Final Report Summary - ATLAS (Development of laser-based technologies and prototype instruments for genome-wide chromatin immunoprecipitation analyses)
Executive summary:
The knowledge of mechanisms leading the interactions among bio-molecules in living cells represents one of the main goal in molecular biology in order to define the dynamics of (direct and indirect) bindings. Currently, the establishment of a stable inter-play between nucleic acid and proteins (in particular DNA proteins crosslink) is mainly obtained through the conventional chemical methods involving the use of a bifuctional reagent, for instance, the formaldehyde.
Protein-Deoxyribonucleic acid (DNA) interactions play an important role in DNA replication, recombination, repair and, consequently, in transcriptional and translational gene regulation. The modulation of chromatin structure is a complex and dynamic process regulated at multiple levels though distinct mechanisms such as histone posttranslational modifications, non-coding RNA and DNA methylation. Aberrations of gene regulation might lie in pathological chromatin status. The establishment of a standing covalent bond between proteins and nucleic acids (crosslinking) open access to the study of interactions between bio-molecules: this is a crucial task for understanding functions and deregulation of gene expression. Enzymatic methods to examine protein-DNA interactions, have been developed in vivo and in vitro though genomic foot-printing. Recently, Chromatin immunoprecipitation (ChIP) assay allows to identify both the binding patterns between transcription factors and chromatin and to evaluate the occurrence of histone modifications. Despite the needs, the current ChIP technology does not allow to discriminate either direct or indirect binding or to study transient chromatin occupancy. Indeed specific bond of transcription factors causes in a large part the connectivity of gene regulatory networks as well as the quantitative level of gene expression. In order to satisfy these requirements, a new and more efficient crosslink reaction, based on the employment of an UV laser source, was developed. The use of UV laser crosslinking thus represents an innovative way to create a stable covalent bond though the excitation of the electronic state of proteins and DNA. Ultraviolet (UV) irradiation creates covalent bindings between the reactive groups of DNA (thymine and cysteine) and amino acids (serine, methionine, lysine, arginine, histidine, tryptophan, phenylalanine or tyrosine).
Therefore, the occurrence of a crosslink is accompanied by a combination of factors, including the inherent photo-reactivity of the excited nucleotides, the geometrical arrangement and the molecular dynamics involved in the mechanism of crosslink. UV light is a zero length crosslinker and produces less perturbation of the complex than chemical crosslinker such as formaldehyde. UV Laser-based chromatin immunoprecipitation (LChIP) induces photo-mediated crosslink in a very short time allowing the study of transient interactions by varying different parameters such as energy, repetition rate and pulse intensity in time unit. UV irradiation of cells at the wavelength of about 260 nm produces covalent bonds in nucleic acids and proteins and, in particular, may preferentially allow bonds between TFs and histones associated within chromatin.
So, LChIP is able to characterise the dynamics of the transcription factors binding on chromatin in living cells because the required time for Laser-mediated crosslink is several orders of magnitude lower than the conventional methods. This technique makes available the study of temporal and spatial bindings of proteins on DNA and so, it becomes a useful tool for understanding regulation of gene expression and maintenance. Although living human cells non-linearly respond to irradiation with intense fs-UV-laser pulses, among the phenomena triggered by such pulses, a highly efficient DNA-proteins relationship occurs.
Definitively, the development of LChIP technique and its applications have the powerful potential in detecting the behaviour of transient DNA-proteins bindings in vivo. The study of these transient interactions in a short time scale, a parameter undetectable with conventional methodologies, is now possible corroborating the extraordinary biomedical potential of LChIP for discriminating chromatin epigenome and TF bindings dynamically.
Project context and objectives:
The availability of the DNA sequence of many eukaryotic genomes and the generation of high density tiling arrays covering such entire genomes has made it possible to decipher the regulatory principles that are based on the interplay of trans-regulatory TFs and their cognate cis-regulatory DNA recognition sequences at genome-wide level. At chromatin level a mutual interplay between transcription factors and epigenetic modifiers (DNA and histone modifying machinaries) sets up determinants of gene (in)activity. The ensemble of histone modifications at a given gene locus has been proposed to establish an epigenetic code of great complexity. Irrespectively of whether such a code does exist or whether chromatin modifications rather constitute a step in signal transduction, it has become increasingly clear that chromatin modifications constitute docking sites for regulatory factors. Thus, description of the information encoded by genomes requires a deconvolution of the genetic and epigenetic programs and the interplay between these two regulatory levels. In addition to their enormous potential and power for the study of gene regulation mechanisms such analyses also provide important tools for diagnosis, prognosis, and therapy of diseases. Decoding chromatinembedded information at the whole genome level by combining ChIP with global analysis such as DNA tiling arrays (ChIP-chip or ChIP-on-chip) or parallel single molecule sequencing, ChIP-seq, have become powerful approaches currently applied to mammalian genomes to analyse gene regulation programmes.
Global ChIP analyses allow converting genomic information into a dynamic regulatory network that operates in a time, cell, development and environmentdependent manner to coordinate cell homeostasis, proliferation, survival and death, as well as cell-cell communication. ChIP is a powerful approach to determine in vivo the chronology of transcription factor recruitment during activation or repression in the context of chromatin and changes in epigenetic signatures and chromatin remodeling. In many applications, protein and DNA are cross-linked using formaldehyde, chromatin is fragmented, and the protein of interest is immunoprecipitated with specific antibodies (XChIP). Alternatively, native ChIP (NChIP) i.e. without chemical crosslinkers, can be used in epigenetic profiling studies. The relative amount of a particular DNA fragment cross-linked to the protein (and therefore present in the precipitate) is determined by qPCR, and is a measure of the occupancy of the factor at that particular position in the genome. Despite providing significant insight in transcription regulation and chromatin-mediated effects, NChIP and XChIP technologies have a number of intrinsic technical limitations. These comprise:
- Selectivity: Formaldehyde introduces covalent bonds between protein-DNA and protein-protein. TFs are generally embedded within complexes/machineries that display multiple interaction surfaces (with DNA or between subunits). This has two consequences: (i) factors that bind DNA directly or indirectly via proteinprotein interactions with chromatin components will be crosslinked with widely different efficiencies depending on the stability of the interaction and (ii) a single protein/complex can be crosslinked to more than one site in the genome e.g. in the case of enhancer-promoter looping. NChIP can only be employed efficiently for very stable interactions, predominantly histone (nucleosome)-DNA interactions.
- Accuracy and efficiency: The overall performance of XChIP depends critically on the efficiency of the first step, the crosslink of factors to DNA by formaldehyde exposure of living cells. However, chemical crosslinking is diffusion-controlled and varies considerably with cell type, manipulation, storage and purity of the chemical. Due to the progressively increasing non-specific crosslink by chemical crosslinkers, formaldehyde crosslinking has to be stopped a long time before a maximal crosslinking of DNA to protein has been reached. This implies low efficiency of the subsequent immunoprecipitation.
- Dynamics and half-lives of DNA interactions: TFs bind with highly different kinetics/half-lives to DNA / chromatin. Current technologies do not allow studying very short-live interactions. It is, for example, not possible to resolve the DNA / chromatin scanningmodel of a transcription factor as opposed to a direct recruitment model. Data obtained in XChIP (co)factor binding studies e.g. on the time-resolved Estrogen Receptor (ER) activation poorly correlates with photo-bleaching experiments resulting in a controversy about the actual molecular mechanisms of TF action. Sensitivity: Conventional XChIP requires large amounts of cells (about 10^6 to 10^8 cells) as starting material. A significantly higher sensitivity (down to 1 000 cells) has been reported for carrier ChIP (CChIP) that involves NChIP procedure for the analysis of histone modification. However, this procedure is not applicable to TF interactions with DNA/chromatin.
- Single cell, tissue crosslinking and sorting: Current ChIP technologies do not permit crosslink of selected individual cells or frozen tissue slides and subsequent analyses of defined population. Epitope masking and modification: Formaldehyde crosslinking alters lysine residues, which can be part of antibody epitopes in targeted DNA binding proteins, particularly in modified histones. This has several serious drawbacks: (i) it reduces substantially the IP efficacy in epigenetic studies and yields <<1% are commonly observed, (ii) due to the covalent stabilisation of entire complexes, antibody epitopes may reside inside a complex, which would become accessible due to dissociation if only DNA crosslinks would occur, (iii) pitope-tagging is limited to peptides that are not modified by formaldehyde.
To overcome most of the above mentioned limitations, ATLAS consortium established novel ChIP technologies (LChIP) consisting in the development and subsequent creation of an experimental platform that, through the employing of a laser source, is able to induce cross-link reactions between DNA and proteins within living cells to study the evolution in time and space of both the transcriptional machinery and epigenetic code. Laser technology, therefore, represents a valid alternative to traditional methods, as it is able to induce bonds between DNA and proteins in very short intervals of time and with high efficiency. The driving idea of ATLAS is the observation, confirmed by experimental data and evidences in the literature, that an ultrashort UV laser can excite the electronic states of nitrogenous bases composing the DNA and amino acid side residues of proteins that bind DNA itself. Once absorbed energy, the system DNA-protein relaxes to the ground state via the formation of a covalent bond. The creation of this link, obtained by a method 'fotophysic', opens up a range of possibilities for study of interaction in 'real time' and rapid decoding of 'cross talk' between transcription factors, enhancers, epigenetic modulators and DNA. The reducing time scale of crosslink induction increases the number of proteins-DNA interaction that can be analysed by 'omic' science, and in general helps to clarify the temporal sequence of the events constituting the biological phenomenon (for example the epigenetic code variation). Comparing the conventional method with LChIP some drawbacks might be overcome. So, the novel LChIP technology identifies direct protein-DNA contacts preferentially if not exclusively under certain conditions (increasing of selectivity).
