scale integration. Science 298, 580-584, 2002) at the EPFL center for microtechnology. Briefly, designs were first transferred to chrome masks using a VPG200 pattern generator (Heidelberg Instruments). Subsequently, microfluidic molds were assembled on silicon wafers with SU8 photoresist for the control layer and AZ9260 positive resist for the flow layer using a SUSS ACS200 Gen3 system (SUSS MicroTec). Microfluidics chips were fabricated by first separately casting PDMS onto the SU8 and the AZ9260 wafers with two different PDMS/curing agent ratios (20:1 and 5:1, respectively), partially curing for 30 minutes at 80°C, peeling off the PDMS from the AZ9260 wafer and aligning it to the SU8 wafer in order the reconstitute the wanted pattern. The chips were finally fully cured at 80°C for one hour and half, peeled off, holed and plasma-bonded to clean glass slides or to PDMS- coated petri dishes.

Automated control of the FloChIP experimental workflow is obtained by a system of exemplary components including: 1) MATLAB software SW, 2) a standard laptop 101, 3) a WAGO fieldbus controller (ModBus 750-881) 102, 4) FESTO 3/2 way 24V miniature solenoid valves 104, 5) compressed air building supply (Fig lib) and 6) a PCR machine. Tygon tubing and western blot tips are used to interface the microfluidic chip 108 and the solenoid valves.

As previously explained, FloChIP is, in essence, a method consisting of the sequential introduction of different reagents into a custom-designed microfluidic chip 108. This sequence of reagents can be programmed with simple scripting commands which are, in turn, translated into sequences of solenoid valve actuations and releases. The concerted action of the solenoid valves, belonging to both the control layer and the flow layer of the chip 108, realizes in an automated fashion the required surface chemistry, immunoprecipitation and tagmentation reactions.

On-chip temperature control can be achieved by placing the microfluidic device 108 on top of a PCR machine with flat heat-block and starting a pre-programmed temperature sequence in sync with the MATLAB script.

In an exemplary operation, a FloChIP method or experiment starts by pre-loading the control lines with distilled water and activating all valves (at a pressure of 25-30 PSI for the control lines and 2.5-5 PSI for the flow valves). Subsequently, all the reagents for the surface chemistry (i.e. biotin-BSA, neutravidin, PBS and biotin-protein A/G, antibodies), IP (chromatin), washes (low-salt, high-salt and LiCL buffers), tagmentation (Tn5 buffer) and elution (SDS buffer), are loaded into pipette tips and inserted into the inlets of the microfluidic device 108. At this stage, all valves are closed and there is no possible cross-talk between any of the reagents above.

BO Immediately after completing the insertion of the tips, the automated protocol is launched by running the respective script. The exemplary protocol entails, in sequential order, the following steps: 20 minutes of BSA-biotin (2mg/ml), 30 seconds of PBS wash, 20 minutes of Neutravidin (lmg/ml), 30 seconds of PBS wash, 20 minutes of biotin-protein A/G (2mg/ml) and 30 seconds of PBS wash. Depending on whether direct or indirect ChIP is performed, immunoprecipitation is carried out in two different ways: loading the chromatin mix into the IP units pre-functionalized with antibodies (direct ChIP) or loading of the pre incubated antibody/chromatin mix (indirect ChIP).

For direct ChIP, following biotin-protein A/G, the antibody or antibodies of choice are loaded on chip for 20 minutes. Moreover, within direct ChIP, it is possible to operate the chip in two distinct multiplexing modes, either antibody multiplex, in which microvalves are actuated in such a way that every IP unit is functionalized with a different antibody, or sample multiplex, in which all IP units are functionalized with the same antibody (Fig.8). The antibodies use in these studies were Abeam antibodies: anti-FI3k27ac ab4729, anti-FI3k4me3 ab8580, anti-FI3k4mel ab8895, anti-FI3k9me3 ab8898, anti-FI3k27me3 ab6147; Santa-cruz antibodies: anti-PU.l sc-390405 and anti-MEF2a anti-MEF2A sc-17785. Following antibody loading and quick PBS wash, chromatin samples are loaded on chip by opening and closing the respective microvalves. These ON/OFF cycles, usually of 2 or 5 minutes, are performed in order to ensure that the chromatin spends enough time inside the micropillar array for the epitopes to be efficiently recognized by the corresponding antibody.

