© 1992 Oxford University Press
`
`Nucleic Acids Research, Vol. 20, No. 7
`
`1691-1696
`
`Strand displacement amplification—an isothermal, in vitro
`DNA amplification technique
`
`G.Terrance Walker*, Melinda S.Fraiser, James L.Schram, Michael C.Little, James G.Nadeau
`and Douglas P.Malinowski
`Department of Molecular Biology, Becton Dickinson Research Center, PO Box 12016,
`Research Triangle Park, NC 27709, USA
`
`Received December 4, 1991; Revised and Accepted March 10, 1992
`
`Downloaded from
`
`http://nar.oxfordjournals.org/
`
` by Carla Askey on March 31, 2013
`
`INTRODUCTION
`
`Strand Displacement Amplification (SDA) is an isothermal, in
`vitro method of amplifying DNA. The original SDA design
`provides powerful amplification compatible with diagnostic
`nucleic acid assays (1). However, it requires restriction enzyme
`cleavage of the DNA sample prior to amplification in order to
`generate an amplifiable target fragment with defined 5'- and
`3'-ends. The target restriction step not only complicates the
`experimental protocol, but it also limits the choice of target DNA
`sequences because the sequence must be flanked by convenient
`enzyme restriction sites. In addition, the target DNA typically
`must be double-stranded for restriction enzyme cleavage. We
`have simplified the SDA protocol by replacing the target
`restriction step with a novel method of generating amplifiable
`fragments. SDA reactions can now be performed in three simple
`steps: 1) the target DNA sample is heat denatured in the presence
`of primers and other reagents, 2) Hindi and exo~ klenow are
`added, and 3) the sample is incubated at 37°C.
`As in the original report on SDA (1), we applied this improved
`version
`to genomic DNA samples from Mycobacterium
`tuberculosis (M. tb), the causative agent of tuberculosis. Enhanced
`amplification was achieved by virtue of the new target generation
`scheme and
`improved
`reaction conditions which
`lower
`background reactions that attentuate target amplification.
`The new target generation scheme is a novel means of
`producing a double-stranded fragment with 5'- and 3'-ends
`modified as desired. We propose a few general applications for
`the method.
`
`MATERIALS AND METHODS
`Materials
`The large fragment of E. coli DNA polymerase I (klenow) was
`purchased from Boehringer Mannheim (sequencing grade).
`Exonuclease deficient large fragment of E. coli DNA polymerase
`I (exo~ klenow) (2) was purchased from United States
`Biochemical. HincII was purchased from New England Biolabs
`at a concentration of 75 units//xL. Ultra pure human placental
`DNA (Sigma) was phenol/chloroform extracted and ethanol
`precipitated
`to
`remove Na2EDTA. Deoxynucleoside
`
`ABSTRACT
`Is an
`(SDA)
`Strand Displacement Amplification
`isothermal, in vitro nucleic acid amplification technique
`based upon the ability of Hindi 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. Scl 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
`107-fold amplification of a genomic sequence from
`Mycobacterium tuberculosis is achieved in 2 hours at
`37° C even in the presence of as much as 10 ng of
`human DNA per 50 yL 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.
`
`* To whom correspondence should be addressed
`
`Ariosa Exhibit 1024, pg. 1
`IPR2013-00276
`
`

`

`Downloaded from
`
`http://nar.oxfordjournals.org/
`
` by Carla Askey on March 31, 2013
`
`denature tirget
`bind prlmeri
`
`primer extension and
`displacement by exo"
`klenow using dGTP,
`dCTP, TTP and dATPaS
`
`**•*»
`
`N;
`
`S^eit
`
`bind opposite primers
`to Sj-ext and S2-ext
`
`displacement by exo"
`klenow
`binding of opposite
`primers
`exteniion by exo"
`klenow
`
`Figure 1. Target generation scheme for SDA. This figure depicts the initial steps
`in an SDA reaction which transform the original target sequence into the
`amplification cycle depicted in Figure 2. A target DNA sample is heat denatured.
