FR2755147A1 - Determining presence or localisation of specific DNA sequences - Google Patents

Abstract

Method for determining presence or localisation of one or more genes or specific DNA sequences (A), or molecules (I) reactive with DNA, on a DNA (B), comprises: (i) fixing and 'combing' a quantity of (B) on a combing surface; (ii) reacting (B) with probes (II), linked to (I) or (A), and (iii) collecting information about the position of (II), distance between (II) and/or amount of (II) (for quantifying the number of hybridised (II)) so as to deduce the presence, localisation and/or quantity of genes or (A). Also claimed are: (1) a kit for carrying out the process above; (2) a genomic DNA, or fragment, able to react, under the conditions of molecular combing, with a (II) corresponding to a translation, transcription or regulatory product, and (3) molecules (III) reactive with genomic DNA that can be combed.

Description

 The present invention relates in particular to a method for detecting and positioning genes in a genome or part of a genome using the technique known as molecular combing.

 The technique of molecular combing, as described in the following references: PCT / FR95 / 00164 of 10/02/95 and PCT / FR95 / 00165 of 10/02/95, applied to nucleic acids, and more particularly to the Genomic DNA, allows the visualization of DNA or RNA in the form of rectilinear and substantially aligned filaments.

 The present invention is based on the fact that, by virtue of probes, i.e., nucleotide sequences such as labeled DNA molecules that specifically recognize aligned portions of the DNA), that hybridized on the combed DNA, the position of the complementary sequence can be directly visualized on the combed genome.

 Under these conditions, it is possible, for example by using two probes labeled with different chromophores as visualizable with a color, red and green for example, to measure the distance between them. But it is also possible, using different probes or a series of contiguous probes (hereinafter referred to as "contig"), to directly measure the length of the zone of interest, and to measure any alterations in the case of a abnormal genome.

 The present invention thus relates, in particular, to the diagnosis of genetic diseases characterized, preferably, by important alterations of the genome, either in its structure, deletion, translocation for example, or in the number of copies of certain sequences (trisomy for example where the sequence represents the entire chromosome), as well as methods for rapidly locating and mapping genes.

The genetic diagnosis can be divided into several areas:
- prenatal,
- pathologies with a genetic component,
- cancer and susceptibility to cancer.

Pre-natal diagnosis
The majority (95%) of fetal abnormalities are due to trisomies of chromosomes 21, 18, 13, X or Y. Their diagnosis is relatively late (17th week of amenorrhea). It requires a large puncture of amniotic fluid (a few tens of milliliters) from which fetal cells in suspension are extracted and cultured for several days (see the technique described by S. Mercier and JL Bresson (1995) Ann Gen., X, 151 -157). A karyotype of these cells is made by microscopic observation and counting of chromosomes by highly specialized personnel.

 Other abnormalities such as translocations or deletions / insertions of large portions of chromosomes can be detected at this stage, or by employing techniques such as fluorescent in situ hybridization (FISH). However, again, this type of diagnosis can only be performed by highly qualified personnel.

 Current prenatal diagnoses therefore have many disadvantages: they can only be performed at a relatively late stage of embryo development; they are not totally without risk to the fetus or the mother; the results are obtained after a fairly long time (about 3 weeks) and are expensive. Finally, a number of chromosomal abnormalities go unnoticed.

Diagnosis of pathologies with a genetic component
Many diseases have a recognized genetic component (diabetes, hypertension, obesity, etc.) that is the result of deletions, insertions and / or chromosome rearrangements of varying sizes. Culturing cells is not a problem at this stage, but FISH techniques, described by Lichter GD et al. (1993), Genomics, 16, 320-324; Brandritt B. et al. (1991), Genomics, iQ 75-82 and Van den Hengh
G. et al. (1992), Science, 257, 1410-1412), have limited resolution and use highly qualified personnel, making these tests inaccessible.

 The development of a more efficient and inexpensive test would allow the general adoption of appropriate therapies, at an early stage of the pathologies involved, likely to improve their remission.

Cancer diagnosis
Among the pathologies with a genetic component, the cancerous diseases constitute an important class affecting a growing part of the population. The current understanding of the process of occurrence of a cancerous condition involves a proto-oncogene proliferation step (mutations in the genome of the cells) that precedes the transformation of the cell into a cancer cell. This proliferation step is unfortunately not detectable, although the possibility of treatment at this stage would certainly increase the chances of remission and decrease the disability of patients.

 In each of these areas, molecular combing can make a major contribution, either by the speed and the small quantity of biological material required, or by the quantitative precision of the results.

 Molecular combing can improve the diagnostic possibilities of genetic diseases, but it can also allow the study and demonstration of the genomic sequences responsible for said diseases. Moreover, currently, the development of a "kit" or diagnostic kit begins with the search for the gene involved in the pathology.

The search for genes involved in pathologies (human or otherwise) is generally carried out in several stages:
(i) Establishment of a target population of individuals with pathologies, their descendants, ascending and collateral, and blood and / or cell sample collection for the storage of genetic material (in the form of DNA or cell strains).

 (ii) Genetic localization by cosegregation probability analysis with genetic markers (linkage analysis). At this stage of the study, we have some close markers located on one (or more) chromosome (s) given (s) that allow to move to the physical location step.

 (iii) Physical mapping: from the genetic markers obtained in the previous step, a screening of human DNA clone libraries (YACs, BACs, cosmids or others) specific for the region or regions determined in the previous step is made. A number of clones containing the foregoing markers are thus obtained. The area of interest can then be accurately mapped using clones of decreasing size. Cloning of the part of the genome under consideration may also be performed again from human DNA.

 (iv) Gene research: several techniques can be used at this stage: exon trapping "(use of cDNA libraries (coupling DNA obtained from messenger RNA)), CpG islands, conservation of inter-specific sequences, etc., which make it possible to assign a coding sequence to one (or more) of the clones selected in the preceding steps.

 This whole strategy represents a lot of work (up to several years, eventually). Also, any technique that makes it possible to reach stage (iv) more quickly is an asset for gene research but also for diagnosis.

 In the current state of the art, when the gene has been localized, for example by the preceding method, its detection is generally carried out by means of specific probes corresponding to the sequence in question, this being amplified by methods of PCR type for example or LCR type (as described in patent EP 0 439 182) or a NASBA type of technique (kit marketed by Organon Tecknika).

 However, the amplification techniques are not totally satisfactory, especially for heterozygotes, since there is a normal copy of the gene in the genome, as well as in the case of large deletion or of repeated-sequence diseases where PCR n is not satisfactory either.

 The diagnosis of a large number of genetic diseases can now be envisaged by using molecular combing and DNA labeling.

 Molecular combing is a technique of anchoring DNA molecules by their ends on surfaces under well-defined physicochemical conditions, followed by stretching with a meniscus in recession. DNA aligned in parallel. The purified DNA used may be of any size, and therefore in particular genomic DNA extracted from human cells.

