Several genes have been identified which, if mutated, are associated with increased predisposition to tumours (1).

Carriers of mutations in these genes show a risk of cancer from a few percent to as high as 90%. Selected genes associated with high inherited predisposition to tumours and those most frequently examined in clinical practice are summarised in table 1 (2-6).

Several molecular methods have been devised aimed at detecting mutations. They can be subdivided into methods detecting:

? new mutations,

? known mutations.

Diagnostics of such mutations in cases appropriately preselected using pedigree and clinical data is justified in clinical practice, even though such techniques are still complex, time-consuming and expensive.

The main kinds of analyses:

? DNA isolation,

? amplification of gene fragments, usually of coding sequences,

? preliminary detection of changes within amplification products using screening techniques,

? sequencing and pyrosequencing,

? southern method and multiplex ligation-dependent probe amplification,

? high resolution melting analysis.

Material for DNA isolation is usually taken from blood leukocytes or less frequently from other tissues. Analyses allow detection of mutations which are constitutional, i.e. present in all cells of the patients. It is especially effective if the biological material for analysis is collected immediately before isolation. However, good results can be achieved even after a few days of storage of blood at room temperature or even storage for a few years at temperature below zero. If fresh tissue is not available, DNA isolation can be performed from tissues fixed in formalin and embedded in paraffin blocks, although the achievement of unequivocal results using such material is more difficult and sometimes even impossible. DNA isolation requires elimination of proteins from cellular lysate. In the phenol-chloroform method this is achieved by digestion with proteinase K and extraction in a mixture of phenol and chloroform. Finally, nuclear acids are extracted using ethyl- or isopropyl- alcohols. At present, the above technique is used only exceptionally although it produces clean and non-degraded DNA (in practice it is used only for DNA isolation from paraffin blocks). Instead, other techniques are used which are less laborious and easier for automation. They are based on selected labelling of DNA with template (e.g. with dyne beads) and then washing in order to separate the DNA.

DNA fragments are amplified using polymerase chain reaction. The reaction mixture includes: DNA template (usually genomic DNA), DNA polymerase, a pair of specific primers, deoxyribonucleotide triphosphates and the reaction buffer. This mixture is exposed to cyclic changes of temperature. Each cycle includes: denaturation, starters annealing and synthesis. After 22 cycles, assuming 100% effectiveness, the copy number of the amplified fragment is increased one million-fold.

DNA-SSCP (single strand conformational polymorphism) was in the past the most popular technique of initial detection of changes in amplification products (7).

Other techniques of this kind include HET (heteroduplex analyses) (8), CMC (chemical mismatch cleavage) (9), DHPLC (denaturing high-performance liquid chromatography) (10) and DGGE (denaturing gradient gel electrophoresis) (11).

At present, the best and the most frequently applied technique of initial detection of changes is DHPLC (10, 12-15). This is a kind of HET based on high resolution of modern chromatographic columns. Analysed DNA fragments are separated in a gradient of denaturing agent. (The key to DHPLC is the solid phase, which has differential affinity for single and double-stranded DNA). Under sub-denaturing conditions, heteroduplexes show lower affinity than homoduplexes to the solid phase of the column and it is easier to elute them. Separation is monitored by UV absorption measured at 260 nm. The elution profile (fig. 1) is characteristic and replicable for a given change and allows differentiation between new changes and already known mutations or polymorphisms.

Based on literature data (16) and our own experience (17) we can say that DHPLC combines the advantages of several methods. Its sensitivity is close to 100% (10, 14, 15). At the same time the cost is relatively low (cost of reagents per sample is 5-10 Euro). The method is quick, and if an autosampler is used, it allows 200 samples to be analysed per day.

Sequencing is the most sensitive technique for detection of changes in genomic material, allowing at the same time their full characterisation.

In the nineties, significant progress in sequencing technologies has been achieved by application of automated machines, for which the identification of particular nucleotides is based on fluorescence induced by laser. Each nucleotide (A, C, G, T) can be labelled with a different fluorescent dye. The most convenient technique is the cyclic sequencing method (18).