For indirect ChIP, the antibody and chromatin are incubated for 2 or 4 hours in a PCR tube prior the loading on-chip. During the IP step, the antibody/chromatin mixes are loaded into the chip in separate IP units by utilizing the same ON/OFF cycles as mentioned above.

Both for direct and indirect ChIP, the overall immunoprecipitation is performed at room temperature time spans between 30 and 60 minutes, depending on the amount of chromatin mix to be processed.

Following immunoprecipitation, rapid salt washes are performed to eliminate non-specific binding: 5 minutes of low-salt buffer (20 mM Tris pH 8.0, 150 mM NaCI, 2mM EDTA, 1% TritonX-100, 0.1% SDS), 5 minutes of high-salt buffer (20 mM Tris pH 8.0, 500 mM NaCI, 2mM EDTA, 1% TritonX-100, 0.1% SDS) and 5 minutes of LiCI buffer (20 mM Tris pH 8.0, 250 mM LiCI, 2mM EDTA, 1% NP40, 1% Na-Deoxycholate).

Following washes, Tn5 buffer (10 mM Tris pH 8.0, 5 mM MgCh) is slowly flown on-chip at 37°C for 45 minutes. This step ensures the complete tagmentation of the immunoprecipitated chromatin. Following Tn5 buffer and a 5-minutes low-salt wash to remove excess adapters, SDS buffer (10 mM Tris pH 8.0, 200 mM NaCI, ImM EDTA, 1% SDS) is loaded on-chip at 65°C for 10 minutes in order to elute the antibody-bound chromatin from the device. The eluate is independently collected from each IP lane into PCR tubes and decrosslinked at 65°C for 4 hours. Following decrosslinking, DNA is purified in Qiagen EB buffer using Qiagen MinElute purification kits.

In FloChIP operation for sequential ChIP, instead of eluting the chromatin in SDS buffer, elution is performed by saturating the antibody with a given elution peptide (abl342 for H3k4me3, abl782 for H3k27me3 and ab24404 for H3k27ac, Abeam - Peptide elution buffer: 20mI of IP buffer, 2pg of an antibody-specific peptide). This way, the eluted chromatin from a given IP lane can be directly re- immunoprecipitated in the subsequent IP lane. Following elution, the chromatin is collection into a western-blot tip inserted in the specific chip outlet. Subsequently, by closing the microvalves connecting the first IP lane and the outlets while opening the ones connecting the outlet and the second IP lane, the chromatin is re-flown on-chip for the second immunprecipitation (Fig 9). This second immunprecipitation is also performed using ON/OFF cycles of 2 minutes each. The total time for the second ChIP is also between 30 and 60 minutes. Finally, after all the chromatin has been re-flown on-chip, the salt washes are repeated, and elution is achieved using the standard SDS buffer.

ChIP-qPCR: Following FloChIP, qPCR was used to evaluate IP efficiency prior to next generation sequencing. qPCR was performed on a StepOnePlus™ (primer sequences: H3k27ac_FW CCACCCTGCACTT ACG ATG, H3k27ac_RV TGAGCTCCCTGTCTCTCCTC, H3k4me3_FW