LChIP involves ultra-fast physical crosslinking by femtosecond UV lasers specifically designed for highly efficient DNA-protein crosslinking. Once calibrated it operates fully reproducible and highly accurate, and is cell and experimenter independent. Due to its ultra-fast crosslinking, LChIP allows fundamental studies on the mechanisms of DNA/chromatin recognition by DNA-binding regulatory factors. Given to its high efficiency documented in proof-of-principle experiments (see sections below), LChIP linked to microscopic and cell-sorting platforms using microfluidic systems permits to photo crosslink subsets of cells on-the-fly, crosslink cells in particular phases of the cell cycle, or leukemia blasts for genomewide analyses. In addition, LChIP does not affect protein epitopes and hence has a dramatically higher IP efficacy allowing the use of epitope-tagging approaches with highly efficient antibodies independent of the presence of lysines. To reach these aims industrial partners, mathematicians, physicists, molecular biologists, chemists and MDs worked together up to construct a prototype of a complex device, including Laser apparatus and microfluidic station, to perform induction of crosslink in living cells in an easy and not time consuming manner. The prototype is additionally governed by software, made for the ATLAS purpose, with a friendly interface and it easy to understand. In order to realise the prototype, the efforts of ATLAS project were centred on two main different research lines:
- A technological one in which the construction of dedicated femtosecond double-pulse, two colours tunable laser apparatus was made, as such as the development of microfluidic and cell sorting machinery (see first part of S&T results / foreground description).
- An experimental one aimed to discover the chemical principles and mechanism of photo-crosslink and the development and validation of LChIP in cell-based assays (see second part of S&T results / foreground description).
The step-by-step progression of the ATLAS project was allowed through a series of objectives achievement. In the list below the most important ones are summarised:
- Development and validation of a custom femtosecond laser prototype for optimised UV-laser photo-induced crosslinking with high efficiency. In the first 12 months of project, a new dedicated ultrashort laser system has been developed that is be able to deliver high-energy, UV, fs-pulses (in the energy range of several microjoule per pulse) at a repetition rate variable within six orders of magnitude, between 1Hz and 1MHz. The wavelength of such pulses is variable in the range 250-280 nm, to match the first UV absorption band range of DNA bases. The combined features of high energy / pulse and high repetition rate were not commercially available at the moment of ATLAS project start and until now, for any known ultrashort laser source. This is so the first prototype released in the context of ATLAS project. Its realisation involved the collaboration of industrial partners specialised in UV-laser source construction and commercialisation as such as some university partners devoted to the management, control and improvement of Laser apparatus.
- Development and implementation of laser system upgrades with dedicated Optical parametric amplifier (OPA) that allows performing two-colour, double pulse irradiation of the sample, thereby significantly minimising cell damage after the irradiation of biomolecules or living systems with UV laser source (however ensuring and optimising DNA-protein crosslink). The developed OPA has unique features in terms of high energy/pulse together with the high pulse repetition rate and has a wide tunability, covering the 200 nm 2600 nm spectral window. The OPA technology, although commercially existent, has been applied for the first time to laser apparatus described above, expanding much the possible wavelengths window.
- Determination of the chemical nature of laser-induced DNA-protein cross-linking, using the synergy between theoretical calculations and study of increasingly complex chemical models, was carried out. Factors effecting the excitation of the DNA and the protein, the formation of intermediate reactive species, and the creation of the new DNA-protein bonds have been addressed in a sequential manner, from the isolated building blocks of each macromolecule to reacting adjacent monomers and the long-range effects that can be found in larger systems with tertiary structure, in particular when proximity (and affinity) is enforced. This multi-level detailed description of the photochemical behaviour of DNA in a protein environment has provided the knowledge for rational design and optimisation of the crosslinking experimental conditions in cell based assays.
- Evaluation of macroscopic parameters characterising the DNA-laser interaction has been carried out in cell based assays in order to indicate the ones (such as time, geometry, cell behaviour) that mainly affect the induction of crosslink between DNA-proteins in living cells. The final benefit has consisted in indicating the values or the restricted range of values for an optimal interaction between the laser and the molecular system.
- First results of LChIP were been obtained on single gene analysis and then the skills arising from these experiments were applied to perform genomewide studies of LChIP (LChIP-seq).
- The feasibility of LChIP was estimated by looking at particular genomic regions and well-defined transcriptional factors (or enhancers of transcription) to better decipher the evolution in time and space of transcriptional pathways and networks and the difference between direct and indirect binding of proteins to DNA.
- Integration of LChIP with microfluidic-based cell sorting platforms was obtained. Dedicated laser system has been integrated into the fluorescence-activated microfluidic cell sorter. This setup allows analysis of selective subsets using much lower number of cells than required for conventional approaches. Moreover, this approach allows the dynamic crosslinking of cells in a flux and number dependent manner.
- Application of LChIP with i) microscopic manipulations of frozen tissue slices; ii) microfluidic station for experiments in cellular subpopulations; iii) global study of epi-modifications and TFs binding in time resolved manner upon selected treatments (f.e. Epi-drugs) was performed.
- Optimisation of amplification methods of very small amounts of DNA obtained by LChIP (#6), allowing the use of low numbers of cells as starting material, has been done. In parallel, another industrial partner carried out the development and characterisation of antibodies to obtain LChILL and LChIP grade antibodies; in particular emphasis was on the production of antibodies against epitope peptides that will be masked by formaldehyde in order to improve the efficiency of IP stage. At the same manner, kit(s) for LChIP (and for DNA amplification for LChIP) was developed. The LChIP contains controls such as LChIP grade antibodies and physically cross-linked cells as well as optimised buffers, PCR primers and adapted protocol. As described previously, in the context of ATLAS project, the borders among different scientific fields have been overcome, by joining Physics and Bio-medical approaches to answer to a typical biological question (such as 'what binds what in a cellular system') and to clarify the principles underpinning the interaction between biomolecules and UV radiation. The multidisciplinary aspect characterised strongly this project and maybe it was the keystone to develop the new LChIP technique.
ATLAS integrated dispersed capabilities of partners from 8 European countries and assembled the critical mass required to enable new global approaches, by networking the necessary expertise to secure European excellence and competitiveness, and to explore new directions in the research field. Consortium delivers new knowledge on basic biological processes relevant to health and disease. With the new LChIp is now possible to better identify:
- Selective factor-DNA crosslink repertories. Using ATLAS technology non-specific protein-protein crosslink is negligible (under appropriate conditions) allowing, for the first time, the study of bona fide pure bindings to chromatin with unmet accuracy. Consequently, highly precise transcription factor (or epigenetic modification) maps can be generated, potentially reflecting dynamics and kinetics of bindings. Comparison with conventional chemical crosslink facilitates to distinguish between direct and indirect (through protein-protein contacts) DNA binding. Such information is of extraordinary mechanistic value.
- Novel strategy of experimental design-important mechanistic question can be addressed. The ATLAS technology could potentially change the actual view of promoter function and gene regulation, leading to novel concepts of gene expression. Indeed, the comparison with data from formaldehyde crosslinking will most likely lead to novel classification/distinction of binding sites in enhancers versus promoters and/or to a completely revised view of enhancer and promoter function. Moreover, the efficiency and specificity of this approach allows to define cellular context specific regulation in a dynamic time frame, which is currently unthinkable due to technical limitations. The applications of this technology most likely open a new strategy of experimental design with very wide applicabilities.
Project results:
The overall objective of the ATLAS consortium was to develop novel types of femtosecond (fs) tunable UV lasers to induce highly efficient DNA-protein crosslinking for LChIP analyses. LChIP overcomes the current limitations of chemical crosslink. Specific, technological objectives were to fuse a microfluidic station with the ultrashort laser source in a unique, new machine to perform dynamic crosslinking in a time and cell type and/or differentiation-specific dependent manner. The LChIP technology could also be applied on frozen tissue slices for global applications. Specific biological objectives were to adapt and optimise LChIP technology to global genome analysis or massive, parallel single molecule sequencing using Solexa - LChIP-seq. We further applied LChIP technologies to selected cell populations (down to the single cell level); we achieved this goal by connecting the Laser with a microfluidic system, which may sort out specific cell populations, and by selectively focusing on cell populations in solid tissues or specimens. Finally, we developed software to automate the laser and microfluidic station system thereby constructing a versatile and user-friendly new-dedicated machine of high commercial interest.
The development of this technology / machinery has been assessed trough technical efforts and scientific ideas that will be discussed separately in this document.
The main goal of this activity was to develop customised laser system based on femtosecond Yb:KGW laser, optical parametric amplifier (OPA) and harmonic generators delivering two synchronised pulses at two different wavelengths from two independent channels. The UV pulse is produced by fourth harmonic (optionally third harmonic) generator installed in the first system channel. A second channel contains a continuously tuneable collinear OPA, equipped with additional tuneable frequency converters, covering the 210-515 nm range. In the first period of ATLAS project the R&D activity was concentrated on: (a) general design of collinear pumped OPA with harmonic generators, (b) study of broadband seed formation in a continuum generation stage, (c) OPA build up and investigation of its performance, (d) investigation of OPA signal conversion to UV. Here just some of these are summarised.
OPA design: OPA was designed as an two stage parametric amplifier seeded by white light continuum (WLC) and pumped by SH of approximately 180 fs pulses generated by Yb:KGW laser. SH generator, continuum generator, OPA stages and harmonic generators are integrated in single unit. A small portion of the incoming pulse is split off and used to generate WLC in a bulk medium. The rest of the pulse is frequency doubled and used as the pump for two OPA stages. The first OPA stage serves for pre-amplification of the spectral portion of broadband signal coming from continuum generator. The selection of which part to amplify is performed by tuning motorised translation and rotation stages. The signal from the output of first OPA is used as a seed for the power amplifier second OPA stage. The tuning of OPA output wavelength is performed by adjusting crystal angles and time delay between pump and seed pulses.
