Nucleic acid hybridization assays using cloned DNA probes to screen uncloned nucleic acid populations

Numerous applications in molecular genetics involve taking an individual DNA clone and using it as a hybridization probe to screen for the presence of related sequences within a complex target of uncloned DNA or RNA. Sometimes the assay is restricted to simply checking for presence or absence of sequences related to the probe. In other cases, useful information can be obtained regarding the size of the complementary sequences, their subchromosomal location or their locations within specific tissues or groups of cells.

The general procedure of dot-blotting involves taking an aqueous solution of target DNA, for example total human genomic DNA, and simply spotting it on to a nitrocellulose or nylon membrane then allowing it to dry. The variant technique of slot-blotting involves pipetting the DNA through an individual slot in a suitable template. In both methods the target DNA sequences are denatured, either by previously exposing to heat, or by exposure of the filter containing them to alkali. The denatured target DNA sequences now immobilized on the membrane are exposed to a solution containing single-stranded labeled probe sequences. After allowing sufficient time for probe-target heteroduplex formation, the probe solution is decanted, and the membrane is washed to remove excess probe that may have become nonspecifically bound to the filter. It is then dried and exposed to an autoradiographic film.

A useful application of dot-blotting involves distinguishing between alleles that differ by even a single nucleotide substitution. To do this allele-specific oligonucleotide (ASO) probes are constructed from sequences spanning the variant nucleotide site. ASO probes are typically 1520 nucleotides long and are normally employed under hybridization conditions at which the DNA duplex between probe and target is stable only if there is perfect base complementarity between them: a single mismatch between probe and target sequence is sufficient to render the short heteroduplex unstable (Figure 5.10). Typically, this involves designing the oligonucleotides so that the single nucleotide difference between alleles occurs in a central segment of the oligonucleotide sequence, thereby maximizing the thermodynamic instability of a mismatched duplex. Such discrimination can be employed for a variety of research and diagnostic purposes. Although ASOs can be used in conventional Southern blot hybridization (see below), it is more convenient to use them in dot-blot assays (see Figure 5.11).

Another method of ASO dot blotting uses a reverse dot-blotting approach. This means that the oligonucleotide probes are not labeled and are fixed on a filter or membrane whereas the target DNA is labeled and provided in solution. Positive binding of labeled target DNA to a specific oligonucleotide on the membrane is taken to mean that the target has that specific sequence. This approach, and related DNA microarray methods, have many diagnostic applications (see Section 17.1.4).

In this procedure, the target DNA is digested with one or more restriction endonucleases, size-fractionated by agarose gel electrophoresis, denatured and transferred to a nitrocellulose or nylon membrane for hybridization (Figure 5.12). During the electrophoresis, DNA fragments, which are negatively charged because of the phosphate groups, are repelled from the negative electrode towards the positive electrode, and sieved through the porous gel. Smaller DNA fragments move faster. For fragments between 0.1 and 20 kb long, the migration speed depends on fragment length, but scarcely at all on the base composition. Thus, fragments in this size range are fractionated by size in a conventional agarose gel electrophoresis system. To achieve efficient size-fractionation of large fragments (40 kb to several megabases), a more specialized system is required, such as a pulsed-field gel electrophoresis apparatus (see Section 10.2.2).

Following electrophoresis, the test DNA fragments are denatured in strong alkali. As agarose gels are fragile, and the DNA in them can diffuse within the gel, it is usual to transfer the denatured DNA fragments by blotting on to a durable nitrocellulose or nylon membrane, to which single-stranded DNA binds readily. The individual DNA fragments become immobilized on the membrane at positions which are a faithful record of the size separation achieved by agarose gel electrophoresis. Subsequently, the immobilized single-stranded target DNA sequences are allowed to associate with labeled single-stranded probe DNA. The probe will bind only to related DNA sequences in the target DNA, and their position on the membrane can be related back to the original gel in order to estimate their size.

An important application of Southern blot hybridization in mammalian genetics is the ability to identify a given DNA probe as a member of a repetitive DNA family. Many important mammalian genes belong to multigene families, and many other DNA sequences show varying degrees of repetition. Once a newly isolated probe is demonstrated to be related to other uncharacterized sequences, attempts can then be made to isolate the other members of the family by screening genomic DNA libraries. Additionally, screening can also be conducted on genomic DNA samples from different species to identify interspecific related sequences. An important route to identifying coding DNA involves identifying sequences that are highly conserved in evolution (Section 10.4.1).