 The DNA molecules thus combed may be denatured before being hybridized with labeled nucleic acid probes by any appropriate means, (in particular with biotin-dUTP nucleotides, or digoxygenin-dUTP), which are then revealed, for example, using fluorescent antibody systems.

 Since molecular combing is characterized by a constant extension of the combed molecules, the measurement of the lengths of the fluorescent fragments observed with the aid of an epifluorescence microscope (for example) thus directly gives the size of the hybridized probe fragments.

 The degree of extension depends on the type of surface, but can be accurately measured, for example it is 2 kilobases (kb) per micrometer (crm) in the case of silanized surfaces according to the protocol described in reference (1) and set implemented in the examples.

 When necessary, it is possible to provide an internal standard, that is to say a DNA of known length, which will calibrate the manipulation.

The present invention, which comprises different modes of implementation, essentially concerns a method of detecting the presence or location of one or more genes or of one or more specific DNA A sequences on a DNA B, characterized in that: (a) a certain amount of said DNA B is fixed and combed on a
combing surface, (b) the combing product B is reacted with one or more probes
labeled, linked to the specific gene (s) or DNA sequences A, (c) the information corresponding to at least one of the
following categories
(1) the position of the probes,
(2) the distance between probes,
(3) the size of the probes (the sum total of the sizes allowing
quantify the number of hybridized probes)
to deduce the presence, location and / or quantity of
genes or specific DNA sequences A.

 In the present description, the combing technology refers to the technology described in the aforementioned documents, as well as the concept of "combing surface" which corresponds to a treated surface allowing the anchoring of the DNA and its stretching by a meniscus in recession.

 It should be noted that the combing surface is preferably a flat surface on which the readings are more convenient.

 By "reaction between the labeled probes and the combed DNA" is meant any chemical or biochemical reaction, in particular immunological type reactions (for example antibodies directed against methylated DNA), protein / DNA or nucleic acid reactions. / DNA (eg hybridization between complementary segments).

 Hybridization is generally preceded by denaturation of the fixed and combed DNA, this technique is known and will not be described in detail.

 By "probe" is meant both a polynucleotide having at least 20 nucleotides corresponding to a single probe that a "contig", that is to say a set of probes that are contiguous or overlapping and covers the area in question. cause, or several separate probes.

The length of the probes used is, for example, between 5 kb and 4050 kb.

 Finally, in a general way, in what will be called "genome", we must understand that this is a simplification, any DNA or nucleic acid sequence that can be fixed on a combing surface is encompassed in this terminology.

 In addition, the term "gene" will sometimes be used interchangeably to designate a portion of gene "of genomic origin or a specific" synthetic polynucleotide sequence ".

 In a first embodiment, the method according to the invention is used to enable the diagnosis of breaks in a genome, as well as for the positional cloning of such breaks. It should be noted that the term "break" covers a large number of local modifications of the genome, the list of which will be explained later.

 The method according to the present invention consists in determining the position of any break points involved in a pathology of genetic origin by hybridization on combed genomic DNA of patients suffering from said pathology, a genomic probe of known size (cloned or otherwise ) located in the region of the desired gene. These breakpoints consist of points in the genetic sequence whose environment changes over several kilobases (kb) between the healthy individual and the sick individual.

More particularly, the present invention relates to a method for demonstrating a genetic abnormality of breakage in a genome, characterized in that (a) a certain amount of said genome is fixed and combed on a
combing surface, (b) the combing product is hybridized with one or more probes
labeled specific corresponding to the genomic sequence
we look for the anomaly, (c) we measure the size of the fragments corresponding to the signals
hybridizations and possibly their repetition, and (d) we deduce the presence of a break, either by direct measurement or
by comparison with a standard.

 By way of illustration, measuring the size of the fragments leads to a histogram, that is to say a graphical representation of the lengths of the fragments observed.

 In order to make a histogram of the probe, the number of clones having a determined probe length is evaluated. In principle the histogram has only one or two peaks depending on the type of breakage analyzed, two peaks when the probe hybridizes into two separate fragments and a single peak when it hybridizes a single fragment.

 This method can also be used to perform positional cloning, i.e., to determine the position of one or more unknown genes involved in a pathology. The principle consists, as before, in hybridizing clones of human or animal or plant DNA, thus serving as a probe on the combed genomic DNA of one or more patients with the pathology studied. The revelation of these hybridizations makes it possible to measure the size of the hybridized fragments and to construct a histogram of the different sizes observed. If the clone used as a probe covers a breakpoint of the gene, the signature of this phenomenon will be readable over the (smaller) length of the hybridized fragment.

 The use of a limited number of specific clones of the incriminated region that may have been deduced by linkage analysis ", will thus allow a rapid and accurate determination of the position of a breakpoint. a deletion or any other genetic reworking.

 In this case, of course, we search for the break in order to map it, in the diagnosis the break is known, it is its presence or absence that is sought.

 Two cases may occur (assuming a breakpoint in the region of the genome involved in the pathology): (i) the probe does not overlap the breakpoint, (ii) the probe overlaps the break point.

 In the case (i), the measurement of the lengths of the fluorescent probes is comparable to that which would be obtained with the same hybridized probe on a non-pathological genomic DNA of the same kind (that is to say essentially, of the same size and prepared in the same conditions).

 In the case (ii), on the contrary, the probe being systematically hybridized on two (or more) separate pieces in the combed genomic DNA (by definition of the existence of a breakpoint), the measurement of the lengths of the probes hybridized fluorescents is distinguished from the result obtained by hybridization on non-pathological genomic DNA. In addition, the size of the hybridized fragments on the pathological DNA makes it possible to estimate the position of the breakpoint within the clone with an accuracy of a few kb.

 As a result, all that remains is to search for the gene in this clone. In practice, it will be a matter of repeating these measurements for all the clones capable of partially covering the zone corresponding to the gene. The number of hybridization slides can be reduced by simultaneously hybridizing several differently labeled probes, or by using a color combination coding method, as will be described hereinafter.

 This technique makes it possible to determine the position of the possible break points of the region of the genome involved in a genetic pathology by hybridization of genomic DNA cloned on combed genomic DNA from patients. This technique therefore applies to the search for regions of the genome responsible for pathologies due to: the deletion of part or all of this region of the genome.

translocation of all or part of this region of the genome.

- the duplication or presence of several copies of all or part of this
genomic region within or within the genome.

- the insertion of any genetic sequence within that region of
genome.

 In a second embodiment and in particular cases, especially when the desired genetic abnormality involves significant deletions or duplications (for example trisomies), the method which is the subject of the present invention can be modified since It then acts to dose the genes or a particular sequence.

More particularly, the present invention relates to a method for assaying a genomic sequence determined in a genome, characterized in that (a) a certain amount of said genome is fixed and combed on a
combing surface, (b) the combing product is hybridized with a control probe marked with
length it corresponding to a so-called control genomic sequence,
that is to say, whose number of copies in the genome is known, and
with a specific probe marked with length lc corresponding to the
genomic sequence to be assayed, said probes can be identified
(c) then measure the total length of the hybridization signals for the
two probes, Lc and Lt, (d) we calculate for each the number of copies of the sequence
corresponding by the report
LT Lc Nt ~~~~~~~ = and = ~~~~~~~~
lt lc
and we deduce the number of copies of the sequence to be assayed with respect to
the control sequence.

