WO2012059563A1 - Genomic analysis method - Google Patents

Abstract

The invention relates to a method for obtaining a specific signature of a DNA molecule. This method makes it possible in particular to obtain signatures for long DNA fragments. It also makes it possible to analyse and compare genomic differences between various individuals.

Description

 Genomic analysis method

The invention relates to a method for obtaining a specific signature of a DNA molecule. This process notably makes it possible to obtain signatures for long DNA fragments. It also allows the analysis and comparison of genomic differences between different individuals.

Introduction:

 At the present time, in the field of medical diagnosis related to structural copy genetic alterations (CNV), the needs are mainly met by two complementary approaches: comparative genomic hybridization technology (or "Comparative Genomic Hybridization"). ", (CGH)) or the individual genome sequencing technology. CGH on metaphase chromosomes is relatively inexpensive but has limitations. The flow rate and accuracy of this technique are low. Only unbalanced chromosomal changes of more than 1 million bases are detected. Structural chromosomal aberrations such as reciprocal translocations or inversions will not be detected. In addition, this technique does not provide information about ploidy. Thus, tetraploids without unbalanced rearrangements will appear normal. There is also a CGH technique on DNA fragments attached to a chip (or CGH array). This technology is more resolutive (up to 5 or 10 kb) but it can not discover any anomaly involving DNA that is not present on the chip, and it is insensitive to translocations and inversions.

More recently, very high throughput sequencing technologies have made it possible, in isolated cases, in basic research, to sequence the genome DNA of patients. The fragments obtained by sequencing are very short (from 35 to 300 bp) and must be aligned on a reference genome to find their respective positions. CNV abnormalities are identified by counting the number of aligned fragments at one position and comparing it to the average number of reads aligned per base in the entire genome. If the number of fragments aligned to a region is 0% or 50% of this average, a complete (homozygous) or partial (heterozygous) deletion of the region will be diagnosed. If the number corresponds to 150% or more, we will conclude to a duplication. In practice this This approach is limited by the difficulty of taking into account sequencing biases, which locally bias the number of fragments obtained. In addition, only about 60% of the regions of the genome corresponding to the genes and 45% of the other regions can be "seen" by this technology. Finally, the cost of sequencing a human genome is still prohibitive.

Methods based on the use of enzymes recognizing specific DNA sequences have also been proposed to produce DNA molecule signatures. One particular example is the study by Neely et al, 2010. This article concerns a method for mapping bacteriophage λ DNA by fluorescent labeling at the level of consensus sequences recognized by the methyltransferase M. Hhal. The labeled DNA is then immobilized and the fluorescence is observed under a microscope to establish a signature termed "fluorocode" by the authors of the article. This technique does not make it possible to obtain a specific signature of a DNA molecule, since the same fluorocode could be obtained for different DNA molecules, for example DNA molecules with consensus sequences in equivalent relative positions. . It further requires several "reads" of the labeled DNA molecule to obtain a fluorocode.

It would therefore be advantageous to have a method for obtaining a specific signature of a given DNA. Preferably, such a method would allow the study of long DNA molecules. It would also be advantageous to have techniques for sequencing at the level of the single molecule.

Summary of the invention

 The present application describes a method for obtaining a specific signature of a DNA molecule. According to the invention, two different molecules will have different signatures. A DNA sequence corresponds to a unique specific signature. The signature obtained by the method according to the invention thus makes it possible to specifically and reproducibly recognize identical DNA molecules and to distinguish different DNA molecules. A first object of the invention therefore relates to a method for obtaining a specific signature of a DNA molecule, comprising:

a) the labeling of a DNA molecule by incorporating at least one modified nucleotide during a DNA polymerization reaction, the modified nucleotide being detectable, or being capable of being modified so as to be detectable; or

 by hybridization of one or more labeled oligonucleotides, said oligonucleotides being between 10 and 15 nucleotides in size; or

 by contacting the DNA with a fluorescent intercalating molecule;

b) immobilizing and stretching the labeled DNA on a solid surface;

c) detecting the labeling along the DNA by means of a suitable optical device;

whereby a specific signature of said DNA sequence is obtained.

Another subject of the invention relates to a method for the identification of a DNA molecule, the method comprising:

i) obtaining the signature of said DNA molecule according to the method described above; and ii) comparing the signature obtained in step i) with predetermined signatures of reference DNA molecules;

the comparison to identify the DNA molecule.

The invention also relates to a method for identifying a genomic anomaly in a subject, said method comprising:

i) obtaining the molecular signature of a genomic DNA molecule of a tissue or cell of a subject, according to the obtaining method described above;

ii) comparing the signature obtained in step i) with the predetermined signatures of reference DNA molecules from a healthy subject to identify said genomic DNA molecule;

a genomic abnormality being detected if the signature of the genomic DNA of the subject differs from the signatures of the reference DNA molecules.

According to a first variant, the method for identifying a genomic anomaly mentioned above further comprises:

obtaining in step i) the signature of several genomic DNA fragments and their concatenation on the basis of overlapping signatures;

the concatenate (s) being then compared with the reference sequences to identify the said concatenate (s); a genomic abnormality being detected if a concatenate differs from the corresponding reference DNA molecule.

According to a second variant, the method for identifying a genomic anomaly mentioned above is used to identify a complementary DNA molecule, the signature of said complementary DNA being compared to the signatures of a set of complementary DNA molecule. reference.

In addition, the subject of the invention is a process for obtaining the nucleotide sequence of a DNA molecule comprising:

a) the labeling of a DNA molecule by incorporation of at least one modified nucleotide during one or more independent DNA polymerization reactions, the modified nucleotide (s) being detectable ( s), or likely to be modified to be detectable;

b) immobilizing and stretching the labeled DNA on a solid surface;

c) the detection of the marking by means of an optical device and a processing means adapted to obtain a resolution information less than or equal to 0.5 nm.

According to a particular embodiment of this object, the detection means used is the TIRF technique coupled to a STORM system.

The invention furthermore relates to a method for sorting different DNA molecules contained in a mixture, the method comprising:

a) the labeling of the DNA molecules

 by incorporating during the course of a polymerization reaction at least one modified nucleotide detectable or capable of being modified so as to be detectable; or

 by hybridization of one or more labeled oligonucleotides, said oligonucleotides being between 10 and 15 nucleotides in size; or

 by contacting the DNA with a fluorescent intercalating molecule;

b) immobilizing and stretching the labeled DNA molecules on a solid surface;

c) detecting the labeling along the DNA molecules by means of a suitable optical device; whereby a set of light profiles corresponding to the different DNA molecules is obtained, the different DNA molecules being discriminated according to their light profile. Detailed description of the invention

 According to a first aspect, the invention relates to a method for obtaining a specific signature of a DNA molecule. This process comprises:

a) the labeling of a DNA molecule

 by incorporating during the course of a polymerization reaction at least one modified nucleotide detectable or capable of being modified so as to be detectable; or

 by hybridization of one or more labeled oligonucleotides, said oligonucleotides being between 10 and 15 nucleotides in size; or

 by contacting the DNA with a fluorescent intercalating molecule;

b) immobilizing and stretching the labeled DNA on a solid surface;

c) detecting the labeling along the DNA by means of a suitable optical device;

whereby a specific signature of said DNA sequence is obtained.

According to the marking and immobilization methods used, the sequence of steps a) and b) can be reversed. Thus, in one embodiment, the labeling precedes immobilization and stretching of the DNA molecule. In a second variant, the immobilization and stretching of the DNA molecule precedes the labeling of said DNA molecule, in particular in the case of a labeling of the DNA molecule by hybridization with labeled oligonucleotides.

In the context of the present invention, the term "signature of a DNA" designates the specific marking obtained along said DNA after optical reading of the signal emitted by the marker employed in step a) of the process according to the invention. As we describe below, the signature can include:

 at the labeling density along the DNA molecule, converted into fluorescence units, or

the sequence of successive distances between the positions of the fluorescent molecules along the DNA molecules determined by using a PSF function.

Said signature obtained according to the invention is "specific" in that, depending on the primary sequence of said DNA molecule, the probability that two DNA molecules can have the same signature is weak, or even null, unlike the signatures obtained by techniques based on the recognition of consensus sequences.