CGGGGGCTGCCCAAAGTTTCA, H3k4me3_RV ATTGGGGAAATTGCAGAGCGAGC, H3k27me3_FW, H3k4mel_FW, H39me3_FW GTCCGGGTCTGACTGTCTTG, H3k27me3_RV, H3k4mel_RV, H39me3_RV ACTGCACTGGGTTCACGAAG). Each qPCR reaction was composed of 10mI Applied Biosystems™ PowerUp™ SYBR™ Green Master Mix, 0.8mI of a 10mM forward primer solution, 0.8mI of a 10mM reverse primer solution, 2mI of DNA and water to a final volume of 20mI. The cycling program was the following: 2 minutes at 50°C, 2 minutes at 95°C and [5 seconds at 95°C, 20 seconds at 60°C]x60 cycles. Fold enrichment values were obtained as ratios between the percent of input of the expected positive and negative regions genomic regions.

NGS Library preparation: NGS library were prepared by mixing 20mI of purified DNA with 2.5mI of forward Nextera adapter, 2.5mI of reverse Nextera adapter, 32.5mI of NebNext master mix, 0.5mI of lx SYBR green and water to 65mI. First, 5 pre-amplification cycles are run as follows: 5 minutes at 72°C, 30 seconds at 98°C and [10 seconds at 98°C, 30 seconds at 63°C, 60 seconds at 72°C]x5 cycles. Subsequently, 15mI out of the original 65mI are separated and amplified for 20 more cycles in order to estimate the optimal number of amplification cycles: 30 seconds at 98°C and [10 seconds at 98°C, 30 seconds at 63°C, 60 seconds at 72°C]x20 cycles. Finally, the remaining 50mI were amplified for N cycles, where N is the rounded-up Ct value determined in the previous reaction.

DNA was size selected using AMPure XP beads in order to obtain a size distribution between 150bp and 500bp. Concentrations were measure with Qubit (ThermoFisher), size distribution was profiled with Fragment analyzer (AATI) and libraries were sequenced on an lllumina NextSeq 500.

FloChIP reads mapping and processing:

Sequencing reads were mapped to the human (hg38 and hgl9) and mouse (mmlO) genomes using STAR (Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15-21 (2013)) with default parameters. Uniquely mapped reads were used to call peaks using FIOMER (Fleinz, S. et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol. Cell 38, 576-589 (2010)) command findPeaks.pl with the appropriate flag, i.e. -histone for histone marks and -factor for transcription factors. FRiP scores we calculated using FIOMER's command annotatePeaks.pl and divided the total number of reads that fall within peaks by the total number of mapped reads. Correlation plots were generated using annotatePeaks.pl on a common peak file, either Encode's peak files or, alternatively, the overlapping set of peaks between Encode and FloChIP datasets.

Bivalency score calculation:

The bvScore is assigned to each promoter and is intended to take into account both the intersection between two ChIP-seq datasets as well as the agreement between the respective sequential-ChIP-seq datasets. Accordingly, the bvScore can be expressed as the product of the co-occurrence score (cScore), which measures the relative coverage of the two ChIP-seq tracks, and the agreement score (aScore), which measures the relative coverage of the two sequential-ChIP-seq tracks. We define the cScore as (nmr4i+nmr27i)/(nmr4i-nmr27i), where nmr₄, and nmr27i are the normalized number of mapped reads in promoter / for FI3K4me3 and FI3K27me3, respectively. The higher the value of the cScore for a promoter, the more similar is the occupancy of the two marks on that promoter. A positive cScore indicates prevalence of FI3K4me3 while a negative cScore indicates prevalence of FI3K27me3. We define the aScore as the absolute value of (nmr₄/_27j+nmr₂₇/_4i)/(nmr₄/₂₇rnmr₂₇/_4i), where nmr₄/_27i and nmr₂₇/₄,· are the number of mapped reads in promoter / for the two sequential-ChIP-seq experiments. The higher the aScore of a promoter, the more similar is the coverage of the two sequential-ChIP-seq datasets on that promoter. Finally, the bivalency score is thus defined as bvScore=abs(cScore* aScore) or equivalently bvScore=log ( abs (cScore * a Score)) . Gene ontology analysis was performed using the online tool

enrichment-anaivsis.