Northern blot hybridization is a variant of Southern blotting in which the target nucleic acid is RNA instead of DNA. A principal use of this method is to obtain information on the expression patterns of specific genes. Once a gene has been cloned, it can be used as a probe and hybridized against a Northern blot containing, in different lanes, samples of RNA isolated from a variety of different tissues (see Figure 5.13). The data obtained can provide information on the range of cell types in which the gene is expressed, and the relative abundance of transcripts. Additionally, by revealing transcripts of different sizes, it may provide evidence for the use of alternative promoters, splice sites or polyadenylation sites.

Southern blot hybridization has been used extensively in molecular genetic studies as a means of genomic restriction mapping: a labeled DNA probe from one genome can be used to infer the structure of related sequences in the same or different genomes. Because the genomic DNA samples are fractionated by separation of restriction fragments according to size, mutations which alter a restriction site, and significantly large insertions or deletions occurring between neighboring restriction sites, can be typed. Such mutations will result in altered restriction fragment lengths, that is restriction fragment length polymorphisms (RFLPs).

Very occasionally, a pathogenic mutation directly abolishes or creates a restriction site, enabling direct screening for the pathogenic mutation. For example, the sickle cell mutation is a single nucleotide substitution (A T) at codon 6 in the b-globin gene, which causes a missense mutation (Glu Val), and at the same time abolishes an MstII restriction site which spans codons 5 to 7. The nearest flanking restriction sites for MstII, located 1.2 kb upstream in the 5-flanking region and 0.2 kb downstream at the 3 end of the first intron, are well conserved. Consequently, a b-globin DNA probe can differentiate the normal b^A-globin and the mutant b^S-globin alleles in MstII-digested human DNA: the former exhibits 1.2 kb and 0.2 kb MstII fragments, whereas the sickle cell allele exhibits a 1.4 kb MstII fragment (Figure 5.14).

The great majority of mutations are not associated with disease; instead, they often occur within noncoding DNA sequences. As a large number of recognition sequences are known for type II restriction endonucleases, many point mutation polymorphisms will be characterized by alleles which possess or lack a recognition site for a specific restriction endonuclease and therefore display restriction site polymorphism (RSP). Accordingly, individual RSPs normally have two detectable alleles (one lacking and one possessing the specific restriction site). RSPs can be assayed by digesting genomic DNA samples with the relevant restriction endonuclease and identifying specific restriction fragments whose lengths are characteristic of the two alleles, so-called RFLPs (Figure 5.15).

DNA probes can also be used to monitor VNTR polymorphisms where alleles differ by a variable number of tandem repeats. To do this, genomic DNA samples are digested with a restriction endonuclease which recognizes well-conserved restriction sites flanking a specific VNTR locus. The resulting restriction fragments are separated according to size on agarose gels, transferred to a suitable membrane and hybridized with a probe representing a unique sequence from the corresponding locus. The resulting pattern of locus-specific RFLPs does not reflect RSP: instead, the differences in sizes of the restriction fragments represent integral numbers of the tandemly repeated unit.

Although the term VNTR could, in theory, encompass a wide range of repeat lengths, in practice the term is usually reserved for moderately large arrays of a repeat unit which is typically in the 564 bp region (so-called hypervariable minisatellite DNA, distinguishing it from simple tandem repeat polymorphism (where the repeat unit length is from 1 to 4 bp; i.e. microsatellite DNA) and tandem repeat polymorphism associated with very large arrays of satellite DNA.

If the VNTR locus is a member of a repeated DNA family, the use of a VNTR repeat probe, rather than a unique flanking probe, will produce a complex polymorphic pattern. For example, hypervariable minisatellite DNA clones have been used as probes against Southern blots of appropriately digested genomic DNA. Cross-hybridization of such probes with the other members of this highly repeated DNA family results in a pattern of hybridization bands representing the summed contributions of two alleles at each of many hypervariable loci scattered throughout the genome. Consequently, the overall polymorphism of the multilocus hybridization patterns is uniquely high. Because it permits distinction between any two individuals who are not identical twins, probing with hypervariable minisatellites has been termed DNA fingerprinting (see Sections 17.4.217.4.4).

Certain diseases are associated with a high frequency of deletion of all or part of a gene. If a partial restriction map has been established for the gene under investigation, deletions can be screened by Southern blot hybridization using an appropriate intragenic DNA probe. If the deletion is a small one, for example a few hundred base pairs, it is often apparent as a consistent reduction in size of normal restriction fragments in the gene. An individual who is homozygous for this mutation, or is a heterozygote with one normal allele and another with a small deletion, can easily be identified by detecting the aberrant size restriction fragments.