`Four primers (B,, B2, S, and S2), present in excess, bind the target strands at
`positions flanking the sequence to be amplified. Primers Si and Sj have HincII
`recognition sequences ( GTTGAC) located 5' to the target complementary
`sequences. The four primers are simultaneously extended by exo" klenow using
`dGTP, dCTP, TTP and dATPS. Extension of B, displaces the S, primer
`extension product, S|-ext. Likewise, extension of B; displaces S^-ext. B2 and
`S2 bind to displaced S,-ext. B, and S, bind to displaced S^-ext. Extension and
`displacement reactions on templates S,-ext and S^-ext produce two fragments
`with a hemiphosphorothioate HincII at each end and two longer fragments with
`a hemiphosphorothioate HincII site at just one end. HincII nicking and exo"
`klenow extension/displacement reactions initiate at these four fragments,
`automatically entering the SDA reaction cycle depicted in Figure 2. Sense and
`antisense DNA strands are differentiated by thin and thick lines. HincII recognition
`sequences are depicted by ( —HH— ).
`
`at 37°C for 10 minutes. For Figure 4, 5'-32P-labelled SDA
`primers were used directly in amplification reactions. Samples
`were mixed with an equal volume of 50% (w/v) urea, 20 mM
`Na2EDTA, 0.5xTBE (3), 0.05% bromophenol blue/xylene
`cylanol and analyzed by denaturing gel electrophoresis (10%
`acrylamide:bis (18:1), 50% (w/v) urea, 0.5xTBE). X-ray film
`(Kodak X-OMAT AR) was exposed for 30 minutes at room
`
`1692 Nucleic Acids Research, Vol. 20, No. 7
`
`Table I. Amplification of M. tb genomic DNA
`
`Initial
`M. tb
`genome
`copies
`(molecules)
`
`500000
`50000
`5000
`500
`50
`zero+
`100
`100
`100
`
`Amplified
`Product
`(molecules)
`
`> l x l O12
`> l x l 012
`7 x 1 0"
`2 x 1 0"
`2 x l 010
`4x10"
`4 x 1 0"
`7xlO10
`2 x l 010
`
`Amplification
`Factor*
`
`Human
`DNA Oig)
`
`> 2 x l O5
`>2xlu*
`lxlO7
`4xlO7
`4X107
`
`4X108
`7 x l 07
`2 x l 07
`
`0.5
`0.5
`0.5
`0.5
`0.5
`0.5
`0.1
`1
`10
`
`•Amplification factors were calculated considering there are — 10 copies of the
`target IS6110 sequence per M. tb genome (1).
`+Signal due to accidental contamination with the equivalent of ~\ M. tb genome
`copy.
`
`5'-triphosphates (dGTP, dCTP, TTP) and 2'-deoxyadenosine
`5'-O-(l-thiotriphosphate)
`(dATPS) were purchased
`from
`Pharmacia. Oligodeoxynucleotides were synthesized on an
`Applied Biosystems, Inc. 380B instrument and purified by
`denaturing polyacrylamide gel electrophoresis. M. tb cultures
`were generously provided by Salmon Siddiqi, Becton Dickinson
`Diagnostic Instrument Systems, Sparks MD. Daryl Shank (Becton
`Dickinson Research Center) kindly prepared genomic DNA from
`M. tb as previously described (1). Aerosol Resistant Tips (ART)
`(Continental Laboratory Products) were used
`to reduce
`contamination of SDA reactions with previously amplified
`products.
`
`SDA Reactions
`Genomic M. tb DNA was serially diluted in 50 mM KjPO.,, pH
`7.4 containing 0.01 or 0.1 ng/fiL human placental DNA. SDA
`was performed in 50 /iL reactions containing varying amounts
`of M. tb DNA, 0.1-10 fig human placental DNA as indicated,
`150 units HincII, 5 units exo~ klenow, 1 mM each dGTP, dC-
`TP, TTP and dATPS, 50 mM KiPO4, pH 7.4, 6 mM
`MgAcetate2, 3% (v/v) 1-methyl 2-pyrrolidinone (NMP)
`(Sigma), 3% glycerol (from stock enzyme solutions), 500 nM
`SDA primers S, and S2 and 50 nM primers B, and B2. Prior
`to addition of Hindi and exo" klenow, reaction samples were
`heated 4 minutes at 95 °C to denature target DNA followed by
`1 minute at 37°C to anneal primers. Upon addition of HincII
`and exo" klenow, amplification proceeded 2 hours at 37°C and
`was then terminated by heating 2 minutes at 95 °C.
