This example pertains to the synthesis of a plasmid, pIR8, useful for preparing "panhandle" adapter molecules. pIR8 is only one example of a suitable plasmid for making adapters. It will be appreciated by persons skilled in the art that other plasmids can be employed using similar methodology. It will also be appreciated that suitable adapters can be produced using automated nucleotide synthesizer technology known, in the art.

The pIR8 plasmid was constructed by first cutting a pUC19 plasmid cloning vector (Yanisch-Perron et al., Gene 33:103 [14] 119 (1985); ATCC Accession No. 37254; 2686 base pairs in length) at its unique NarI and EcoRI sites. (NarI cuts pUC19 at nucleotide 235 and EcoRI cuts at nucleotide 396.) The resulting restriction fragments were blunted using the Klenow fragment of E. cold DNA polymerase I plus all four dNTPs, and ligated using T4 DNA ligase. This created a molecule "BDA I" which had a regenerated EcoRI site at the former NarI site. BDA I includes all the nucleotides of pUC19 except nucleotides 235 396. It also contains the polylinker cloning site of pUC19 having the following sequence (Seq. ID NO.:1):



wherein restriction enzyme cleavage sites are indicated. The polylinker cloning site spans from nucleotide 396 to nucleotide 451.

BDA I was cut at its unique HinDIII and Scal sites. (HinDIII cuts pUC19 at nucleotide 447 and ScaI cuts at nucleotide 2177.) The two resulting fragments were blunted using the Klenow fragment of DNA polymerase I and all four dNTPs. The smaller fragment thus generated was separated from the larger fragment by electrophoresis in low-melt agarose. Similarly, a pUC18 plasmid (Yanisch-Perron et al., id.), which differs from pUC19 only in having its polylinker cloning site oriented in reverse, was cut at its unique NarI and Scal sites. The two resulting fragments were blunted using the Klenow fragment and all four dNTPs. The larger fragment was purified by electrophoresis in low-melt agarose. The larger fragment of pUC18 and the smaller fragment of BDA I thus prepared were ligated together using T4 DNA ligase and subsequently transformed into E. coli. Selection on Amp plates yielded transformants containing a recombinant plasmid pIR8.

The pIR8 plasmid has the polylinkers of pUC18 and pUC19 placed in an inverted orientation relative to each other. As shown in Figure 7, cutting pIR8 with any of the restriction enzymes that cut within the polylinker (except HinDIII) generates a large fragment and a small fragment each with identical ends. For better efficiency in the cycling reactions, the smaller molecules are preferred for use as adapters in BDA. As detailed elsewhere herein, the larger fragment is useful for a non-cycling BDA reaction because it contains both a replication origin for E. cold and a selectable marker. Due to the inverted orientation of the polylinkers, inverted repeats capable of pairing within the single strands of the small fragment range from 11 bases (SphI-cut) to 60 bases (EcoRI-cut).

The adapters were purified by cloning a fragment containing the self complementary polylinker sequences of pIR8 into phagemid DNA with subsequent generation of single strands using M13-derived phage, restriction enzyme digestion, and isolation on acrylamide gels. The protocol was as follows:

The small EcoRI fragment of pIR8 was ligated into the unique EcoRI site of a pIB plasmid to create the plasmid pMIRl. The pIB plasmid was constructed by: (a) cutting the commercially available phagemid pIB130 (available from International Biotechnologies, Inc., New Haven, Conn.) with SacI and HinDIII; (b) blunting the ends of-the fragments using the Klenow fragment and all four dNTPs; and (c) ligating with T4 DNA ligase. The pMIR1 DNA was transformed into MV1190 cells (containing the F pilus) which were subsequently allowed to grow to early log phase and then superinfected with the M13-based vector designated M13KO7 (available from International Biotechnologies, Inc., New Haven, Conn.). The mixture was allowed to shake for 45 minutes at 37° C. Kanamycin was added to select for cells containing both the M13KO7 and pMIR1 plasmids. After 16 hours, single-stranded DNA was purified from the cells. In this DNA, the ratio of pMIR1 DNA to M13KO7 DNA was about ten to one. This DNA mixture was heated to 75° C. in BamHI buffer (150 mM NaCl, 10 mM Tris-HCI pH 7.9, 10 mM MgCl2, 1 mM dithiothreitol, 100 µg/mL BSA) and allowed to cool slowly to 37° C. to allow the inverted polylinker sequences in pMIR1 to self-anneal. Three units of BamHI were added per microgram of pMIR1 DNA and the resulting 224-base adapter molecule was purified from 5% acrylamide gels. Approximately 5 µg of adapter (as determined via fluorescence with ethidium bromide) was recovered in this manner. The concentration of the adapter was about 105 ng/µL.