Large deletions will lead to absence of specific restriction fragments. Homozygous deletion of large DNA segments can easily be detected as complete absence of appropriate restriction fragments associated with the gene. If, however, an individual is heterozygous for a relatively large gene deletion, the deletion may still be detected by demonstrating comparatively reduced intensity of specific gene fragments. For example, patients with 21-hydroxylase deficiency often have deletions of about 30 kb of the 21-hydroxylase/C4 gene cluster. Such pathological deletions eliminate the functional 21-hydroxylase gene, CYP21, and an adjacent C4B gene, leaving the related CYP21P pseudogene and C4A genes. Patients with homozygous deletions will show absence of diagnostic restriction fragments associated with CYP21 and C4B, while carriers of the deletion will show a 2:1 ratio of CYP21P:CYP21 and of C4A:C4B (Collier et al., 1989; Figure 5.16).

A simple procedure for mapping genes and other DNA sequences is to hybridize a suitable labeled DNA probe against chromosomal DNA that has been denatured in situ. To do this, an air-dried microscope slide preparation of metaphase or prometaphase chromosomes is made, usually from peripheral blood lymphocytes or lymphoblastoid cell lines. Treatment with RNase and proteinase K results in partially purified chromosomal DNA, which is denatured by exposure to formamide. The denatured DNA is then available for in situ hybridization with an added solution containing a labeled nucleic acid probe, overlaid with a coverslip. Depending on the particular technique that is used, chromosome banding of the chromosomes can be arranged either before or after the hybridization step. As a result, the signal obtained after removal of excess probe can be correlated with the chromosome band pattern in order to identify a map location for the DNA sequences recognized by the probe. Chromosome in situ hybridization has been revolutionized by the use of fluorescence in situ hybridization (FISH) techniques (see Section 10.1.4).

In this procedure, a labeled probe is hybridized against RNA in tissue sections (Wilkinson, 1998). Tissue sections are made from either paraffin-embedded or frozen tissue using a cryostat, and then mounted on to glass slides. A hybridization mix including the probe is applied to the section on the slide and covered with a glass coverslip. Typically, the hybridization mix has formamide at a concentration of 50% in order to reduce the hybridization temperature and minimize evaporation problems.

Although double-stranded cDNAs have been used as probes, single-stranded complementary RNA probes (riboprobes) are preferred: the sensitivity of initially single-stranded probes is generally higher than that of double-stranded probes, presumably because a proportion of the denatured double-stranded probe renatures to form probe homoduplexes. cRNA riboprobes that are complementary to the mRNA of a gene are known as antisense riboprobes and can be obtained by cloning a gene in the reverse orientation in a suitable vector such as pSP64 (see Figure 5.4). In such cases, the phage polymerase will synthesize labeled transcripts from the opposite DNA strand to that which is normally transcribed in vivo. Useful controls for such reactions include sense riboprobes which should not hybridize to mRNA except in rare occurrences where both DNA strands of a gene are transcribed.

Labeling of probes is performed using either selected radioisotopes, notably ³⁵S, or by nonisotopic labeling. In the former case, the hybridized probe is visualized using autoradiographic procedures. The localization of the silver grains is often visualized using only dark-field microscopy (direct light is not allowed to reach the objective; instead, the illuminating rays of light are directed from the side so that only scattered light enters the microscopic lenses and the signal appears as an illuminated object against a black background). However, bright-field microscopy (where the image is obtained by direct transmission of light through the sample) provides better signal detection (see Figure 5.17). Fluorescence labeling is a popular nonisotopic labeling approach and detection is accomplished by fluorescence microscopy (see Box 5.2, Figure 5.5A).

Figure 5.17. Tissue in situ hybridization. The example shows the pattern of hybridization produced using a ³⁵S-labeled b-myosin heavy chain antisense riboprobe against a transverse section of tissue from a 13 day embryonic mouse. The dark areas represent strong labeling, notably in the ventricles of the heart. Kindly supplied by Dr David Wilson, University of Newcastle upon Tyne.

Figure 5.4. Riboprobes are generated by run-off transcription from cloned DNA inserts in specialized plasmid vectors. The plasmid vector pSP64 contains a promoter sequence for phage SP6 RNA polymerase linked to the multiple cloning site (MCS) in addition to an origin of replication (ori) and ampicillin resistance gene (amp). After cloning a suitable DNA fragment in one of the 11 unique restriction sites of the MCS, the purified recombinant DNA is linearized by cutting with a restriction enzyme at a unique restriction site just distal to the insert DNA (Pvu II in this example). Thereafter labeled insert-specific RNA transcripts can be generated using SP6 RNA polymerase and a cocktail of NTPs, at least one of which is labeled (UTP in this case).