`Amplified products were detected using a oligodeoxynucleotide
`detector probe (Figure 3 and Table I) or SDA primers (Figure 4)
`which were 5'-32P-labelled using T4 polynucleotide kinase in a
`50 nL reaction containing 6.9 mM Tricine (pH 7.6), 50 mM
`TRIS-HC1 (pH 8), 10 mM MgCl2 5 mM DTT, 2.3 /iM 7-32P-
`ATP (3000 Ci/mmol, 10 mCi/mL), 50 units kinase (New England
`Biolabs) and either 1 /iM oligodeoxynucleotide detector probe
`or 10 /iM SDA primer. 32P-labelling was carried out for 1 hour
`at 37 °C and terminated by heating 2 minutes at 95 °C. For
`Figure 3, aliquots (2 /tL) of the 32P-detector probe kinase mix
`were added to 10 /iL aliquots from SDA reactions, and the
`mixture was heated 2 minutes at 95 °C and then 2 minutes at
`37°C. The 5'-32P-detector probe was extended to a diagnostic
`length upon addition of 2 /xL of 1 unit//*L klenow and incubation
`
`Ariosa Exhibit 1024, pg. 2
`IPR2013-00276
`
`

`

`Downloaded from
`
`http://nar.oxfordjournals.org/
`
` by Carla Askey on March 31, 2013
`
`Nucleic Acids Research, Vol. 20, No. 7 1693
`
`the sequence to be amplified. B, and B2 bind upstream of S| and
`S2, respectively. Exo~ klenow (present in excess over the
`number of target sequences) simultaneously extends all 4 primers
`using dGTP, dCTP, TTP and dATPaS. Extension of S, and S2
`forms two products, S]-ext and S2-ext. Extension of B| and B2
`results in displacement of S|-ext and S2-ext from the original
`target templates. Displaced and single-stranded S|-ext then
`serves as target for primers S2 and B2. Likewise, displaced
`S2-ext is target for S| and B,. All four primers are extended on
`templates Srext and S2-ext. Combined extension and
`displacement steps produce 2 double-stranded fragments with
`hemiphosphorothioate HincII recognition sites at each end and
`two
`longer
`double-stranded
`fragments with
`a
`hemiphosophorothioate HincII site at only one end. In the
`presence of HincII (which can be added at the top of Figure 1
`along with exo~ klenow), the four target fragments at the
`bottom of Figure 1 automatically enter the strand displacement
`amplification cycle (Figure 2), where HincII nicks
`the
`hemiphosphorothioate recognition site, and exo~ klenow
`extends the 3'-end at the nick while displacing the downstream
`strand that in turn serves as target for the opposite SDA primer.
`These steps, repeated continuously over the course of the reaction,
`produce exponential growth in the number of target sequences.
`107-Fold amplfication (Table I) theoretically derives from ~23
`repetitions or cycles of the steps in Figure 2 (223= 107). Each
`displaced strand in Figure 2 contains the sequence 5'GAC at the
`5'-end and the complement 5GTC at the 3'-end. 5GAC is the
`portion of the HincII recognition sequence located 3' to the nick
`site.
`The transition between Figures 1 and 2 may become clearer
`with a detailed account for the last double-stranded fragment at
`the bottom of Figure 1. If HincII nicks the hemiphosphorothioate
`site on the left end of the fragment containing the S) primer
`sequence, exo~ klenow initiates replication at the nick and
`displaces the downstream strand containing 5GAC at the 5'-end
`and the full complement of the primer S2 sequence at the 3'-end.
`S2 binds to this displaced strand, and exo~ klenow extends S2
`forming the double-stranded fragment depicted as the product
`of the step labelled 'polymerize using dGTP, dCTP, TTP and
`dATP(aS)' on the right side of Figure 2. Similar reactions occur
`with the entire set of fragments depicted at the bottom of Figure 1.
