Preparation of nucleic acid probes

Preparation of nucleic acid probes

In standard nucleic acid hybridization assays the probe is labeled in some way. Nucleic acid probes may be made as single-stranded or double-stranded molecules (see Figure 5.1), but the working probe must be in the form of single strands.

Conventional DNA probes are isolated by cell-based DNA cloning or by PCR. In the former case, the starting DNA may range in size from 0.1 kb to hundreds of kilobases in length and is usually (but not always) originally double-stranded. PCR-derived DNA probes have often been less than 10 kb long and are usually, but not always, originally double-stranded. Conventional DNA probes are usually labeled by incorporating labeled dNTPs during an in vitro DNA synthesis reaction (see Section 5.1.1).

RNA probes can conveniently be generated from DNA which has been cloned in a specialized plasmid vector ( Melton et al., 1984). Such vectors normally contain a phage promoter sequence immediately adjacent to the multiple cloning site. An RNA synthesis reaction is employed using the relevant phage RNA polymerase and the four rNTPs, at least one of which is labeled. Specific labeled RNA transcripts can then be generated from the cloned insert (seeSection 5.1.1).

Oligonucleotide probes are short (typically 1550 nucleotides) single-stranded pieces of DNA made by chemical synthesis: mononucleotides are added, one at a time, to a starting mononucleotide, conventionally the 3 end nucleotide, which is bound to a solid support. Generally, oligonucleotide probes are designed with a specific sequence chosen in response to prior information about the target DNA. Sometimes, however, oligonucleotide probes are used which are degenerate in sequence. Typically this involves parallel syntheses of a set of oligonucleotides which are identical at certain nucleotide positions but different at others. Oligonucleotide probes are often labeled by incorporating a ³²P atom or other labeled group at the 5 end (see next section).

5.1.1. DNA and RNA can conveniently be labeled in vitro by incorporation of nucleotides (or nucleotide components) containing a labeled atom or chemical group

Although, in principle, DNA and RNA can be labeled in vivo, by supplying labeled deoxynucleotides to tissue culture cells, this procedure is of limited general use; it has been restricted largely to preparing labeled viral DNA from virus-infected cells, and studying RNA processing events. A much more versatile method involves in vitro labeling: the purified DNA, RNA or oligonucleotide is labeled in vitro by using a suitable enzyme to incorporate labeled nucleotides. Two major types of procedure have been widely used:

• Labeling of new strands duringin vitroDNA or RNA synthesis. In this type of procedure, DNA or RNA polymerase is used to make labeled DNA or RNA copies of a starting DNA. The in vitro DNA or RNA synthesis reaction requires that at least one of the four nucleotide precursors carries a labeled group. Labeling of DNA by in vitro DNA synthesis is normally accomplished using one of three methods: nick-translation (this section); random primed labeling (this section); or PCR-mediated labeling (see Section 6.1.1). Labeling of RNA generally is carried out using an in vitro transcription system (see further on in this section).

• End-labeling. This type of procedure involves addition of a labeled group to one or a few terminal nucleotides. It is less widely used, but is useful for a number of procedures, including labeling of single-stranded oligonucleotides (see below) and restriction mapping. Inevitably, because only one or a very few labeled groups are incorporated, the specific activity (the amount of radioactivity incorporated divided by the total mass) of the labeled DNA is much less than that for probes in which there has been incorporation of several labeled nucleotides along the length of the DNA.

Labeling DNA by nick translation

The nick-translation procedure involves introducing single-strand breaks (nicks) in the DNA, leaving exposed 3 hydroxyl termini and 5 phosphate termini. The nicking can be achieved by adding a suitable endonuclease such as pancreatic deoxyribonuclease I (DNase I). The exposed nick can then serve as a start point for introducing new nucleotides at the 3 hydroxyl side of the nick using the DNA polymerase activity of E. coli DNA polymerase I at the same time as existing nucleotides are removed from the other side of the nick by the 5 3exonuclease activity of the same enzyme. As a result, the nick will be moved progressively along the DNA ('translated') in the 5 3 direction (see Figure 5.2A). If the reaction is carried out at a relatively low temperature (about 15° C), the reaction proceeds no further than one complete renewal of the existing nucleotide sequence. Although there is no net DNA synthesis at these temperatures, the synthesis reaction allows the incorporation of labeled nucleotides in place of the previously existing unlabeled ones.