DNA molecule studied

The DNA molecule subjected to the process according to the invention may be any DNA molecule for which obtaining a specific signature is desirable. It can come from a cell of a unicellular or multicellular organism, and in particular be extracted from a sample of cells or tissue of a multicellular organism (for example a sample of hair, skin, blood, etc.). of DNA to be studied may in particular be a molecule of unknown sequence.

The DNA subjected to the process according to the invention may be a DNA molecule, in particular a genomic DNA molecule, extracted from a eukaryotic cell (for example a plant, fungus or animal cell, in particular a mammalian cell). , in particular a human) or prokaryotic cell, or a virus, for example a mammalian virus with a DNA genome, in particular a human virus, or a bacteriophage with a DNA genome. The method according to the invention can also be applied to a complementary DNA molecule. The DNA molecule to be studied may also be a DNA molecule isolated or contained in a DNA library.

The genomic DNA can in particular come from a patient likely to be suffering from a disease involving an abnormality in his genome, for example a genetic disease or a cancer. The genomic DNA may also come from a healthy subject, in particular with a view to carrying out the process according to the invention to obtain the signatures of "reference" DNA molecules (see below).

The method according to the invention may in particular comprise, in a particular embodiment, the prior extraction of the DNA from the cells of the organism whose genome is to be analyzed. As an illustration, in the case of mammals (eg humans or other animals), the DNA can be extracted from lymphocytes contained in a blood sample. Tissue or cell harvesting techniques (especially by biopsy, smear, blood sampling, etc.) and DNA extraction are well known to those skilled in the art. The genomic DNA can in particular be extracted by alkaline lysis of the cells by taking the necessary precautions during handling to minimize the fragmentation of the DNA molecules. According to a particular embodiment, the DNA molecules subjected to the process according to the invention are molecules of high molecular weight, preferably of size greater than 50 kb, in particular greater than 60 kb, in particular greater than 100 kb, more particularly greater than 200 kb. Those skilled in the art can take into account the density of the marking obtained to determine the most suitable fragment size. Thus, a specific signature can in particular be obtained for a 60 kb fragment for a labeling density of 50% of one of the four bases of the DNA (A, T, C or G), that is to say for a labeling of 12.5% of the DNA molecule on average. Those skilled in the art are aware of the techniques for obtaining DNA molecules of suitable size. For example, this operation can be carried out by isolating the cells in agarose blocks in which the solutions of cell lysis and DNA digestion are diffused, followed by a separation of the DNA molecules by electrophoresis, the cutting out of the cells. an agarose block corresponding to the desired size and agarose digestion with agarase.

The DNA to be studied can be anchored or not anchored to the solid surface before stretching. In a particular embodiment, the DNA to be studied is modified so as to allow it to be immobilized on a solid surface (for example by exploiting the biotin-streptavidin system). Anchoring at one end may advantageously make it possible to obtain a particular distribution of the molecules on the solid surface.

According to another embodiment, the DNA molecules are combed "as is" from an unmodified DNA solution at one of its ends (it being understood that the DNA may be DNA modified at the one of its ends, but the modification is not carried out for anchoring on the surface).

DNA labeling

 According to a particular embodiment, the DNA is labeled by incorporation of nucleotide analogues during a polymerization reaction, according to techniques well known to those skilled in the art. These techniques use a DNA polymerase.

The skilled person has several techniques at his disposal to carry out this marking. In particular, it is possible to label the DNA by in vitro synthesis using a recombinant DNA polymerase. Examples of usable DNA polymerases are the Klenow fragment of Escherichia coli DNA polymerase I, Taq and Pfu heat-resistant polymerases and their commercial variants, Phi29 polymerase. Preferably, a sufficiently processive DNA polymerase is used to incorporate modified nucleotides into high molecular weight DNA. Preferably, the DNA polymerase used is a high fidelity polymerase. The Phi29 DNA polymerase available for sale, in particular from New England Biolabs, is an example of a polymerase which combines these properties and which can be used in the context of the present invention. According to another advantageous embodiment, the incorporation of modified nucleotides in double-stranded DNA, especially nucleotides coupled to a fluorophore, is implemented using a DNA polymerase particularly tolerant to modified nucleotides. For example, the mutated form of the polB family polymerase extracted from the bacterium Pyrococcus furiousus (Pfu), designated E10, is capable of substituting up to 100% of cytosines by cytosines coupled to the Cy3 fluorophore, as described in Ramsay and collar. This E10 Pfu polymerase can of course be used to incorporate other labeled nucleotides, especially nucleotides labeled with other fluorescent markers than Cy3, especially the other markers mentioned in the present application. The incorporation of modified nucleotides in the double-stranded DNA molecules (for example a circular plasmid, circular bacterial chromosome, genomic or linearized plasmid DNA or the like) can thus be carried out according to one of the detailed protocols provided in Examples 1. According to a particular embodiment, the in vitro synthesis is carried out by carrying out a single synthesis cycle, so as to obtain labeled DNA molecules on only one of their two strands.

Incorporation can be carried out from cytosol extracts of cells in culture, preferably human cells, containing the components of the replicative machinery. Typically, these components are extracted from cells (HeLa cell type) in culture, are supplemented with the modified nucleotides that it is desired to incorporate into the DNA and brought into contact with DNA, for example DNA genomics, which one wishes to mark. These techniques are known to those skilled in the art (eg Marheineke et al., 2009). According to another embodiment of the invention, the DNA which one wishes to label is DNA located in the nucleus of cells in culture, in which fluorescent nucleotides are introduced by microinjection. The modified nucleotides are then incorporated into the DNA in vivo by the native replicative machinery of the cells. These techniques are known to those skilled in the art (Zink et al., 1998).

The polymerization reaction is carried out in the presence of the four bases of dATP DNA, dTTP, dGTP and dCTP. According to one embodiment, one or more of the nucleotides may be replaced by an analogue. For example, the dUTP can replace the dTTP. According to the invention, the labeling step may be carried out by incorporating at least one nucleotide analogue, said analogue being either a fluorescent analogue or an analogue capable of being modified to make it fluorescent. According to one variant, the polymerization is carried out in the presence of the four bases of the DNA, supplemented with the analogue (s) to be incorporated in the DNA. According to another variant, the analog (s) to be incorporated are used by completely replacing the corresponding unmodified base in the reaction mixture.

Thus, according to a particular embodiment, the method according to the invention comprises the labeling of the DNA with at least one fluorescent analogue of a nucleotide selected from dATP, dTTP (or dUTP), dGTP and dCTP. According to a specific variant of this embodiment, a single fluorescent analogue is employed. When several different analogs of different nucleotides are used, the fluorescent label used is identical or different for each of these analogs. Preferably the fluorescent label is different for each of the analogs used.

Fluorescent nucleotide analogs are well known to those skilled in the art and are commercially available. Any fluorophore capable of being bound to a nucleotide either directly or via a spacer (e.g. (CH ₂ ) _n ) can be used in the present invention. Mention may in particular be made of nucleotide analogs labeled with diethylaminocoumarin, fluorescein, a cyanine, especially cyanine 3 or cyanine 5, tetramethylrhodamine, Lissamine®, Texas Red®, alexa or a marker of the ATTO family (in particular ATT0647N, ATTO 550, ATT0565 available from Jena BioSciences coupled to nucleotides (or the like) according to different variants). Different vendors market this type of fluorescent analogs, including Perkin Elmer, Amersham, etc.

According to one particular embodiment, the markers are photoactivatable molecules as a function of the oxido-reducing conditions of the reaction medium. This embodiment is particularly advantageous in the context of the implementation of the STORM technology described below.

The nucleotide analogue used may alternatively not itself be fluorescent, but may carry a group that can be modified to make it fluorescent. The coupling of fluorophores can be covalent or transient (to overcome any steric limitations). In the case of transient coupling, fluorescent molecules having affinity for the nucleotide analog incorporated in the DNA are used. Bromo-, iodo-, fluoro-nucleotides (especially bromo-, iodo-, fluoro-UTP, -ATP or -CTP), oxo-nucleotides (such as oxo-GTP), nucleotides containing, for example, an alkylene group (for example 5-ethynyl-dUTP, C8-Alkyne-dUTP, γ-Propargyl-ATP) and nucleotides modified with a group containing an amino function, in particular an aminoallyl function such as aminoallyl-dUTP, and aminoallyl-dCTP .