`Duplexes containing two HincII sites in Figure 1 may also
`undergo more complex reactions in which the two restriction sites
`are nicked simultaneously. Following dissociation of HincII from
`the nicked sites, strand extension/displacement may proceed from
`one or both sites. Although we have not yet determined the
`relative importance of these different pathways, all such processes
`ultimately lead to amplifiable species found in Figure 2.
`
`Amplification of M. tb DNA
`A set of samples containing varying amounts of target genomic
`M. tb DNA and the indicated amounts of human DNA was
`subjected to SDA. The two SDA primers (S| and S2) contain
`target binding regions at their 3'-ends that are complementary
`to opposite strands of the IS6110 sequence of M. tb (4) at
`nucleotide positions 972-984 and 1011-1023 (S, = 5'dTTG-
`AATAGTCGGTTACTTG7TG/iajGCGTACTCGACC, S? =
`5'dTTGAAGTAACCGACTATTG77U/iCACTGAGATCCCC-
`T). (HincII recognition sites are italicized). The two outside
`primers (B| and B2) bind to opposite strands at nucleotide
`positions 954-966 and 1032-1044 [B, = 5'dTGGACCCGC-
`CAAC.-B2 = 5'dCGCTGAACCGGAT|. After
`the SDA
`
`5' s,
`
`. S2 5'
`.
`,
`5'
`. polymerize using
`I dGTP, dCTP. TTP
`and d T f t S)
`
`hybridize SDA primeri to displaced strands
`
`Figure 2. The SDA reaction cycle. These reaction steps continuously cycle during
`the course of amplification. Present in excess are two SDA primers (S, and S2).
`The 3'-end of S, binds to the 3'-end of the displaced target strand T,, forming
`a duplex with 5'-overhangs. Likewise, S2 binds T2. The 5'-overhangs of S, and
`S^ contain the Hinc II recognition sequence (5 GTTGAC). Exo~ klenow extends
`the 3'-ends of the duplexes using dGTP, dCTP, TTP and dATPS, which produces
`hemiphosphorothioate recognition sites on S, -T, and S2T2. HincII nicks the
`unmodified primer strands of the hemiphosphorothioate recognition sites, leaving
`intact the modified complementary strands. Exo~ klenow extends the 3'-end at
`the nick on S| T, and displaces the downstream strand that is equivalent to T2.
`Likewise, extension at the nick on S2T2 results in displacement of T,. Nicking
`and polymerization/displacement steps cycle continuously on S| -T, and S2T2
`because extension at a nick regenerates a nickable HincII recognition site. Target
`amplification is exponential because strands displaced from S( -T, serve as target
`for S2 while strands displaced from SyT2 serve as target for S|. Sense and
`antisense DNA strands are differentiated by thin and thick lines. Intact and rucked
`HincII recognition sequences are depicted by _ ^ H_
`and
`-•
`•-
`,
`respectively. The partial HincII recognition sequence 3 GAC and its complement
`3 GTC are present at the 5'- and 3'-ends of displaced strands as represented by
`•_ and —•.
`
`temperature using intensifying screens. 32P-labelled gel
`electrophoresis bands were excised and quantified by liquid
`scintillation counting using appropriate gel electrophoresis
`background controls. The number of amplified product molecules
`(Table I) was calculated based upon the specific activity of the
`5'-32P-detector probe which was measured from aliquots
`purified by gel electrophoresis. Amplification factors were
`calculated considering there are ~ 10 copies of the IS6110 target
`sequence per M. tb genome (1).
`
`RESULTS
`
`A limitation of the original SDA design is the requirement for
`restriction enzyme cleavage of the target DNA sample prior to
`amplification (1). The restriction step is necessary to provide
`target fragments with defined 5'- and 3'-ends. Subsequent binding
`of the SDA primers to the 3'-ends of these target fragments and
`extension by exo~ klenow create hemiphosphorothioate HincII
`recognition sites from which amplification initiates.