An adapter molecule suitable for BDA was synthesized chemically. The sequence of the entire 121-nucleotide molecule (there is no complementary strand) was as follows Seq. ID NO.:2):

5' GATCCCGGGTACCATGGCCAAGCTTAAGTACTCGCTTTTG
GGTTAGGAGAGCAGCATCTGACGACGGAGATGACGGAAAT
GAAAACGACGGCGAGTACTTAAGCTTGGCCATGGTACCCGG 3'

Significant features of this molecule include a duplex of 31 base-pairs with unique recognition sites for XmaI (SmaI), KpnI, NcoI, EaeI, MscI, HindIII, AflII, and ScaI; an overhanging 5' end sticky for BamHI, BglII, MboI, or BclI; and a 54-base singly-stranded loop sequence.

A 341-bp Sau3A I duplex fragment from pUC18 (spanning nucleotides 1662 to 2003) was purified from low-melt agarose for use as a template usable for BDA (sequence containing an SOI).

The sequence (Seq. ID NO.:3) of one of the strands of the fragment is as follows (complementary strand not shown):

5' gatctgtctatttcgttcatccatagttgcctgactccccgtcgtgtaga taactacgatacgggagggcttaccatctggccccagtgctgcaatgata ccgcgagacccacgctcaccggctccagatttatcagcaataaaccagcc agccggaagggccgagcgcagaagtggtcctgcaactttatccgcctcca tccagtctattaattattaccagaaagctagagtaagtagttcgccagtt aatagtttgcgcaacgttgttgccattgctacaggcatcgtggtgtcacg ctcgtcgtttggtatggcttcattcagctccggttcccaac 3'

wherein a 5-base diagnostic restriction site for AvaII is underlined. Also, a 21 base primer target site and a 30-base primer target site are underlined. The initial gate, of course, represents an overhanging sticky end.

The ends of this fragment were "sticky" to the ends of the 224-base BamHI cut adapter of Example 1. Concentration of the 341-bp fragment was determined by fluorescence in the presence of ethidium bromide to be 200 ng per µL.

EXAMPLES 4-10

An important step in BDA is ligation of adapters to the templates. Too few adapters present during ligation will lead to increased ligation of templates to each other rather than to adapters.

Three ligation products were prepared utilizing 0.2 pg (Ligation A), 2 pg (Ligation B), and 20 pg (Ligation C) of the 341-bp template of Example 3, each in the presence of 0.2 µg of the adapter of Example 1, wherein 0.2 pg of 341 bp template corresponds to about 535,000 copies thereof. Ligation reactions were performed at 16° C. in 20 µL volume of 1x T4 ligase buffer (50 mM Tris-HCI pH 7.8, 10 mM MgCl2, 20 mM DTT, 1 mm ATP, 50 µg/mL BSA) and 400 units of T4 DNA ligase from New England BioLabs, Beverly, Mass. The molar ratios of adapter to 341-bp template in Ligations A, B, and C were 3,000,000 to 1, 300,000 to 1, and 30,000 to 1, respectively. No extraneous DNA (i.e., no non-target DNA) was included in these ligations to minimize effects arising from non specific priming of replication. The ligation reaction conditions were chosen with two aims: (1) to maximize likelihood of ligating the 341-bp template to the adapters using high molar ratios of adapter to template; and (2) to mimic the amount of DNA typically used in a ligation reaction (wherein 0.2 µg=2.6 µM of 5' ends).

Another variable to be controlled in BDA is the concentration of primer because, unlike with PCR, complementary strands are held in close proximity, even after denaturation.

This is due to the "closed loop" structure of the starting substrates and the fold-back structure of the replication products made in subsequent cycles of replication.

A 30-base primer (5 AACTACTTACTCTAGCTTCCCGGCAACAAT3) was synthesized. This sequence (Seq. ID NO. :4) is complementary to the region from nucleotide 1903 to nucleotide 1874 of pUC18 within the 341-bp template of Example 3. This is exactly 100 bases from one end of the template. Thus, extension of the primer from the 3' end thereof would yield a complementary 241-bp sequence (including the primer).

Seven different BDA reactions were set up using the ligation products as described above and various molar ratios of primers relative to the corresponding ligation product. (Normally, PCR utilizes 25 pmol primer to 456,000 copies of template (or 3.3x10e7 primers relative per template)). Examples, with corresponding amounts of primers and ligation products, are tabulated below:



Examples 4, 5, 6, 7, and 8 were all performed using Vent(TM) DNA Polymerase (from New England BioLabs, Beverly, Mass.), using the manufacturer's recommended reaction conditions (10 mM KCL, 20 mM Tris-HCI pH 8.8 at 25° C., 10 mM ammonium sulfate, 2 mM magnesium sulfate, 0.1% Triton X-100, 100 pg/ml BSA, 200 µM of each dNTP, 10 units of enzyme) in 100 µL volumes under mineral oil. Examples 9 and 10 were performed using about two Units of Taq DNA Polymerase (from Promega Biotech, Madison, Wis.) and according to the manufacturer's recommended conditions in 100 pL volumes under mineral oil. All BDA reaction mixtures were prepared on ice prior to initiation of the reactions on a Perkin-Elmer-Cetus PCR machine. BDA cycles were defined as follows:

1. Heat to 96° C. Hold 5 minutes.

2. Cool to 55° C. Hold for I minute.

3. Heat to 72° C. Hold for 45 seconds.

4. Heat to 96° C. Hold for I minute.

5. Repeat steps 2-4 for a total of 45 cycles.

6. On the last cycle, the 72° C. reaction was incubated for 2 minutes rather than 45 seconds and was not subsequently heated to 96° C. Afterward, the reactions were cooled to 10° C. for storage until analysis.