Coupling a fluorescent label to a modified nucleotide is well known to those skilled in the art.

 By way of illustration, the fluorescent labeling of a modified nucleotide carrying an amino function such as aminoallyl-dUTP or aminoallyl-dCTP may be carried out by reaction of the modified DNA with a fluorescent group carrying an N-function. hydroxysuccinimide (NHS), sulfo-NHS or sulfotetrafluorophenyl (STP) which will allow coupling at the amino function. The fluorescent label employed can thus be an NHS ester of a fluorescent molecule such as Cy3-NHS, Cy5-NHS, ATTO-NHS, Alexa-NHS, fluorescein-NHS, rhodamine-NHS or others.

Fluorescent labeling of a modified nucleotide carrying an alkyl function such as 5-ethynyl-dUTP can be achieved by the "Click Chemistry", a reaction of the modified DNA with a solution of copper bromide (CuBr), tris (benzyltriazolylmethyl) amine and a fluorescent molecule containing an azide group (such as fluoresceinazide, cyanine3-azide, cyanine-azide, etc.). In a particular embodiment, the coupling with a fluorescent marker may be carried out on a single-stranded DNA molecule comprising a nucleotide modified to increase the efficiency of this coupling. The methods for obtaining a single-stranded DNA are well known to those skilled in the art. In particular, the addition of a concentration of concentrated urea to the double-stranded DNA solution may be mentioned. The fluorescent labeling is then carried out after purification of the single-stranded DNA.

According to one particular embodiment, the labeling is carried out by replication of a DNA molecule (template DNA) as described above with the Phi29 DNA polymerase, in the presence of dATP, dTTP (or alternatively dUTP), dCTP. and unmodified dGTPs and one or more fluorescent or fluorescent nucleotide analogues. According to a variant of this embodiment, the nucleotide analogue completely replaces the corresponding unmodified base in the reaction medium. The polymerization reaction is carried out at the temperature required for the operation of the enzyme according to the supplier's recommendations (30 ° C. for Phi29) and for a time sufficient to allow the incorporation of modified nucleotides over several tens of kilobases over a period of time. at least part of the matrix molecules. According to a second embodiment of the invention, the labeling is carried out by hybridization of the DNA molecule with one or more labeled single-stranded oligonucleotides of size between 10 and 15 nucleotides. The sequence of these oligonucleotides is chosen so that it is distributed approximately uniformly in the genome of interest, while allowing to obtain a specific labeling of a given DNA molecule. In the case of the human genome, these oligonucleotides are chosen according to the composition of the genome, and their distribution. The sequence of the oligonucleotides may be of the form NNX1X2X3X4 5 6 7 8NN, wherein X1X2X3X4X5X6X7X8 is a defined sequence and N is a random base. In this way the determined sequence of the oligonucleotide will be sufficiently short (for example 8 bases) to have numerous occurrences in the genome, and sufficiently long by addition of random bases (for example 8 + 4 = 12 bases here) for hybridize stably. The oligonucleotides may in particular be labeled by incorporation of one or more fluorescent molecules at their end, according to procedures well known to those skilled in the art. According to a particular aspect of this second embodiment, the hybridization labeling is carried out after immobilization and stretching of the DNA molecule to be studied.

According to a third particular embodiment, the labeling of the DNA molecule is carried out by contacting said DNA with an intercalating agent capable of binding to the double-stranded DNA and which emits a fluorescent signal when it is so bound. . Many intercalating agents are known to those skilled in the art. There may be mentioned, for example, ethidium bromide, propidium iodide, Sybr Green, TOTO (1,1 '- (4,4,7,7-tetramethyl-4,7-diazaundecamethylene) -bis- 4- [3-methyl-2,3-dihydro- (benzo-1,3-thiazole) -2-methylidene] quinolinium tetraiodide) or orange oxazole (YOYO).

The intercalating agent can be used to label the DNA molecule before or after immobilization on a solid support. Preferably, the labeling is carried out before immobilization of the DNA molecule, in particular using the molecule YOYO. Immobilization and stretching of DNA

 Step b) of the method for obtaining a specific signature of a DNA molecule comprises immobilizing and stretching said DNA on a solid support.

Preferably, the labeled DNA is immobilized and stretched by the molecular combing technique. This technique is well known to those skilled in the art. A detailed description will be found in particular in application WO9521939, incorporated herein by reference. In summary, this technique exploits the movement of a solvent meniscus that stretches and orients the DNA molecules in solution in the direction of motion, while locally applying a force that clumps the DNA molecules onto the solid surface. A stretch rate of 1 base / 0.5 nm can be obtained with this technique.

The surface on which the DNA molecules are immobilized is rigid (for example glass) or semi-rigid (for example a flexible polymer sheet). The surface can in particular be composed or covered with an organic or inorganic polymer, a metal, metal oxide, semiconductor, or a combination of these materials.

According to a particular embodiment, the surface is covered, in whole or in part, with groups capable of reacting with one end of the DNA molecules in solution and thus of fixing it. Of course, one skilled in the art knows the different systems that can be used to fix a DNA molecule and will know in which cases the DNA molecule must be functionalized in parallel with the surface. By way of example, mention may be made of the streptavidin-biotin system, an antibody-antigen system, and the like.

According to another particular embodiment, the DNA is not modified in view of its anchoring. For example, silanized glass slides are used (for example by soaking the glass slide in a bath of 80% trimethoxy (7-octen-1-yl) silane) which has not undergone additional treatment and the DNA molecule is then immobilized by application of the molecular combing technique. According to a variant of this embodiment, the support may also be a support composed of or covered with a synthetic polymer, polyacrylamide, polycarbonates, etc.

In a particular embodiment, the labeled DNA is immobilized in an individualized manner. In other words, the labeled DNA is diluted to an adequate concentration to obtain, after immobilization, labeled DNA molecules separated from each other. By way of illustration, DNA fragments of 50 to 200 kb in length are used at a concentration of 0.2 to 1 μM. The dilution may be carried out in 50 mM Tris buffer, pH 8 or 50 mM MES, pH 5.5, for example.

The molecular combing can be carried out in a suitable apparatus for moving the carrier at a constant speed out of a reservoir containing the labeled DNA to be fixed and combed. The speed of movement of the support out of the tank can be adapted by those skilled in the art. By way of illustration, the examples below show the implementation of a movement at a constant speed of 300 μιη / s out of the tank.

Observation

After immobilization, the stretched DNA molecules are observed using an optical microscope to generate a fluorescent signal. The microscope may include a Total Internal Reflection Fluorescence (TIRF) system to minimize noise around the DNA molecule to be studied. In a particular embodiment, the microscope is equipped with a magnifying objective chosen so as to optimize the observation of the labeled DNA molecule. The objective may especially have a magnification of at least 40 ×, preferably at least 50 ×, at least 60 × or even at least 100 ×. In another particular embodiment, the microscope is equipped with a high resolution CCD camera (for example 6.5 μιη × 6.5 μπι, or 65 nm per pixel for a 100 × objective) with high quantum efficiency (preferably greater than 65%). Under these conditions, one pixel of the camera covers a length of about 130 base pairs of DNA stretched and immobilized on the surface.

According to a preferred embodiment, the microscope used comprises a TIRF system and is equipped with a magnification objective of at least 100 × and a high resolution CCD camera (for example 6.5 μπι × 6.5 μπι, or 65 nm per pixel) with high quantum efficiency (preferably greater than 65%).

Fluorescent molecules are excited by a laser of suitable wavelength, known to those skilled in the art. The light emitted is captured by the camera attached to the microscope whose lens is equipped with a suitable optical filter. Suitable means for obtaining a fluorescence are therefore used. It will be possible to mention the use of appropriate excitation, dichroic and emission filters depending on the fluorophore used to label the DNA and the use of a laser excitation at a suitable wavelength. By way of illustration, the examples describe filters that can be used for the detection of the Cy3 marker or the use of a 633 nm wavelength laser to excite the Cy5 or 532 nm to excite the Cy3, the a person skilled in the art knowing the means necessary to obtain a fluorescence from other markers.

