The present invention relates to a method for digesting cells or tissues in order to provide a solution comprising DNA in a form suitable for hybridisation.

Prior art methods for preparing cellular DNA in a form suitable for hybridisation involve a complex series of steps including disruption of the cells, enzymatic digestion of various macromolecules, precipitation, separation and dissolution steps in order to remove unwanted cellular components and leave a more or less pure solution of DNA. Reagents used in this procedure may interfere with hybridisation and must also be removed. A typical example of such a process can be found in Blin, N. and Stafford, D. . 1976, Nucleic Acids Research, 3_, 2303.

Brandsma and Miller, Proc. Natl. Acad. Sci. USA; 77 (11), 6851-6855, (1980) and Wagner e_t al. , Identification of Human Papillomavirus in Cervical Swabs by Deoxoyribonucleic Acid j Situ Hybridisation, Obstetrics and Gynaecology 6±, 767-772 (1984) describe method for detecting DNA in un-fractionated cells but these methods are probably not sensitive enough for the detection of single copy genesin less than 10 ⁵ cells. The method of Brandsma and Miller is carried out on a nitrocellulose filter and comprises treating the cells with 0.5 M sodium hydroxide for 7 minutes at ambient temperature and washing with neutralising buffers and solvents. Such a process

partially digests the cell contents, releasing DNA, but will not destroy cellular macromolecules sufficiently to prevent interference with a hybridisation test.

It is has now, surprisingly, been discovered that a solution of DNA suitable for use in hybridisation experiments may be obtained simply by digesting whole cells or tissue in a strongly alkaline medium at a sufficient temperature and for a sufficient time to achieve substantial destruction of all macromolecules except for D Ao This technique may be applied to fresh, fixed or archival material.

Accordingly the present invention provides a process for producing a solution of cellular DNA from cells or tissue which process consists of treating the cells or tissue with a strongly alkaline aqueous medium for a time and at a temperature sufficient to form a solution suitable for hybridisation for instance a clear solution. Solutions ©f DNA prepared according to the present invention may be used for hybridisation by application directly to a suitable filter support. Alternatively the solution may be used, after neutralisation of the excess hydroxide ions for analysis by hybridisation directly in solution, by enzymatic amplification of target sequences by the DNA polymerase chain reaction (R.K. Saiki et al, Nature, 324, , 163 (1986)) or by similiar methods.

This process has advantages in being extremely simple and quick by comparison with earlier techniques. As

there are no separation steps and hence minimal consequential loss of DNA it is also beneficial in increased sensitivity and is quantitative.

Without wishing to be bound by a particular theory, it is believed that DNA is more resistant to degradation by strongly alkaline media than any other cellular acromolecule and the present invention exploits the differential rates of such degradation.

The process of the invention will normally be applied to whole (live or fixed) cells or tissues but is equally applicable to disrupted cells. The benefits of simplicity of procedure and of sensitivity are lost if separation steps precede the digestion in hydroxide medium.

It is the concentration of hydroxide ions in the aqueous medium which is significant in achieving the required degradation and the nature of the cation is less critical. The concentration of hydroxide ions, the duration of the reaction and the temperature at which the reaction is conducted may be varied as convenient though it will be appreciated that less concentrated solutions of hydroxide ions will require longer time and/or higher temperatures; shorter times can be achieved using more concentrated hydroxide ions and/or higher temperatures and lower temperatures are possible when using more concentrated hydroxide ions and/or a longer duration of reaction.

In practical terms the temperature of the

reaction should be above the freezing point and at, or preferably below, the boiling point of the reaction mixture. Conveniently the reaction is performed at from about 20°C to about 95°C, more preferably at about 80°C. The concentration of hydroxide ions is conveniently from about 0.1M to about IM, more preferably from about 0.25M to about 0.5M and most preferably about 0.4M. Such concentrations may be achieved using a variety of alkaline or basic materials, preferably using an alkali metal or alkaline earth metal hydroxide, more preferably sodium or potassium and most preferably sodium hydroxide.

Suitably the duration of the reaction is selected having regard to any constraints on the temperature and/or concentration of hydroxide ions to be used. Typically the duration will be in the range of from a few minutes to a few hours preferably from 5 minutes to 1 hour. A duration of, for instance, the order of half an hour, particularly 30 minutes is convenient but a reaction time of about 10 minutes is considered optimal.

In a particularly preferred embodiment of the present invention the cells or tissue are digested using 0.4M sodium hydroxide at a temperature of about 80°C for about half an hour. This affords a solution containing DNA suitable for hybridisation in denatured form and which will generally have been degraded only to limited extent so that it is still suitable for analysis by hybridisation.

With archival material such as fixed and paraffin-

embedded tissue samples it may be necessary to clean up the sample before digestion. Thus for example with paraffin embedded samples, the paraffin wax may be dissolved using a solvent such as xylene and then ethanol.