All of Examples 4-10 were simultaneously treated under the above conditions. Upon completion of the BDA, the reaction mixtures were extracted with chloroform to remove the mineral oil and 15 µL from each Example were electrophoresed on a 5% acrylamide gel stained with ethidium bromide. (Whenever a band resulting from a BDA reaction is evident on a gel stained with ethidium bromide, it is an indication that DNA replication occurred in the corresponding BDA reaction. This is because there is an insufficient amount of starting DNA in any of the reactions to produce any bands on the gel.) Example 4, 5, 7 and 8 produced multiple bands in the range of 500 base pairs (bp) to 1 kilobase pairs (kbp). Example 8 produced a single band of about 700 bp. The predicted length of a full duplex arising from BDA amplification of the 341 bp fragment is 713 bp. (The presence of single-stranded loops in any BDA-produced molecule adds a degree of uncertainty in predicting molecular sizes. Such loops are understood by persons skilled in the art to exhibit anomalous migration behavior in gels.) Examples 4, 5, and 7 also exhibited faint bands of about 700 bp as well as other bands.

To more accurately characterize the structure of these BDA products, the reaction of each Example was first eluted through Centricon 100 filters (available from Amicon, Beverly, Mass.) to separate low molecular-weight compounds such as buffer and primers from the BDA products. One third of the eluted volume of each Example was either digested with SalI or left uncut as a control. SalI cleaves within the polylinker of the adapter and was expected to release the replicated template portion of the BDA product from the adapter portion. Thus, SalI digestion of full-duplex was expected to produce molecular fragment sizes of about: 254 bp (template portion) and about 205 bp (adapter portion). SalI digestion of looped-duplex BDA products was expected to produce fragment sizes of about 254 bp (template portion) and a fragment of indeterminate length probably having an apparent size of about 330 bp due to the ' presence of the looped adapter.

SalI digestion of Example 4 yielded a prominent band at about 205 bp, a less intense band at 225 bp, a faint band near 250 bp, another faint band near 290 bp, a moderately intense band near 330 bp, several faint bands of less than 150 bp, and several bands between 500 bp and 1 kbp that appeared to be uncut. SalI digestion of Example 5 yielded a quantity of uncut DNAs between 500 bp and I kbp, a band near 330 bp, and a band near 290 bp. SalI digestion of Example 7 yielded results very similar to SalI-cut Example 5. SalI digestion of Example 8 exhibited very faint bands near 330 bp and possibly 210 bp.

These data indicate that:

(a) BDA reaction conditions utilizing about 2 pg of template DNA produce detectable BDA products;

(b) Template concentration (relative to adapter concentration) can be varied to some extent in the ligation of adapters to templates and still result in BDA products. Even though the exact composition of the BDA products produced in Examples 4-10 was not clear from the limited analyses performed, these Examples did provide ranges for certain BDA reaction conditions that were useful in subsequent Examples; and

(c) Some non-specific priming occurred in these reactions, probably as a result of high primer concentration, thereby generating certain artifacts. Such binding would give rise to artificially high amounts of adapter DNA. Raising the annealing temperature or incorporating trimethyl ammonium chloride into the reaction would be expected to ameliorate these problems.

An adapter (Seq. ID NO.:5) was synthesized having the following sequence (wherein the polylinker regions are indicated by capital letters):


A 6-base diagnostic restriction site for PvuII is underlined. Also, a 4-base restriction site for Sau3A is underlined.

Although not used in any of the Examples disclosed herein, this adapter is useful for BDA.

An adapter (Seq. ID NO.:6) was synthesized having the following sequence (wherein the polylinker regions are indicated by capital letters):


A 6-base diagnostic restriction site for PvuII is underlined. Also, a 4-base restriction site for SaU3A is underlined.

The following primers were synthesized: a 30-base primer having the sequence (Seq. ID NO.:7):

5'aaaacrtactmagatcccggcaacaat3'

and a 21-base primer having the sequence (Seq. ID NO.:8):

5 gatagtctatttcgttcatc3'

When these primers are used together in a PCR reaction with the template of Example 3, the product is a 24 -bp sequence that serves as an internal control. Production of this 241-bp fragment indicates that conditions are appropriate for PCR to occur.