Of course, those skilled in the art will be able to adapt the equipment to obtain the best possible resolution. According to a preferred embodiment, the microscopy implemented is a super-resolution fluorescence microscopy. In a preferred aspect, the super-resolution microscopy used implements a successive localization of individual fluorophores. Among the available technologies, mention may be made of PALM (photoactivated localization microscopy) and STORM (stochastic optical reconstruction microscopy) which make it possible to overcome the resolution limit imposed by diffraction. These techniques are described in the articles by Betzig et al, 2006 and Rust et al., 2006, which are incorporated herein by reference. These techniques implement photoactivation of the fluorophores incorporated in the DNA by flashes of light that randomly excite a small number of fluorescent molecules, rather than continuously excite them. These fluorescent molecules are then imaged until the fluorophore is destroyed or reverts to an inactive state.

According to a particular embodiment, the microscope used comprises a TIRF system and a STORM or PALM system, preferably STORM.

The light signals collected in the form of images are then processed by a computerized process for the purpose of locating the relative position of the fluorophores (and thus the modified bases in the DNA) with respect to each other along the linear axis DNA molecules, and therefore their relative distance from each other.

The determination of the light profile of each molecule present on an image can be performed automatically by an algorithm implemented in a computer program. According to one particular aspect, the algorithm can in particular be carried out in two stages. The first includes identifying the peaks of maximum intensity on the foreground pixels along a ridge line. A peak line is a high curvature surface (maximum curvature in at least one direction) produced by the image of the DNA molecule, defined by high absolute values of the smallest eigenvalues of the symmetric matrix of the partial derivatives of second order (Hessian matrix) of each pixel. According to a particular embodiment, a Gaussian blur (sigma = 1 pixel) can be applied in order to reduce the influence of the noise. The second step comprises the determination of a spline curve by interpolation of the peak curvature peak points established in the first step. The curve is sampled at equidistant points (for example at each pixel, or at each pixel). half-pixel) to approximate the value of the surface at these points. The approximation can be done by directly taking into account the value of the pixel, by Taylor's approximation of the surface at the chosen point, or by a Richardson-Lucy deconvolution based on a "point spread function". The successive list of these values constitutes the fluorescent profile of the DNA molecule (FIG. 3). In one embodiment of the invention based on STORM or PALM technology, a Point Spread Function (PSF) is a mathematical function describing the signal emitted by a fluorescent molecule. unique, will be adjusted on the light signal to break it down into as many PSFs as single emitting molecules. This operation makes it possible to point the center of each PSF and thus to position the emitting molecules relative to one another along the axis of the immobilized DNA molecule and stretched (preferably by molecular combing). The pointing operation is carried out by means of algorithms well known to those skilled in the art (for example, Mortensen et al, 2010). The center of each PSF, corresponding to the position of each fluorescent molecule, can thus be located with a precision depending on the number of photons acquired, and not on the diffraction limit of the emitted light. The sequence of successive distances between the positions of the fluorescent molecules along the DNA molecules thus provides a specific signature of each of the molecules, expressed as a base pair of DNA. In another embodiment of the invention, the pointing of the individual molecules producing the light signal is not required, the linear light signal, expressed in units of arbitrary intensity per nanometer, directly representing the specific signature of each molecule of light. DNA. In this case a signal decomposition operation can be performed to increase the resolution of the signature. This operation can be performed by methods well known to those skilled in the art, such as wavelet-based methods (Donoho et al., 1995) or deconvolution (Bertero and Boccacci, 1998).

It is possible that different fragments representing different copies of the same DNA sequence are immobilized on the surface in the process according to the invention. The method for obtaining a specific signature of a DNA molecule may then further comprise, after step c), a concatenation step where signature overlaps are identified between several fragments, indicating that the same region of the genome is observed. In this case, the two signatures are then merged to generate a longer signature. Other objects of the invention

The method described above makes it possible to obtain the signature of a DNA molecule. This signature can advantageously be implemented to constitute reference libraries of DNA molecules.

Thus, according to one of its aspects, the invention relates to a specific signature of a given DNA obtained according to the method detailed above.

The signatures of the reference DNA molecules can advantageously be used to identify unknown DNA molecules. Accordingly, another aspect of the invention relates to a method for identifying a DNA molecule comprising: i) obtaining the signature of said DNA molecule according to the method of obtaining described above ; and

ii) comparing the signature obtained in step i) with predetermined signatures of reference DNA molecules;

said comparison to identify the DNA molecule.

The signature obtained in step i) is compared by a computer process based on statistical correlations.

It is possible that different fragments from the same DNA molecule are immobilized on the surface in the process according to the invention. As mentioned above, the identification method may then further comprise, between steps i) and ii) a concatenation step where signature overlaps are identified between several fragments, indicating that the same region of the genome is observed. In this case, the two signatures are then merged to generate a longer signature.

The reference signatures may in particular correspond to the signatures of DNA molecules obtained from healthy subjects. There may be mentioned genomic DNA of a healthy subject, or complementary DNA obtained from the transcriptome of a cell or tissue of a healthy subject. In the case of complementary DNA, reference signatures may be produced for each tissue / cell type. By "healthy subject" is meant according to the invention a subject for which no genomic alteration is suspected. Preferentially, the DNA molecule signatures of healthy subjects are representative of the diversity of the populations considered, especially human. The reference DNA molecules may correspond to the signature of chromosomes, for example to the signature of the 23 human chromosomes. The complete signature of a chromosome will of course have been obtained by concatenation of the overlapping signals corresponding to this chromosome, according to the modalities described above. The reference signatures may also correspond to DNA molecules of patients suffering from genetic diseases diagnosed or suffering from a documented cancer. These signatures can allow the identification of the pathologies concerned.

Of course, as mentioned above, the applications of the method according to the invention are not limited to the study of the human genome. The DNA studied can in particular be extracted from an animal, in particular a mammal (including man), but also from a plant, a fungus, a prokaryote, or a virus, for example a virus. mammalian DNA genome, in particular a human virus, or a bacteriophage with DNA genome. In particular, the reference signatures may correspond to DNA molecules from each of these organisms.

The method according to the invention may therefore in particular be a method of identifying a DNA extracted from a plant that is to be identified. In this context, the signature of the DNA studied will be compared with reference signatures from different plants or varieties of plants. According to a particular embodiment, the reference signatures comprise all or part of DNA signatures of genetically modified plants. Thus, the method according to the invention can also make it possible to determine whether a given plant is genetically modified. In another aspect, the invention provides a method for identifying a genomic abnormality in a subject comprising:

i) obtaining the molecular signature of a genomic DNA molecule of a tissue or cell of a subject; ii) comparing the signature obtained in step i) with the predetermined signatures of reference DNA molecules from a healthy subject to identify said genomic DNA molecule;

a genomic abnormality being detected if the signature of the genomic DNA of the subject differs from the signatures of the reference DNA molecules. These differences include deletions, inversions, duplications of chromosome segments.

According to another embodiment, the DNA obtained in step i) is compared with predetermined signatures of reference DNA molecules from diseased subjects whose disease is the result of an alteration in the genome of one or more cells of said patient, a genomic anomaly that may be suspected if the signature obtained in step i) corresponds to one of the reference signatures. The invention therefore also relates to a method for identifying a genomic abnormality in a subject comprising:

i) obtaining the molecular signature of a genomic DNA molecule of a tissue or cell of a subject;

ii) comparing the signature obtained in step i) with the predetermined signatures of reference DNA molecules from diseased subjects, whose disease is the result of genetic alteration, in order to identify said DNA molecule genomics;

a genomic abnormality being detected if the signature of the genomic DNA of the test subject corresponds to at least one of the reference signatures.

Moreover, the invention may advantageously be used to diagnose genetic damage in the context of amniocentesis. This method allows for a finer identification of genetic alterations (inversion, deletions, duplications, translocations at one or more chromosomes), and cytogenetic karyotyping. The method for identifying a genomic anomaly according to the invention can be implemented independently or in addition to such karyotyping.