While the invention has been disclosed with reference to certain preferred embodiments, numerous modifications, alterations, and changes to the described embodiments, and equivalents thereof, are possible without departing from the sphere and scope of the invention. The features of any one of the described embodiments may be included in any other of the described embodiments. The methods steps are not necessary carried out in the exact order presented above and can be carried out in a different order. Accordingly, it is intended that the invention not be limited to the described embodiments, and be given the broadest reasonable interpretation in accordance with the language of the appended claims.

REFERENCES

1. Johnson, DS; Mortazavi, A; et al. "Genome-wide mapping of in vivo protein-DNA interactions".

Science. 316: 1497-1502, (2007).

2. Daria Shlyueva et al. "Genome-Wide Ultrabithorax Binding Analysis Reveals Highly Targeted Genomic Loci at Developmental Regulators and a Potential Connection to Polycomb-Mediated Regulation" Plos One, doi.org/10.1371, (2016).

3. Niu Wl, Lu ZJ, Zhong M, et al. "Diverse transcription factor binding features revealed by genome wide ChIP-seq in C. elegans" doi: 10.1101/gr.ll4587.110. Genome Res. (2011)

4. The ENCODE Project Consortium. "An integrated encyclopedia of DNA elements in the human genome" Nature 489, 57-74, (2012).

5. van Galen, P. et al. "A multiplexed system for quantitative comparisons of chromatin landscapes".

Mol. Cell 61, 170-180 (2016).

6. Julie Brind'Amour et al., "An ultra-low-input native ChIP-seq protocol for genome-wide profiling of rare cell populations" Nature Communications. doi:10.1038/ncomms7033, (2015).

7. Lara-Astiaso D et al. Immunogenetics. Chromatin state dynamics during blood formation. Science.

2014 Aug 22;345(6199):943-9. doi: 10.1126/science.l256271.

8. Landt SG, et al. "ChIP-seq guidelines and practices used by the ENCODE and modENCODE consortia". Genome research;22(9) doi: 10.1101 (2012)

9. https y/www.dlagenod&.com/en/p/sx-g^g-jp-star-compact-automated-systeni-l-unjt

10. Peter J. Park. "ChIP-Seq: advantages and challenges of a maturing technology", Nat Rev Genet doi: 10.1038/nrg2641 (2009)

11. Zhenning Cao et al. A microfluidic device for epigenomic profiling using 100 cells. Nature Methods 12, 959-962 (2015) doi:10.1038/nmeth.3488

3^· 12. Christian Schmidl, et al. "ChIPmentation: fast, robust, low-input ChIP-seq for histones and transcription factors", Nat Methods. doi:10.1038/nmeth.3542 (2015)

13. Unger MA, Chou HP, Thorsen T, Scherer A, Quake SR, "Monolithic Microfabricated Valves and Pumps by Multilayer Soft Lithography", Science 288: 113-116 (2000).

14. Gasper, VV. C. et al. Fully automated high-throughput chromatin immunoprecipitation for ChIP- seq: Identifying ChIP-quality p300 monoclonal antibodies. Sci. Rep. 4, 5152 (2014).

15. Aldridge, S. et al. AHT-ChIP-seq: a completely automated robotic protocol for high-throughput chromatin immunoprecipitation. Genome Biol., 14, p. R124 (2013).

16. Van Galen, P. et ai. A multiplexed system for quantitative comparisons of chromatin landscapes. Mol. Cell 61, 170-180 (2015).

17. Chabbert, C. D. et al. A high-throughput ChiP-Seq for large-scale chromatin studies. Moi. Syst.

Biol. 11, 777 (2015).

18. Ma, S. et al. Low-input and multiplexed microfluidic assay reveals epigenomic variation across cerebellum and prefrontal cortex. Science Advances. 4, 4 (2018).