Random primed DNA labeling

The random primed DNA labeling method (sometimes known as oligolabeling) (Feinberg and Vogelstein, 1983) is based on hybridization of a mixture of all possible hexanucleotides: the starting DNA is denatured and then cooled slowly so that the individual hexanucleotides can bind to suitably complementary sequences within the DNA strands. Synthesis of new complementary DNA strands is primed by the bound hexanucleotides and is catalyzed by the Klenow subunit of DNA polymerase I (which contains the polymerase activity in the absence of associated exonuclease activities). DNA synthesis occurs in the presence of the four dNTPs, at least one of which has a labeled group (seeFigure 5.2B). This method produces labeled DNAs of high specific activity. Because all sequence combinations are represented in the hexanucleotide mixture, binding of primer to template DNA occurs in a random manner, and labeling is uniform across the length of the DNA.

End-labeling of DNA

Single-stranded oligonucleotides are usually end-labeled using polynucleotide kinase (kinase end-labeling). Typically, the label is provided in the form of a ³²P at the g-phosphate position of ATP and the polynucleotide kinase catalyses an exchange reaction with the 5-terminal phosphates (see Figure 5.3A). The same procedure can also be used for labeling double-stranded DNA. In this case, fragments carrying label at one end only can then be generated by cleavage at an internal restriction site, generating two differently sized fragments which can be separated by gel electrophoresis and purified.

Larger DNA fragments can be end-labeled by various alternative methods. Fill-in end-labeling (Figure 5.3B) is one popular approach, and uses the Klenow subunit of E. coli DNA polymerase. Again, fragments carrying label at one end only can be generated by restriction cleavage and size fractionation. An alternative PCR-based method is primer-mediated 5end-labeling (see Section 6.1.1).

Labeling of RNA

The preparation of labeled RNA probes (riboprobes) is most easily achieved byin vitro transcription of insert DNA cloned in a suitable plasmid vector. The vector is designed so that adjacent to the multiple cloning site is a phage promoter sequence, which can be recognized by the corresponding phage RNA polymerase. For example, the plasmid vector pSP64 contains the bacteriophage SP6 promoter sequence immediately adjacent to a multiple cloning site (see Figure 5.4). The SP6 RNA polymerase can then be used to initiate transcription from a specific start point in the SP6 promoter sequence, transcribing through any DNA sequence that has been inserted into the multiple cloning site. By using a mix of NTPs, at least one of which is labeled, high specific activity radiolabeled transcripts can be generated (Figure 5.4). Bacteriophage T3 and T7 promoter/RNA polymerase systems are also used commonly for generating riboprobes. Labeled sense and antisense riboprobes can be generated from any gene cloned in such vectors (the gene can be cloned in either of the two orientations) and are widely used in tissue in situhybridization (Section 5.3.4).

5.1.2. Nucleic acids can be labeled by isotopic and nonisotopic methods

Isotopic labeling and detection

Traditionally, labeling of nucleic acids has been conducted by incorporating nucleotides containing radioisotopes. Such radiolabeled probes contain nucleotides with a radioisotope (often ³²P, ³³P, ³⁵S or ³H), which can be detected specifically in solution or, much more commonly, within a solid specimen (autoradiography - see Box 5.1).

The intensity of an autoradiographic signal is dependent on the intensity of the radiation emitted by the radioisotope, and the time of exposure, which may often be long (one or more days, or even weeks in some applications). ³²P has been used widely in Southern blot hybridization, dot-blot hybridization, colony and plaque hybridization (see below) because it emits high energy b-particles which afford a high degree of sensitivity of detection. It has the disadvantage, however, that it is relatively unstable (see Table 5.1). Additionally, its high energy b-particle emission can be a disadvantage under circumstances when fine physical resolution is required to interpret the resulting image unambiguously. For this reason, radionuclides which provide less energetic b-particle radiation have been preferred in certain procedures, for example ³⁵S-labeled and ³³P-labeled nucleotides for DNA sequencing and tissue in situ hybridization, and³H-labeled nucleotides for chromosome in situ hybridization. ³⁵S and ³³P have moderate half-lives while ³H has a very long half-life. However, the latter isotope is disadvantaged by its comparatively low energy b-particle emission which necessitates very long exposure times.

³²P-labeled and ³³P-labeled nucleotides used in DNA strand synthesis labeling reactions have the radioisotope at the a-phosphate position, because the b- and g-phosphates from dNTP precursors are not incorporated into the growing DNA chain. Kinase-mediated end-labeling, however, uses [g-³²P]ATP (see Figure 5.3A). In the case of ³⁵S-labeled nucleotides which are incorporated during the synthesis of DNA or RNA strands, the NTP or dNTP carries a ³⁵S isotope in place of the O^- of the a-phosphate group. ³H-labeled nucleotides carry the radioisotope at several positions. Specific detection of molecules carrying a radioisotope is most often performed by autoradiography (see Box 5.1).