The invention also relates to a method for sorting molecules. This process is applied to a mixture of different DNA molecules. According to this aspect, the invention therefore relates to a method of sorting different DNA molecules included in a mixture, the method comprising:

a) the labeling of the DNA molecules by incorporating during the course of a polymerization reaction at least one modified nucleotide detectable or capable of being modified so as to be detectable; or

 by hybridization of one or more labeled oligonucleotides, said oligonucleotides being between 10 and 15 nucleotides in size; or

 by contacting the DNA with a fluorescent intercalating molecule;

b) immobilizing and stretching the labeled DNA molecules on a solid surface;

c) detecting the labeling along the DNA molecules by means of a suitable optical device;

whereby a set of light profiles corresponding to the different DNA molecules is obtained, the different DNA molecules being sorted according to their light profile.

The sorting of molecules is of interest in the case where a DNA extract represents very small quantities of molecules (for example obtained at a crime scene, or on traces in a container having contained food potentially contaminated with a GMO , etc). If the amount of DNA is too small to consider existing DNA characterization techniques such as sequencing, but it is nevertheless desirable to recognize the presence of a particular redundant sequence in the mixture (eg to identify the presence of Human DNA through the repeated elements of the genome, or a multi-copy plasmid containing a transgene in a potentially genetically modified organism), the molecule sorting is a very interesting alternative. In this case the DNA molecules in the mixture typically have some redundancy due to the existence in the mixture of molecules having the same DNA sequence, in part or in part. In the latter case, the molecules are said to be overlapping. Thus, the mixture may comprise n subgroups of DNA molecules, each subgroup having one or more properties specific thereto (eg, a biological origin or a structural property). In this case, the sorting method according to the invention can also be followed by a comparison step 2 to 2 of the different light profiles obtained, thus making it possible to assemble profiles on the basis of the overlap observed between two profiles. According to a specific embodiment, the profiles of each molecule are cut into blocks to make the comparison. This approach has the advantage of avoiding any possible problem of discontinuity in the similarity of the profiles along the molecules, since the pairing of the molecules that share a common sequence part will be done locally (by means of the blocks). rather than global (through complete profiles). Thus, it is possible to cut the light profiles obtained in blocks of size t overlapping by p pixels, with p <t. A similarity measure (for example a Euclidean distance) is calculated between all possible pairs of blocks. We then look for situations where a succession of blocks belonging to a given molecule show a strong similarity to a succession of blocks of another molecule. This situation is characteristic of two molecules sharing a region of the same DNA sequence.

For example, it is possible to determine the Euclidean distance between two blocks bl and b2 of size N, given by the following equation.

We define for each block of each molecule, its k best neighbors (for example with k = 50, 100, 150, etc.) among all the blocks of all the other molecules. The k nearest neighbors (k Nearest Neighbors, kNN) of a given block are those which are most similar to it because they are removed by a small Euclidean distance. According to an alternative embodiment, other strategies for calculating the similarity between two profiles or sub-segments of the profile of molecules include:

 - standardized cross-correlations

 - standardized cross-correlations with partial overlap

 - Euclidean distance on Fourier transforms at term

- distance from Hellinger

 - Minimum Variance Matching (MVM) according to Keogh

 the Euclidean distance on thresholded wavelet coefficients

In a particular embodiment, the similar blocks are then represented on a matrix. For example, for all the possible pairs of molecules {m _1} m ₂ ), among the kNNs of each block of m _}, the blocks belonging to m ₂ , which can be represented on a two-dimensional graph, whose axes are the positions of the respective blocks of m _} and m ₂ (see Figure 4 for illustrative purposes).

A correlation coefficient between two molecules can also be determined. In particular, in order to quantify the comparison of the mi and m ₂ molecules, the Pearson correlation coefficient of the positions of the blocks represented in the matrix (m 1,1112). The Pearson correlation coefficient classifies all possible pairs of molecules (mi, m ₂ ) from most similar to least similar. If necessary, the calculation of the Pearson coefficient can be replaced by the calculation of the Kendall coefficient.

Of course, these calculation methods can also be applied to the other methods according to the invention requiring a profile comparison.

After assembling the different light profiles identified on the support, it is possible to obtain a specific signature (which corresponds to the succession of profiles obtained by the assembly) for each different DNA molecule present in the initial mixture. These specific profiles can then be used to identify the DNA molecules present in the mixture, according to the procedures developed above.

In view of the resolutions obtained using superresolution microscopy techniques, there is also described a method for obtaining the nucleotide sequence of a DNA molecule, comprising:

a) the labeling of a DNA molecule by incorporation of at least one modified nucleotide during one or more independent DNA polymerization reactions, the modified nucleotide (s) being detectable ( s), or likely to be modified to be detectable;

b) immobilizing and stretching the labeled DNA on a solid surface;

c) the detection of the marking by means of an optical device and a processing means adapted to obtain a resolution information less than or equal to 0.5 nm.

This resolution can in particular be achieved by the use of TIRF technologies coupled to a STORJVI or PALM system described above, and analyzed by pointing PSFs as also described above. The resolution of an optical microscope is commonly considered to be λ / 2, about 250 nm. However, the center of the image of a point (for example a single fluorophore) can be located with much greater accuracy. Indeed, if the observation is performed by a microscope equipped with a CCD camera whose pixels measure a nm side, and whose effective noise value of each pixel is b (in number of counted photons), then the uncertainty σ on the position of the photon emitter (Thompson, RE, Larson, DR, Webb, WW Precise nanometer localization analysis for individual fluorescent probes Biophys J. (2002) 82: 2775-2783) is given by s ² + a ² l \ 2 8TTS ⁴ b ²

σ = 1 ~

N ² N

with s (in nm) the standard deviation of PSF and N the total number of photons collected. It is found that this uncertainty is in theory limited only by the number N of photon (the other parameters being constants), and thus makes it possible to reach a precision less than one nanometer. With methods of correcting the movement of the microscope drift well known to those skilled in the art, and after calibration of the quantum response of each pixel, such resolution has been achieved in practice (Pertsinidis, A., Zhang, Y., Chu, S. (2010) Subnanometer single-molecule localization, registration and distance measurements, Nature 466: 647-51). According to the method for obtaining the nucleotide sequence of a DNA molecule according to the invention, the DNA studied can be a genomic DNA or a complementary DNA.

Legends of the figures

 Figure 1: Example of image obtained after labeling a DNA with an extract of Xenopus egg cytoplasm, combing and acquisition of fluorescence

 Figure 2: Light signal converted into arbitrary fluorescence units by image analysis software (ImageJ, pink curve (top curve)) while a representative sample of the surrounding background noise is represented by the blue curve (curve bottom).

Figure 3. Example of the comparison of the profile of two lambda phage molecules. The molecule 1 on the abscissa has 208 blocks, and the molecule 2 on the y-axis has 147 blocks.

A blue dot is drawn at coordinates when a block i of molecule 1 has a block j of molecule 2 among its k best neighbors (kNN.

 Figure 4. Histogram of the number of pairs recognized as being similar by the (ordered) method as a function of the minimum correlation coefficient threshold required to consider two pairs as being similar (abscissa).

Figure 5. Distribution of the correlation coefficients obtained by comparing the profiles of two phage lambda molecules (solid line curve) and the profiles of a lambda phage molecule and a mouse DNA molecule (dotted line curve). Examples

 Example 1. Marking with Xenopus laevis replicative egg system extracts

1 °) Marking of DNA

The labeling technique is described in detail in Marheineke et al, 2009 and is based on a protocol originally described in Blow et al., 1986. Briefly, the protocol requires the extraction of cytosol from xenopal eggs. For this, eggs are collected on female xenopus and rinsed in deionized water for 5 minutes. The eggs are then dissociated with an appropriate solution ("thaw" solution) and activated with an appropriate solution (Barth solution) supplemented with 0.25 μg / ml calcium ionophore A23187. After rinsing, the eggs are centrifuged at 2 ° C at a rate of 350 x g for 1 minute to allow removal of any trace of liquid solution and then at 20000 x g for 15 minutes to break them. Of the three liquid phases that result, the central phase is the cytoplasmic extract, which is collected with a syringe and transferred to a tube placed on ice until use. For labeling, a mixture of 1/20 of cytosol extract, 1/20 of creatine phosphate solution, 1/50 of cycloheximide solution, 1/50 of dUTP-rhodamine solution is prepared. After completion of the reaction, it is stopped by the addition of ice cold PBS solution. The DNA is centrifuged at 1000 × g for 7 minutes and resuspended in 50 μl of PBS.