These Examples comprise experiments that were performed to examine the contribution to a BDA reaction of each of the components thereof. These Examples utilized standard conditions for BDA, annealing, elongation, and denaturation, as outlined below.

At time of use, BDA samples were extracted once with 120 µL of chloroform/isoamyl alcohol (24/1). 15 µL of each sample were electrophoresed in gels for size analysis using standard methods.

Ligations for all BDA reactions were as follows: 0.2 µg of the adapter of Example 12 cut with BamHI was ligated to 200 pg of Sau3A-digested DNA. Ligation of DNA was assayed by production of dimers of the Example-12 Adapter. Typically, the template to be amplified was the 341-bp Sau3A I fragment of Example 3. However, Arabidopsis genomic DNA was used where indicated. In all ligations, the same amount of T4 DNA ligase was used as described in Examples 4-10. The same ligation conditions were used throughout.

All BDA and PCR reactions, unless noted otherwise, utilized 2 pg of template DNA, 2 µg of 21-base primer (Example 13), and/or 2.7 µg of 30-base primer (Example 13). Reactions were performed using either the Vent (trademark) DNA polymerase (obtained from New England BioLabs, Beverly, Mass.) or Taq DNA polymerase (obtained from Promega, Madison, Wis.) according to manufacturer's specifications No apparent difference was noted in substrate specificity or amount of product produced by either enzyme.

PCR reactions were performed using conventional protocol.

The BDA reaction cycle profile was as follows:

1. Heat to 96° C. for 5 minutes.

2. Cool to 58° C. for 2 minutes.

3. Heat to 72° C. for 45 seconds.

4. Heat to 96° C. for I minute.

5. Cool to 58° C. for 20 seconds.

6. Repeat steps 3-5 45 times.

7. Heat to 96° C. for I minute.

8. Cool to 58° C. for 20 seconds.

9. Heat to 72° C. for 2 minutes.

10. Store at 10° C. overnight or until use.

When restriction mapping was performed, the BDA samples were purified by centrifugation on Centricon 100 filters and eluted with three 1-mL washings of water to remove small impurities and primers. Restriction analysis utilized 15 µL of each sample and 4 to 5 Units of each corresponding restriction endonuclease.

Specifically, each Example was performed, and results obtained, as follows:

Example 14: A standard PCR reaction performed using the template of Example 3 and the 30-base and 21-base primers of Example 13. A strong 241 -bp band was produced. This indicated that PCR works using this template and these primers under the conditions employed below for BDA.

Example 15: BDA was attempted using a non-ligated mixture of the Example 12 adapter, the Example-3 template, and the 21-base primer of Example 13. No BDA products were evident on the gel, indicating that ligation of adapters to the template is essential for BDA.

Example 16: BDA was attempted using a ligated mixture of the Example-3 template and 0.2 µg of Sau3A I-digested Arabidopsis genomic DNA (instead of the Example-12 Adapter). The 30-base primer of Example 13 was used. No BDA products were evident on the gel, indicating that a looped adapter is essential for BDA.

Example 17: As in Example 16 except that the 21-base primer of Example 13 was used. No BDA products were evident on the gel, indicating again that a looped adapter is essential for BDA.

Example 18: BDA reaction using the Example- 12 adapter ligated to the Example-3 template, and using the 30-base primer of Example 13. No BDA products were evident on the gel, indicating that, like PCR, BDA can sometimes exhibit variable priming.

Example 19: As in Example 18 except that the 21-base primer of Example 13 was used. Several bands of BDA products were seen on the gel, indicating that the BDA reaction occurred with specific priming.

Example 20: BDA reaction using the Example-12 adapter ligated to Arabidopsis DNA digested with Sau3A I; the 30- base primer of Example 13 was also used. Faint bands of various sizes on the gel indicated either that the primers non- specifically bound to the Arabidopsis DNA or that the Arabidopsis DNA contained small sequences substantially homologous to the primers.

Example 21: As in Example 20 except that the 30-base primer of Example 13 was used. No BDA products were evident on the gel, indicating that the Arabidopsis DNA contained no primer target sites for the 30-base primers.

Example 22: BDA reaction of Example-12 adapters ligated together without any intervening template; the 21 -base primer of Example 13 was used. A number of bands appeared on the gel, indicating that the primers non specifically bound to one or more locations on the adapters.

Example 23: As in Example 22 except that the 30-base primer of Example 13 was used. A number of bands having different sizes than the fragments seen in Example 22 appeared on the gel. Again, this indicates that the 30-base primer non-specifically bound to one or more locations on the adapters.

In an effort to better understand the results in Examples 14-23, the reactants in each of the Example 14-23 reactions were individually electrophoresed in a 5% acrylamide gel. The object was to ascertain whether or not ligation had occurred prior to actually beginning BDA.

The following results were obtained as visualized on the gel:

Example 24: Corresponding to Example 15; a single band appeared on the gel which corresponded to the unligated Adapter molecules.