During the analysis the sequences of PCR products for both DNA strands are assessed. The real change is detected in both DNA strands. The sequencing procedure comprises several stages:

? preparative PCR – based on the amplification of a chosen fragment of the gene using pairs of specific starters,

? asymmetric PCR – separate amplification with each of the starters using fluorescent dye-labelled dideoxynucleotides,

? electrophoresis in denaturing polyacrylamide gel with simultaneous detection and registration of products,

? analysis of the results using computer programs.

During asymmetric PCR all possible oligonucleotides of different length, complementary to the template and containing fluorochromes at the 3'-end are created. They are separated during electrophoresis and the particular order of coloured nucleotides can be read as the sequence complementary to the template. The detected DNA sequence is compared with the wild sequence (fig. 2) available in databases such as GenBank and EMBL, and the type of change can be precisely described.

Currently, the leading companies are offering modern automated DNA sequencing instruments (DNA sequencers) allowing simultaneous sequencing of 96 samples based on capillary electrophoresis of products using the cyclic method using fluorescent dye-labelled dideoxynucleotides. Recent progress in this discipline has based not only on increasing the number of simultaneously analysed samples but also on improved "chemistry” such as gel compositions, which enable accurate sequencing of almost one thousand bases of one fragment.

New sequencers ("GS FLX machines”) rely on real time sequencing by simultaneous synthesis of many short DNA fragments (around 400 hundred base pairs in length). They apply pyrosequencing with detection of luminescence occurring during ATP degradation. These machines allow analyses of 400-600 million base pairs per instrument run. Pyrosequencing (19, 20) uses an one strand DNA fragment as the template on which synthesis of a complementary strand is performed through addition of 4 different deoxynucleotides triphosphate (dNTPs). Addition of each base is associated with liberation of pyrophosphate which is transformed into ATP using sulfurylase and adenosine-5'-phosphosulfate. ATP is used by luciferase for the transformation of luciferin into oxyluciferin. During this reaction light is generated with an intensity corresponding to the amount of pyrophosphate produced early. This light is registered by CCD and transformed into peaks on a pyrogram (fig. 3). The same reaction scheme is valuable for different dNTPs. If the added nucleotide is not complementary to the template it is not included in the newly synthesised strand and pyrophosphate is not created. The presence of a light signal is the requisite for adding a new nucleotide to a given sequence.

Based on our own experience the utmost disadvantage of pyrosequencing technique is the difficulty in determining the number of incorporated nucleotides in regions with base pairs repetitions. In homopolymeric regions of more than 5 nucleotides, estimation of complementary joined nucleotides is impossible due to nonlinear light response. However the problem can be solved by using nucleotide reversible terminators (NRTs), analogues modified by attaching a cleavable fluorophore to the base and a chemically reversible moiety to the 3' cap (3'- O -allyl or 3'- O -(2-nitrobenzyl)). During the extension step when the complementary NRT is incorporated, the reaction is temporarily terminated and resumed only when the capping moiety is removed (by deallyation or laser irradiation). In this way, when analysing the results on the pyrogram, each peak correspond to each incorporated nucleotide and homopolymeric regions can be clearly identified (21).

DNA mutation covering the sites of primers annealing or other fragments outside of amplified ones are not detected using DNA tests based on analyses described above. Some of such changes are large rearrangements.

A technique very popular in the past for detection of large rearrangements was Southern blotting, described for the first time by E. M. Southern in 1975.

At present the method which has almost completely replaced detection of DNA rearrangements by Southern blotting is MLPA (multiplex ligation-dependent probe amplification) (22). This technique is based on the ligation of specific probes and their subsequent amplification and allows an assessment of the exon copy numbers to be made. On this basis, conclusions can be drawn concerning deletions or duplications of gene fragments or of whole genes.

In this technique many probes are used simultaneously in "one tube”. Probes matching the sequences complementary to exon sequences also contain primer sequences and one of each pairs additionally a unique insertion sequence called a stuffer sequence. Hybridising sequences of each pair of probes match neighbouring DNA fragments and only if hybridisation is complete can ligation take place. After probe hybridisation to the template, the DNA fragments are ligated, then denatured. The dissociated ligated probe containing primer sequences is then amplified using PCR. The presence of stuffers of different lengths allows differentiation of products labelling different targets, and the amount of product is proportional to the copy number in the template. Each peak corresponds to the product of amplification of specific ligated pairs of probes (fig. 4). Relative differences in the height or area of the peak indicate quantitative (sometimes qualitative) changes of a target sequence for the probe.