 In the case of pre-natal diagnosis of trisomy 21, the method may consist in the hybridization of a cosmid probe specific for a control chromosome (chromosome 1, for example-length probe It) labeled with biotinylated nucleotides, and hybridization of a chromosome 21-specific cosmid probe (digoxygenin-labeled probe length lc) to combed genomic DNA extracted from amniotic specimens.

For example, an avidin-Texas Red revelation system (red color) may be used for the control and antidigoxygenin probe
FITC (green color) for the specific probe: the total length of the red hybridization signals observed in a given area of the surface, LT, and the total length of green hybridization signals observed in the same area, or in an area equivalent of the surface, LC, thus lead to the numbers Nt and Nc defined above.

 The ratio Nc / Nt close to 1 will signal a normal genotype (2 chromosomes 21 for 2 chromosomes 1), while a ratio close to 1.5 will indicate a trisomic genotype (3 chromosomes 21 for 2 chromosomes 1).

 In the case of the diagnosis of oncogenes or proto-oncogenes, the same method can be used: a control probe will be hybridized and revealed in red for example and a probe corresponding to the gene or part of the desired gene will be hybridized and revealed in green for example.

After the measurements carried out as above, the Nc / Nt ratio will give the relative abundance of the gene with respect to the frequency of two copies per diploid genome.

 The aberrant methylation of GpC islands frequently observed in many cancers (92% of colon cancers) can also be detected by the process according to the invention by reaction between the combed DNA and fluorescent antibodies directed against methylated cytosines.

 Indeed, the loss of heterozygosity on chromosome 9p21 is one of the most common genetic alterations identified in human cancers. The tumor suppressor gene CDKN2 / p16 / MTS1 located in this region is frequently inactivated in many human cancers by homozygous deletion. However, another mode of inactivation has been reported which involves the loss of transcription associated with de novo methylation of 5 'GpC islands of CDKN2 / pl6 in lung cancers, gliomas and carcinomas of desquamation of the head and by the neck.

These aberrant GpC islet methylations also frequently occur in breast cancer (33%), prostate (60%), kidney (23%) and colon (92%) cell lines.

 The precise location of methylation areas on a gene is of great importance for understanding the mechanism of cancer development and for a possible screening test. Molecular combing can detect with a precision of some Kbp the location of such GpC islands involved in the development of cancer.

 This technique for determining the number of copies of a gene in a genome can optionally be used to detect the absence of part of the genome.

 In the case of a pathology characterized by the deletion of a large part of a chromosome, it is indeed necessary to take as a target sequence a clone contained in the deleted area, and for control sequence a clone outside this area. It is thus possible to detect deletions the size of a cosmid clone (30-50 kb) or greater.

 If there is a sufficient density of combed molecules, it is possible to detect smaller deletions (a few kb), corresponding to a part of the target sequence used.

 The statistical uncertainty on the ratio Nc / Nt is of order 1 // Nc + 1 // Nt. It is therefore important to have a large number of hybridized probes, typically Nc, Nt> 100.

 However, in practice, it is also possible to increase the accuracy of these measurements by using not one but several types of control probes and target probes, without necessarily trying to distinguish between these types of probes. ie by revealing them all in the same way.

 The possibility of obtaining such a number of signals has been demonstrated: it is possible to count a hundred signals on a silanized glass surface of 20x20 mm of useful area. This density can be considerably increased when there is a large amount of DNA.

 It appears that a sufficient number of genomes per area of 20 x 20 mm of useful area is around 100, when only one probe is used. In the case of the use of several probes, or larger probes, it is conceivable to be able to reduce either: the combing surface and thus the analyzed surface.

the density of DNA used, therefore the number of genomes combed on the surface.

 Depending on the main constraint (speed required, or DNA in limited quantity), one or the other of these two ways could be used.

 The exposed technique involves the use of strict preparation protocols but without particular technical difficulties. At the signal analysis level, no particular qualification is needed, thus making the technique generalizable to all labs with staff with minimal skills in molecular biology.

 Hundreds of thousands of cells should in principle be sufficient to prepare a genomic DNA solution leading to a high density of combed molecules on the analysis surfaces. It is therefore no longer necessary in principle to carry out cell cultures in most cases. The collection-analysis package should be able to be realized in a few days.

 The simplicity of the signals to be analyzed (parallel and distinct from the hybridisation background noise) makes it possible to envisage a complete automation of the signal analysis process (surface scanning, acquisition and operation of the measurements). Integration with a system of storage of surfaces corresponding to different patients makes it possible to consider important returns, giving the possibility to provide various types of diagnoses in a few days.

 The method described above can allow different types of diagnosis: chromosome counting (trisomy, monosomy, etc.), counting copies of a gene, detection of known deletions, or other chromosomal modifications resulting in a modification of the hybridized length of the gene. a genomic probe given by genome.

 It should be noted that it is also possible to detect a partial deletion on a single allele.

 These different types of diagnostics can be combined through the use of multiple visualization systems (multiple colors, or combination of colors), or any other method allowing the distinction between hybridization signals from separate probes and intended for diagnosis. specific.

The present invention also relates to "kits" for diagnosis comprising at least one of the following elements - a combing surface, - probes marked or intended to be marked, corresponding to the
abnormalities to be detected; a device for combing the DNA; a control genome and the corresponding control probes;
genome possibly being fixed on the surface to be combed.

 The principle of the technique based on the combing of the DNA of the patient, this step of preparation of the DNA requires extraction protocols, combing and the corresponding material (treated surfaces, molecular combing apparatus).

 The diagnosis itself requires the hybridization of specific nucleotide probes and the discovery of these probes by antibody systems. Furthermore, color coding can be performed in the case of combined diagnostics, so it is possible to provide pre-marked probe batches corresponding to a particular diagnostic catalog.

 Since the analysis requires the measurement of the length of the signals, a system for analyzing these signals (software and automatic equipment) also emerges from this invention.

 Finally, in a third mode of implementation, the present invention relates to a method allowing in particular the physical mapping of a genome.

 Since the goal of physical mapping is clone scheduling within a genome, molecular combing is naturally applicable to this goal by simply hybridizing clones to the combed genome (for example, in the case of a YAC , the entire genome of the yeast can be combed, to avoid the separation of the artificial chromosome from the natural chromosomes of the yeast).

 The position of the clones is obtained by direct measurement of their distance to a reference clone, or to any other reference hybridization signal on the combed genome. The constant extension of the combed DNA then makes it possible to directly establish in kilobases (kb) the respective position of the clones as well as their size.

 The method we propose makes it possible to minimize the number of hybridizations necessary for the scheduling of a given number of clones, given a fixed number of hybridization revealing colors.