Example 25: A control containing the Example-12 Adapters ligated together; a single band appeared on the gel which corresponded to the size of the ligated Adapters.

Example 26: Corresponding to Examples 22 and 23; a single band was seen on the gel as in Example 25.

Example 27: Corresponding to Examples 20 and 21; a single band was seen on the gel as in Example 25.

Example 28: Corresponding to Examples 18 and 19; a single band was on the gel as in Example 25.

Example 29: Corresponding to Examples 16 and 17; no discrete bands were visible on the gel due to the presence of a multitude of differently sized fragments. This was as expected because Sau3A I-cut duplex Arabidopsis DNA generates a large number of differently sized fragments.

Examples 14-29 indicate that BDA is similar to PCR in that non-specific priming can sometimes occur. Nevertheless, priming does appear to result in actual amplification of DNA.

These Examples comprise an evaluation of BDA amplification of the Example-3 template using the Example- 12 adapters and either the 30-base or 21-base primer of Example 13. BDA protocols were as described in Examples 14- 23.

The following results were obtained, as visualized on a 1% agarose gel:

Example 30: A "negative PCR control" on an Arabidopsis target DNA, performed using a 1.6-kb template from Arabidopsis with corresponding primers, but containing no DNA polymerase. A single diffuse band was seen on the gel corresponding to the size of the primers.

Example 31: A "positive PCR control" on an Arabidopsis target DNA, performed as in Example 30 but including DNA polymerase. A strong band on the gel at 1.6 kb indicated that PCR occurred.

Example 32: A "negative PCR control" on the Example-3 template, performed using the 21-base and 30-base primers of Example 13 but lacking DNA polymerase. No distinct bands were evident on the gel.

Example 33: A "negative BDA control" on a BDA template, performed using the Example-3 template ligated to the Example-12 adapters, and including the 21-base and 30-base primers of Example 13, but no DNA polymerase. No apparent bands were evident on the agarose gel.

Example 34: A "negative BDA control" for the 21-base primer on a BDA template, performed as in Example 33 but including only the 21-base primer of Example 13. No apparent bands were evident on the agarose gel.

Example 35: A "negative BDA control" for the 30-base primer on a BDA template, performed as in Example 33, but including only the 30-base primer of Example 13. No bands were evident on the agarose gel.

Example 36: A PCR reaction on a BDA template involving 2.0 pg of the Example-3 template ligated to Example-12 adapters and the 21-base and 30-base primers of Example 13. The primary product on the agarose gel was a band having an apparent size of about 263 bp, which substantially agrees with the expected PCR product of such a reaction. Subsequent electrophoresis of this product on 5% polyacrylamide exhibited a band of about 239 bp. The difference in apparent size between the different gels is within a reasonable variability expected for a fragment of 241 bp under these conditions.

Example 37: BDA reaction as in Example 36 but employing only the 21-base primers of Example 13. Two bands were seen on the agarose gel at about 378 and 717 bp.

Example 38: BDA reaction as in Example 36 but employing only the 30-base primers of Example 13. Two bands were seen on the agarose gel at about 473 and 666 bp.

Example 39: A "PCR control" reaction using 0.2 pg of the Example-3 template and the 21-base and 30-base primers of Example 13. An extremely faint band was evident on agarose gel at about 241 bp.

Example 40: A "PCR control" reaction as in Example 39 but using tenfold more template DNA (2.0 pa). This reaction produced a more pronounced band at about 241 bp on agarose gel than did Example 39.

Example 41: PCR reaction on the Example-3 template as in Example 36 but using both the 21-base and 30-base primers and one-tenth the amount of template. Only one faint band at about 241 bp on the agarose gel was evident.

Example 42: BDA reaction as in Example 36 but using the 21-base primer and one-tenth the amount of template. No 30- base primers were used. No bands were evident on the agarose gel.

Example 43: BDA reaction as in Example 37 but using the 30-base primer and one-tenth the amount of template. No bands were evident on the agarose gel.

These Examples (30-43) indicate that BDA produces distinct products having molecular sizes that correspond to the particular primer(s) were used.

The products of Examples 30-43 were further characterized using Southern blotting. The bands from the agarose gel were blotted onto nitrocellulose and probed with the 30-base primer of Example 13 labeled on its 5' end with 32p.

None of the controls (Examples 30-35) produced any labeled bands on the autoradiogram.

The single band on the gel of Example 36 produced a corresponding strongly labeled band on the autoradiogram.

Each of the two bands seen on the gel of Example 37 produced a corresponding strongly labeled band on the autoradiogram. Since none of Examples 30-35 produced a band on the autoradiogram, BDA has apparently amplified the Example-3 template in Example 37. Moreover, since the region that was probed in the Southern blot was internal to the 21-base primer target site on the template, it was concluded that BDA in Example 37 extended the 21 base primer in the correct direction and faithfully copied the template because the BDA products included the 30-base primer target site as well. The 30-base primer target site is 241 bases away from the 21-base primer target site on the template.