19. Rotem, A. et ai. Single-cell ChIP-seq reveals cell suhpopu!ations defined by chromatin state. Nat. Biotechnol. 33, 1165-1172 (2015).

20. Weiner, A. et al. Co-ChIP enables genome-wide mapping of histone mark co-occurrence at single-molecule resolution. Nat. Biotechnol. 34, 953-961 (2016).

21. Mikkelsen, T. S. et al. Genome-wide maps of chromatin state in pluripotent and lineage- committed cells. Nature 448, 553-560 (2007).

PI. US 20150105287 Al. Microfluidic Systems and Methods for Chromatin Immunoprecipitation (ChIP). Chang Lu, Zhenning Cao

P2. WO 2017025594 Al. Methods for studying nucleic acids. Christoph Bock, Christian SCHMIDL

P3. WO 2001001025 A2. Quake Microfabricated elastomeric valve and pump systems Marc A.

Unger, Hou-Pu Chou, Todd A. Thorsen, Axel Scherer, Stephen R.

The entire contents of each of the above references are herewith incorporated by reference.

Claims

1. System (1) for immunoprecipitation of a protein-DNA complex or for capture of a protein-DNA complex including:

- a microfluidic device (108) including a plurality of antibody units (111, 201), wherein each antibody unit (111,201) includes at least one microfluidic channel (MFC), wherein the at least one microfluidic channel (MFC) includes at least one surface configured to receive a multi-layered assembly (MLA) of biochemical species for capturing a target molecule to form a biofunctionalized surface;

- at least one inlet for providing at least one sample containing DNA binding proteins to the microfluidic device (108); and

- at least one outlet for providing DNA or eluted DNA for sequencing.

2. System (1) according to the previous claim, wherein the at least one microfluidic channel (MFC) includes the multi-layered assembly (MLA) of biochemical species for capturing a target molecule provided on or attached to the least one microfluidic channel surface to form the biofunctionalized surface.

3. System (1) according to the previous claim 1 or 2, wherein the at least one microfluidic channel (MFC) includes an array of elongated structures (MP).

4. System (1) according to the previous claim, wherein the array of elongated structures (MP) is composed of a repeating pattern (304a, 304b) of elongated structures (MP), wherein each elongated structure (MP) includes at least one surface configured to receive a multi-layered assembly (MLA) of biochemical species for capturing a target molecule provided on or attached to the least one surface.

5. System (1) according to the previous claim, wherein each elongated structure (MP) includes at least one surface and a multi-layered assembly (MLA) of biochemical species for capturing a target molecule provided on or attached to the least one biofunctionalized surface.

6. System (1) according to any one of the previous claims, wherein the array of elongated structures (MP) defines a plurality of microfluidic canals or passages (MCA) defined between adjacent elongated structures (MP).

7. System (1) according to any one of the previous claims, wherein the elongated structures (MP) extend from at least one surface of the microfluidic channel (MFC) and extend inside the at least one microfluidic channel (MFC).

8. System (1) according to any one of the previous claims, wherein the elongated structure (MP) is or defines a micropillar.

9. System (1) according to any one of the previous claims, wherein the elongated structure (MP) has or defines a rhomboidal (304a), squared (304b), or circular cross-section.

10. System (1) according to any one of the previous claims, wherein the at least one microfluidic channel (MFC) is elongated structure-less.

11. System (1) according to any one of the previous claims, wherein the at least one microfluidic channel (MFC) and/or the elongated structures (MP) includes a hydrophobic surface configured to receive a multi-layered assembly (MLA) of biochemical species.

12. System (1) according to any one of the previous claims, wherein the at least one microfluidic channel (MFC) and/or the elongated structures (MP) include or consist solely of a polymer, or include or consists of at least one polymer surface.

13. System (1) according to the previous claim, wherein the least one polymer surface is biofunctionalized and includes at least one antibody and/or at least one molecular capturing agent attached thereto.