2 °) Preparation of glass slides for molecular combing

 The preparation of the blades for combing is described in detail in Marheineke et al. First, the coverslips are carefully cleaned by treatment in a plasma oven to remove any traces of organic matter. The coverslips are then sonicated in pure heptane for 10 minutes and then a silane layer is deposited by incubation for 16 hours in a solution of 100 ml of heptane supplemented with 100 μl of octenyltrichlorosilane. The coverslips are then sonicated for 5 minutes each time in a succession of liquid: heptane, water, 50% methanol / water, water, chloroform, water, 50% methanol / water, water and then 2 minutes in chloroform. The slats are finally air dried.

3 °) Molecular combing

The DNA labeled in step 1 in a volume of 50 μl is incubated for 30 minutes at 65 ° C., cooled and then incubated for 30 minutes at room temperature. The DNA is diluted in 1.2 ml of MES solution at pH 5.7-6.2. The mixture is introduced into the tank of the combing apparatus. A coverslip prepared in step 2 °) is fixed at the location provided in the apparatus to combing and immersed in the DNA solution for 5 minutes. The coverslip is removed at a constant speed of 300 μιη / s, which causes the DNA to be combed on both sides of the coverslip. This is then dried in air and fixed on a glass slide. The combing of the DNA is visualized with an optical microscope equipped with a 100X objective, an immersion in oil and a FITC filter.

4 °) Acquisition of images

The slats are placed under the objective 100X of an optical microscope equipped with a CCD camera (type CoolSnap HQ, Photometrics) having a grid of 512 x 512 photosensors of 6.5 μπι ² . An excitation light of 510 nm is applied by a laser light, and a matched filter is used to collect the signals. 16-bit TIFF images are captured and transmitted for analysis to computer systems (Figure 1).

5 °) Image analysis

 An example of an image obtained by this method is given in FIG. 1, and an example of an image analysis result is given in FIG. 2. The luminous trace converted into fluorescence units represents the digital signature of this molecule. DNA.

A vertical rectilinear trace corresponding to an individual molecule of a DNA fragment of about 30,000 bases, labeled with fluorescent rhodamine molecules covalently fixed on UTP bases, analogous to thymidine, is observed (FIG. 1). The scattered light spots correspond to background noise.

 In Figure 2, the light signal has been converted into arbitrary fluorescence units by image analysis software (ImageJ, pink curve - top curve) while a representative sample of the surrounding background noise is represented by the blue curve (bottom curve). Example 2. Hybridization Labeling of Fluorescent Oligonucleotides

 A specific signature of a DNA molecule can be obtained by hybridization on these molecules of short DNA sequences (oligomers) labeled during their synthesis and thus covalently by a fluorescent molecule. The process comprises four main steps:

1) combing high molecular weight DNA (eg genomic DNA) on glass slides as described above (this DNA is not labeled with fluorescent molecules.

2) the hybridization of this DNA by single-stranded oligomers of 10 to 15 bases, associating known and random sequences as described above, and known to be distributed approximately nearly uniformly in the genome of interest. These oligomers have at one of their ends one or more fluorescent molecules. Hybridization results in base complement attachment of the oligomers to the combed DNA molecules at sequence-dependent positions.

3) Images of the combed DNA molecules are made in hyper resolution (TIRF and STORM system).

 4) A computer analysis of the images converts the sequence of the light signals generated by the oligomers, and the distance that separates them, into a specific signature of each molecule of combed DNA. The use of different types of fluorophores, each associated with a specific oligonucleotide sequence, can provide a signal of increased complexity, which in turn increases the specificity of the signatures.

EXAMPLE 3 Marking of DNA with Amino-Allyl Nucleotides The incorporation of modified nucleotides into double-stranded DNA molecules (for example a circular plasmid, a circular bacterial chromosome or genomic or linearized plasmid DNA or other) is carried out by in vitro synthesis using a Phi29 DNA-dependent polymerase (eg Fermentas # E0402, 10 U / μΙ). Briefly, a solution of 25 ng / μΐ of DNA is prepared in 10 μl of pure water supplemented with 0.5 μΐ of cresol red. The DNA is denatured by the addition of 1 μl of 0.2 M KOH buffer and incubation for 3 minutes at room temperature. The mixture is neutralized by the addition of 1 μl of Tris-Cl buffer. The polymerization reaction is conducted by preparing the following mixture: 1 μΐ _^ random oligomers of 6 bases, 3 μΐ denatured DNA prepared above, 6 μΐ of pure water, 1 μΐ of pyrophosphatase (10 U / μΙ) , 2 μl of 10mM dCTP amino allyl, 1 μl of each dATP nucleotide, dTTP, 25mM dGTP, 2 μl of 2.5mM dCTP, 5 μl of reaction buffer provided with the commercial enzyme , 1 μΐ of Phi29 enzyme, 26 μΐ of pure water. The mixture is incubated for 16 hours at 37 ° C. and then the DNA is denatured for 20 minutes at 70 ° C. the DNA is purified on a filtration column (for example Macherey-nagel Nucleospin Extractll).

The coupling between the modified nucleotides bearing amino-allyl groups (for example dUMP-aa or dCMP-aa) is achieved by a chemical reaction known to those skilled in the art. Briefly a solution of 10 μl of 0.1 M carbonate buffer and pH 9.3 containing 200 mg of dissolved fluorescent group (for example Cy3-NHS or Cy5-NHS) is added with 1 μg of DNA containing the modified nucleotide and 80 μl of pure water. The mixture is incubated overnight at 25 ° C with gentle and constant agitation. The uncoupled fluorophores are then removed by precipitation, for example by means of an acetate buffer or by a commercial purification kit (for example Promega ref # A7280 or Macherey-nagel Nucleospin Extractll). The purified DNA is redissolved in pure water (for example 50 μl). The labeling efficiency is achieved by spectrophotometric measurement known to those skilled in the art, by calculating the absorption ratio between the DNA (260 nm) and the fluorophore (at the specific wavelength of the fluorophore).

If necessary, more efficient coupling can be achieved on single-stranded DNA containing the modified nucleotides. After the polymerization step in the presence of modified nucleotides bearing an amino-allyl group, the DNA is denatured by methods known to those skilled in the art, for example by addition of concentrated urea. After purification the DNA is treated as described above to obtain fluorescent labeling.

Example 4: DNA labeling with ethynyl-dUTP nucleotides and "click chemistry"

Nucleotide 5-Ethynyl-dUTP (5-EdUTP) can also be incorporated into double-stranded DNA by the same protocol as described in Example 3. The coupling between the 5-EdUTPs and the fluorescent group follows a different protocol. as following. A solution of 5 μΐ of DNA (2 mM) in pure water is added with 2 μΐ of a solution of azide (50 mM), 3 μl of a solution of 0.1 M CuBr, 0, 1 M tris buffer (benzyltriazolylmethyl) amine. The mixture is incubated at 25 ° C for 3 hours. The DNA is purified by acetate precipitation by a protocol well known to those skilled in the art.

If necessary, more efficient coupling can be achieved on single-stranded DNA containing the modified nucleotides. After the polymerization step in the presence of modified nucleotides bearing an ethynyl group, the DNA is denatured by methods known to those skilled in the art, for example by addition of concentrated urea. After purification the DNA is treated as described in the previous paragraph to carry out the fluorescent labeling. Example 5 Incorporation of Fluorescent Nucleotides with the E10 Mutant of the Pfu Polymerase

The direct incorporation of fluorophore-coupled nucleotides into double-stranded DNA is possible through the use of particularly modified nucleotide tolerant DNA polymerases. For example, the mutated form of the polB family polymerase extracted from the bacterium Pyrococcus furiousus (Pfu), designated E10, is capable of substituting up to 100% of cytosines by cytosines coupled to the Cy3 fluorophore, as described in Ramsay and collar. (J. AM CHEM SOC, 2010, 132, 5096-5104).