Each of the two bands seen on the gel of Example 38 produced a corresponding strongly labeled band on the autoradiogram. Again, since none of Examples 30-35 produced a band on the autoradiogram, BDA has apparently amplified the template in Example 38.

Each of the single bands seen on the gel of Examples 39-41 produced a corresponding radiolabeled band on the autoradiogram.

From these Examples (30-43) it was concluded that:

1. BDA resulted in replication (and, therefore, amplification) of at least portions of the 341-bp template of Example 3. Control reactions (Examples 32 35) containing starting amounts of the Example-3 template (0.2 pg or 2.0 pa) produced either no products or such small amounts of products that were undetectable as either ethidium bromide stained bands in agarose or by Southern blotting. Only when the template was actually amplified by PCR or BDA were products detectable. BDA reactions utilizing only a single primer directed the synthesis of easily detectable amounts of fragments that contained sequences of the template.

2. It is known that, in order to visualize a DNA band on an ethidium bromide-stained agarose gel, about 15-30 ng of DNA are required. In the above Examples, the amount of DNA loaded onto the gel was about 15% of the total of the BDA reaction. Thus, the total BDA product was at least 100 ng. If 50 percent of the BDA product was due to sequences of the 341-bp Sau3A I fragment, then the magnitude of amplification seen in these experiments was at least 25,000 fold (50 ng product per 2 pg of starting template).

3. Replication from the 21-base primer extended at least 241 bases in the direction of the 30-base primer target site and faithfully copied the template.

Other experiments were performed to further characterize the products of Examples 30-43. The BDA products were purified from any primers present therewith by centrifugation with three 1-mL washes on Centricon 100 filters and subjected to restriction-mapping analysis on a 5% acrylamide gel. The fragment sizes seen on the acrylamide gel were found to correlate with fragment sizes seen on the 1% agarose gel. Actual fragment lengths on both the 1 % agarose and 5% acrylamide gels were determined in a conventional way by plotting the inverse logarithm of the molecular weight of length standards (a l-kb "ladder") against mobility in each gel. Unknown fragment sizes were interpolated from the nearest known- molecular-weight bands comprising the ladder. The unusual panhandle structure of BDA adapters resulted in an unusual electrophoretic mobility for both monomer and dimer (ligated) forms of the adapter. For example, although the adapters used in these Examples were 7 each 224 bases long, they had an apparent mobility corresponding to about 340 bp. Dimeric forms of these adapters appeared to migrate at rates corresponding to a molecular weight of about twice that of the adapter monomers.

Three restriction endonucleases (PvuII, Sau3A I, and AvaII) were used to further study the BDA products of Examples 37 and 38. Cleavage sites on a closed-loop structure 240 formed by ligating the Example-3 template with the Example- 12 adapters are shown in Figure 8. (Although the maps in Figure 8 depict, for clarity, the location of the PvuII and Sau3AI recognition sites in the spacer region 242, it will be understood by persons skilled in the art that these enzymes do not efficiently cleave recognition sites unless they are present in duplex DNA.) These restriction endonucleases were selected because each cuts at one or more loci on the closed loop structure 240. A single PvuII site is located in the loop region 242 of the adapter 244. No PvuII sites are located in the template 246. A single AvaII site is present in the template 246, but not in the adapter 244. A Sau3A I site is present where the panhandle region 248 of each adapter 244 is joined to the corresponding end of the template 246 and within the loop 242 of each adapter 244

The unique Avaii site in the template 246 is 176 bases from the 5'-end of the 21-base primer target site (arrow 250) and 63 bases from the 5'-end of the 30 base primer target site (arrow 252). (These two lengths do not add up exactly to 241 bp because of the locations of the Avaii cutting sites.) The Sau3A sites in the duplex region 248 of the adapter 244 are 341 bp from the 5'-end of the 21 base primer target site 250 and 241 bp from the 5'-end of the 30-base primer target site 252. As shown in FIG. 8, a BDA reaction resulting from priming with a 30-base primer complementary to the 30-base primer target site 252 would be expected to yield the BDA product 254. A BDA reaction resulting from priming with a 21-base primer complementary to the 21-base primer target site 250 would be expected to yield the BDA product 256.

Each Example was prepared as follows, with results as noted:

Example 44: Product from Example 38 cut with Sau3A I.

Example 45: Product from Example 38 cut with PvuII.

Example 46: Product from Example 38 cut with AvaII.

Example 47: Product from Example 38 uncut.

Example 48: Product from Example 37 cut with Sau3A I.

Example 49: Product from Example 37 cut with PvuII.

Example 50: Product from Example 37 cut with AvaII.

Example 51: Product from Example 37 uncut.

Example 52: (PCR control) product from Example 36 uncut.