`We have developed a new method of target generation which
`eliminates the need for restriction enzyme cleavage of the target
`prior to SDA (Figure 1). The target DNA sample is heat
`denatured in the presence of an excess of 4 primers. Two primers
`(S| and SJ are typical SDA primers, with target binding regions
`at their 3'-ends and HincII recognition sequences (5GTTGAC)
`located 5' to the target complementary sequences. The other two
`primers (B, and B2) consist simply of target binding regions. S,
`and S2 bind to opposite strands of the target at positions flanking
`
`Ariosa Exhibit 1024, pg. 3
`IPR2013-00276
`
`

`

`io3n
`
`0.1 10
`
`zeron
`
`0.1 10
`
`M.W.
`ladder
`1
`
`Downloaded from
`
`—- 40-mer
`
`-
`
`33-mer
`
`—
`
`20-mer
`
`c
`
`io'n
`
`01 10
`
`tb
`M
`genome
`copies
`human
`DNA
`
`79-mers
`
`32P-SDA
`primers
`(37-mers)
`
`21-mers
`
`Figure 4. Denaturing gel electrophoresis analysis of SDA reactions using 5'-32P-
`labelled SDA primers. SDA reactions were performed using 5'-32P-labelled SDA
`primers (S, and S2) in the presence of 0.1 or 10 ^g of human DNA. Aliquots
`(10 nL) from each 50 #iL SDA reaction were analyzed. The number of M. tb
`genome copies present in each 50 yL reaction is shown above each lane. Unreacted
`5'-32P-labelled SDA primers are 37-mers. Target specific S'-^P-labelled
`amplification products are a set of 21- and 79-mers. Background reaction products
`appear as a high molecular weight smear and more distinct bands over 55- to
`110-mers. The 21-mer bands also derive in part from background reactions (see
`text). Reactions were not allowed to proceed below 37°C in order to minimize
`background reactions.
`
`http://nar.oxfordjournals.org/
`
` by Carla Askey on March 31, 2013
`
`estimate each sample is contaminated with the equivalent of — 1
`genome copy of M. tb. Although this contamination problem
`hinders accurate determination of sensitivity, SDA enables 32P-
`detection of as few as 10—50 initial target DNA molecules (1 —5
`genome copies containing - 10 copies of the target IS6110
`sequence) (Table I and unpublished observations).
`
`Background Amplification
`As previously discussed (1), background amplification reactions
`compete with target specific amplification and are more prevalent
`at higher non-target (human) DNA concentrations. The sensitivity
`of SDA therefore depends upon the amount of non-target DNA
`present (Table I). Background reactions presumably begin with
`extension of primers spuriously bound to non-target sequences.
`In processes similar to those depicted in Figure 1, these
`background extension products are
`then displaced and
`subsequently hybridized to a second SDA primer, which is in
`turn extended and displaced. In background reactions, these initial
`priming events will generally
`involve imperfectly paired
`sequences and therefore should occur far less efficiently than the
`carefully designed priming/extension reactions leading to bona-
`fide target generation. However, once a non-target sequence is
`attached to two primer-derived nickable HincII sites (see bottom
`of Figure 1), it will undergo amplification as effectively as the
`genuine target sequence.
`
`1694 Nucleic Acids Research, Vol. 20, No. 7
`
`35-mer
`
`P-detector
`probe
`(15-mer)
`
`Figure 3. Denaturing gel electrophoresis analysis of SDA reactions using a
`5'- P-labelled detector probe. SDA products were detected by extension of a
`5'-32P-labelled probe (15-mer) to a length of 35 or 56 nucleotides. Aliquots (10
`fiL) from each 50 nL SDA reaction were analyzed. The number of M. tb genome
`copies present in each 50 pL reaction is shown above each lane. Each sample
`contained 0.5 /ig of human DNA. Longer autoradiography exposures reveal target
`specific bands in the zero target lane. Band intensities are nearly constant for
`samples containing >5000 intial targets due to the limited amount of 32P-labelled
`detector probe. Unextended ^P-labelled detector probe appears as multiple bands
`due to degradation by the 3'-5' cxonuclease activity of klenow used in the detection
`step. DNA polymerase extension reactions of the 5'-32P-probe were not allowed
`to proceed below 37"C in order to minimize background detection reactions.