Example 6 Molecular Combing

Preparation of the support

 A glass microscope slide having a surface area of 22 × 22 mm and a thickness of 0.17 mm is washed twice successively in acetone, isopropanol and pure water, dried and placed immediately in a plasma furnace (eg Diener electronics Femto) for a treatment of 2 minutes at 80% intensity. The slide is then immediately soaked in a bath composed of 100 ml of heptane and 150 μl of Trimethoxy (7-octen-1-yl) silane at 80% for a few seconds and then dried for 12 hours at room temperature. The slide is finally sonicated successively in a bath of heptane, of pure water and finally of chloroform, for 5 minutes each time.

combing

 Molecular combing is carried out in an ad hoc apparatus, making it possible to move a glass slide out of a reservoir at a constant speed. A DNA solution to be combed is prepared as follows: 150 ng of DNA suspended in water, 150 μΐ of MES buffer (for example euromedex ref # EU0033), completion to 1.5 ml with water pure. The solution is placed in the tank of the device. A glass slide 22x 22 mm and 0.17 mm thick prepared as described in 2.1 is placed on the suitable support and incubated for 5 minutes in the DNA solution. The strip is then moved at a constant speed of 300 μιη / s out of the tank.

Example 7 Image Acquisition by Light Microscopy 7.1. High resolution optical microscopy

 A microscope slide on which the DNA molecules are immobilized is mounted on an optical microscope (eg Nikon Eclipse Ti) equipped with a high resolution camera (for example Coolsnap HQ2 from Roper Scientific Photometrics or Ixon 897 from ANDOR Technology). Anti-whitening mounting liquid is used (Spector lab). Images are acquired through an objective of 60X (for example Plan Apo Vc, 60x / l, 40 OIL; ∞ / 0, 17 WD 0, 13 of Nikon) or 100X (for example Plan APO Vc; 100x / 1, 40 OIL; ∞ / 0, 17 Nikon WD DIC N2) by immersion in oil (eg type F; Ne = 1.518 ue = 41 23 ° C from Olympus). Appropriate excitation, dichroic and emission filters are used depending on the fluorophore used to label the DNA. For example, an excitation filter FF01-531 / 40-25, a dichroic filter FF562-Di02-25x36 and a transmission filter FF593 / 40-25 from SEMROCK Brightline will be used for the Cy3. For the Cy5 an excitation filter FF01-628 / 40-25, a dichroic filter FF660-D101-25x36 and a transmission filter FF01 -692 / 40-25 from SEMROCK Brightline will be used.

7.2. Implementation of the TIRF system

 An alternative method to 7.1 is to use the Total Internal Reflection Microscopy (TIRF) technique. The changes are as follows: a lens with a large numerical aperture Nikon APO TIRF; 100x / l, 49 OIL; ∞ / 0.13-0.20 DIC N2) and the preferred mounting oil is, for example, Nikon Immersion oil type F; Nd = 1.515 - vd (dispersion) = 43 at 23 ° C - viscosity = 800cSt. The excitation is produced by a wavelength laser adapted to the chosen fluorophore, for example for Cy5 the excitation is produced at 633 nm (Vortran Laser technology (class IIIB)) or for Cy3 at 532 nm (Melles Griot 43 Laser ion series). The laser beam excites the medium with an angle of incidence greater than the critical angle thanks to the "TIRF" mirror present in the microscope.

Example 8 Obtaining a Light Intensity Profile on Individual DNA Molecules

Get a profile by peak detection

The determination of the light profile of each molecule present on an image is performed automatically by an algorithm implemented in a computer program. The algorithm takes place in two stages. A ridge line is a high curvature surface (Maximum curvature in at least one direction) produced by the image of the DNA molecule, defined by high absolute values of the smallest eigenvalues of the symmetric matrix of the second order partial derivatives (Hessian matrix) of each pixel. According to a particular embodiment, a Gaussian blur (sigma = 1 pixel) can be applied in order to reduce the influence of the noise. The second step includes determining a spline curve by interpolation of the maximum curvature points detected on the peak established in the first step. The curve is sampled at equidistant points (for example at each pixel, or at each half-pixel) in order to approximate the value of the surface at these points. The approximation can be done by directly taking into account the value of the pixel, by the Taylor approximation of the surface at the chosen point, or by a Richardson-Lucy deconvolution based on a "point spread function". The successive list of these values constitutes the fluorescent profile of the DNA molecule.

Example 9 Sorting Molecules on the Basis of Their Light Profiles

The purpose of this application is to sort a group of molecules composed of n different types of molecules in n distinct subgroups, each type of molecule being able to be represented a large number of times. The (real) example below is to show that it is possible to distinguish lambda phage molecules on the one hand, and molecules from the E genome. coli on the other hand.

9.1. Cutting profiles into blocks

 The profiles of each molecule are cut into blocks of size t overlapping by p pixels, with p <t. A Euclidean distance is calculated between all possible pairs of blocks. We then look for situations where a succession of blocks belonging to a given molecule show a small distance from a succession of blocks of another molecule. This situation is characteristic of two molecules sharing a region of the same DNA sequence.

9.2. Calculating the Euclidean distance between each block

On définit pour chaque bloc de chaque molécule, ses k meilleurs voisins (par exemple avec k= 50, 100, 150, etc) parmi tous les blocs de toutes les autres molécules. Les k meilleurs voisins (k Nearest Neighbors, kNN) d'un bloc donné sont ceux qui lui sont le plus similaires car ils en sont éloignés par une petite distance euclidienne. The Euclidean distance between two blocks bl and b2 of size N is given by the following equation:

We define for each block of each molecule, its k best neighbors (for example with k = 50, 100, 150, etc.) among all the blocks of all the other molecules. The k nearest neighbors (k Nearest Neighbors, kNN) of a given block are those which are most similar to it because they are removed by a small Euclidean distance.

Other strategies are also possible to calculate the distance between two profiles or sub-segments of the molecule profile:

 - Standard cross-correlations

 - Standard cross-correlations with partial overlap

 - Euclidean distance on Fourier transforms at term

 - Distance from Hellinger

 - Minimum Variance Matching (MVM) according to Keogh

 - Euclidean distance on thresholded wavelet coefficients

 - Weyl gap

9.3. Matrix representation of similar blocks

For all the possible pairs of molecules {m _1} m ₂ ), from among the kNNs of each block of m _}, the blocks belonging to m ₂ are selected and represented on a two-dimensional graph whose axes are the positions respective blocks of m _} and m ₂ (Figure 3).

9.4. Calculation of a correlation coefficient between two molecules

In order to quantify the comparison of the mi and m ₂ molecules, the Pearson correlation coefficient of the positions of the blocks represented in the matrix (m _1} m ₂ ) is calculated. The Pearson correlation coefficient classifies all possible pairs of molecules (m _1} m ₂ ) from most similar to least similar. If necessary, the calculation of the Pearson coefficient can be replaced by the calculation of the Kendall coefficient.

9.5. Sample result on 52 lambda molecules and 52 E. coli molecules

In this example, 52 lambda phage molecules (30 to 40 kb) and 52 molecules of the E. coli (30 to 40 kb) were identified by the method described above (steps 9.1 to 9.4). The molecules were labeled by substituting 17% of the native thymidines with an uracil coupled to a fluorescent molecule on either strand of DNA (amino-allyl labeling + Cy3-NHS). The genome of the lambda phage is about 48 kb while the genome of E. coli is about 4.6 Mb, a hundred times more. So the starting hypothesis postulates that it is It is likely that many lambda molecules share sequence segments with each other, whereas because of its very large size relative to lambda phage, it is very unlikely that two molecules of E. coli share a sequence segment. The comparison of the profiles by the method described above must therefore make it possible to distinguish the lambda molecules, which will have a similar profile to each other, molecules of E. coli that will have a profile different from all others. Figure 4 illustrates a comparison of two different lambda molecule profiles by the method described above. In this comparison the correlation coefficient of Pearson is 0.99. The results show that the method is clearly able to sort the pairs of lambda molecules in net excess over the pairs of E. coli molecules or with respect to the pairs of E. coli-lambda molecules (Figure 4). It can be seen that the higher the correlation coefficient threshold increases - that is, the more the specificity increases - the better the process can distinguish the pairs of lambda molecules, to the point of totally eliminating the artefactual lambda-E. coli pairs.