Results obtained were:

1. Examples 45 and 47 each produced two differently sized bands on a 5% acrylamide gel. The Example-45 bands appeared to be about 712 and 476 bp long, respectively, which is exactly the same size as the two Example-47 bands. Therefore, it was concluded that the two Example-45 bands were apparently not cut by PvuII, indicating either that they lacked the CAGCTG sequence recognized by PvuII or such sequence was present solely in a single-stranded form such as a loop where the enzyme cannot cut. Likewise, Examples 49 and 51 each yielded exactly the same two bands on the gel, having apparent sizes of 712 and 476 bp long, respectively. Therefore, it was concluded that, as in Example 45, the two bands in Example 49 also represented DNA apparently uncuttable by Pvuii. Since PvuII-cutting indicates the presence of the loop sequences in double- stranded form, the absence of PvuII cutting in Examples 45 and 49 indicated either the absence of the loop sequences in these Examples or that the loop portion was present only in single-stranded form.

2. The Example 46 reaction exhibited two bands on the gel, having apparent sizes of about 643 and 418 bp, that were smaller than two corresponding bands exhibited by Example 47, one by about 61 bp and the other by about 69 bp. Thus, it was concluded that the two bands in Example 47 each contained an AvaII site, which is unique to the Example-3 template. In addition, Example 46 produced a third very faint band at about 68 bp and the two bands produced in Example 47 were larger than the corresponding bands of Example 46 by about 68 bp. These AvaII-cutting results are consistent with a BDA product structure being linear at one end and terminating at the linear end at the 5' end of the corresponding primer target site. These results are also consistent with BDA amplification of the 341-bp template and are consistent with the 1% agarose analysis and the Southern blotting experiments of Examples 30-43.

3. Example 50 produced a prominent band on the gel

sized at about 185 bp. Since the AvaII site is located 176 bp from the 5'-end of the 21-base primer target site, the 185 bp value determined here is well within the margin of error for this type of analysis. Two other substantially fainter but larger bands were seen, but their structure has not yet been determined.

4. Example 44 produced no discernable bands on the gel.

In contrast, Example 48 produced a somewhat wide band sized at about 341 bp. This is exactly the size that would be expected for a BDA product primed by the 21-base primer. There does not appear to be sufficient product from Example 44 to ascertain the sizes of any products formed therein.

This Example illustrates how a representative adapter molecule, as illustrated as item 192 in FIG. 5 and described herein above, can be synthesized.

It will be recalled in Example I that a plasmid pMIR1 was constructed by ligating the small EcoRI fragment of pIR8 into the unique EcoRI site of a pIB plasmid. The pMIR1 plasmid was transformed into MV1190 cells which were superinfected with M13K07. Single-stranded pMIR1 DNA was subsequently isolated from the cells, allowed to self-anneal, and cleaved with BamHI. A 224 base adapter was used further in Example 1. However, the BamHI digestion also yielded a larger fragment which contains an E. cold replication origin and a gene encoding ampicillin resistance. Cleavage of this single-stranded pMIR1 product with PstI yields a panhandled molecule having a duplex panhandle of about 45-50 bp and a single-stranded loop of about 2.7 kb. The single-stranded loop contains the E. cold origin of replication and the ampicillin-resistance gene.

This Example comprises a test of the ability of BDA to amplify a specific DNA sequence of interest from a mixture of sequences obtained from genomic DNA.

Arabidopsis genomic DNA (0.1 µg; genome size is about 50,000,000 bp) was completely digested with 10 Units of Sau3A I restriction endonuclease. The resulting fragments, having ends sticky for BamHI, were ligated to 0.3 ug of Example-12 adapters to produce looped templates. BDA was performed using a 27-base primer having the following sequence (Seq. ID NO.:9):

5' AAACGACGGCGAGTAATGAACTAAACG3'

This primer was employed because it was known to be effective for identifying individual plaques, plotted on a bacterial "lawn," carrying a portion of the Arabidopsis genome. (Thus, the Arabidopsis genome contains at least one copy of this primer or a sequence homologous to it.) In, this Example, it was desired to clone the portion of the Arabidopsis genome containing this sequence or its close homolog.

After a usual number of BDA cycles, examination by gel electrophoresis followed by ethidium bromide staining revealed the presence of two faint bands on the gel. Southern blot analysis using the 27-base primer labeled at the 5' end with 32p indicated that both bands visible on the gel included a target site for this primer. A control reaction using 0.1 µg of Arabidopsis genomic DNA cut with Sau3A I but not replicated by BDA did not produce any bands detectable on either an ethidium bromide-stained gel or on a Southern blot.

The BDA-amplified genomic DNA was purified on Centricon 100 filters by washing three times with 1 mL distilled water. Approximately one-fourth of the BDA- amplified DNA was then digested using Sau3A I to cleave off the adapters. DNA amplified by BDA was expected to have a first end sticky for BamHI-cut DNA and a second, blunt, end terminating with the primer sequence. Any other DNAs in the mixture (e.g., adapters and genomic fragments not amplified by BDA) were expected to have ends sticky for BamHI-cut DNA only.