The invention relates to a method, characterized in that (a) a certain amount of said genome is fixed and combed on a
combing surface, (b) the combing product is hybridized with colored probes
corresponding to each clone, (c) the corresponding information is taken instead of each clone
as well as the corresponding sizes and distances on the genome, (d) the operations b) and c) n times are repeated by modifying the colors of the
clones, knowing that with p different colors after n
hybridizations one can position pn clones.

 In the context of standard mapping methods, the number of hybridizations I, necessary to map N clones using p colors, increases linearly with the number of clones, N.

 Thus, with 3 colors, it is now necessary to perform at least 15 hybridizations of the above type to map 30 clones.

This number of hybridization is imp
A = Red then Red
B = Green then Green
C = Red then Green
D = Green then Red
From 5 to 8 clones, 3 hybridizations will be necessary, to distinguish the clones by a succession of 3 colors.

More generally, using p colors, to map
N clones, it will suffice a number of hybridizations l such that: N = pI
By comparison with the standard method, 30 clones can be mapped in 5 hybridizations (instead of 30) if only 2 colors are available, and in only 4 hybridizations (instead of 15) if there are 3 colors.

 The mapping principle presented here is simple. However, in order to overcome certain possible experimental artefacts (signal size dispersion, saturation variability, molecule breakage, etc.), it will be advantageous to use image processing software and adequate statistical analysis.

 The following examples will better understand other features and advantages of the present invention.

Legends of the figures
Figure la: illustration of the case of a clone (red) overlapping the gene (and therefore the point of breakage) sought. The hatched areas must be imagined as absent, which implies in particular that when two segments of probes are on both sides, they are actually only a segment of a length equal to the sum of their lengths. The underlying chromosome (s) are represented by white bars.

(a) situation in a healthy patient.

 (bl) deletion (s) of a part of the gene. (b2) deletion of the entire gene.

 (c1) and (c2) translocation of a part of the gene. (c3) translocation of the entire gene.

(dl) duplication (s) of a part of the gene (possibly with inversion).

(d2) duplication (s) of the totality of the gene (possibly with inversion).

(d3) repetition (s) of a part of the gene in another part of the genome. (e) insertion (s) into the gene of a different part of the gene (and the cloned genome).

Figure lb: ideal histograms lengths of hybridized clones corresponding to the previous situations. In all cases (one or two peaks), the situation is clearly distinguishable from that of the whole clone (a).

Figure 2a: illustration of the case of a clone partially overlapping the desired gene.

(a) situation in a healthy patient.

 (bl) deletion (s) of a part of the gene. Another situation has not been shown, where the deleted part is included entirely in the clone (cf corresponding case of diagram 1a).

(b2) deletion of the entire gene.

 (c1) and (c2) translocation (s) of a part of the gene. Another situation, corresponding to the case (c2) of the scheme la has not been represented, where the translocated part is entirely included in the clone.

(c3) translocation (s) of the entire gene.

(d1) duplication (s) of a part of the gene identified by a double arrow (the duplicated part is entirely included in the clone). (d2) duplication (s) of a part of the gene with inversion. (d3) duplication (s) of a part of the gene (the duplicated part is partially included in the clone).

(d4) duplication (s) of the entire gene (without inversion).

(d5) repetition (s) of a part of the gene in another part of the genome.

(e) insertion (s) into the gene of a different part of the gene (and the cloned genome).

FIG. 2b: ideal histograms of the lengths of hybridized clones corresponding to the preceding situations. In all cases (one or two peaks), the situation is clearly distinguishable from that of the whole clone (a).

Figure 3a: illustration of the case of a clone included totally in the desired gene.

(a) situation in a healthy patient.

(bl) deletion (s) of a part of the gene. Another situation has not been shown, where the deleted part is included entirely in the clone (cf corresponding case of diagram 1a).

(b2) deletion of the entire gene.

 (c1) and (c2) translocation (s) of a part of the gene. Another situation, corresponding to the case (c2) of the scheme la has not been represented, where the translocated part is entirely included in the clone.

(c3) translocation (s) of the entire gene.

(d1) duplication (s) of a part of the gene identified by a double arrow (the duplicated part is entirely included in the clone). (d2) duplication (s) of a part of the gene with inversion. (d3) duplication (s) of a part of the gene (the duplicated part is partially included in the clone).

(d4) duplication (s) of the entire gene (without inversion).

(d5) repetition (s) of a part of the gene in another part of the genome.

(e) insertion (s) into the gene of a different part of the gene (and the cloned genome).

FIG. 3b: ideal histograms of the lengths of hybridized clones corresponding to the preceding situations. In all cases (one or two peaks), the situation is clearly distinguishable from that of the whole clone (a).

Figure 4: Histograms of the lengths of the staircase model. Study of the influence of the characteristic size.

Figure 5: Histograms of the lengths of the staircase model. Study of the influence of the breakage rate.

Figure 6: Histograms of the lengths of the Gaussian model. Study of the influence of the characteristic size.

Figure 7: Histograms of the lengths of the Gaussian model. Study of the influence of the breakage rate.

Figure 8: Simulation of histograms in the case of hybridization of a BAC (125 kb) on a breakpoint: (a) fragments of 35 and 90 kb, (b) fragments of 50 and 75 kb, (c ) 60 and 65 kb fragments, (d) control situation.

Figure 9: illustration of the principle of mapping without color coding.

Each line represents the possible positioning of the new mapped subclone compared to the one with which it is hybridized. The orientation of the first two clones is arbitrary.

At each stage, two positions on either side of the old clone used are possible for the new hybridized clone.

Figure 10: illustration of the principle of color coded cartography.

At each hybridization, the same subclones are hybridized, giving rise to the same pattern (pattern) of hybridization if the color information is ignored. The two color palettes used make it possible to unambiguously identify each of the subclones (coding principle).

MATERIAL AND METHODS
Histowrams method
The diagram of FIG. 1a presents a certain number of these situations in the case of a probe completely covering the region of the genome sought (region containing any genomic sequence whose modification causes the appearance of the pathology). Preferably, the probe covers a part of the genome not involved in the break, that is to say on either side of the deletion for example.

 The corresponding theoretical histograms (assuming that the combed genomic DNA is intact, which is naturally unrealistic, and which will be discussed in the third part) are presented in the diagram of Figure 1b. Similar situations are represented in the case where the clone used does not fully cover the gene (or the region of the genome) sought on the diagrams of FIGS. 2a and 3a.

 In the following, we will indifferently use the expression this region of genome "(implied in the pathology) or" the gene ", since the technique does not allow to determine strictly speaking a gene, but a region of the genome undergoing alterations (resulting in a "breakpoint").

 The set of situations represented in the preceding diagrams gives an idea of the expected difference between a situation where the cloned probe used hybridizes to a region of the genome of a healthy individual that is modified in the case of a sick individual.

In all the cases presented (except that of a translocation of all or part of the gene containing the entire clone - see scheme 3a (c3)),
The (theoretical) histogram differs from the "normal" histogram: - either by the presence of a single peak but situated at a value different from
the length of the clone (superior - case (d2) for example, or lower
- case (bl)).