Solutions of DNA produced according to the present invention may be used for hybridisation according to a variety of techniques. Most conveniently, the DNA is bonded to a substrate such as a nitrocellulose or, preferably, nylon membrane to which the DNA becomes convalently bound (Reed, K.C. and Mann, D.A, (1986) Nucl. Acid, Res 13, 7207-7221). After binding the DNA to the substrate, conventional hybridisation methods may be applied. Because the DNA preparation is a simple one-step process the losses of material are minimised, it is possible to work with very small samples and the absence of interference from extraneous cell debris permits detection of a single copy of a DNA sequence per cell in a sample of only 10 ⁴ cells. This in turn allows the detection of, for instance, viral DNA from human immunodeficiency virus (HIV), hepatitis virus etc. in blood samples and mycoplasma contamination of cell cultures with greater sensitivity than previously possible.

Accordingly, the invention also provides a process for producing a substrate having cellular DNA bound thereto which process comprises treating cells or tissue with a strongly alkaline aqueous medium for a time and at a temperature sufficient to produce a solution of DNA

suitable for hybridisation and contacting a hybridisation substrate with the solution under conditions permitting DNA to bind to the substrate.

Conditions permitting DNA to bind to hybridisation substrates are well known. Preferably the substrate is a nylon membrane in which case the binding is suitably conducted in the presence of about 0.4M sodium hydroxide and at about 20°C. Thus in a particularly preferred embodiment of the present invention the DNA solution is produced by treatment of cells with 0.4M sodium hydroxide at 80°C for about 10 minutes and the solution, after cooling to about 20°C, is applied directly to the nylon hybridisation membrane which, after binding the DNA, is washed with neutralising buffer. Hybridisation may then be conducted according to standard techniques.

The present invention further provides a process for detecting a DNA sequence within cells suspected to contain such DNA which comprises producing an aqueous solution of DNA by treating the cells with a strongly alkaline aqueous medium for a time and at a temperature sufficient to produce a solution of DNA suitable for hybridisation and contacting the DNA thus produced with a probe, comprising nucleic acid hybridisable to the DNA sequence to be detected and a detectable label, under conditions permitting hybridisation between the DNA sequence to be detected and the probe.

In a particular embodiment the DNA solution

produced in the first step is neutralised with suitable acid or buffer solution and hybridisation is conducted in the solution thus obtained. In an alternative embodiment a

* hybridisation substrate is contacted with the solution under conditions permitting DNA to bind to the substrate, the substrate is washed and hybridisation is conducted by contacting the substrate with a solution of the probe.

Conditions for use in hybridisation procedures are well known and the present process may be used at high or low stringency as required. The probe and label may be any probe/label combination, of which a large number are readily available. One advantage of the process of the present invention is that interference in the hybridisation by extraneous cell debris is minimised or eliminated. The hybridisation substrate can be washed to remove residues from the cells other than the DNA before hybridisation and the substrate with bound DNA is therefore suitable for re-use in further hybridisation experiments optionally after removal of the first probe.

The invention will now be described with reference to the figures of the accompanying drawings and to the following Examples which are not intended to limit the scope of the invention in any way.

Figure 1; Hydrolysis of HeLa DNA or BHK cells in 0.4N NaOH at 80°C lug samples of HeLa DNA in 50ul of PBS were made

up to 250ul with 0.5M NaOH resulting in a final NaOH concentration of 0.4M. After heating to 80°C for various lengths of time, duplicate samples were aliquoted onto Genescreen plus nylon membrane using a slot blot manifold and the filter was probed with total HeLa geno ic DNA and autoradiographed as described below.

Samples of 2.5 x 10 ⁵ BHK clone 9 cells in 50ul of PBS were made up to 250ul with 0„5M NaOH resulting in a final NaOH concentration of 0.4M. After heating to 80°C for various lengths of time, duplicate samples were aliquoted onto Genescreen plus as above and the filter was probed with total BHK genomic DNA and autoradiographed as described below. Figure 2: Hydrolysis of Yeast RNA in alkali

Total RNA f om Saccharomyces ceravisiae was hydrolysed as described in Figure 1 and the equivalent of lug was loaded into duplicate slots for each time point. Figure 3s Detection of CAD gene in normal BHK cells

BHK cells suspended in 50ul of PBS were made up to 250ul with 0.5M NaOH resulting in a final NaOH concentration of 0.4M. After heating to 80°C for 30 minutes, duplicate samples were aliquoted onto Genescreen plus as above and the filter was probed with pCADl42 cDNA (6.5 kb) (Shigesada et _^ al. , Mo . Cell. Biol., 5_, 1735-1742 (1985)). labelled by random priming as described below. Figure At Detection of pUC13 DNA in the presence of normal BHK cells.