OPA performance: OPA performances have been tested using a standard PHAROS laser source producing approximately 300 fs pulses at 6 W average power. In order to test OPA at lowest energy limits the laser was operated at 100 kHz repetition rate, that gives 60 ?J per pulse. The tuning range is limited by the long-wave transparency limit of BBO crystal that is around 3 ?m. Therefore, the lowest possible wavelength of the signal wave is around 620 nm. The maximum signal pulse energy is obtained at approximately 650 nm and is of approximately 20 % when calculating the ratio of signal to pump pulse (515 nm) energies.
OPA output conversion to UV: OPA output wavelength tuning range can be extended with additional frequency conversion stages, for example by frequency doubling the signal and idler waves. In this way the tunable wavelength range has been extended into the ultraviolet down to 315 nm; the efficiency of the signal wave conversion to its second harmonic is 35 %, for the idler wave; this value decreases to 25 % for the longer wavelengths (around 1 250 nm). The wavelength range can be further extended into ultraviolet by generating the fourth harmonic of the signal and idler waves, and wavelengths down to 210 nm can be achieved, limited by the properties of the BBO nonlinear crystal. Efficiency, measured as the power ratio of the fourth harmonic radiation to the second harmonic radiation, varies from 5 to 23 % for the signal wave and from 10 up to 40 % for the idler wave. Such a high variation of conversion efficiency is limited by the properties of BBO crystal at shorter wavelength: in range < 230 nm of FHS, effective non-linear coefficient of BBO crystal decreases sharply. On the other hand, efficiency of the generation of fourth harmonic of idler wave is lower for wavelengths > 330 nm because of lower pulse energy generated in previous harmonic leading to unsaturated generation of higher harmonic using one crystal for the range. Once assessed the technical points discussed above, the main goal of second period of ATLAS project was the development and implementation of dedicated laser system delivering two synchronised pulses at two different wavelengths in UV region from two independent channels. This part consisted of: (a) development femtosecond diode laser pumped Yb:KGW laser system operating at different repetition rates and providing generation of high spatio-temporal quality pulses at 1030 nm with energy up to 1.3 mJ; (b) development of module for different harmonic pulse generation at 515 nm, 314 nm and 257 nm; (c) development of femtosecond parametric light amplifier (OPA) and harmonic generators based pulse source delivering femtosecond pulses continuously tunable in 210-2 600 nm wavelength range; (d) full system installation at premises of Università degli Studi di Napoli Federico II.
Versatile laser system for pumping of frequency convertors: In the frame of the ATLAS project on the base of the technologies and design principles worked out in Light Conversion the dedicated Yb:KGW laser system PHAROS optimised for its effective application in the research on UV-laser photo-induced cross-linking has been developed. The laser is designed using principle of chirped pulse amplification and is manufactured as a single unit comprising of fs pulse oscillator, regenerative amplifier, pulse stretcher-compressor. The compact and robust opto-mechanical laser design includes easy to replace modules (oscillator, amplifier, stretcher / compressor, electronic modules) with temperature stabilised and sealed housings ensuring stable laser operation within varying environments. The oscillator (OSC) produces a train femtosecond pulses that are used as seed for regenerative amplifier (RA). A series of experimental tests and design innovations related to improving of OSC operational characteristics (starting of mode locking, long term stability, OSC cavity dispersion control by implementation of chirped mirrors) has been accomplished. The output OSC oscillator produces the femtosecond pulses and exhibits smooth approximately 18 nm FWHM spectrum, that corresponds to approximately 80 fs pulse width. Average OSC output power is approximately 1 W power and long-term pulse energy stability is better than 5 %. One of the main objectives in laser development was an enhancement of laser output energies. Another characteristic of high importance is a laser ability to operate at different repetition rates maintaining low values of energy variations for long periods of time. Improvement has been achieved by careful RA cavity design, implementation of computer controlled cooling of RA frame and 'power lock' function in pump module driver operation.
The maximum pulse energy of 1.3 mJ is produced at repetition rates of 2 kHz. With increasing repetition rate the average output power rises reaching more than 6 W, at the expense of drop of energy per pulse (to 0.5 mJ at 10 kHz and to approximately 30 ?J at 200 kHz). The stability of laser operation during 15 hours of operation was also evaluated. The change in mean of output pulse power is less than 0.2 %. Standard deviation of short term power variation is below 0.5 %.
Laser output pulse duration is defined by RA gain bandwidth
Search for pump conditions and cavity geometry has been performed for realisation of broadest amplification bandwidth and providing at the same time the pulse amplification to pulse energies > 1 mJ. Pulse FWHM deduced from of output pulse autocorrelation is 176 fs. At 200 kHz repetition rate the pulse width is around 194 fs. PHAROS pumped harmonics generator usually acts as temporal pulse cleaner and even at highest laser output pulse energy the harmonic pulse has smooth, wings-free profile. The spatial parameters of the laser output beam are rather good. Beam asymmetry is negligible and M2 parameter for both directions through the range of laser repetition rates from 1 to 200 kHz. Laser system is equipped with pulse picker based on electro-optic Pockels cell (PC). The special PC control functions enabling users for prompt change of irradiation conditions in application experiments have been developed. They allow for: (a) selection of every nth pulse from the output train of pulses, (b) selection of portion of pulses consisting of n pulses. The laser is equipped with an extensive software package, which ensures its smooth hands-free operation and allows for fast and easy Pharos integration into various processing devices.
Fixed wavelength UV pulse generation by harmonics generators: Harmonics generator is an essential option for femtosecond Yb:KGW laser and gives a possibility to convert infrared laser radiation into VIS and UV range making it as effective tool for a research on UV-laser photo-induced cross-linking. Harmonic generator was designed as single unit 'HIRO 'allowing for generation of second, third and fourth harmonics of fundamental approximately 1 030 nm radiation of Yb:KGW laser. In all the frequency conversion stages the BBO crystals featuring wide transparency and phase matching ranges, large non-linear coefficient, high damage threshold and excellent optical homogeneity of different orientations are employed. The orientations and the length of the crystals are optimised for every particular frequency conversion stage. HIRO unit have a separate output port for each harmonic and is equipped with Output selector, which redirects theinternal beams for the generation of appropriate harmonics. The measured UV pulse energies at the output of Chanel 1 of the systems was 325 ?J (at 515 nm), 122 ?J (at 343 nm) and 75 ?J (at 257 nm) when pumped with 625 ?J energy pulses at 1030 nm what corresponds to half of maximum available pump pulse energy.
Tunable UV pulse generation by OPA and harmonic generators: The core element of the 2nd system channel that provides the generation of wavelength tunable UV pulses is the optical parametric amplifier. OPA is designed as a two stage parametric amplifier seeded by White light continuum (WLC) and pumped by SH of approximately 180 fs pulses generated by PHAROS. The generation of tunable UV pulses is achieved by employment of two additional harmonic generation stages which converts the tuning range of signal OPA wave to the UV range. SH generator, continuum generator and OPA stages are integrated in single unit ORPHEUS.
A small portion of the incoming pulse is split off and is used to generate WLC in a bulk medium. The rest of the pulse is frequency doubled and used as the pump for two OPA stages. The first OPA stage serves for preamplification of the portion of spectrum of broadband signal coming from continuum generator. The selection of which part to amplify is performed by tuning motorised translation and rotation stages. The signal from the output of first OPA is used as a seed for the power amplifier - second OPA stage. The tuning of the OPA output wavelength is performed by setting of proper angular position of both OPA crystals and adjusting time delay between pump and seed pulses.
Set-up conditions: Parametric amplifiers are pumped by the second harmonic of PHAROS pulses. The SH generator with computer controlled angle adjustment is installed inside ORPHEUS. OPA of ORPHEUS produces the femtosecond pulses tunable in the range of 630-2 600 nm. Fresh/residual fundamental and second harmonic radiation (1030 nm and 515 nm respectively) are accessible from dedicated output ports. The tuning curves of ORPHEUS that was installed in laboratory of Università degli Studi di Napoli Federico II has been measured using a pump radiation of 4.2 W of average power at 7 kHz repetition rate. This corresponds to pump pulse energy of 0.6 mJ. The built-in SH generator provided 515 nm pulse generation with up to 55 % energy conversion efficiency delivering OPA pump pulse with energy up to 0.33 mJ. The overall tuning range is limited by the long-wave transparency limit of BBO crystal, that is around 2.6 ?m. Therefore, the lowest possible wavelength of the signal wave when using a pump at 515 nm is around 620 nm. The energy conversion into parametric radiation is most of tuning range is around 20 %. The energy content for signal and idler waves is governed by Manley-Rowe relation. The drop of energy conversion in short wavelength region is caused by the rising idler absorption. In the wavelength range close to the degeneracy the lowering of parametric pulse energy is the result of lower continuum signal intensity in this wavelength region. OPA output wavelength tuning range has been extended to short-wavelength visible and UV region by using additional frequency conversion stages.
Using SH generator for frequency doubling of OPA signal and idler waves the tuning range of 305-630 nm has been covered with maximum energy conversion efficiency of approximately 38 %. The second frequency conversion stage installed at the output of Orpheus ensured generation of femtosecond pulses with wavelengths continuously tunable down to 210 nm.
During first and second period of ATLAS project, in parallel with Laser development as discussed above, also the microfluidic device was constructed and tested in order to perform experiment with low amount of material and on selected cell population (see experimental part below).