14. System (1) according to the previous claim, wherein the antibody is specific to a DNA binding protein.

15. System (1) according to any one of the previous claim 12 to 14, wherein the at least one microfluidic channel (MFC) and/or the elongated structures (MP) include, attached to or provided on the least one polymer surface, a first layer (401) assembled by deposition of a first protein species to the at least one polymer surface of the microfluidic channel (M FC) and/or the elongated structures (MP).

16. System (1) according to any one of the the previous claims 11 to 15, wherein the first layer (401) is assembled by adsorption of a first protein species to the at least one hydrophobic or hydrophilic surface or the at least one polymer surface.

17. System (1) according to the previous claim, wherein the polymer is a hydrophobic polymer or polydimethylsiloxane (PDMS), and/or the first protein species is BSA or biotinylated-BSA.

18. System (1) according to any one of the previous claims 16 to 17, further including a second layer

(402) composed of a second protein species having high biophysical affinity to the first protein species.

19. System (1) according to the previous claim, wherein the second protein species is a protein of the avidin family or a glutaraldehyde.

20. System (1) according to any one of the previous claims 18 to 19, further including a third layer

(403) composed of a third protein species with high biophysical affinity to the second layer or assembly (402), or an antibody or biotinylated antibody.

21. System (1) according to the previous claim, wherein the third protein species is a protein A or protein G or protein A/G, or a biotinylated protein A or protein G or protein A/G.

22. System (1) according to any one of the previous claims 19 to 20, further including a fourth layer

(404) composed of an antibody or capturing agent.

23. System (1) according to any one of the previous claims 16 to 22, wherein the first (401), second (402), third (403), and/or fourth (404) layers are assembled by flowing a concentrated protein solution through the at least one microfluidic channel of the antibody unit (111, 201).

24. System (1) according to any one of the previous claims, wherein the multi-layered assembly is configured to target any molecule to which a specific antibody can bind.

25. System (1) according to the previous claim, wherein the multi-layered assembly is configured to target protein/DNA complexes, transcription factors, histone modifications or DNA methylation.

26. System (1) according to anyone of the previous claims, wherein the system is a bead-less system or a magnetic bead-less system.

27. System (1) according to anyone of the previous claims, wherein the system (1) is a system for chromatin immunoprecipitation.

28. System (1) according to anyone of the previous claims, wherein the at least one inlet is configured for providing a cross-linked sheared or sonicated chromatin to the microfluidic device (108).

29. System (1) according to anyone of the previous claims, wherein each antibody unit (111) is configured to capture a different target molecule or to carry out a different immunoprecipitation reaction.

30. System (1) according to anyone of the previous claims, wherein the system (1) is configured to collect a captured target molecule and provide the captured target molecule to a different antibody unit (111) to carry out a sequential immunoprecipitation.

31. System (1) according to anyone of the previous claims, wherein the microfluidic device (108) includes at most 7425 microfluidic canals or passages (MCA), or at most 4400 elongated structures.

32. System (1) according to anyone of the previous claims, wherein each antibody unit (111) includes at most 135 microfluidic canals or passages (MCA), or at most 80 elongated structures (MP).

33. System (1) according to anyone of the previous claims, wherein each microfluidic canal or passage (MCA) has a channel height between 5pm and 50pm, and/or a channel width between 5pm and 50pm, and/or a channel length between 5pm and 50pm; and/or wherein each elongated structure (MP) has a height between 5pm and 50pm, and/or a width between 5pm and 50pm, and/or a length between 5pm and 50pm.

34. System (1) according to anyone of the previous claims, wherein each antibody unit (111) defines a surface area to volume ratio of between 0.5 micrometer ¹ and 5 micrometer ¹, or each microfluidic channel (MFC) defines a surface area to volume ratio of between 0.5 micrometer ¹ and 5 micrometer ¹.