2.5 x 10 ⁴ BHK cells suspended in 50ul of PBS were made up to 250ul with 0.5M NaOH resulting in a final NaOH concentration of 0.4M. After heating to 80°C for 30 minutes, duplicate samples were aliquoted onto Genescreen plus as above and the filter was probed with pUC13 DNA labelled by random priming as described below. Figure 5: Detection of Human papilloma virus 16 in infected cells.

SiHa and C331 cells were suspended at 2.5 x 10 ⁴ cells in 50ul of PBS and were hydrolysed in 250ul 0.4M NaOH at 80°C for 30 minutes then duplicate samples were aliquoted onto Genescreen plus as above and the filter was probed with HPV16 DNA labelled by random priming as described below.

EXAMPLES

EXAMPLE 1

Abbreviations:

1 x SSCs 150mM sodium chloride, 15mM sodium citrate,pH7.4

1 x SSPE: 150mM sodium chloride, 15mM sodium phosphate, pH

7.4

Debhardt's solution: 0.1% Ficoll 400 (Pharmacia), 0.1% BSA

(Sigma) 0.1% polyvinylpyrrolidone (Sigma).

SDSs sodium dodecyl sulphate.

Genescreen plus membrane was from NEN. All cells were grown in Dulbecco's modified Eagles medium (E4). C331

(Auersperg N, J. Nat. Cancer Inst. 3_2, 135-164, 1964) and SiHa (Friedl F e_t al^ Proc. Soc. Exp. Biol. Med. 135, 543-545, 1970) are human cervical carcinoma cells kindly supplied by Dr Lionel Crawford. HeLa DNA was prepared by the Proteinase K method (Maniatis e_t al. , "Molecular Cloning, a Laboratory Manual" ColdSpring Harbour Laboratory (1982)). BHK cells are baby Syrian hamster kidney cells. DNA probes were prepared by random priming (Feinberg and Vogelstein, Anal. Bioche . ^32, 6-13 (1983)) using labelled dCTP (3000 Ci/mmol, Amersham) . RNA probes were prepared using the SP6 promoter/polymerase system (SP6 RNA polymerase was from Promega Biotech) and labelled UTP (400 Ci/mmol, Amersham) according to the methods of Melton et al., Nucl. Acid. Res. 12, 7035-7056 (1984)) after subcloning DNA fragments into plasmids pSP64 or pSP65. Whole cell alkaline hydrolysis and blotting was performed as follows:

Cells were grown to confluence on normal plastic dishes or used as a washed cell pellet. The limit of detection for a single copy gene with a good probe is 5 xlO ³ cells.

Cells were washed twice in PBS and stored at -20°C before use. A typical experiment would involve a sample of cells resuspended in 50ul of PBS in a screw capped Eppendorf tube to which 200ul of 0.5M NaOH is added to give 250ul of cells in 0..4M NaOH. This mixture is heated to 80°C for 30 minutes.

A piece of Genescreen plus filter or Zetaprobe nylon membrane and a piece of 3MM (Whatman) were prepared for the blotting apparatus, by wetting in 0.4M sodium hydroxide and the slot blotter was assembled. Aliquots of lOOul of the clear DNA solution were placed into duplicate slots of the slot blotter. The solution was drawn through by capillary action when a capillary type slot blotter was used (relying on capillary action or absorbent towels). Wells were rinsed with lOOul of the 0.4M NaOH solution, the apparatus dismantled and the filter placed in 250ml of 2x SSC for 5 minutes to neutralise it. The filter was blotted and allowed to dry at room temperature for at least one hour before use.

The filter was evenly wetted with 2 x SSC and placed in a suitable container or thick plastic bag.

Prehybridisation was for at least 15 minutes at 42°C in 50% deionised formamide, 5 x SSPE, 5 x Denhardt's 1% SDS, lOOug/ml sheared and denatured herring or salmon sperm DNA after removal of air bubbles.

If necessary to denature the probe, 1ml of hybridisation solution was added to it, the probe was heated to 65°C for 10 minutes and cooled on ice before adding it to the filter.

Hybridisation was in the same solution containing up to 10 ⁷ dpm/ml of labelled denatured probe at 42°C (on a rocker if available) for 2-16 hours after removal of air bubbles. Probe was removed and the filter washed as

follows: 5—15 minutes in 250ml of 2 x SSC at room temperature, 30 minutes in 250ml of 2 x SSC, 1% SDS at 65°C and finally, if high stringency was required, 30 minutes in 250ml of 0.2 x SSC, 1% SDS at 65°C. Excess buffer was blotted off on 3MM paper but the filter was not dried if it was intended to strip off the probe and reuse the blot. For autoradiography, the filter was wrapped in plastic.