Potential impact:
ATLAS is a multidisciplinary research platform with the aim of developing a fully new technological approach to allow the global screening of DNA bound proteins and chromatin modifications using dynamic and time ranges currently unreachable. This consortium integrated a wide variety of competences to develop and apply a new technical approach to understand and to model biological processes at different levels of organisation (genome, transcriptome, proteome, phosphoproteome, interactome, regulatory networks, physiological processes). Actually no tools are present which reach these targets in a dynamic manner and so ATLAS works to answer to this specific requirement. ATLAS integrated dispersed capabilities of partners from 8 European countries and assembled the critical mass required to enable new global approaches, by networking the necessary expertise to secure European excellence and competitiveness, and to explore new directions in the research field.
Consortium planned to deliver new knowledge on basic biological processes relevant to health and disease. The quantitative data delivered served as the basis to design robust models using computational biology approaches.
The overall objective of ATLAS was to establish a novel global ChIP technology based on highly efficient, precise, robust and reproducible laser-assisted DNA protein crosslinking. Development of this innovative technology overcame the limitations imposed by the conventional formaldehyde-based chemical crosslinking and allowed the deconvolution of genomic information into transcription factor and epigenetic mediator-modulated executor networks with unprecedented accuracy, sensitivity and precision at a dynamic range which is several times larger than that possible when using conventional ChIP methodology. The technique was established to allow genome-wide studies of human genome.
In addition, the ATLAS consortium developed this technology further to make it compatible with fluorescence-activated cell sorting. Thus, individual cells with pre-defined 'marker' characteristics, such as cells of a particular differentiation status, for example of the hematopoietic lineage, leukemic blasts, or cells that have characteristics of cancer stem cells, can be selectively laser-crosslinked and their epigenome and genome-encoded information analysed. In addition to studies done for ATLAS purpose and technique development, this characteristic will allow a multitude of studies, including mechanisms of transcription regulation by factor-DNA recognition (stochastic vs. productive DNA interaction within time frames covering second to minute scales), as well as genetic and epigenetic aberrations of gene expression in cancer (stem) cells, cell lineage analysis, and studies to distinguish gene regulation originating from direct and indirect binding to 'response elements' , respectively.
In close collaboration with one of the Small and medium-sized enterprise (SME) partners, a specialist in laser production, the ATLAS consortium developed both 'open' and 'closed' (pre-adjusted) laser prototypes for the various possible ChIP technologies and had as perspective to provide such laser systems to the scientific community, either through the industrial partner or through a newly established spin-off company. The expected high efficiency, accuracy, selectivity and the extraordinary large dynamic range of laser-ChIP and laser-ChIP-chip technology allowed to perform genome-wide deconvolution experiments, which are impossible to date.
Among the possible scientific advancements obtained during ATLAS project, following are the most important:
Selective factor-DNA crosslink allowed precision mapping of binding sites repertoires. Using ATLAS technology non-specific protein-protein crosslink will be negligible and for the first time it will be possible to study pure factor binding to DNA with unmet accuracy, thus high precision factor (or epigenetic modification) maps can be generated. Comparison with conventional chemical crosslink will facilitate to distinguish between direct and indirect (through protein-protein contacts) DNA binding. Such information is of high mechanistic value. By modulating the parameters (wavelength, delay time to first pulse, circularisation) of the second pulse, we will as well explore the possibility of photocrosslinking also protein-protein complexes.
Novel strategy of experimental design - important mechanistic questions can be addressed. The ATLAS technology will potentially change the actual view of promoter function and gene regulation, leading to novel concepts of gene expression. Indeed, the comparison with data from formaldehyde crosslinking will most likely lead to novel classification/distinction of binding sites in enhancers versus promoters and / or to a completely revised view of enhancer and promoter function. Moreover, the efficiency and specificity of this approach will allow to define cellular context specific regulation in a dynamic time frame, which is currently unthinkable due to technical limitations. The applications of this technology will most likely open a new strategy of experimental design with very wide applications.
Knowledge transfer and enforcement of other EU projects. The ATLAS technology will re-enforce the contacts among different multidisciplinary European groups and will establish a core technological centre not only for ATLAS members but also for partners of several EU IP/STREP programs as indicated above.
This mode of tight consortia linkage will undoubtedly strengthen the scientific competitiveness of the European partners involved.
New dimension of ChIP analyses with SORT-ChIP. The combination of FACS and ChIP will open an entirely new dimension of genome-wide analysis. At the moment, complex biological samples such as blood or bone marrow can only be analysed as a whole. FACS-laser-ChIP will allow selective crosslinking at the single cell level and combine the potential of FACS-based cell sorting with the power of genome-wide transcription factor binding and epigenetic mapping analysis by ChIP and ChIP-chip. Moreover, features of minor but highly important cell populations (e.g. cancer stem cells) will become amenable to ChIP-chip analysis due to the selection potential of the FACS and the highly efficient crosslink by the dual pulse fs laser.
Economic impact
The development of the laser-ChIP and FACS-laser-ChIP-based technologies will have an important economic impact given the tremendous rapid expansion of the field and the use of ChIP for basic, but also applied and industrial/pharmaceutical, research. The project allowed the development of closed systems that can be marketed either by the industrial partner, or in the context of a spin-off company that can be created from the accumulated know-how.
Moreover, the project will strengthen the competitiveness of the participant LightCon company which had the opportunity to improve its expertise within the project and to profit from the transfer of know-how from the collaboration with outstanding experts from different fields of life science and technology. Finally, ATLAS-developed technology increased LightCon competence and facilitated its efforts to enter in a biological area of application.
Strategy
We have entered a phase of genome function analysis by which we can map regulatory proteins and machineries as well as epigenetically relevant DNA and chromatin modifications (i.e. post-translational modifications of histones that may constitute an 'epigenetic code', presence of histone variants, nucleosome densities) at the genome wide level. To overcome the major limitations for the use of the conventional ChIP-chip technology which are nearly exclusively due to the limitations of the initial chemical crosslinking step, the ATLAS consortium has gathered an international multidisciplinary consortium of high calibre mathematicians, physicists, chemists, molecular biologists (both array specialists and signal transduction specialists) and molecular oncologists with the aim to replace this step by laser-assisted photo-crosslinking and establish a ChIP-chip technology in which the precision and reliability of the ChIP matches that of the array technology. In addition the novel laser-ChIP approach opens a large spectrum of additional experimental options that cannot be addressed with current technologies.
Step by step, the ATLAS activity could be summarised as follows: At the first level ATLAS developed a laser system that is specifically adapted to DNA-protein crosslinking. We have already provided proof-of-principle that a rather 'primitive' UV laser can be used for laser-ChIP at efficiencies, which are several times higher than those of conventional formaldehyde-based crosslinking (as reported in annex 1 document).
In the second phase this novel laser system-ChIP was set up and validated in two modes of operation in parallel. On the one hand laser-ChIP was applied to individual promoters of high scientific interest to explore novel options of the laser-ChIP technology (as explained in detail in S&T results/foregrounds section), on the other hand laser-ChIP-chip was set up and used to address genome function-related questions.
A particular strength of the ATLAS project is its tight link to three other European consortia; this collection of top specialists used laser-ChIP technologies to address very distinct aspects of genome complexity and deconvolution of genome-encoded information. This configuration is of mutual benefit: ATLAS provides them with a novel technology that enhances their experimental capacities, while these consortia provide ATLAS with very expensive materials (genomic tiling arrays) for validation of its technology.
In its third phase ATLAS developed an even more challenging technology, which allows performing genome-wide deconvolution studies in pre-defined subsets of complex biological cell populations. This technology comprises the combination of Fluorescence-activated cell sorting (FACS) with the laser ChIP. The technology was developed in two steps: First ATLAS used a discontinuous mode of operation to determine key parameters, in particular the sensitivity (minimal number of cells necessary for ChIP) and selectivity (multidimensional selection of cells). The second step was technically most challenging but offers the highest scientific potential. ATLAS constructed the prototype of a FACS containing a dual pulse femtosecond laser for in situ crosslinking at the single cell level during sorting. With this FACS-laser set-up it will be for the first time possible to crosslink in a heterogeneous population of cells selectively those that are recognised by the FACS according to pre-defined parameters (e.g. cell dimensions, cell surface markers).
Potential applications of such technologies are obviously not only widespread but future development of the field will demand such an option to apply ChIP and ChIP-chip assays to 'real' biological samples, which are inherently composed of multiple cell populations. In this field, some experiments were just done (as reported in results section) and resulted interesting for further investigations. Contribution that requires a European approach. It is clear that neither of individual partners can address this complex multidisciplinary approach either on local or on national levels. Only a unique combination of biologists, chemists, physicists, clinicians, mathematicians, medical doctors and persons from biotechnology industries (SMEs) together with a great contribution from the European Union were able to achieve the expected impact. In the past, some of the partners established a tight collaboration and got preliminary results which could not be followed up without the creation of this consortium. Although all partners are getting some support from their National Foundations, only joined efforts along with a support from the European Union significantly influenced the quality of this research project. We are not aware of any other projects involving experimental biology, data mining, mathematical modelling, informaticians and molecular medicine applied to chemistry and physics. Moreover, the central participation of SMEs allows the development of commercial tools and implements the connection between the industrial and academic scientific work; SMEs enjoyed benefits from the research approach and its application for the creation of LChIP technology; academic partners received from the direct binding with industries for an immediate commercial view of the products of this consortium.
The participation in the consortium of partners performing different, but complementary, investigations enhanced the possibility of exploiting the novel findings of the project.
ATLAS improved Europe's innovation performance by stimulating a better integration between research and innovation and by working towards a more innovation-friendly policy.
On the bases of our good results, one of our long-term goals is to enhance the inclination for turning research into useful and commercially valuable innovations. We believe that successful development of this project based on tight collaboration between groups will help us in the future to develop the 'European Centre of Excellence' for technological excellence.
All partners represent the leading European groups in their field of research and had all potential to achieve the main goal of this very ambitious project. Only the synergy of all partners (which include many disfavoured European countries and country regions) helped us to reach the defined goal but without the financial support the achievements of all impacts (listed at now) would not have been possible.