The Sau3A I-cut DNA was co-precipitated with 0.1 µg of pUC18 DNA cleaved with HincII and BamHI. Such digestion leaves the pUC18 DNA with a blunt end and an end sticky for Bamhi-cut DNA, just like the BDA-amplified DNA. The precipitated DNAs were resuspended in buffer, ligated using T4 ligase overnight at 16° C., and transformed into E. cold cells. Selection on plates containing ampicillin yielded 23 colonies.

Restriction analysis of the first seventeen colonies on 1% agarose gels revealed that four colonies contained apparently identically sized inserts (about 300-500 bp). One other colony appeared to contain a rearrangement of pUC18 and the remaining twelve appeared to contain pUC18 without any insert therein.

DNA sequence analysis of the four clones containing identically sized inserts revealed that the inserts in all four had identical sequences. Moreover, each of the inserts yielded DNA fragments having the exact structure of DNA amplified by BDA. Specifically, one end of each fragment was blunt and the other end was sticky for Sau3A I (or BamHI). Sequence analysis from the blunt end of the fragment yielded the following sequence data (Seq. ID NO.: 10):


Sequence analysis from the sticky end yielded the following sequence data (Seq. ID NO.:11):


On each end, the sequence was determined far enough to unambiguously reveal the sequence information sought. The portions of these fragments between the sequenced ends were not determined. Nevertheless, the fragments had the following important features:

(1) The blunt end comprised twenty-seven nucleotides having the exact sequence of the primer. It should be noted that it is possible for a sequence similar but not exactly identical to the primer to be extended in the BDA reaction. However, since the DNA polymerase does not efficiently edit mismatches between the primer and the primer target site, the BDA-amplified DNA will have the exact sequence of the primer, regardless of the downstream sequence. (PCR has the same uncertainty associated with it.) However, as in this Example, production of an amplified DNA product having a blunt end terminating with the primer sequence is a strong indication that the desired SOI was amplified by BDA, since there is virtually no other way for such a product to be produced other than by BDA.

(2) The sequence at the sticky end of the BDA-amplified DNA is exactly as expected of a fragment excised from adapters by cutting with Sau3A I.

Conclusions from this Example were as follows:

(a) BDA will occur when performed using a primer specific for a particular sequence of interest in a genomic DNA such as from Arabidopsis. BDA will generate amplified DNA sequences containing the primer sequence (or a sequence substantially homologous thereto) at a concentration sufficient for detection on ethidium bromide-stained: gels. The amplified sequences can be cloned into appropriately cleaved plasmid vectors.

(b) Amplification of DNA on these looped genomic templates resulted from primer binding to primer target sites, as indicated by the binding of labeled primer to specific bands on Southern blots. In addition, the precise alignment of the bands identified in the Southern blots with the ethidium bromide stained bands on the gels suggests strongly that the ethidium bromide-stained products are the result of BDA specifically initiated by the primer sequence. Sequence analysis of clones derived from the BDA products confirmed the presence of the priming sequence in all cloned fragments.

(b) The BDA products of this Example had structures and sequences that were as predicted for BDA products. At one end, the DNAs were blunt and had a sequence identical to that of the primer. At the other end, the DNA had an end sticky for DNA fragments cut with the same restriction enzyme used to cleave away the adapters. Unamplified DNA molecules or primers would not have blunt first ends and would not terminate on the blunt end with the primer sequence. Since the primer was single-stranded, it was not possible for the DNA ligase to attach the primer fortuitously onto a random DNA fragment. Moreover, there was no evidence of a Sau3A I site at the 3' end of the primer, which is what would be expected for a random fragment (not amplified by BDA) ligated to a single-stranded primer. The only Sau3A I site in the BDA product was at the sticky end thereof.

(d) If these analyses had produced several different clones each with a different DNA sequence, one might conclude that either the priming sequence was present in several copies in the genome (like a repetitive sequence) or the BDA process used to generate them was prone to artifactual amplification of sequences with no relation to the primer. But, the isolation of four identical clones, and no others, containing genomic DNA amplified by BDA suggests strongly that the BDA reaction performed here targeted a single genomic sequence of interest and that part of the sequence of interest was at least homologous to (if not identical to) the primer used in the BDA reaction.

(e) The four identical cloned fragments reported here were apparently produced by a BDA reaction directed by specific priming of a template hybridized to a homologous (or identical) sequence in the Arabidopsis genome.

While the invention has been described in connection with preferred embodiments and numerous examples, it will be understood that it is not limited to these embodiments and examples. On the contrary, it is intended to cover all alternatives, modifications, and equivalents as may be included within the true spirit and scope of the invention as defined by the appended claims.




Return to BDA Home Page

Return to BIOTECHNOLOGY SOFTWARE JOURNAL Page

Return to Kevin Ahern's Home Page

Return to Kevin Ahern's CV