Nonisotopic labeling and detection

Nonisotopic labeling systems involve the use of nonradioactive probes. Although developed only comparatively recently, they are becoming increasingly popular and are finding increasing applications in a variety of different areas (Kricka, 1992). Two types of non-radioactive labeling are conducted:

• Direct nonisotopic labeling, where a nucleotide which contains the label that will be detected is incorporated. Often such systems involve incorporation of modified nucleotides containing a fluorophore (Figure 5.5A), a chemical group which can fluoresce when exposed to light of a certain wavelength (fluorescence labeling - see Box 5.2).

• Indirect nonisotopic labeling, usually featuring the chemical coupling of a modified reporter molecule to a nucleotide precursor. After incorporation into DNA, the reporter groups can be specifically bound by an affinity molecule, a protein or other ligand which has a very high affinity for the reporter group. Conjugated to the latter is a marker molecule or group which can be detected in a suitable assay (Figure 5.6). The reporter molecules on modified nucleotides need to protrude sufficiently far from the nucleic acid backbone to facilitate their detection by the affinity molecule and so a long carbon atom spacer is required to separate the nucleotide from the reporter group.

Two indirect nonisotopic labeling systems are widely used:

• The biotin-streptavidin system utilizes the extremely high affinity of two ligands: biotin (a naturally occurring vitamin) which acts as the reporter, and the bacterial protein streptavidin, which is the affinity molecule. Biotin and streptavidin bind together extemely tightly with an affinity constant of 10^-14, one of the strongest known in biology. Biotinylated probes can be made easily by including a suitable biotinylated nucleotide in the labeling reaction (see Figure 5.7).

• Digoxigenin is a plant steroid (obtained from Digi-talis plants) to which a specific antibody has been raised. The digoxigenin-specific antibody permits detection of nucleic acid molecules which have incorporated nucleotides containing the digoxigenin reporter group (see Figure 5.7).

A variety of different marker groups or molecules can be conjugated to affinity molecules such as streptavidin or the digoxigenin-specific antibody. They include various fluorophores (see Box 5.2), or enzymes such as alkaline phosphatase and peroxidase which can permit detection via colorimetric assays or chemical luminescence assays, etc.

Figure 5.1. Origin and characteristics of nucleic acid hybridization probes.

Figure 5.2. DNA labeling by in vitro DNA strand synthesis. (A) Nick translation. Pancreatic DNase I introduces single-stranded nicks by cleaving internal phosphodiester bonds (p), generating a 5 phosphate group and a 3 hydroxyl terminus. Addition of the multisubunit enzyme E. coli DNA polymerase I contributes two enzyme activities: (i) a 5 3 exonuclease attacks the exposed 5 termini of a nick and sequentially removes nucleotides in the 5 3 direction; (ii) a DNA polymerase adds new nucleotides to the exposed 3 hydroxyl group, continuing in the 5 3 direction, thereby replacing nucleotides removed by the exonuclease and causing lateral displacement (translation) of the nick. (B) Random primed labeling. The Klenow subunit of E. coli DNA polymerase I can synthesize new radiolabeled DNA strands using as a template separated strands of DNA, and random hexanucleotide primers.

Figure 5.3. End-labeling of DNA. (A) Kinase end-labeling of oligonucleotides. The 5-terminal phosphate of the oligonucleotide is replaced in an exchange reaction by the ³²P-labeled g-phosphate of [g-³²P]ATP. The same procedure can be used to label the two 5 termini of double-stranded DNA. (B) Fill-in end-labeling by Klenow. The DNA of interest is cleaved with a suitable restriction nuclease to generate 5 overhangs. The overhangs act as a primer for Klenow DNA polymerase to incorporate labeled nucleotides complementary to the overhang. Fragments labeled at one end only can be generated by internal cleavage with a suitable restriction site to generate two differently sized fragments which can easily be size-fractionated.

Figure 5.7. Structure of digoxigenin- and biotin-modified nucleotides. Note that the digoxigenin and biotin groups in these examples are linked to the 5 carbon atom of the uridine of dUTP by spacer groups consisting respectively of a total of 11 carbon atoms (digoxigenin-11-UTP) or 16 carbon atoms (biotin-16-dUTP). The digoxigenin and biotin groups are reporter groups: after incorporation into a nucleic acid they are bound by specific ligands containing an attached marker such as a fluorophore.