`
`reaction is terminated, all species shown in Figure 2 are present
`and detectable. In the present study, amplified target fragments
`were detected by DNA polymerase extension of a 5'-32P-
`labelled probe (5'-32P-dCGTTATCCACCATAC)
`(IS6110
`nucleotide positions 992 — 1006) that is complementary to an
`internal segment of the target sequence. The 5'-32P-labelled
`detector probe is extended to a length of either 35 or 56
`nucleotides when hybridized to one or the other of the following
`two strands produced during SDA: 5'dG/lCACTGAGATCCC-
`CTATCCGTATGGTGGATAACGTCTTTCAGGTCGAGT-
`ACGCCGrC and 5'dTTGAATAGTCGGTTACTTG77G/lCA-
`CTGAGATCCCCTATCCGTATGGTGGATAACGTCTTTC-
`AGGTCGAGTACGCCG7TC. (Partial and complete Hindi
`recognition sequences are italicized). The first strand is one of
`the strands displaced during SDA and corresponds to IS6110
`nucleotide positions 972-1023 with 5GAC and 5GTC
`appended to the 5'- and 3'-ends. The second longer strand is the
`same strand before nicking by HincII and as such has a 5'-end
`identical to primer S|.
`Results from amplification of genomic M. tb DNA are shown
`in Figure 3 and Table I. The amount of amplified product is
`dependent upon the number of initial targets. (However, levels
`of amplified product are nearly constant for > 5000 initial M.
`tb genome copies due to the limited amount of the 32P-detector
`probe used.) Longer autoradiography exposures reveal target
`specific 32P-bands in the zero target lane of Figure 3; we
`
`Ariosa Exhibit 1024, pg. 4
`IPR2013-00276
`
`

`

`Downloaded from
`
`http://nar.oxfordjournals.org/
`
` by Carla Askey on March 31, 2013
`
`Nucleic Acids Research, Vol. 20, No. 7 1695
`
`Hindi sites is generated from each original double-stranded
`target. The previous method of target generation produces two
`double-stranded fragments each with a hemiphosphorothioate
`Hindi site at one end (1), whereas the current method produces
`four double-stranded
`fragments with a
`total of six
`hemiphosphorothioate Hindi sites (Figure 1). In addition, target
`sequences can now be chosen without regard for convenient
`flanking restriction sites. Furthermore, SDA can now be applied
`to either double- or single-stranded target DNA target samples.
`Previously, targets generally had to be double-stranded for target
`generation by restriction enzyme cleavage. SDA is now
`compatible with sample preparation methods which render target
`genomic DNA single-stranded prior to amplification.
`Primers B, and B2 in Figure 1 can also be typical SDA
`primers containing target binding regions at their 3'-ends and
`Hindi recognition sites located toward their 5'-ends. The result
`is an enhancement of amplification from additional nicking and
`extension/displacement
`reactions
`at
`these
`outside
`hemiphosphorothioate Hindi sites (data not shown). Displaced
`strands from these outside sites serve as target for both opposite
`primers.
`Performance of SDA with just one SDA primer (S| or S2)
`generates single-stranded displaced strands
`in a
`linear
`amplification mode. In a related sense, SDA could be performed
`with an excess of one SDA primer over the other (e.g.,
`[S|]> >[S2]) which should produce exponential amplification
`of a target with a predominance of one amplified strand over
`the other in a manner analogous to 'asymmetric' PCR. In either
`case, the resultant single-stranded displaced strands could serve
`as either detector probes (e.g., Southern blots) or templates for
`DNA sequence analysis.
`The ability of SDA to amplify low target numbers at 37°C
`is quite remarkable when compared with PCR, which achieves
`exquisite sensitivity only when non-specific priming reactions are
`diminished by stringent (high temperature) reaction conditions
`(5). Although the absence of temperature cycling in SDA reduces
`the number of recurrent opportunities for errant hybridization
`and primer extension, considerable non-specific priming surely
`occurs at the 37°C reaction temperature. However, initiation of
`SDA for any sequence (target or background) must take place
`via multiple steps similar to those shown in Figure 1. The
`efficiency of this concerted series of events will be greatly
`diminished if the primer extension and strand displacement steps
`must wait for the adventitious, nonspecific hybridization of
`primers, as will generally be the case for background reactions.