Example 10: Identification of Specific Signatures of Molecules

We show below that two lambda DNA molecule profiles labeled at 20% each generate a specific signature because their maximum correlation coefficient is obtained only when the two sequences are perfectly aligned. Moreover, this maximum coefficient is much greater than those produced by aligning a profile of a lambda molecule with that of an E molecule. coli, in all possible combinations. 10.1. Obtaining images of intensity profiles of DNA sequences.

A DNA molecule image is generated as follows. A Spread Function Point (PSF), a mathematical function describing the distribution of photons emitted by a single fluorescent molecule, is determined by approximating by a Gaussian function the points obtained empirically on individual fluorophores by the microscopic system described in Example 5.1. . This PSF is assigned to a given percentage (percentage of labeled bases) of adenine and thymidine bases randomly selected on a DNA molecule of interest. Each base of the DNA molecule is assigned to a pixel of size 0.5 nm (the size of a base when a DNA molecule is stretched) on a pixel line in the image. A random number of these pixels corresponding to the marked base (for example adenine) is selected, depending on the percentage of desired labeling, and the maximum value of one pixel is set at the constant value of 100 photons. A convolution in two dimensions is then carried out by integration of the PSF, then a noise of random Poisson is added (function of the intensity of the pixels) as well as a slight Gaussian noise (of average 0 and standard deviation 10).

10.2. Calculation of the intensity profile

 The profile of each molecule is determined by the peak detection method as described in Example 8.1 above.

10.3. Calculation of the similarity between different intensity profiles by correlation coefficient Comparison of two molecules of lambda by slip

 The profile of each lambda molecule (complete and fragment) derived from the images shown in Figure 6 is compared by sliding the two profiles relative to each other, pixel by pixel, and calculating a correlation coefficient of Pearson to each position. This experiment shows that the maximum coefficient of 0.669 stands out remarkably well from the other values obtained and is obtained precisely at the position corresponding to the alignment of the two sequences (Figure 5, solid line curve). The overall shape of the curve also testifies to the specific character of the signature, by the gradual and symmetrical increase of the correlation coefficients on either side of the position corresponding to the exact alignment.

Comparison of a lambda molecule and a mouse DNA molecule

The following negative control reinforces the notion of specific signature. A molecule from the 212 kb mouse genome also served as a basis for producing an image as described in 10.1 (80% randomization), followed by a profile as described in 8.1. This profile was compared to that of lambda phage, by progressive sliding and calculation of a correlation coefficient at each position (Figure 5, dashed curve). It is unequivocally clear that the coefficients resulting from this comparison of two different molecules show a maximum at 0.27, which is very much less than the maximum coefficient obtained when comparing two identical molecules as described above.

References Bertero and P. Boccacci, 1998, "Introduction to Inverse Problems in Imaging" (IOP Publishing, Bristol)

 Betzig et al, "Imaging Intracellular Fluorescent Proteins at Nanometer Resolution", Science, 2006, 1642-1645

Blow et al, "Initiation of DNA replication in nuclei and purified DNA by cell-free extract of Xenopus eggs", Cell, 1986, 47, 577-587

 Donoho et al, "Nonlinear Solution of Inverse Linear Problem by Wavelet-Vaguelette Decomposition", Appl. Comput. Harmon. Anal., 2 (1995), 101-126

 Marheineke, et al., Methods in Molecular Biology, DNA Replication: Methods and Protocols, 2009, 521, 575-60.

 Mortensen et al, "Optimized localization analysis for single-molecule tracking and super-resolution microscopy", Nature Methods 7, 2010, 377 - 381

 Neely et al, "DNA fluorode ocode: a single molecule, optical map of DNA with nanometer resolution", Chem. Sci., 2010, 1, 453-460

 Pertsinis et al. "Subnanometer single-molecule localization, registration and distance measurement", Nature 2010, 466: 647-51

 Ramsay et al. CyDNA: Synthesis and Replication of Highly Cy-Dye Substituted DNA by an Evolved Polymerase "J. Am., Chem Soc., 2010, 132, 5096-5104)

Rust et al., "Sub-diffraction-limiting imaging by stochastic optical reconstruction microscopy (STORM)", Nat Methods, 2006, 793-795

 Thompson, R.E., Larson, D.R., Webb, W.W. Precise nanometer localization analysis for individual fluorescent probes. Biophys. J. (2002) 82: 2775-2783

 Zink et al., "Structure and dynamics of human interphase chromosome territories in vivo"; Hum Genet; 1998 Feb; 102 (2): 241-51

Claims

A method for obtaining a signature specific for a DNA molecule, comprising:  a) the labeling of a DNA molecule by incorporation of at least one modified nucleotide during a DNA polymerization reaction, the modified nucleotide being detectable, or being capable of being modified so as to be detectable ;  b) immobilizing and stretching the labeled DNA on a solid surface;  c) detecting the labeling along the DNA by means of a suitable optical device, whereby a specific signature of said DNA sequence is obtained. 2. The method according to claim 1, wherein the DNA molecule has a size greater than 50 kb, preferably greater than 60 kb, more preferably greater than 100 kb, even more preferably greater than 200 kb. 3. The method of claim 1 or 2, the labeling step a) comprising incorporating at least one modified nucleotide during a DNA polymerization reaction in vitro with Phi29 polymerase. 4. Method according to any one of claims 1 to 3, the detection in step c) using a TIRF system. 5. Method according to any one of claims 1 to 4, the detection in step c) implementing a STORJVI system. The method of any one of claims 1 to 5, wherein the immobilizing and stretching step b) precedes the marking step a). A method for identifying a DNA molecule comprising:  i) obtaining the signature of said DNA molecule according to the method of any one of claims 1 to 6; and  ii) comparing the signature obtained in step i) with predetermined signatures of reference DNA molecules; the comparison to identify the DNA molecule. A method for identifying a genomic abnormality in a subject comprising:  i) obtaining the molecular signature of a genomic DNA molecule of a tissue or cell of a subject, according to the method of any one of claims 1 to 6;  ii) comparing the signature obtained in step i) with the predetermined signatures of reference DNA molecules from  -of a healthy subject and / or  - a sick person, their illness resulting from an alteration of the genome of one or more of their cells;  to identify said genomic DNA molecule; a genomic anomaly being detected  if the signature of the genomic DNA of the subject differs from the signatures of the reference DNA molecules in the case of reference DNA molecules originating from a healthy subject, or  if the signature of the genomic DNA tested corresponds to at least one of the reference signatures originating from a diseased subject. The method of claim 8, further comprising  obtaining in step i) the signature of several genomic DNA fragments and their concatenation on the basis of overlapping signatures;  the concatenate (s) being then compared with the reference sequences to identify the said concatenate (s); a genomic abnormality is detected if a concatenate differs from the reference DNA molecule that made it possible to identify it. The method of claim 8 for identifying a complementary DNA molecule, the signature of said complementary DNA being compared to the signatures of a reference complementary DNA molecule set. 11. A method for obtaining the nucleotide sequence of a DNA molecule comprising: a) the labeling of a DNA molecule by incorporation of at least one modified nucleotide during one or more independent polymerization reactions DNA, the modified nucleotide (s) being detectable, or likely to be modified (s) to be detectable; b) immobilizing and stretching the labeled DNA on a solid surface; c) the detection of the marking by means of an optical device and a processing means adapted to obtain a resolution information less than or equal to 0.5 nm. The method of claim 11, wherein the DNA molecule is genomic DNA or complementary DNA. The method of any one of claims 1 to 12, wherein the step immobilization and stretching is performed using the molecular combing technique. 14. A method of sorting different DNA molecules comprised in a mixture, the method comprising:  a) the labeling of the DNA molecules by incorporation during a polymerization reaction of at least one modified nucleotide detectable or capable of being modified so as to be detectable; b) immobilizing and stretching the labeled DNA molecules on a solid surface; c) detecting the labeling along the DNA molecules by means of a suitable optical device; whereby a set of light profiles corresponding to the different DNA molecules is obtained, the different DNA molecules being discriminated according to their light profile.