If the filters were to be reused, probe was removed by boiling in ImM EDTA pH 8.0, 1% SDS for 30 minutes. Autoradiography was carried out at -70°C with intensifying screens using Kodak XAR5 film, preflashed to give a linear response. Results

In order to establish a convenient protocol for the preparation of DNA from whole cells in a form suitable for blotting, 0.4M NaOH was chosen because this is used in the standard alkaline Southern blot procedure and the time and temperature were varied in a preliminary experiment. It was determined that treatment at 80°C results in little loss of DNA hybridisation signal after about 30 minutes without significant evaporative loss of hydrolysate so this temperature was selected as a convenient one for future experiments.

The stability of pure DNA was tested using HeLa DNA and cell hydrolysis was tested with several cells types (Figure 1). The rate of hydrolysis of DNA in whole cells is the same as for pure DNA but the time course of DNA

binding to the nylon is initially lower in whole cell hydrolysates due to competition by other macromolecules with DNA for binding. As expected, hydrolysis of DNA in all cell types tested showed the same kinetics.

In the initial period of hydrolysis much of the signal detected could be due to RNA, for this reason the kinetics of RNA hydrolysis were investigated. Figure 2 shows that using Yeast whole RNA, essentially no RNA - dependant signal remains after hydrolysis for 5 minutes in 0.4M NaOH at 80°C. Thus, for the purpose of quantitation of DNA sequence copy number, it is necessary to hydrolyse cells for at least 5 minutes, to avoid the need for an RNAase step to be included after binding nucleic acids to the filter. The standard conditions initially adopted for quantitating gene copy number were to hydrolyse cells in 0.4M NaOH at 80°C for 30 minutes. After further investigation of the reaction time an optimal duration of 10 minutes was adopted.

Sensitivity in detecting genes of mammalian, viral and bacterial origin was estimated. A single copy gene, CAD, could be detected in 5 x 10 ³ BHK cells with a 24 hour film exposure using preflashed Kodak XAR5 film and two intensifying screens (Figure 3). The 6.5Kb of sequences corresponding to the amount of CAD cDNA in the probe could be detected in as few as 5 x 10 ³ cells and corresponding to 30fg of target DNA. Amplified genes of interest in cells can certainly be detected in fewer than 10 ⁴ cells although

the practical limit has not yet been determined. The limits of the sensitivity for detection of a plasmid sequence diluted in mammalian cells was determined using pUCl3 DNA diluted in BHK clone 9.4 cells (10 ⁴ cells per slot). 25fg of the plasmid DNA could be detected readily using a DNA probe, approaching the lower limit of 3 fg claimed by Church and Gilbert Proc. Natl. Acad. Sci 81, 1991-1995 (1983)) for RNA/DNA hybridisation. It is possible that by changing the specific activity of the probe or the hybridisation and washing conditions that this could be improved. About 5 x 10 ⁵ cells can be loaded per slot before saturating the support. Thus it should be possible to detect a sequence that represents less than 25fg in 5 x 10 ⁵ cells, or about one cell with a 6Kb single copy gene in 100 cells which do not contain that sequence. Non-quantitative uses for the assay were tested where a simple plus or minus answer was needed. Two human cervical carcinoma cell lines, SiHa cells containing about 50 copies of HPV 16 were used to show that the method was useful for detecting viral DNA. As a negative control, C331 cervical carcinoma cells were used. However, these cells turned out to habor HPV16 which was only detected with this sensitive assay and had not been observed before. The virus could be detected easily in 10 ⁵ cells after an overnight exposure. The method has also been used to assay for mycoplasma infection of cultured cell lines in our laboratory.

Discussion

The main advantage of this technique is that is enables the quantification of DNA sequences in cells by nucleic acid hybridisation without the need to first isolate pure DNA. Taken together with the high sensitivity of the hybridisation to nylon membrane, this affords a considerable improvement in the ability to detect sequences in very small numbers of cells as well as an improved ability to handle large numbers of samples.

A short list of other uses would include the analysis of gene copy number in gene amplification, detection of DNA viruses and pro-retroviruses in tissues and cells lines, detection of transfected DNA sequences in cell lines and transgenic animals and DNA-typing blood parasites such as Plasmodium falciparum in humans and Babesia bovis in cattle.

EXAMPLE ^

Method: A single 5 to 10 micron section of approximately lcm ² area (or equivalent number of sections to approximate to this tissue volume) of formalin fixed paraffin embedded tissue is sectioned and placed in an Eppendorf tube. The paraffin wax is removed from the tissue by washing in xylene and then in absolute ethanol. The tissue is then hydrolysed in 0.4M sodium hydroxide for one hour. This time is optimal for this type of sample. The supernatent is then pipetted on to a nylon membrane [Gene Screen Plus] though a slot or dot blot manifold, washed with 0.4M NaOH and rinsed in 2 x SSC before being probed with radiolabelled DNA. Hybridisation, washing and autoradiography are carried out according to Example 1.