- by the presence of two or more peaks.

 The simple data of the histogram of the lengths and the length Lc of the clone (obtained for example by hybridization on the DNA of a healthy individual and obtaining the peak of the corresponding histogram) thus makes it possible to to know if the studied cone covers a point of break (in the broad sense understood here).

The anomalies observed in the histogram can be grouped into several categories: (a) a peak for a length Li <Lc (b) a peak for a length Ls> Lc (c) several peaks of shorter length (s) ( s) at Lc (d) several peaks of length (s) greater than Lc (e) a peak for a length L = Lc and one or more peaks in length (s)
less than Lc (f) a peak for a length L = Lc and one or more peaks in length (s)
greater than Lc (g) one or more peaks of lengths <, = and> to Lc
Previously reperfied anomaly classes are generally not enough, unfortunately, to define the underlying genetic defect.

 The diagram of FIG. 1b shows that it is not possible a priori to distinguish between a partial (cl) or (c2), total (c3) translocation or an insertion in the case of a clone that would cover the the whole gene. It is the same for the case where the clone only partially covers the gene (diagram of Figures 2b and 3b).

 On the other hand, the data of the lengths corresponding to the peaks of the histogram makes it possible in certain cases to specify the position in the clone of the end of the part of the deleted, translocated, duplicated or contacting an insertion gene.

 In any case, the measurement of a histogram distinct from that expected in the case of a normal genome signals the presence in the genome studied, of an abnormality in the sequence contained in the clone used.

 (a: a peak for a length Li <Lc): it is almost certainly one (or more) deletion (s). However, as noted, this may be a case where multiple peaks are confounded.

We will see that it is generally possible to distinguish between these two cases by comparison with a probe corresponding to a different zone of the genome (reference probe).

 In the case of a deletion, it is possible to deduce the size by subtraction of the two remaining lengths of the intact clone length (only in the case where the clone covers all of the deleted portion of the gene). The position relative to the clone can not be known, however, only additional hybridizations with other clones or subclones that can allow this determination.

 However, the same type of histogram can correspond to several deletions, in which case it is only possible to measure the total size of the deletions (always assuming that the clone covers the entire deleted portion of the gene).

 In cases where the clone overlaps only partially the gene, the detected deletion may represent only a part of the effective deletion of the gene (see diagram 2a (bl) for example).

 (b: a peak for a length Ls> Lc): in this case, it is one or more duplication (s) inside the gene or at its ends, of a part or of the totality of it. The clone can completely or partially cover the gene. The sequences may be inverted, but this can not be detected in the case where the clone completely covers the gene. The position of the duplication (s) can not be determined.

 (c: several peaks of length (s) below (s) at Lc): it may be in this case partial translocations of the gene or the insertion of a sequence within the gene. In the hypothesis of a translocation, it is possible to specify the two ends of the gene within the clone (in which case the clone completely covers the gene), or to advance two possibilities for these ends (in the case where the clone partially overlaps the gene).

 (d: several peaks length (s) greater (s) to Lc): it is in this case several duplication of all or part of the gene in the region of the genome covered by the clone.

 (e: a peak for a length L = Lc and one or more peaks of length (s) lower (s) than Lc): it is a (or more) duplication (s) of all or part of the gene in an area of the genome distinct from the area covered by the clone, this in the case where the clone covers all or part of the gene.

 (f: a peak for a length L = Lc and one or more peaks of length (s) greater than Lc): although this situation has not been represented in diagrams 1 to 3, it is a duplication case of type (d) associated with a normal gene.

 (g: one or more peaks of lengths <, = and> to Lc): combinations of (e) and (f).

 In a number of situations, the analysis of the hybridization signals and their distance, in case they are aligned, should also provide additional information.

 The example of the scheme (e) will illustrate this idea: in this case of an insertion, it is necessary to imagine that any segment not hybridized, but surrounded by two colinear hybridizations (of any sizes) is likely to represent an insertion: one will build therefore a histogram of these measurements: if there is actually an insertion, it should appear as a peak, just like any hybridization.

 This measure will also be used in the case of deletions or translocations, for example.

 The previously discussed technique assumes that hybrid combed genomic DNA is intact. Indeed, the reference situation, that of a healthy individual whose DNA corresponding to the clone used for the hybridization is not modified, is represented theoretically by a peak at the value of the length of the clone.

 In practice, the DNA is subjected to hydromechanical constraints during its preparation, and also during combing, which leads to numerous random breaks. These random breaks can take place within the target sequence of the probe, which will result, after combing, in a separation into several pieces of said sequence. During hybridization with the probe, these segments will be hybridized, and after revelation, measured as fragments of size smaller than that of the target sequence.

 These breaks will not necessarily affect the set of target sequences in the set of genomes used for combing. Depending on the number of human genomes combed on the surface studied, and depending on the conditions of preparation, the proportion of broken target sequences will remain reasonably low. In this case, the ideal histogram consisting of a single peak at the length of the probe will be replaced by a smaller peak accompanied by a distribution tail at the lower lengths.

 Experience and basic practical considerations show that an optimal number of genomes combed per surface area studied is of the order of 100 to 200.

 In what follows, we will study simulation of the process of breakage (consecutive to the preparation and the combing of the DNA) depending on two parameters, allowing to evaluate the influence on the theoretical histogram of the lengths of these phenomena.

 Two parameters seem important in the breakout phenomenon: the characteristic size and the breakage rate.

 The characteristic size approximately defines the size limit reached by molecules subjected to a large number of manipulations. An example of possible manipulation is pipetting: depending on the diameter of the cone of the pipette, the maximum size limit of the molecules will be more or less important. Another example of manipulation is vortexing the DNA solution.

 The rate of breakage is supposed to model the number of manipulations: for example, the more the number of pipetting increases, the more the molecules may be broken.

A. Model in step of staircase.

 The simplest way to model DNA manipulations in solution is to assign a zero break probability to fragments smaller than the characteristic size LO, and a break probability equal to 1 for fragments larger than LO.

The break of a fragment of size L is also performed randomly in any two fragments: L <LO => Po (L) = OL> lo => Po (L) = 1
In order to introduce a break rate in this model, a parameter sets the number of breaking processes to which a molecule and its successive fragments are subjected.

 This model makes it possible to simulate the histograms expected during a hybridization of a probe of known length Lc according to the parameters of the model. The conditions chosen are: probe hybridizing on a chromosome of 50 Mb, 200 genomes combed, Lc = 50 kb.

A.1. Influence of the characteristic size on the piston.

 FIG. 4 represents, for different rates (or number of stages) of breakage, the appearance of the theoretical histograms obtained for characteristic size values (LO = Lbreak) of 1, 2, 4 or 8 times the size of the probe. It is clear that the larger the characteristic size is close to the size of the probe, the more the histogram differs from the single peak described in part 2, the tail of the distribution at lengths less than the length Lc of the probe taking importance, until practically disappearing the peak for the expected value Lc.