`Once non-target amplicons are generated through this inefficient
`process, however, they will be amplified as readily as bona-fide
`target sequences. Consequently, significant accumulation of
`background product can result if non-target DNA is present in
`large excess over specific target sequences. To reduce the
`potential for these background reactions, it is important to
`maintain the highest level of stringency compatible with enzyme
`stability. Accordingly, SDA reactions are not allowed to proceed
`at temperatures less than 37 °C, and we increased stringency by
`adding organic co-solvents
`to
`the reaction. With
`these
`modifications to our original reaction conditions (1), we can now
`amplify specific target sequences > 107-fold even in the
`presence of as much as 10 /xg of human DNA per 50 /tL reaction.
`SDA and PCR are affected by common experimental
`parameters (e.g., target length, target dG-dC content, primer
`sequences, and organic solvents such as 1-methyl 2-pyrrolidinone,
`glycerol and formamide). Because it operates under conditions
`
`To examine the background reactions, SDA was performed
`using 5'-32P-labelled primers (S, and S2) at various target levels
`and in the presence of either 0.1 or 10 /tg of human DNA
`(Figure 4). Excess unreacted 5'-32P-labelled SDA primers are
`37-mers. Target specific SDA products appear as
`two
`32P-21-mers and a set of four 32P-79-mers. The 21-mers are the
`nucleotide sequences located 5' to the Hindi nicking sites on
`the SDA primers (the short strands indicated in Figure 2 above
`the step labelled 'polymerize and displace strand')- The set of
`four 79-mers are the unnicked amplification strands shown in
`Figure 2 above the step labelled 'nick with Hindi'. In Figure 4,
`background amplification products appear only in lanes for
`reactions containing high (10 ng) human DNA, where they run
`as a high molecular weight smear and more distinct bands ranging
`in size from 55-110 nucleotides. The ~43-mer bands, which
`are present in all lanes in Figure 4, appear in the absence of
`Hindi and as such are not SDA background reaction products
`(e.g., 'primer-dimers'). (The 43-mer probably represents exo~
`klenow extension of a minor hairpin conformation of primer
`Si.) The 21-mer bands in Figure 4 also arise from background
`reactions as well as from target amplification because they derive
`from the target-independent sequences at the 5'-ends of the SDA
`primers; notice that 21-mers are apparent in the zero target sample
`with 10 /xg of human DNA although no target specific 79-mers
`are discernible.
`The apparent lack of background products in the low (0.1 /ig)
`human DNA lanes (Figure 4) supports the hypothesis that
`background reactions arise from spurious amplification of human
`DNA sequences. Background amplification predominates in
`reaction mixtures containing high human DNA and low numbers
`of initial target sequences (< 104 genome copies). Under these
`conditions, target amplification proceeds efficiently for only ~2
`hours before high levels of background product accumulate and
`dramatically reduce target amplification by competing for limiting
`components like Hindi. The more target copies initially present,
`the more target-specific amplification that can occur before
`essential components are depleted. Consequently, in reactions
`containing low initial target numbers, the final level of amplified
`target depends on the initial copy number (1). At very high initial
`levels (rMO6 M.
`target
`tb genome copies), however,
`background reactions are attenuated by target amplification,
`which predominates under these conditions, and the final level
`of amplified target is independent of the number of targets initially
`present (data not shown).
`
`DISCUSSION
`
`We have streamlined the original SDA experimental protocol by
`eliminating the need to cleave a target DNA sample by a
`restriction enzyme(s) prior to amplification (1). The new target
`generation scheme uses 4 primers in a concerted series of DNA
`polymerase extension and displacement reactions. Although the
`overall reaction mechanism of SDA appears complex as depicted
`in Figures 1 and 2, individual reaction steps are invisible to the
`experimenter. In fact, the experimental protocol consists simply
`of 1) heat denaturing a target DNA sample in the presence of
`primers and other reagents 2) addition of Hindi and exo~
`klenow, and 3) incubation at 37°C.