35. System (1) according to anyone of the previous claims, wherein the at least one microfluidic channel (MFC) defines a cavity having a substantially square or rectangular cross-sectional profile, or honeycomb or circular cross-sectional profile.

36. System (1) according to anyone of the previous claims, wherein the pattern comprises a plurality of elongated structures (MP) disposed as a matrix or grid array.

37. System (1) according to anyone of the previous claims, further including:

a calculator or calculating means (101);

a plurality of electromechanical (EM) valves (104) configured to allow or block the passage of a pressurized fluid to the microfluidic device (108) to provide samples to be processed to the microfluidic device (108); a microcontroller unit (102) connected to the calculator or calculating means (101) and to the plurality of electromechanical (EM) valves (104), the microcontroller unit (102) being configured to action the plurality of electromechanical (EM) valves (104); and

inlet collection tubes (106) for holding samples to be processed, the collection tubes being connected to a plurality of EM valves (104) and to the microfluidic device (108).

38. System (1) according to the previous claims, further including outlet collection tubes (109) connected to the microfluidic device (108) for collecting processed samples from the microfluidic device (108).

39. System (1) according to any one of the previous claims, wherein the system (1) is an immunoprecipitation system configured to determine protein/DNA interactions in vivo.

40. System (1) according to any one of the previous claims, wherein the system (1) is configured to immunoprecipitate or capture a protein-DNA complex

41. System (1) according to any one of the previous claims, wherein the system (1) is a chromatin immunoprecipitation system.

42. Method for immunoprecipitation of a protein-DNA complex or for capture of a protein-DNA complex including the steps of:

- providing the system (1) according to any one of the previous claims;

- providing at least one sample containing DNA binding proteins to the microfluidic device (108); and

- recuperating, at the least one outlet (110), DNA or eluted DNA for sequencing.

43. Method according to the previous claim, wherein the method is a chromatin immunoprecipitation method.

44. Method according to the previous claim 42 or 43, wherein the providing step includes providing a cross-linked sheared or sonicated chromatin to the microfluidic device (108).

45. Method according to the previous claim, further including the step of:

- preparing the DNA for sequencing by providing a transposase loaded with NGS adapters in the microfluidic chip (108) for attachment to the DNA.

46. Method according to the previous claim, further including the step of:

- eluting the DNA in the microfluidic chip (108).

47. Method according to the previous claim, wherein eluting the DNA in the microfluidic chip (108) is carried out by applying thermal energy and a salt buffer.

48. Method according to the previous claim 40 or 41, further including the step of:

- carrying out sequencing of the DNA.

49. Method according to the previous claim, wherein sequencing is carried out using next generation sequencing (NGS) processing.

50. Method according to any one of the previous claims 42 to 49, further including the step of:

- carrying out on-chip library indexing to decrease downstream library preparation time.

51. Method according to the previous claim wherein carrying out on-chip library indexing to decrease downstream library preparation time comprises flowing, in the microfluidic device (108), a transposase loaded with NGS adapters on top of the immunoprecipitated chromatin.

52. Method according to any one of the previous claims 44 to 51, wherein in the step of providing a sheared or sonicated chromatin to the microfluidic device (108), samples containing less than 1000, or less than 500 cells are provided.

53. Method for fabricating a system for immunoprecipitation of a protein-DNA complex or for capture of a protein-DNA complex according to any one of claims 1 to 41 including the steps of:

- providing the system according to any one of the previous claims 1 to 40 wherein the at least one microfluidic channel (MFC) and/or elongated structures (MP) are non-biofunctionalized; and

- flowing a concentrated protein solution through the at least one microfluidic channel (MFC) of the antibody unit or units (111, 201) to assemble a plurality of layers (401, 402, 403, 403) on at least one surface of the at least one microfluidic channel (MFC) and/or the elongated structures (MP) to biofunctionalize the at least one microfluidic channel (MFC) and/or the elongated structures (MP).

54. Apparatus configured to carry out the method of any one of the previous claims 42 to 53.