List of websites: http://www.atlas-eu.com(opens in new window)
The knowledge of mechanisms leading the interactions among bio-molecules in living cells represents one of the main goal in molecular biology in order to define the dynamics of (direct and indirect) bindings. Currently, the establishment of a stable inter-play between nucleic acid and proteins (in particular DNA proteins crosslink) is mainly obtained through the conventional chemical methods involving the use of a bifuctional reagent, for instance, the formaldehyde.
Protein-Deoxyribonucleic acid (DNA) interactions play an important role in DNA replication, recombination, repair and, consequently, in transcriptional and translational gene regulation. The modulation of chromatin structure is a complex and dynamic process regulated at multiple levels though distinct mechanisms such as histone posttranslational modifications, non-coding RNA and DNA methylation. Aberrations of gene regulation might lie in pathological chromatin status. The establishment of a standing covalent bond between proteins and nucleic acids (crosslinking) open access to the study of interactions between bio-molecules: this is a crucial task for understanding functions and deregulation of gene expression. Enzymatic methods to examine protein-DNA interactions, have been developed in vivo and in vitro though genomic foot-printing. Recently, Chromatin immunoprecipitation (ChIP) assay allows to identify both the binding patterns between transcription factors and chromatin and to evaluate the occurrence of histone modifications. Despite the needs, the current ChIP technology does not allow to discriminate either direct or indirect binding or to study transient chromatin occupancy. Indeed specific bond of transcription factors causes in a large part the connectivity of gene regulatory networks as well as the quantitative level of gene expression. In order to satisfy these requirements, a new and more efficient crosslink reaction, based on the employment of an UV laser source, was developed. The use of UV laser crosslinking thus represents an innovative way to create a stable covalent bond though the excitation of the electronic state of proteins and DNA. Ultraviolet (UV) irradiation creates covalent bindings between the reactive groups of DNA (thymine and cysteine) and amino acids (serine, methionine, lysine, arginine, histidine, tryptophan, phenylalanine or tyrosine).
Therefore, the occurrence of a crosslink is accompanied by a combination of factors, including the inherent photo-reactivity of the excited nucleotides, the geometrical arrangement and the molecular dynamics involved in the mechanism of crosslink. UV light is a zero length crosslinker and produces less perturbation of the complex than chemical crosslinker such as formaldehyde. UV Laser-based chromatin immunoprecipitation (LChIP) induces photo-mediated crosslink in a very short time allowing the study of transient interactions by varying different parameters such as energy, repetition rate and pulse intensity in time unit. UV irradiation of cells at the wavelength of about 260 nm produces covalent bonds in nucleic acids and proteins and, in particular, may preferentially allow bonds between TFs and histones associated within chromatin.
So, LChIP is able to characterise the dynamics of the transcription factors binding on chromatin in living cells because the required time for Laser-mediated crosslink is several orders of magnitude lower than the conventional methods. This technique makes available the study of temporal and spatial bindings of proteins on DNA and so, it becomes a useful tool for understanding regulation of gene expression and maintenance. Although living human cells non-linearly respond to irradiation with intense fs-UV-laser pulses, among the phenomena triggered by such pulses, a highly efficient DNA-proteins relationship occurs.
Definitively, the development of LChIP technique and its applications have the powerful potential in detecting the behaviour of transient DNA-proteins bindings in vivo. The study of these transient interactions in a short time scale, a parameter undetectable with conventional methodologies, is now possible corroborating the extraordinary biomedical potential of LChIP for discriminating chromatin epigenome and TF bindings dynamically.
Project context and objectives:
The availability of the DNA sequence of many eukaryotic genomes and the generation of high density tiling arrays covering such entire genomes has made it possible to decipher the regulatory principles that are based on the interplay of trans-regulatory TFs and their cognate cis-regulatory DNA recognition sequences at genome-wide level. At chromatin level a mutual interplay between transcription factors and epigenetic modifiers (DNA and histone modifying machinaries) sets up determinants of gene (in)activity. The ensemble of histone modifications at a given gene locus has been proposed to establish an epigenetic code of great complexity. Irrespectively of whether such a code does exist or whether chromatin modifications rather constitute a step in signal transduction, it has become increasingly clear that chromatin modifications constitute docking sites for regulatory factors. Thus, description of the information encoded by genomes requires a deconvolution of the genetic and epigenetic programs and the interplay between these two regulatory levels. In addition to their enormous potential and power for the study of gene regulation mechanisms such analyses also provide important tools for diagnosis, prognosis, and therapy of diseases. Decoding chromatinembedded information at the whole genome level by combining ChIP with global analysis such as DNA tiling arrays (ChIP-chip or ChIP-on-chip) or parallel single molecule sequencing, ChIP-seq, have become powerful approaches currently applied to mammalian genomes to analyse gene regulation programmes.
Global ChIP analyses allow converting genomic information into a dynamic regulatory network that operates in a time, cell, development and environmentdependent manner to coordinate cell homeostasis, proliferation, survival and death, as well as cell-cell communication. ChIP is a powerful approach to determine in vivo the chronology of transcription factor recruitment during activation or repression in the context of chromatin and changes in epigenetic signatures and chromatin remodeling. In many applications, protein and DNA are cross-linked using formaldehyde, chromatin is fragmented, and the protein of interest is immunoprecipitated with specific antibodies (XChIP). Alternatively, native ChIP (NChIP) i.e. without chemical crosslinkers, can be used in epigenetic profiling studies. The relative amount of a particular DNA fragment cross-linked to the protein (and therefore present in the precipitate) is determined by qPCR, and is a measure of the occupancy of the factor at that particular position in the genome. Despite providing significant insight in transcription regulation and chromatin-mediated effects, NChIP and XChIP technologies have a number of intrinsic technical limitations. These comprise:
- Selectivity: Formaldehyde introduces covalent bonds between protein-DNA and protein-protein. TFs are generally embedded within complexes/machineries that display multiple interaction surfaces (with DNA or between subunits). This has two consequences: (i) factors that bind DNA directly or indirectly via proteinprotein interactions with chromatin components will be crosslinked with widely different efficiencies depending on the stability of the interaction and (ii) a single protein/complex can be crosslinked to more than one site in the genome e.g. in the case of enhancer-promoter looping. NChIP can only be employed efficiently for very stable interactions, predominantly histone (nucleosome)-DNA interactions.
- Accuracy and efficiency: The overall performance of XChIP depends critically on the efficiency of the first step, the crosslink of factors to DNA by formaldehyde exposure of living cells. However, chemical crosslinking is diffusion-controlled and varies considerably with cell type, manipulation, storage and purity of the chemical. Due to the progressively increasing non-specific crosslink by chemical crosslinkers, formaldehyde crosslinking has to be stopped a long time before a maximal crosslinking of DNA to protein has been reached. This implies low efficiency of the subsequent immunoprecipitation.
- Dynamics and half-lives of DNA interactions: TFs bind with highly different kinetics/half-lives to DNA / chromatin. Current technologies do not allow studying very short-live interactions. It is, for example, not possible to resolve the DNA / chromatin scanningmodel of a transcription factor as opposed to a direct recruitment model. Data obtained in XChIP (co)factor binding studies e.g. on the time-resolved Estrogen Receptor (ER) activation poorly correlates with photo-bleaching experiments resulting in a controversy about the actual molecular mechanisms of TF action. Sensitivity: Conventional XChIP requires large amounts of cells (about 10^6 to 10^8 cells) as starting material. A significantly higher sensitivity (down to 1 000 cells) has been reported for carrier ChIP (CChIP) that involves NChIP procedure for the analysis of histone modification. However, this procedure is not applicable to TF interactions with DNA/chromatin.
- Single cell, tissue crosslinking and sorting: Current ChIP technologies do not permit crosslink of selected individual cells or frozen tissue slides and subsequent analyses of defined population. Epitope masking and modification: Formaldehyde crosslinking alters lysine residues, which can be part of antibody epitopes in targeted DNA binding proteins, particularly in modified histones. This has several serious drawbacks: (i) it reduces substantially the IP efficacy in epigenetic studies and yields <<1% are commonly observed, (ii) due to the covalent stabilisation of entire complexes, antibody epitopes may reside inside a complex, which would become accessible due to dissociation if only DNA crosslinks would occur, (iii) pitope-tagging is limited to peptides that are not modified by formaldehyde.
To overcome most of the above mentioned limitations, ATLAS consortium established novel ChIP technologies (LChIP) consisting in the development and subsequent creation of an experimental platform that, through the employing of a laser source, is able to induce cross-link reactions between DNA and proteins within living cells to study the evolution in time and space of both the transcriptional machinery and epigenetic code. Laser technology, therefore, represents a valid alternative to traditional methods, as it is able to induce bonds between DNA and proteins in very short intervals of time and with high efficiency. The driving idea of ATLAS is the observation, confirmed by experimental data and evidences in the literature, that an ultrashort UV laser can excite the electronic states of nitrogenous bases composing the DNA and amino acid side residues of proteins that bind DNA itself. Once absorbed energy, the system DNA-protein relaxes to the ground state via the formation of a covalent bond. The creation of this link, obtained by a method 'fotophysic', opens up a range of possibilities for study of interaction in 'real time' and rapid decoding of 'cross talk' between transcription factors, enhancers, epigenetic modulators and DNA. The reducing time scale of crosslink induction increases the number of proteins-DNA interaction that can be analysed by 'omic' science, and in general helps to clarify the temporal sequence of the events constituting the biological phenomenon (for example the epigenetic code variation). Comparing the conventional method with LChIP some drawbacks might be overcome. So, the novel LChIP technology identifies direct protein-DNA contacts preferentially if not exclusively under certain conditions (increasing of selectivity).