A.2. Influence of the fracture rate on the histogram.

Figure 5 shows, for different characteristic sizes,
The evolution of the theoretical histogram as a function of the rate of break, that is to say the number of stages of random break (N time steps). Naturally, it appears that the peak at the expected length Lc decreases as the rate of breakage increases.

 In this model, however, the peak does not disappear completely, even when the breaking rate increases indefinitely, since no break occurs when all remaining fragments are smaller than Lc.

A3. Interest of the model.

 The main interest of the previous model is to provide a quick way to evaluate the probable shape of the histogram of lengths of hybridized probe fragments, so that the results of the observations can be interpreted.

B. Eussian model.

A slightly more realistic model consists of a similar model of successive stages of breakage, but with a break law different from the "all or nothing" law of the staircase model. A simple way to "round off" this law is to replace the previous probability with a Gaussian break probability law:
B.1. Influence of the characteristic size on the historramme.

 FIG. 6 represents, for different breaking rates, the appearance of the theoretical histograms obtained for characteristic size values (LO = Lbreak) d 1, 2, 4 or 8 times the size of the probe, identical to those used in the FIG. simulation performed as part of the staircase model. In the same way, the more the characteristic size is close to the size of the probe, the more the histogram differs from the single peak described in part 2, the tail of the distribution at lengths less than the length Lc of the probe taking the importance, until practically remove the peak for the expected value Lc.

B.2. Influence of the breakout rate on the histogram.

Figure 5 shows, for different characteristic sizes,
The evolution of the theoretical histogram as a function of the rate of break, that is to say the number of stages of random break (N time steps). It naturally appears that the peak at the expected length Lc decreases as the rate of breakage increases.

However, this model has a more realistic behavior than the previous one, since when the breaking rate increases, the peak corresponding to
Lc decreases regardless of the value of the characteristic size, to disappear completely when the break rate increases indefinitely.

B.3. Interest of the model.

 The Gaussian model, being more realistic, will be the one we will use later to simulate histograms of lengths of hybridized probe fragments.

C. Simulation of hybridization at a break point.

 To study the conditions of the applicability of our technique, we present the simulation of three types of situations characteristic of the hybridization of a clone on a point of breakage. We will take for example the case of a BAC (Bacterial Artificial Chromosome) of 125 kb covering a breakpoint involved in a translocation (cf for example diagram 2a (c1)).

 The simulations presented in FIG. 8d correspond to a situation where the clone is hybridized on a normal genomic DNA: under these conditions, most of the hybridizations have a length of Lc = 125 kb, apart from the broken fragments because of the manipulations. We chose two values of the characteristic size LO: LO = 100 (<Lc) and LO = 200 (> Lc).

The breakage rate is unique and equal to N = 10.

 These histograms show that it is possible to distinguish a peak at 125 kb in most cases, even quite unfavorable as in the caseLO = 100, N = 10.

 C. 1. Fragments of 35 and 90 kb.

 FIG. 8a presents the results of the simulation of a histogram of hybridized fragment lengths in the case where the 125 kb sthybrid BAC in two unrelated parts of the genome, of very different sizes (35 and 90 kb), as in FIG. case described in Figure 2a (cul).

 In both cases (LO = 100, LO = 200), the two peaks at 35 and 90 kb are recognizable, although the situation is much more favorable in the case where the characteristic size is clearly greater than the largest fragment.

 It should also be noted that, even in the case of less sampling, which would lead to a not very distinct 90 kb peak, the peak at 35 kb would remain, signaling an anomaly with respect to the normal size of the clone.

C.2. Fragments of 50 and 75 kb.

 FIG. 8b represents the result of simulations in the case of two fragments of size which differ little from each other. Again, the corresponding peaks are sharp and distinct from each other. In this case, it can be expected that with less sampling, it will be possible to identify these two peaks.

C.3. Fragments of 60 and 65 kb.

 FIG. 8c represents the result of simulations in the case of two fragments of very similar sizes. In both cases, the peaks clearly stand out from the histogram of the small fragments, but it is to be feared that the limited precision of the measurements will confuse them into one.

 Again, however, such information will be sufficient to ensure the existence of a breakpoint midway away from the hybrid BAC.

C.4. conclusions
Previous simulations show that, under conditions that avoid over-breaking the DNA before and during combing, it is possible to unambiguously distinguish between a control situation where the probe hybridizes to a normal human genome, and a situation where the probe hybridizes on a genome modified by the presence of a breakpoint.

 The heuristic parameters characterizing the combing show that it is, however, preferable to have a characteristic size greater than or equal to the maximum length of the hybridization. Since this is unknown, a simple criterion is therefore a characteristic size in the order of or greater than the length of the clone used as a probe.

 To judge experimentally of the achievement of this criterion, it therefore seems desirable to use, in addition to probes intended to search for an anomaly in the genome, a control probe of comparable size that hybridizes to a different part of the genome. genome.

The corresponding histogram thus makes it possible to qualitatively determine the parameters of the model reflecting the experimental situation.

Surface treatment
22 × 22 mm slides are prepared and coated with silane using the method described in PCT / FR95 / 00164 and PCT / FR95 / 00165.

Preparation of cosmid probes
The cosmid DNA is marked by extension from a statistical primer with nucleotides modified with digoxygenin or which are biotinylated. For biotin-labeled probes, the BioprimeTM DNA labeling kit (Gibco-BRL) containing statistical octamer primers and biotin-14dCTP is used. For digoxigenin labeling, a different mixture of dNTPs is used which contains dCTP, dATP and dGTP (0.1 mM), dTTP (0.065 mM) and dig-11-dUTP (0.033 mM).

The size and concentration of the labeled DNA fragments are confirmed by 0.6% agarose gel electrophoresis and band densitometry. The efficiency of the labeling is evaluated from spots deposited on nylon membranes using successive dilutions of digoxigenin probes and biotin probes. After incubation with alkaline phosphatase anti-dig (AP) (Boehringer Mannheim) or streptavidin AP (Gibco
BRL), the spots are revealed with Gibson-BRL's N BT (NitroBlue Tetrazolium) and BCIP (5 Bromo-4-Chloro-3-lndolylphosphate). Positive spots are compared with results obtained from control DNA samples labeled with dig or biotin.

 In order to limit the background noise, the strains are finally purified on Bio-Spin 6 columns (Biorad) by centrifugation for 5 minutes at 1500 g.

Yeast DNA solution
The yeast DNA containing YAC 774G4 (1600 kb) is prepared in 100 μl of 0.8% LMP agarose blocks (1 μg / block) using a standard PFGE block preparation protocol. The blocks are stored in 0.5 M EDTA at 4 ° C and rinsed with 15 ml TE buffer (10 mM Tris / 1 mM EDTA, pH 8) for 2 hours before use Each block is stained with 3.3 μM YOYO-1 (Molecular Probes) in 100 μl T4oE2 (40 mM Tris / 2 mM EDTA, pH 8) for 1 hour at room temperature (RT) The agarose is then melted for 45 minutes at 68 ° C and digested for 2 hours at 40 ° C using 2 U of ss-agarase I in NEB xl agarase (Biolabs) buffer per block.