The advantages of this technique are that only a small amount of DNA is necessary to perform analyses and that efficiently reproducible results may be achieved even from degraded genetic material.

Commercially available probes include those for the most important genes associated with a high risk of tumours, such as: ATM, BRCA1, BRCA2, CHEK1, MLH1, MSH2, MSH6, PMS2, APC, FANCA, FANCD2, PTCH, BMPR1A, SMAD4, TP53, CDH1, MEN1, NF1, NF2, STK11, SMARCB1, RB1, CDKN2A-CDKN2B, WT1.

This real-time PCR based method can be used for detection of SNPs as well as for large rearrangements. All mutations (small and large) can be screened simultaneously in one test, which allows reducing testing time. The basis of the genotyping is a unique pattern of melting curves.

The first step of the analysis is real-time PCR with fluorescent dye, mostly SYBR Green, LCGreen or Syto 9 (23). That allows monitoring the amplification of the DNA template, since fluorescence intensity is proportional to the amount of double-strand DNA (dsDNA). After overheating, the melting behaviour of the PCR products is monitored by plotting the changes in fluorescence that occur by denaturating double-strand DNA (dsDNA) (fig. 5). The pattern of melting temperature (Tm) differences can allow the discrimination of homo- and heterozygotes. The main problem of the HRMA method is that differences in the melting curve shape can easily identify heterozygotes, but may not distinguish all homozygotes (24). However, high sensitivity and specificity, low costs, small amount of DNA required (<5 ng) and the simplicity of the method are prominent features that make HRMA a great candidate as new screening method for cancer predisposition genes (23, 25).

The advantages of RNA analyses are mainly due to the possibility of detecting a mutation using a lower number of reactions (this is related to the shorter length of RNA in comparison to DNA). So far the main disadvantages of these techniques have included difficulties in achieving reproducible results, lower stability of RNA with mutations and troubles in interpretation associated with occurrence of RNA alternative splicing.

RNA analyses include a few stages:

? RNA isolation,

? amplification of coding parts of the genes,

? detection of changes in amplification products.

In the majority of laboratories, RNA is isolated from peripheral blood lymphocytes. RNA isolation is performed similarly to DNA isolation. However, due to the widespread presence of thermostable RNAses in tissues, RNA isolation has to be performed more carefully. It is very popular to use the isolation method by Chomczyński (26) from cellular lysates in a solution of guanidine thiocyanate (RNA inhibitors) followed by extraction in a mixture of phenol and chloroform. Slightly acidic pH of phenol leads to extraction of not only proteins but also of DNA, which under such conditions is practically insoluble.

RNA can be transcribed into cDNA (complementary DNA) using reverse transcriptase and then amplified using PCR. RNA does not include introns, so only a few pairs of starters usually allow amplification of the whole coding parts of selected genes. Analysing cDNA on regular agarose gels usually allows detection of RNA abnormalities which are caused by deletions or insertions of over a few dozen base pairs or splice mutations.

Detection of alterations in amplification products

Products of the RT-PCR reaction can be analysed using all of the techniques mentioned above and by using the in vitro transcription translation (IVTT) assay, also called the protein truncation test (PTT) (27, 28). RT-PCR is the first stage of PTT. In PTT one primer includes sequences not only initiating transcription into cDNA but also sequences initiating translation – it is protein synthesis " in vitro ” on a cDNA template. After electrophoresis and transmission into membranes, the length of the synthesised protein is assessed, which is altered not only by large deletions or insertions within the RNA but also when single nucleotide mutations lead to stop coding (TGA, TAA or TAG) or splicing mutations. The disadvantage of PTT is its limitation in the detection of missense mutations. It is estimated that even if we use all known tests of direct detection of changes their sensitivity does not exceed 70-80%, mainly due to the lack of diagnosis of alterations in sequences regulating the function.

More and more is known about the type and frequency of mutations predisposing to tumours, which are characteristic for different populations, including founder mutations and changes recurrently occurring in families of a given ethnic group. DNA tests aiming detection of all known mutations are highly valuable due to the unusually high economical effectiveness. Such genes as BRCA1, MLH1, MSH2and VHL have been studied intensively in many populations and it is known which mutations should be checked first and in which fragments of genes (29-31).