The specific activity of nucleic acid probes is an important parameter to control, since it determines the sensitivity of nucleic acid detection. Probe specific activity is not only dependent on the specific activity and amount of radiolabeled nucleotide incorporated into the probe, but also on the amount of probe available for hybridization. Therefore, when choosing a method for synthesis of high specific activity probes, one should take into account the ability of the enzymatic reaction to incorporate low concentrations of high specific activity radiolabeled nucleotides (e.g. 800 Ci/mmol, 10 mCi/ml vs. 6000 Ci/mmol, 10 mCi/ml) and what amount of radiolabeled nucleotide can be economically afforded per reaction. These factors should be balanced with the ability to degrade or separate the template used from the probe synthesized so that the template will not decrease the effective amount of probe available for hybridization. Below, four methods for generating labeled nucleic acids are evaluated for their ability to produce probes of high specific activity, taking into account these criteria.

In Vitro Transcription of RNA Probes

In vitro transcription reactions use phage RNA polymerases to synthesize single-stranded, strand-specific RNA probes of a discrete length from a DNA template. There is a trade-off between synthesis of high specific activity probe and the synthesis of full-length probe in such reactions. The concentration of the limiting nucleotide should be >3 µM for most transcripts. The greater the concentration of unlabeled limiting nucleotide added to supplement the radiolabeled nucleotide, the lower the specific activity of the probe transcript. To make very high specific activity probes, the limiting nucleotide should be comprised completely of radioactively labeled nucleotide, omitting any unlabeled form of this nucleotide. The reaction volume should be kept low so that the concentration of the labeled nucleotide can be maximized without becoming prohibitively expensive. Generally, [alpha- 32P] CTP or UTP at 800 Ci/mmol and 10 mCi/ml is used for the synthesis of radioactive RNA probes. Note that using 50 µCi of [ 32P]NTP at 800 Ci/mmol and 10 mCi/ml in a 20 µl reaction results in a 3.125 µM concentration of the [ 32P]NTP and generates RNA probes with specific activities in the range of 4 x 10 8 cpm/µg.

Higher specific activity labeled nucleotides (e.g. 3000 Ci/mmol) may also be used. However, since they are usually also sold at 10 mCi/ml, the same 50 µCi in a 20 µl reaction volume would result in a lower concentration of limiting nucleotide such that the synthesis of full-length transcripts would be greatly reduced. This problem can be overcome using Ambion's CU Minus™ promoter technology. Transcription reactions containing low concentrations of a nucleotide (e.g. 32P-UTP or 32P-CTP) encoded in the first 12 bases of a transcript experience high levels of abortive transcription. Eliminating C and U nucleotides from the first 12 bases that will be incorporated after the promoter sequence greatly reduces abortive transcription. CU Minus vectors effectively incorporate radiolabeled nucleotides at specific activities of 3000 or 6000 Ci/mmol (without addition of cold nucleotides), which translates into higher specific activity probes (up to 7.5 times higher) and thus stronger signals in nuclease protection assays and blot hybridizations. In addition to CU Minus vectors, Ambion also supplies primers to convert existing T7, T3 and SP6 promoters in any vector to CU Minus promoters.

RNA probes produced can be readily separated from the DNA template by DNase treatment of the terminated transcription reaction and/or gel purification. In addition, probe is generated from only one of the template strands. Therefore, there is no template or second probe strand to effectively lower the probe's specific activity by competing with target for hybridization to it.

Random Priming

Random priming reactions use Klenow enzyme to extend random oligomers hybridized to denatured DNA in the presence of deoxynucleotides. The low K m of Klenow enzyme allows efficient incorporation of low molar concentrations of high specific activity radiolabeled nucleotides (e.g. [alpha- 32P]dATP or dCTP at 3000-6000 Ci/mmol, at a final concentration of 0.67 µM). The probes generated by this reaction have specific activities in the range of greater than or equal to 1-3 x 10 9 cpm/µg. Probe yield is dependent on the amount of starting template used in the reaction. The larger the amount of DNA template used in the reaction, the greater the yield of probe. However, since there is no way to separate the template from the probe produced, there is a trade off between probe yield and specific activity when using the random priming method. Large amounts of template result in lower specific activity since the template competes with the labeled probe for hybridization to the target. While the presence of template is the major factor reducing specific activity of such probes, it should also be realized that both template strands are used to generate probe so that there is additional competition by the other labeled strand generated.

5' End-labeling with [gamma-32P] ATP and Polynucleotide Kinase

5' end-labeling incorporates a single [ 32P]phosphate per molecule independent of sequence length. Therefore, a comparison of specific activity expressed in cpm/µg is only meaningful when comparing two probes of roughly equal size, and it is perhaps more realistic to talk about specific activity in terms of cpm/pmol than cpm/µg. Using Ambion's KinaseMaxí 5' End-Labeling Kit, incorporation levels of 95-100% of [gamma- 32P]ATP for the forward reaction and 20% for the exchange reaction are routinely obtained. When ATP with a specific activity of 7000 Ci/mmol is used on the reference day stated on the stock vial, you should expect to get 1.32 x 10 7 cpm/pmol for 100% labeling of the 5' ends. This converts to 2 x 10 9 cpm/µg for a 20-mer oligonucleotide or 8 x 10 7 cpm/µg for a 500 nt single-stranded DNA.