`This new target generation scheme offers a number of
`advantages over the previous method of generating target
`fragments by restriction enzyme cleavage (1). Amplification is
`enhanced because a larger number of hemiphosphorothioate
`
`Ariosa Exhibit 1024, pg. 5
`IPR2013-00276
`
`

`

`Downloaded from
`
`http://nar.oxfordjournals.org/
`
` by Carla Askey on March 31, 2013
`
`thermophilic large fragment of DNA polymerase I from Bacillus
`stearothermophilus (Bio-Rad) strand displaces and lacks 5'-3'
`exonuclease activity.
`
`REFERENCES
`1. Walker, G. T., Little, M. C, Nadeau, J. G. and Shank, D. D. (1992) Proc.
`Nail. Acad. Sci. USA 89, 392-396.
`2. Derbyshire, V., Freemont, P. S., Sanderson, M. R., Beese, L., Friedman,
`J. M., Joyce, C. M. and Steitz, T. A. (1988) Science 240, 199-201.
`3. Maruatis, T., Fritsch, E. F. and Sambrook, J. (1982) Molecular Cloning:
`A Laboratory Manual (Cold Spring Harbor Laboratory, Cold Spring Harbor,
`NY).
`4. Thierry, D., Cave, M. D., Eisenach, K. D., Crawford, J. T., Bates, J. H.,
`Gicquel, B. and Guesdon, J. L. (1990) Nucleic Acids Res. 18, 188.
`5. Mullis, K. B. (1991) PCR 1, 1-4.
`6. Guatelli, J. C, WhitfieW, K. M., Kwoh, D. Y., Barringer, K. J., Richman,
`D. D. and Gingeras, T. R. (1990) Proc. Nail. Acad. Sci. USA 87,
`1874-1878.
`7. Kramer, F. R. and Ljzardi, P. M. (1989) Nature 339, 401-402.
`8. Miller, H. I. (1990) PCT International Patent Publication Number WO
`90/10064.
`9. Kessler, C. and Ruger, R. (1991) PCT International Patent Publication
`Number WO 91/03573.
`
`1696 Nucleic Acids Research, Vol. 20, No. 7
`
`of low stringency (37 °C) through a strand displacement
`mechanism, SDA
`is especially sensitive
`to experimental
`parameters. For example, SDA amplification factors decrease
`— 10-fold for each 50 nucleotide increase in target length, over
`a range of 50 to 200 nucleotides (unpublished observations); this
`may be related to the non-processive nature of the polymerase.
`In addition, the best choice for an organic solvent and its optimal
`concentration appear to depend on the particular sequence to be
`amplified (unpublished observations); a literature survey suggests
`this is also true for PCR. As with PCR, care must be taken to
`choose primer sequences that do not support 'primer-dimer'
`amplification reactions, which can arise in SDA reactions by the
`hybridiztion of primer sequences 3' to their Hindi recognition
`sites.
`The scheme in Figure 1 represents a general and convenient
`method of producing double-stranded fragments with defined 5'-
`and 3'-ends as illustrated by the following potential applications:
`1) The Hindi recognition sequences in Si and S2 could be
`substituted by any restriction sites in order to produce fragments
`for ligation into a cloning vector. (The scheme in Figure 1 can
`be performed with dATP instead of dATPS.) 2) The Hindi
`recognition sequences in S) and S2 could be substituted by RNA
`polymerase promoters for production of transcription templates.
`3) The technique could be applied to synthesis of double-stranded
`cDNA using a 5'-3' exonuclease deficient reverse transcriptase
`possessing strand displacement ability (does not require a reverse
`transcriptase with RNase H activity).
`Various isodiermal DNA amplification techniques require
`formation of a double-stranded DNA fragment containing
`terminal sequences specific for the given amplification technique.
`For example, the self-sustained sequence replication (3SR) system
`produces a double-stranded DNA fragment with a bacteriophage
`RNA polymerase promoter at one or both ends (6). Such a
`fragment can be generated from target DNA through 2 cycles
`of PCR using primers containing promoter sequences at their
`5'-ends. Alternatively, one can first cleave target DNA with a
`restriction enzyme(s), creating a fragment with defined 5'- and
`3'-ends to which 3SR primers can bind after denaturation in a
`manner analogous to the original target generation scheme of
`SDA (1). (It should be noted