 Dilution of the DNA (0.25 μg / ml) in 50 mM MES (pH 5.5) is performed very carefully to avoid breakage of the DNA strands. The solution is then poured into a 4 ml tank of Teflon allowing the introduction of 3 silanized 22x22 mm blades for combing. The DNA solution can also be stored at 4 ° C for several days.

Molecular combing
The silanized slides are dipped in the Teflon® reservoir containing a solution of total yeast DNA (0.25 g / ml in 50 mM MES, pH 5.5), incubated at RT and extracted from the reservoir after 10 minutes of incubation using a simple mechanical device. During the incubation, the DNA molecules are anchored on the surface at their ends. By extracting the surface of the reservoir, this has the same effect as the evaporation provided in the "drop method", the meniscus moves relative to the surface and exerts a constant tensile force on the molecules remaining in the reservoir.

 The surfaces covered with the combed DNA are then observed with an epifluorescence microscope so as to control the characteristics of the combing. A record of the representative fields of view for each slide is made for the control of the posthybridization.

 The surfaces are then glued on microscope slides (cyanocrylate) and heated overnight at 60 ° C. They can be stored for several months if protected from moisture at -20 ° C or at room temperature.

 The detections of the biotin and digoxigenin-labeled probes are performed simultaneously using the same protocol for each detection layer, each antibody layer being incubated for 30 minutes at 37 ° C., the slides being washed after each layer (3 times 5 minutes at RT in 4xSSC / Tween 20 0.05%).

 For biotin-labeled probes, the following layers are used (50 μl of hybridization buffer per slide): (1) 40 mg / ml of Avidin-Texas Red (Vector). (2) 5 mg / ml of biotinylated goat anti-avidin (Vector). (3) 40 mg / ml Avidin-Texas Red (Vector).

 For digoxigenin-labeled probes: (1) 34 mg / ml of a mouse anti-dig congregate with FITC (Jackson). (2) 28 mg / ml donkey-FITC anti-mouse (Jackson). (3) 30 mg / ml mouse anti-rabbit-FITC (Jackson).

 After detection, the slides are rinsed rapidly in PBS xl and mounted with a reagent (Vectashield, Vector) before observation. The blades can be kept for months at 4 C in the dark.

EXAMPLE 1
In the case of combing human genomic DNA, simultaneous hybridizations of two cosmids separated by a gap of several tens of kb (109C8 and 180F1) are carried out: approximately 80 coupled signals (cosmid 1-gap-cosmid 2) could be measured on a single plate of 22x22 mm, corresponding to a total distance of about 120 kb.

* This result ensures that it is possible to observe about a hundred hybridizations of an entire BAC on a normal human genome. This experimental situation is to be compared with the theoretical situations envisaged in FIG. 5: the histograms 5d indeed show that for 100 diploid genomes (thus 200 occurrences of a cloned sequence), and with particular parameters, approximately 10 hybridization signals 125 kB (LO = 100 kb), or 16 intact hybridization signals (LO = 200 kb) are expected. In other words, the theoretical parameters used in the simulations are much more pessimistic than the experimental conditions that we are currently able to achieve.

 All of this theoretical and experimental analysis therefore allows a first validation of the present invention.

 In addition to the search for breakpoints using BACs (or other probes of equivalent or smaller size) mentioned above, it is also conceivable to use YACs (available for the entire human genome, and for other genomes ). Simulations similar to those presented above show that it is conceivable to detect the hybridization of a YAC of 1600 kb into two distinct fragments of 400 and 1200 kb (for example), on the basis of 100 genomes combed in different species. conditions corresponding to LO = 600, N = 10 in the Gaussian model.

 However, unlike observations of BACs or probes of similar or smaller size, which can be performed without difficulty using an epifluorescence microscope equipped with an objective x 100 or x 63 and a camera ( maximum size of a field of view, respectively 123 Fm and 195 Fm approximately), the measurement of fragments of several hundreds of kb requires the use of lower magnification objectives, for which problems of detectability of the signals can possibly occur. ask: x 40: field of view of maximum size 307 llm approximately.

x 20: field of view of maximum size 614 m approximately.

 The advantage of using probes of this size is obviously to reduce the number of hybridizations needed to find an interesting clone.

EXAMPLE 2
Assay of phage-k DNA in the E. coli genome
Genomic DNA from E. Coli (4.7 Mb) containing a copy of the phage A genome (line 5243, 49 kb) was combed on silanized surfaces after being counter stained with a fluorescent molecule (YOYO-1) (Fig. 1 (A).

 The total length of DNA combed by field of view was estimated from 30 to 50 fields of view distributed uniformly over the whole of each surface.

 Biotin-dUTP labeled phage k DNA was then hybridized and revealed using a FITC-coupled antibody system (green).

The total length of DNA hybridized per field of view was estimated from 100 fields of view distributed uniformly over the whole of each surface (Figure 1 (B)).

The results of 4 experiments are shown in the table below in which Nb E coli = Lc and Nb = LT
## EQU1 ## E. ## STR1 ## Coli #R
434 0.19 0.19 1.0 0.6
436 0.25 0.22 0.9 0.5
437 0.31 0.26 0.8 0.4
438 0.20 0.24 1.2 0.7
439 0.26 0.39 1.5 0.8
43 ~ 13 0.57 0.74 1.3 0.4
4314 0.61 0.56 0.9 0.3
These results showed, except for the 43 ~ 9 slide, a ratio reasonably close to the expected value of 1. However, the principle used, consisting in measuring the set of signals before hybridization is impracticable for larger genomes (I '). AR uncertainty is mainly due to the low number of E. coli genomes measured, of the order of 6 to 24).

EXAMPLE 3
Assay of the number of amplicons in the genome of hamster cells.

 The feasibility study has been extended to mammalian genomic DNA. The system chosen is the DNA of 2 lines of fibroblasts of the hamster lung (lines 618 and GMA32) respectively containing 1 and 2 copies of the target gene AMP1.

 Genomic cell DNA supplied in solution was combed on silanized surfaces, with an estimated density of 100 diploid genomes per 22x22 mm area.

 A cosmid probe (D3S1, 40 kb) specific for the target gene region was labeled with dig-dUTP. Another cosmid probe (565.5A1, 40 kb), specific for a region about 1 Mb from the target gene was used as a control, and labeled with biotin-dUTP.

 The probes were hybridized on the previously denatured combed genomic DNA, and revealed in red (biotinylated probes) and green (digoxygenated probes) in one case, and with inverted colors in the other. The hybridization signals of each color were measured on a number of fields of view representative of the surface.

 In the case of the A32 line, the total size obtained for the D3S1 probe (approximately 1400 sm) represents 70 copies of the gene. The size obtained on the same fields of view for the control probe (approximately 1180 sm) represents 60 haploid genomes. This measurement therefore gives a target / control ratio of 1.2 + 0.3 compatible with the presence of a copy of the gene per haploid genome.