LChIP involves ultra-fast physical crosslinking by femtosecond UV lasers specifically designed for highly efficient DNA-protein crosslinking. Once calibrated it operates fully reproducible and highly accurate, and is cell and experimenter independent. Due to its ultra-fast crosslinking, LChIP allows fundamental studies on the mechanisms of DNA/chromatin recognition by DNA-binding regulatory factors. Given to its high efficiency documented in proof-of-principle experiments (see sections below), LChIP linked to microscopic and cell-sorting platforms using microfluidic systems permits to photo crosslink subsets of cells on-the-fly, crosslink cells in particular phases of the cell cycle, or leukemia blasts for genomewide analyses. In addition, LChIP does not affect protein epitopes and hence has a dramatically higher IP efficacy allowing the use of epitope-tagging approaches with highly efficient antibodies independent of the presence of lysines. To reach these aims industrial partners, mathematicians, physicists, molecular biologists, chemists and MDs worked together up to construct a prototype of a complex device, including Laser apparatus and microfluidic station, to perform induction of crosslink in living cells in an easy and not time consuming manner. The prototype is additionally governed by software, made for the ATLAS purpose, with a friendly interface and it easy to understand. In order to realise the prototype, the efforts of ATLAS project were centred on two main different research lines:
- A technological one in which the construction of dedicated femtosecond double-pulse, two colours tunable laser apparatus was made, as such as the development of microfluidic and cell sorting machinery (see first part of S&T results / foreground description).
- An experimental one aimed to discover the chemical principles and mechanism of photo-crosslink and the development and validation of LChIP in cell-based assays (see second part of S&T results / foreground description).
The step-by-step progression of the ATLAS project was allowed through a series of objectives achievement. In the list below the most important ones are summarised:
- Development and validation of a custom femtosecond laser prototype for optimised UV-laser photo-induced crosslinking with high efficiency. In the first 12 months of project, a new dedicated ultrashort laser system has been developed that is be able to deliver high-energy, UV, fs-pulses (in the energy range of several microjoule per pulse) at a repetition rate variable within six orders of magnitude, between 1Hz and 1MHz. The wavelength of such pulses is variable in the range 250-280 nm, to match the first UV absorption band range of DNA bases. The combined features of high energy / pulse and high repetition rate were not commercially available at the moment of ATLAS project start and until now, for any known ultrashort laser source. This is so the first prototype released in the context of ATLAS project. Its realisation involved the collaboration of industrial partners specialised in UV-laser source construction and commercialisation as such as some university partners devoted to the management, control and improvement of Laser apparatus.
- Development and implementation of laser system upgrades with dedicated Optical parametric amplifier (OPA) that allows performing two-colour, double pulse irradiation of the sample, thereby significantly minimising cell damage after the irradiation of biomolecules or living systems with UV laser source (however ensuring and optimising DNA-protein crosslink). The developed OPA has unique features in terms of high energy/pulse together with the high pulse repetition rate and has a wide tunability, covering the 200 nm 2600 nm spectral window. The OPA technology, although commercially existent, has been applied for the first time to laser apparatus described above, expanding much the possible wavelengths window.
- Determination of the chemical nature of laser-induced DNA-protein cross-linking, using the synergy between theoretical calculations and study of increasingly complex chemical models, was carried out. Factors effecting the excitation of the DNA and the protein, the formation of intermediate reactive species, and the creation of the new DNA-protein bonds have been addressed in a sequential manner, from the isolated building blocks of each macromolecule to reacting adjacent monomers and the long-range effects that can be found in larger systems with tertiary structure, in particular when proximity (and affinity) is enforced. This multi-level detailed description of the photochemical behaviour of DNA in a protein environment has provided the knowledge for rational design and optimisation of the crosslinking experimental conditions in cell based assays.
- Evaluation of macroscopic parameters characterising the DNA-laser interaction has been carried out in cell based assays in order to indicate the ones (such as time, geometry, cell behaviour) that mainly affect the induction of crosslink between DNA-proteins in living cells. The final benefit has consisted in indicating the values or the restricted range of values for an optimal interaction between the laser and the molecular system.
- First results of LChIP were been obtained on single gene analysis and then the skills arising from these experiments were applied to perform genomewide studies of LChIP (LChIP-seq).
- The feasibility of LChIP was estimated by looking at particular genomic regions and well-defined transcriptional factors (or enhancers of transcription) to better decipher the evolution in time and space of transcriptional pathways and networks and the difference between direct and indirect binding of proteins to DNA.
- Integration of LChIP with microfluidic-based cell sorting platforms was obtained. Dedicated laser system has been integrated into the fluorescence-activated microfluidic cell sorter. This setup allows analysis of selective subsets using much lower number of cells than required for conventional approaches. Moreover, this approach allows the dynamic crosslinking of cells in a flux and number dependent manner.
- Application of LChIP with i) microscopic manipulations of frozen tissue slices; ii) microfluidic station for experiments in cellular subpopulations; iii) global study of epi-modifications and TFs binding in time resolved manner upon selected treatments (f.e. Epi-drugs) was performed.
- Optimisation of amplification methods of very small amounts of DNA obtained by LChIP (#6), allowing the use of low numbers of cells as starting material, has been done. In parallel, another industrial partner carried out the development and characterisation of antibodies to obtain LChILL and LChIP grade antibodies; in particular emphasis was on the production of antibodies against epitope peptides that will be masked by formaldehyde in order to improve the efficiency of IP stage. At the same manner, kit(s) for LChIP (and for DNA amplification for LChIP) was developed. The LChIP contains controls such as LChIP grade antibodies and physically cross-linked cells as well as optimised buffers, PCR primers and adapted protocol. As described previously, in the context of ATLAS project, the borders among different scientific fields have been overcome, by joining Physics and Bio-medical approaches to answer to a typical biological question (such as 'what binds what in a cellular system') and to clarify the principles underpinning the interaction between biomolecules and UV radiation. The multidisciplinary aspect characterised strongly this project and maybe it was the keystone to develop the new LChIP technique.
ATLAS integrated dispersed capabilities of partners from 8 European countries and assembled the critical mass required to enable new global approaches, by networking the necessary expertise to secure European excellence and competitiveness, and to explore new directions in the research field. Consortium delivers new knowledge on basic biological processes relevant to health and disease. With the new LChIp is now possible to better identify:
- Selective factor-DNA crosslink repertories. Using ATLAS technology non-specific protein-protein crosslink is negligible (under appropriate conditions) allowing, for the first time, the study of bona fide pure bindings to chromatin with unmet accuracy. Consequently, highly precise transcription factor (or epigenetic modification) maps can be generated, potentially reflecting dynamics and kinetics of bindings. Comparison with conventional chemical crosslink facilitates to distinguish between direct and indirect (through protein-protein contacts) DNA binding. Such information is of extraordinary mechanistic value.
- Novel strategy of experimental design-important mechanistic question can be addressed. The ATLAS technology could potentially change the actual view of promoter function and gene regulation, leading to novel concepts of gene expression. Indeed, the comparison with data from formaldehyde crosslinking will most likely lead to novel classification/distinction of binding sites in enhancers versus promoters and/or to a completely revised view of enhancer and promoter function. Moreover, the efficiency and specificity of this approach allows to define cellular context specific regulation in a dynamic time frame, which is currently unthinkable due to technical limitations. The applications of this technology most likely open a new strategy of experimental design with very wide applicabilities.
Project results:
The overall objective of the ATLAS consortium was to develop novel types of femtosecond (fs) tunable UV lasers to induce highly efficient DNA-protein crosslinking for LChIP analyses. LChIP overcomes the current limitations of chemical crosslink. Specific, technological objectives were to fuse a microfluidic station with the ultrashort laser source in a unique, new machine to perform dynamic crosslinking in a time and cell type and/or differentiation-specific dependent manner. The LChIP technology could also be applied on frozen tissue slices for global applications. Specific biological objectives were to adapt and optimise LChIP technology to global genome analysis or massive, parallel single molecule sequencing using Solexa - LChIP-seq. We further applied LChIP technologies to selected cell populations (down to the single cell level); we achieved this goal by connecting the Laser with a microfluidic system, which may sort out specific cell populations, and by selectively focusing on cell populations in solid tissues or specimens. Finally, we developed software to automate the laser and microfluidic station system thereby constructing a versatile and user-friendly new-dedicated machine of high commercial interest.
The development of this technology / machinery has been assessed trough technical efforts and scientific ideas that will be discussed separately in this document.
The main goal of this activity was to develop customised laser system based on femtosecond Yb:KGW laser, optical parametric amplifier (OPA) and harmonic generators delivering two synchronised pulses at two different wavelengths from two independent channels. The UV pulse is produced by fourth harmonic (optionally third harmonic) generator installed in the first system channel. A second channel contains a continuously tuneable collinear OPA, equipped with additional tuneable frequency converters, covering the 210-515 nm range. In the first period of ATLAS project the R&D activity was concentrated on: (a) general design of collinear pumped OPA with harmonic generators, (b) study of broadband seed formation in a continuum generation stage, (c) OPA build up and investigation of its performance, (d) investigation of OPA signal conversion to UV. Here just some of these are summarised.
OPA design: OPA was designed as an two stage parametric amplifier seeded by white light continuum (WLC) and pumped by SH of approximately 180 fs pulses generated by Yb:KGW laser. SH generator, continuum generator, OPA stages and harmonic generators are integrated in single unit. A small portion of the incoming pulse is split off and used to generate WLC in a bulk medium. The rest of the pulse is frequency doubled and used as the pump for two OPA stages. The first OPA stage serves for pre-amplification of the spectral portion of broadband signal coming from continuum generator. The selection of which part to amplify is performed by tuning motorised translation and rotation stages. The signal from the output of first OPA is used as a seed for the power amplifier second OPA stage. The tuning of OPA output wavelength is performed by adjusting crystal angles and time delay between pump and seed pulses.
