G. Terrance Walker

Strand Displacement Amplification (SDA) is an isothermal, in vitro nucleic acid amplification technique based upon the ability of HincII to nick the unmodified strand of a hemiphosphorothioate form of its recognition site, and the ability of exonuclease deficient klenow (exo- klenow) to extend the 3'-end at the nick and displace the downstream DNA strand. Exponential amplification results from coupling sense and antisense reactions in which strands displaced from a sense reaction serve as target for an antisense reaction and vice versa. In the original design (G. T. Walker, M. C. Little, J. G. Nadeau and D. D. Shank (1992) Proc. Natl. Acad. Sci 89, 392-396), the target DNA sample is first cleaved with a restriction enzyme(s) in order to generate a double-stranded target fragment with defined 5'- and 3'-ends that can then undergo SDA. Although effective, target generation by restriction enzyme cleavage presents a number of practical limitations. We report a new target generation scheme that eliminates the requirement for restriction enzyme cleavage of the target sample prior to amplification. The method exploits the strand displacement activity of exo- klenow to generate target DNA copies with defined 5'- and 3'-ends. The new target generation process occurs at a single temperature (after initial heat denaturation of the double-stranded DNA). The target copies generated by this process are then amplified directly by SDA. The new protocol improves overall amplification efficiency. Amplification efficiency is also enhanced by improved reaction conditions that reduce nonspecific binding of SDA primers. Greater than 10(7)-fold amplification of a genomic sequence from Mycobacterium tuberculosis is achieved in 2 hours at 37 degrees C even in the presence of as much as 10 micrograms of human DNA per 50 microL reaction. The new target generation scheme can also be applied to techniques separate from SDA as a means of conveniently producing double-stranded fragments with 5'- and 3'-sequences modified as desired.

An isothermal in vitro DNA amplification method was developed based upon the following sequence of reaction events. Restriction enzyme cleavage and subsequent heat denaturation of a DNA sample generates two single-stranded target DNA fragments (T1 and T2). Present in excess are two DNA amplification primers (P1 and P2). The 3' end of P1 binds to the 3' end of T1, forming a duplex with 5' overhangs. Likewise, P2 binds to T2. The 5' overhangs of P1 and P2 contain a recognition sequence (5'-GTTGAC-3') for the restriction enzyme HincII. An exonuclease-deficient form of the large fragment of Escherichia coli DNA polymerase I (exo- Klenow polymerase) [Derbyshire, V., Freemont, P. S., Sanderson, M. R., Beese, L., Friedman, J. M., Joyce, C. M. & Steitz, T. A. (1988) Science 240, 199-201] extends the 3' ends of the duplexes using dGTP, dCTP, TTP, and deoxyadenosine 5'-[alpha-thio]triphosphate, which produces hemiphosphorothioate recognition sites on P1.T1 and P2.T2. HincII nicks the unprotected primer strands of the hemiphosphorothioate recognition sites, leaving intact the modified complementary strands. The exo- Klenow polymerase extends the 3' end at the nick on P1.T1 and displaces the downstream strand that is functionally equivalent to T2. Likewise, extension at the nick on P2.T2 results in displacement of a downstream strand functionally equivalent to T1. Nicking and polymerization/displacement steps cycle continuously on P1.T1 and P2.T2 because extension at a nick regenerates a nickable HincII recognition site. Target amplification is exponential because strands displaced from P1.T1 serve as targets for P2 and strands displaced from P2.T2 serve as targets for P1. A 10(6)-fold amplification of a genomic sequence from Mycobacterium tuberculosis or Mycobacterium bovis was achieved in 4 h at 37 degrees C.

The equilibrium binding of ethidium to the right-handed (B) and left-handed (Z) forms of poly(dG-dC).poly(dG-dC) and poly(dG-m5dC).poly(dG-m5dC) was investigated by optical and phase partition techniques. Ethidium binds to the polynucleotides in a noncooperative manner under B-form conditions, in sharp contrast to highly cooperative binding under Z-form conditions. Correlation of binding isotherms with circular dichroism (CD) data indicates that the cooperative binding of ethidium under Z-form conditions is associated with a sequential conversion of the polymer from a left-handed to a right-handed conformation. Determination of bound drug concentrations by various titration techniques and the measurement of circular dichroism spectra have enabled us to calculate the number of base pairs of left-handed DNA that adopt a right-handed conformation for each bound drug; 3-4 base pairs of left-handed poly(dG-dC).poly(dG-dC) in 4.4 M NaCl switch to the right-handed form for each bound ethidium, while approximately 25 and 7 base pairs switch conformations for each bound ethidium in complexes with poly(dG-dC).poly(dG-dC) in 40 microM [Co(NH3)6]Cl3 and poly(dG-m5dC).poly(dG-m5dC) in 2 mM MgCl2, respectively. The induced ellipticity at 320 nm for the ethidium-poly(dG-dC).poly(dG-dC) complex in 4.4 M NaCl indicates that the right-handed regions are nearly saturated with ethidium even though the overall level of saturation is very low. The circular dichroism data indicate that ethidium intercalates to form a right-handed-bound drug region, even at low r values where the CD spectra show that the majority of the polymer is in a left-handed conformation.

Strand Displacement Amplification (SDA) is an isothermal, in vitro method of amplifying a DNA target sequence prior to detection [Walker et al (1992) Nucleic Acids Res., 20, 1691-1693]. Here we describe a multiplex form of SDA that allows two target sequences and an internal amplification control to be co-amplified by a single pair of primers after common priming sequences are spontaneously appended to the ends of target fragments. Multiplex SDA operates at a single temperature, under the same simple protocol previously developed for single-target SDA. We applied multiplex SDA to co-amplification of a target sequence (IS6110) that is specific to members of the Mycobacterium tuberculosis-complex and a target (16S ribosomal gene) that is common to most clinically relevant species of mycobacteria. Both targets are amplified 10(8)-fold during a 2 hour, single temperature incubation. The relative sensitivity of the system was evaluated across a number of clinically relevant mycobacteria and checked for crossreactivity against organisms that are closely related to mycobacteria.

The most attractive feature of SDA is its operation at a single temperature, which removes the need for instrumented temperature cycling as with PCR and the ligase chain reaction. Highly reproducible temperature profiles, over a large array of samples, can burden the accuracy and expense of an amplification technique. However, the expense of a temperature cycler is offset somewhat by the cost of additional enzymes used in isothermal techniques. In comparisons with isothermal, transcription-based techniques, SDA requires fewer enzymes and has a simpler mechanism. SDA may also be more robust than transcription-based processes because it is not susceptible to contaminating ribonuclease activity. This is generally more of a concern when using clinical samples. The most significant disadvantage of SDA is its inability to efficiently amplify long target sequences. Until this short-coming is eliminated, SDA will be assigned to the diagnostic laboratory along with the ligase chain reaction. Currently, SDA cannot compete with PCR in research applications such as the isolation of gene sequences. The second disadvantage of SDA is that it operates at relatively low (nonstringent) temperatures, which produces considerable background reactions. Consequently, SDA reaction products cannot be analyzed routinely by ethidium-stained gel electrophoresis, as is used commonly with PCR, unless the target sample contains a large number of initial targets.