 These experiments (Figure 2) thus show the feasibility of the gene assay from the target and control probe hybridization signal sig- nals, provided that a sufficient number of signals can be observed.

EXAMPLE 4 Dar Daires mapping of 6 cosmids on a YAC of 1600 Db contained in yeast genomic DNA
YAC 774G4 contains a human genomic DNA clone of chromosome 15 containing the calpain gene (CANP3), the mutation of which is responsible for belt dystrophy (LGMD 2A). In collaboration with J. Beckman's group of Genethon (Evry), we combed yeast genomic DNA contained in low melting agarose blocks (1 block, 1, ug DNA per block) on surfaces. silanized.

mapping cosmid pairs on human zenomorphic DNA
4 cosmids belonging to 3 neighboring contigs separated by gaps were used in 2 series of hybridizations, carried out in collaboration with the group of S. Povey of the MRC (London). Contigs cover the region of the TSC 1 gene involved in one of the forms of tuberous sclerosis (chromosome 9). The cosmid probes were prepared according to the protocol above.

 The genomic DNA used was extracted from cell cultures and put in low melting agarose block at the rate of 106 cells per block. 3 blocks treated according to the previous protocol (4 ml reservoir, use of a final molar MES pH 5.5 of 150 mM), were used for combing genomic DNA.

 The cosmids were hybridized 2 by 2 on similar slides, a total of several tens of double signals (red / green aligned) per slide being averaged. The same measurement protocol as before has been observed, giving rise to final values having the same precision as in the previous experiment.

 Figure 2 shows some typical images that allowed the measurement of the sizes of the two studied gaps.

Simultaneous mapping of 8 cosmids on YAC
8 cosmids (280A6, 37A1, 149I38, 134A11, 99B4, 140D1, 81F2 / 81H10 and 177B3) were hybridized on a YAC of the region of the TSC 2 gene (35HG8), in collaboration with the S. Povey group of the MRC ( London). The protocols are identical to those of the first experiment.

 As far as preliminary studies are concerned, no precise measurement has yet been made, and Figure 3 simply collects some images illustrating the principle of the reconstruction of the scrim obtained from a simultaneous hybridization of 8 cosmids.

Mapping of restriction segments
This mapping technique is naturally applicable to any other type of combed DNA or subclone. A possible extension of the technique would consist, for example, in mapping no longer subclones of a clone, but directly restriction fragments of the combed DNA (for example a BAC type clone).

 This would avoid the development of intermediate subclones before sequencing, the physical map of the restriction fragments being obtained with sufficient precision to allow a reconstruction of the final sequence. The applicability of this technique is based on the good separation of the restriction bands, and the sufficient size (> 10 kb) of the major fragments, as well as the subsequent subcloning of DNA from these bands (subcloning into small vectors, after additional enzymatic restriction, for sequencing).

Claims (17)

 1) A method for detecting the presence or location of one or more genes or of one or more specific DNA sequence A on a DNA B, characterized in that: (a) one fixes and combs a certain amount of said DNA B on a surface  combing, (b) the combing product B is reacted with one or more probes  labeled, linked to the specific gene (s) or DNA sequences A, (c) the information corresponding to at least one of the  following categories  (1) the position of the probes,  (2) the distance between probes,  (3) the size of the probes (the sum total of the sizes allowing  quantify the number of hybridized probes)  to deduce the presence, location and / or quantity of genes  or specific DNA sequences A.  2) Process according to claim 1 for demonstrating a genetic anomaly of breakage in a genome, characterized in that: (a) a certain amount of said genome is fixed and combed on a  combing surface, (b) the combing product is hybridized with one or more probes  specific labels corresponding to the genomic sequence  look for the anomaly, (c) measure the size of the fragments corresponding to the signals  hybridization, and (d) the presence of a break is deduced either by direct measurement or by  comparison with a witness.  3) Method according to claim 2, characterized in that the probe covers a part of the genome not involved in the break.  4) Method according to one of claims 2 and 3, characterized in that establishes a histogram of the lengths of the measured probes.  5) Method according to claim 4, characterized in that the histogram of the lengths of the probes is compared with a histogram made on a control genome with similar probes.  6) Method according to claim 1 for assaying a genomic sequence determined in a genome, characterized in that (a) a certain amount of said genome is fixed and combed on a  combing surface, (b) the combing product is hybridized with a control probe marked with  length it corresponding to a so-called control genomic sequence,  that is to say, whose number of copies in the genome is known, and  with the help of a specific probe marked length lc  corresponding to the genomic sequence to be assayed, said probes  can be identified separately, (c) the total length of the hybridization signals for  two probes, Lc and Lt, (d) we calculate for each the number of copies of the sequence  corresponding by the report  LT Lc Nt ~~~~~~~~ = and = ~~~~~~~~~  lt Ic  and we deduce the number of copies of the sequence to be compared.  to the control sequence.  7) Method according to claim 6, characterized in that the so-called control genomic sequence is a sequence present at 2 copies per genome and the presence of a gene abnormality is deduced when Nc and Nt are significantly different.  8) Method according to one of claims 6 and 7, characterized in that the sequence to be assayed corresponds to a sequence of a chromosome linked to a trisomy and that the control sequence is constituted by a sequence of another chromosome, the Nc / Nt ratio then being close to 1.5 in case of trisomy.  9) Method according to claim 6, characterized in that the sequence to be assayed corresponds to a deletion of a genome, the control sequence preferably being a sequence of the same chromosome.  10) Method according to claim 1, characterized in that the sequence to be assayed is a repeated sequence, the normal sequence then preferably being a sequence of the same chromosome.  11) Method according to claim 1, characterized in that: (a) a certain amount of said genome is fixed and combed on a  combing surface, (b) the combing product is hybridized with colored probes  corresponding to each clone, (c) the corresponding information is taken instead of each clone  as well as the corresponding sizes and distances on the genome, (d) the operations b) and c) n times are repeated by modifying the colors of the  subclones, knowing that with p different colors after n  hybridizations it is possible to position pn subclones.  12) Method according to one of claims 1 to 11, characterized in that using biotinylation-labeled probes and DIG which are revealed by an avidin-red system and anti-DIG-FITC (green).  13) Method according to one of claims 1 to 12, characterized in that disposed on the combed surface at least 100 copies of the genome.  14) Method according to one of claims 1 to 13, characterized in that the surface comprises a calibration DNA that calibrates each measurement.  15) Method according to one of claims 2 to 14, characterized in that the genome is from a sample of amniotic fluid.  16) Diagnostic kit for carrying out the method according to one of claims 1 to 15, characterized in that it comprises at least one of the following elements - a combing surface, - probes marked or intended for marked, corresponding to  abnormalities to be detected, - a device for combing DNA, - a control genome and the corresponding control probes, said genome  possibly being fixed on the surface to be combed.  17) Diagnostic kit for carrying out the method according to one of claims 1 to 15, characterized in that it comprises standard reference probes and genetic markers for locating a sequence.