OPA performance: OPA performances have been tested using a standard PHAROS laser source producing approximately 300 fs pulses at 6 W average power. In order to test OPA at lowest energy limits the laser was operated at 100 kHz repetition rate, that gives 60 ?J per pulse. The tuning range is limited by the long-wave transparency limit of BBO crystal that is around 3 ?m. Therefore, the lowest possible wavelength of the signal wave is around 620 nm. The maximum signal pulse energy is obtained at approximately 650 nm and is of approximately 20 % when calculating the ratio of signal to pump pulse (515 nm) energies.
OPA output conversion to UV: OPA output wavelength tuning range can be extended with additional frequency conversion stages, for example by frequency doubling the signal and idler waves. In this way the tunable wavelength range has been extended into the ultraviolet down to 315 nm; the efficiency of the signal wave conversion to its second harmonic is 35 %, for the idler wave; this value decreases to 25 % for the longer wavelengths (around 1 250 nm). The wavelength range can be further extended into ultraviolet by generating the fourth harmonic of the signal and idler waves, and wavelengths down to 210 nm can be achieved, limited by the properties of the BBO nonlinear crystal. Efficiency, measured as the power ratio of the fourth harmonic radiation to the second harmonic radiation, varies from 5 to 23 % for the signal wave and from 10 up to 40 % for the idler wave. Such a high variation of conversion efficiency is limited by the properties of BBO crystal at shorter wavelength: in range < 230 nm of FHS, effective non-linear coefficient of BBO crystal decreases sharply. On the other hand, efficiency of the generation of fourth harmonic of idler wave is lower for wavelengths > 330 nm because of lower pulse energy generated in previous harmonic leading to unsaturated generation of higher harmonic using one crystal for the range. Once assessed the technical points discussed above, the main goal of second period of ATLAS project was the development and implementation of dedicated laser system delivering two synchronised pulses at two different wavelengths in UV region from two independent channels. This part consisted of: (a) development femtosecond diode laser pumped Yb:KGW laser system operating at different repetition rates and providing generation of high spatio-temporal quality pulses at 1030 nm with energy up to 1.3 mJ; (b) development of module for different harmonic pulse generation at 515 nm, 314 nm and 257 nm; (c) development of femtosecond parametric light amplifier (OPA) and harmonic generators based pulse source delivering femtosecond pulses continuously tunable in 210-2 600 nm wavelength range; (d) full system installation at premises of Università degli Studi di Napoli Federico II.
Versatile laser system for pumping of frequency convertors: In the frame of the ATLAS project on the base of the technologies and design principles worked out in Light Conversion the dedicated Yb:KGW laser system PHAROS optimised for its effective application in the research on UV-laser photo-induced cross-linking has been developed. The laser is designed using principle of chirped pulse amplification and is manufactured as a single unit comprising of fs pulse oscillator, regenerative amplifier, pulse stretcher-compressor. The compact and robust opto-mechanical laser design includes easy to replace modules (oscillator, amplifier, stretcher / compressor, electronic modules) with temperature stabilised and sealed housings ensuring stable laser operation within varying environments. The oscillator (OSC) produces a train femtosecond pulses that are used as seed for regenerative amplifier (RA). A series of experimental tests and design innovations related to improving of OSC operational characteristics (starting of mode locking, long term stability, OSC cavity dispersion control by implementation of chirped mirrors) has been accomplished. The output OSC oscillator produces the femtosecond pulses and exhibits smooth approximately 18 nm FWHM spectrum, that corresponds to approximately 80 fs pulse width. Average OSC output power is approximately 1 W power and long-term pulse energy stability is better than 5 %. One of the main objectives in laser development was an enhancement of laser output energies. Another characteristic of high importance is a laser ability to operate at different repetition rates maintaining low values of energy variations for long periods of time. Improvement has been achieved by careful RA cavity design, implementation of computer controlled cooling of RA frame and 'power lock' function in pump module driver operation.
The maximum pulse energy of 1.3 mJ is produced at repetition rates of 2 kHz. With increasing repetition rate the average output power rises reaching more than 6 W, at the expense of drop of energy per pulse (to 0.5 mJ at 10 kHz and to approximately 30 ?J at 200 kHz). The stability of laser operation during 15 hours of operation was also evaluated. The change in mean of output pulse power is less than 0.2 %. Standard deviation of short term power variation is below 0.5 %.
Laser output pulse duration is defined by RA gain bandwidth
Search for pump conditions and cavity geometry has been performed for realisation of broadest amplification bandwidth and providing at the same time the pulse amplification to pulse energies > 1 mJ. Pulse FWHM deduced from of output pulse autocorrelation is 176 fs. At 200 kHz repetition rate the pulse width is around 194 fs. PHAROS pumped harmonics generator usually acts as temporal pulse cleaner and even at highest laser output pulse energy the harmonic pulse has smooth, wings-free profile. The spatial parameters of the laser output beam are rather good. Beam asymmetry is negligible and M2 parameter for both directions through the range of laser repetition rates from 1 to 200 kHz. Laser system is equipped with pulse picker based on electro-optic Pockels cell (PC). The special PC control functions enabling users for prompt change of irradiation conditions in application experiments have been developed. They allow for: (a) selection of every nth pulse from the output train of pulses, (b) selection of portion of pulses consisting of n pulses. The laser is equipped with an extensive software package, which ensures its smooth hands-free operation and allows for fast and easy Pharos integration into various processing devices.
Fixed wavelength UV pulse generation by harmonics generators: Harmonics generator is an essential option for femtosecond Yb:KGW laser and gives a possibility to convert infrared laser radiation into VIS and UV range making it as effective tool for a research on UV-laser photo-induced cross-linking. Harmonic generator was designed as single unit 'HIRO 'allowing for generation of second, third and fourth harmonics of fundamental approximately 1 030 nm radiation of Yb:KGW laser. In all the frequency conversion stages the BBO crystals featuring wide transparency and phase matching ranges, large non-linear coefficient, high damage threshold and excellent optical homogeneity of different orientations are employed. The orientations and the length of the crystals are optimised for every particular frequency conversion stage. HIRO unit have a separate output port for each harmonic and is equipped with Output selector, which redirects theinternal beams for the generation of appropriate harmonics. The measured UV pulse energies at the output of Chanel 1 of the systems was 325 ?J (at 515 nm), 122 ?J (at 343 nm) and 75 ?J (at 257 nm) when pumped with 625 ?J energy pulses at 1030 nm what corresponds to half of maximum available pump pulse energy.
Tunable UV pulse generation by OPA and harmonic generators: The core element of the 2nd system channel that provides the generation of wavelength tunable UV pulses is the optical parametric amplifier. OPA is designed as a two stage parametric amplifier seeded by White light continuum (WLC) and pumped by SH of approximately 180 fs pulses generated by PHAROS. The generation of tunable UV pulses is achieved by employment of two additional harmonic generation stages which converts the tuning range of signal OPA wave to the UV range. SH generator, continuum generator and OPA stages are integrated in single unit ORPHEUS.
A small portion of the incoming pulse is split off and is used to generate WLC in a bulk medium. The rest of the pulse is frequency doubled and used as the pump for two OPA stages. The first OPA stage serves for preamplification of the portion of spectrum of broadband signal coming from continuum generator. The selection of which part to amplify is performed by tuning motorised translation and rotation stages. The signal from the output of first OPA is used as a seed for the power amplifier - second OPA stage. The tuning of the OPA output wavelength is performed by setting of proper angular position of both OPA crystals and adjusting time delay between pump and seed pulses.
Set-up conditions: Parametric amplifiers are pumped by the second harmonic of PHAROS pulses. The SH generator with computer controlled angle adjustment is installed inside ORPHEUS. OPA of ORPHEUS produces the femtosecond pulses tunable in the range of 630-2 600 nm. Fresh/residual fundamental and second harmonic radiation (1030 nm and 515 nm respectively) are accessible from dedicated output ports. The tuning curves of ORPHEUS that was installed in laboratory of Università degli Studi di Napoli Federico II has been measured using a pump radiation of 4.2 W of average power at 7 kHz repetition rate. This corresponds to pump pulse energy of 0.6 mJ. The built-in SH generator provided 515 nm pulse generation with up to 55 % energy conversion efficiency delivering OPA pump pulse with energy up to 0.33 mJ. The overall tuning range is limited by the long-wave transparency limit of BBO crystal, that is around 2.6 ?m. Therefore, the lowest possible wavelength of the signal wave when using a pump at 515 nm is around 620 nm. The energy conversion into parametric radiation is most of tuning range is around 20 %. The energy content for signal and idler waves is governed by Manley-Rowe relation. The drop of energy conversion in short wavelength region is caused by the rising idler absorption. In the wavelength range close to the degeneracy the lowering of parametric pulse energy is the result of lower continuum signal intensity in this wavelength region. OPA output wavelength tuning range has been extended to short-wavelength visible and UV region by using additional frequency conversion stages.
Using SH generator for frequency doubling of OPA signal and idler waves the tuning range of 305-630 nm has been covered with maximum energy conversion efficiency of approximately 38 %. The second frequency conversion stage installed at the output of Orpheus ensured generation of femtosecond pulses with wavelengths continuously tunable down to 210 nm.
During first and second period of ATLAS project, in parallel with Laser development as discussed above, also the microfluidic device was constructed and tested in order to perform experiment with low amount of material and on selected cell population (see experimental part below).
Potential impact:
ATLAS is a multidisciplinary research platform with the aim of developing a fully new technological approach to allow the global screening of DNA bound proteins and chromatin modifications using dynamic and time ranges currently unreachable. This consortium integrated a wide variety of competences to develop and apply a new technical approach to understand and to model biological processes at different levels of organisation (genome, transcriptome, proteome, phosphoproteome, interactome, regulatory networks, physiological processes). Actually no tools are present which reach these targets in a dynamic manner and so ATLAS works to answer to this specific requirement. ATLAS integrated dispersed capabilities of partners from 8 European countries and assembled the critical mass required to enable new global approaches, by networking the necessary expertise to secure European excellence and competitiveness, and to explore new directions in the research field.
Consortium planned to deliver new knowledge on basic biological processes relevant to health and disease. The quantitative data delivered served as the basis to design robust models using computational biology approaches.
The overall objective of ATLAS was to establish a novel global ChIP technology based on highly efficient, precise, robust and reproducible laser-assisted DNA protein crosslinking. Development of this innovative technology overcame the limitations imposed by the conventional formaldehyde-based chemical crosslinking and allowed the deconvolution of genomic information into transcription factor and epigenetic mediator-modulated executor networks with unprecedented accuracy, sensitivity and precision at a dynamic range which is several times larger than that possible when using conventional ChIP methodology. The technique was established to allow genome-wide studies of human genome.
In addition, the ATLAS consortium developed this technology further to make it compatible with fluorescence-activated cell sorting. Thus, individual cells with pre-defined 'marker' characteristics, such as cells of a particular differentiation status, for example of the hematopoietic lineage, leukemic blasts, or cells that have characteristics of cancer stem cells, can be selectively laser-crosslinked and their epigenome and genome-encoded information analysed. In addition to studies done for ATLAS purpose and technique development, this characteristic will allow a multitude of studies, including mechanisms of transcription regulation by factor-DNA recognition (stochastic vs. productive DNA interaction within time frames covering second to minute scales), as well as genetic and epigenetic aberrations of gene expression in cancer (stem) cells, cell lineage analysis, and studies to distinguish gene regulation originating from direct and indirect binding to 'response elements' , respectively.
In close collaboration with one of the Small and medium-sized enterprise (SME) partners, a specialist in laser production, the ATLAS consortium developed both 'open' and 'closed' (pre-adjusted) laser prototypes for the various possible ChIP technologies and had as perspective to provide such laser systems to the scientific community, either through the industrial partner or through a newly established spin-off company. The expected high efficiency, accuracy, selectivity and the extraordinary large dynamic range of laser-ChIP and laser-ChIP-chip technology allowed to perform genome-wide deconvolution experiments, which are impossible to date.
Among the possible scientific advancements obtained during ATLAS project, following are the most important:
Selective factor-DNA crosslink allowed precision mapping of binding sites repertoires. Using ATLAS technology non-specific protein-protein crosslink will be negligible and for the first time it will be possible to study pure factor binding to DNA with unmet accuracy, thus high precision factor (or epigenetic modification) maps can be generated. Comparison with conventional chemical crosslink will facilitate to distinguish between direct and indirect (through protein-protein contacts) DNA binding. Such information is of high mechanistic value. By modulating the parameters (wavelength, delay time to first pulse, circularisation) of the second pulse, we will as well explore the possibility of photocrosslinking also protein-protein complexes.
Novel strategy of experimental design - important mechanistic questions can be addressed. The ATLAS technology will potentially change the actual view of promoter function and gene regulation, leading to novel concepts of gene expression. Indeed, the comparison with data from formaldehyde crosslinking will most likely lead to novel classification/distinction of binding sites in enhancers versus promoters and / or to a completely revised view of enhancer and promoter function. Moreover, the efficiency and specificity of this approach will allow to define cellular context specific regulation in a dynamic time frame, which is currently unthinkable due to technical limitations. The applications of this technology will most likely open a new strategy of experimental design with very wide applications.
Knowledge transfer and enforcement of other EU projects. The ATLAS technology will re-enforce the contacts among different multidisciplinary European groups and will establish a core technological centre not only for ATLAS members but also for partners of several EU IP/STREP programs as indicated above.
This mode of tight consortia linkage will undoubtedly strengthen the scientific competitiveness of the European partners involved.
New dimension of ChIP analyses with SORT-ChIP. The combination of FACS and ChIP will open an entirely new dimension of genome-wide analysis. At the moment, complex biological samples such as blood or bone marrow can only be analysed as a whole. FACS-laser-ChIP will allow selective crosslinking at the single cell level and combine the potential of FACS-based cell sorting with the power of genome-wide transcription factor binding and epigenetic mapping analysis by ChIP and ChIP-chip. Moreover, features of minor but highly important cell populations (e.g. cancer stem cells) will become amenable to ChIP-chip analysis due to the selection potential of the FACS and the highly efficient crosslink by the dual pulse fs laser.
Economic impact
The development of the laser-ChIP and FACS-laser-ChIP-based technologies will have an important economic impact given the tremendous rapid expansion of the field and the use of ChIP for basic, but also applied and industrial/pharmaceutical, research. The project allowed the development of closed systems that can be marketed either by the industrial partner, or in the context of a spin-off company that can be created from the accumulated know-how.
Moreover, the project will strengthen the competitiveness of the participant LightCon company which had the opportunity to improve its expertise within the project and to profit from the transfer of know-how from the collaboration with outstanding experts from different fields of life science and technology. Finally, ATLAS-developed technology increased LightCon competence and facilitated its efforts to enter in a biological area of application.
Strategy
We have entered a phase of genome function analysis by which we can map regulatory proteins and machineries as well as epigenetically relevant DNA and chromatin modifications (i.e. post-translational modifications of histones that may constitute an 'epigenetic code', presence of histone variants, nucleosome densities) at the genome wide level. To overcome the major limitations for the use of the conventional ChIP-chip technology which are nearly exclusively due to the limitations of the initial chemical crosslinking step, the ATLAS consortium has gathered an international multidisciplinary consortium of high calibre mathematicians, physicists, chemists, molecular biologists (both array specialists and signal transduction specialists) and molecular oncologists with the aim to replace this step by laser-assisted photo-crosslinking and establish a ChIP-chip technology in which the precision and reliability of the ChIP matches that of the array technology. In addition the novel laser-ChIP approach opens a large spectrum of additional experimental options that cannot be addressed with current technologies.
Step by step, the ATLAS activity could be summarised as follows: At the first level ATLAS developed a laser system that is specifically adapted to DNA-protein crosslinking. We have already provided proof-of-principle that a rather 'primitive' UV laser can be used for laser-ChIP at efficiencies, which are several times higher than those of conventional formaldehyde-based crosslinking (as reported in annex 1 document).
In the second phase this novel laser system-ChIP was set up and validated in two modes of operation in parallel. On the one hand laser-ChIP was applied to individual promoters of high scientific interest to explore novel options of the laser-ChIP technology (as explained in detail in S&T results/foregrounds section), on the other hand laser-ChIP-chip was set up and used to address genome function-related questions.
A particular strength of the ATLAS project is its tight link to three other European consortia; this collection of top specialists used laser-ChIP technologies to address very distinct aspects of genome complexity and deconvolution of genome-encoded information. This configuration is of mutual benefit: ATLAS provides them with a novel technology that enhances their experimental capacities, while these consortia provide ATLAS with very expensive materials (genomic tiling arrays) for validation of its technology.
In its third phase ATLAS developed an even more challenging technology, which allows performing genome-wide deconvolution studies in pre-defined subsets of complex biological cell populations. This technology comprises the combination of Fluorescence-activated cell sorting (FACS) with the laser ChIP. The technology was developed in two steps: First ATLAS used a discontinuous mode of operation to determine key parameters, in particular the sensitivity (minimal number of cells necessary for ChIP) and selectivity (multidimensional selection of cells). The second step was technically most challenging but offers the highest scientific potential. ATLAS constructed the prototype of a FACS containing a dual pulse femtosecond laser for in situ crosslinking at the single cell level during sorting. With this FACS-laser set-up it will be for the first time possible to crosslink in a heterogeneous population of cells selectively those that are recognised by the FACS according to pre-defined parameters (e.g. cell dimensions, cell surface markers).
Potential applications of such technologies are obviously not only widespread but future development of the field will demand such an option to apply ChIP and ChIP-chip assays to 'real' biological samples, which are inherently composed of multiple cell populations. In this field, some experiments were just done (as reported in results section) and resulted interesting for further investigations. Contribution that requires a European approach. It is clear that neither of individual partners can address this complex multidisciplinary approach either on local or on national levels. Only a unique combination of biologists, chemists, physicists, clinicians, mathematicians, medical doctors and persons from biotechnology industries (SMEs) together with a great contribution from the European Union were able to achieve the expected impact. In the past, some of the partners established a tight collaboration and got preliminary results which could not be followed up without the creation of this consortium. Although all partners are getting some support from their National Foundations, only joined efforts along with a support from the European Union significantly influenced the quality of this research project. We are not aware of any other projects involving experimental biology, data mining, mathematical modelling, informaticians and molecular medicine applied to chemistry and physics. Moreover, the central participation of SMEs allows the development of commercial tools and implements the connection between the industrial and academic scientific work; SMEs enjoyed benefits from the research approach and its application for the creation of LChIP technology; academic partners received from the direct binding with industries for an immediate commercial view of the products of this consortium.
The participation in the consortium of partners performing different, but complementary, investigations enhanced the possibility of exploiting the novel findings of the project.
ATLAS improved Europe's innovation performance by stimulating a better integration between research and innovation and by working towards a more innovation-friendly policy.
On the bases of our good results, one of our long-term goals is to enhance the inclination for turning research into useful and commercially valuable innovations. We believe that successful development of this project based on tight collaboration between groups will help us in the future to develop the 'European Centre of Excellence' for technological excellence.
All partners represent the leading European groups in their field of research and had all potential to achieve the main goal of this very ambitious project. Only the synergy of all partners (which include many disfavoured European countries and country regions) helped us to reach the defined goal but without the financial support the achievements of all impacts (listed at now) would not have been possible.
List of websites: http://www.atlas-eu.com(opens in new window)
