

Reaction	Template	180	Template 181

A, complete	325,782	441,823
A, no polymerase	5,187	5,416
A, no primer	4,351	12,386
B, complete	5,674	176,291
B, no polymerase	2,988	1,419
B, no primer	1,889	1,266

As can be seen from these results, specific template-directed labeling of[0096]primer 182 can also be determined by measuring the total radioactivity of the reaction products after washing with magnetic beads to remove unreacted nucleotides. The background in this experiment due to nonspecific label from all other sources was approximately 3-4% (compare

templates

180 and 181 in the “B, complete” reaction). Control experiments (“no polymerase” and “no primer”) showed that the bulk of the background label was probably contributed by unincorporated nucleotides that were not completely removed by the washing step. The “A, complete” reactions showed that, -for both templates, productive template:primer complexes were present.

Example 3

Two amplification primers,[0097]TGL 105 and TGL 106 (FIG. 2), were used to amplify a cloned stretch of bovine DNA containing two DNA sequence polymorphisms: a C or T atposition 114 and an A or G at position 190 (FIG. 2). DNAs containing these polymorphisms were molecularly cloned and available on plasmids, as follows: plasmid p183, C114 and A190; plasmid p624, T114 and A190; plasmid p814, C114 and G190. Four PCR reactions with biotinylated primers were performed to amplify and purify specific strands of these plasmids for use as templates:



Primers	Plasmids	Detection Primers

105 biotinylated,	p183 andp624	TGL	182
106 unbiotinylated
105 unbiotinylated,	p183 andp814	TGL	166
106 biotinylated

The duplex PCR products were bound to magnetic microspheres, denatured with NaOH, and the biotinylated strand purified as described above. Templates prepared with[0098]biotinylated TGL 105 were subjected to analysis by DNA sequencing withunbiotinylated primer TGL 106 in order to measure the amount of template present. Similarly, template prepared usingbiotinylated TGL 106 was analyzed by sequencing withunbiotinylated TGL 105.

Approximately equal amounts of template (2 pmoles) were annealed for 5 min at 65° C. to the polymorphism detection primers,[0099]TGL 182 and TGL 166 (see above and FIG. 2). These primers hydrogen-bond to the templates in a sequence-specific fashion such that their 3′-termini are adjacent to nucleotide

positions

114 and 190, respectively (FIG. 2). Template-directed primer extension reactions (reaction “B” conditions) were carried out on these primer:template complexes in the presence of the four ddNTPs, one of which (ddATP) was labeled. The products of these extension reactions were analyzed by electrophoresis on a 15% polyacrylamide/8M urea gel followed by autoradiography (FIG. 3).

Example 4

[0100]


	Primer oligo TGL391:
	^5′TGTTTTGCACAAAAGCA^3′

	Primer oligo TGL346:
	^5′GTTTTGCACAAAAGCAT^3′

	Template oligo TGL382:
	^3′CACAAAACGTGTTTTCGTAGGA^5′-biotin:
	(streptavidin-bead)

Oligonucleotide TGL382 was purchased from the Midland Certified Reagent Company, Midland, Tex. It was biotinylated using Midland Certified Reagent Company's “Biotin dX” reagent (a biotin derivative phosphoramidite) which is suitable for use in automated DNA synthesis in the 5′ terminal nucleotide position. The biotinylated oligonucleotide was then purified by anion exchange HPLC. Streptavidin-conjugated M-280 Dynabeads were washed in TNET buffer (10 mM Tris-HCl, pH 7.5, 100 mM NaCl, 1 mM EDTA, 0.1% Triton X-100) and resuspended in the same buffer at a concentration of 7×10[0101]⁸beads/ml. 10-100 pmoles of biotinylated oligonucleotide TGL382 was incubated with 100 μl of the Dynabead suspension in TNET for 30 minutes at 20° C. in order to allow the biotin moiety to bind to the streptavidin. The beads were then washed (using a magnet to immobilize them) three times with 200 μl of TNET and resuspended in 100 μl of TNET. For annealing, 25 μl of this suspension of the Dynabeads with the attached template oligonucleotide was immobilized with the magnet, the TNET withdrawn, and 25 μl of 40 mM Tris-HCL, pH 7.5, 20 mM MgCl₂, 50 mM NaCl, containing 2 μM ofoligonucleotide primers 346 or 391, was added. The template and each primer were annealled by incubating them for 5 minutes at 65° C., followed by slow cooling over a period of 20 minutes to room temperature. Beads containing the bound template-primer complexes were washed twice with 200 μl TNET, followed by resuspension in 25 μl of 40 mM Tris-HCl, pH 7.5, 20 mM MgCl₂, 50 mM NaCl.

The following ddNTP mixes were used:[0102]

[0103]³⁵S-labelled dideoxynucleoside triphosphate mixes (labelled nucleotide indicated in the form ddN★TP):



ddG Mix:	5 μM ddG*TP	10 μM ddATP	10 μM ddTTP
	10 μM ddCTP
ddA Mix:	10μM ddGTP	5 μM ddA*TP	10 μM ddTTP
	10 μM ddCTP
ddT Mix:	10 μM ddGTP	10μM ddATP	5 μM ddT*TP
	10 μM ddCTP
ddC Mix:	10 μM ddGTP	10 μM ddATP	10μM ddTTP
	5 μM ddC*TP

For each bead-bound, template-primer complex, four extension reactions were carried out, one reaction for each of the four ddNTP mixes. Extension reactions contained the following components: 5.0 μl bead suspension containing the annealled template-primer complex, 0.5 μl of 100 mM dithiothreitol, 0.5 μl of “Mn[0104]⁺⁺ solution” (100 mM MnCl₂, 150 mM DL-isocitrate, pH 7.0; purchased from U.S. Biochemicals, Cleveland, Ohio), 1.0 μl of ddG, ddA, ddT, or ddC mix, 2.0 μl of H₂O, and 1.0 μl of T7 DNA polymerase (“Sequenase”, version 2.0, US Biochemicals, 1625 units/ml in 50 mM Tris-HCl, pH 7.5, 10 mM 2-mercaptoethanol, 1 mg/ml bovine serum albumin).

Reactions were allowed to proceed for 15 minutes at 20° C., then stopped by washing the magnetically immobilized beads three times with 500 μl TNET. Beads were resuspended in final volume of 25 μl TNET prior to the detection assays.[0105]

Incorporation of labelled dideoxynucleotides by the primer extension reaction was assayed two different ways: gel electrophoresis followed by autoradiography, and direct autoradiographic analysis of labelled DNA.[0106]

1. Gel electrophoresis followed by autoradiography ([0107]³⁵S-labelled material only). Samples of washed, bead-bound DNA were heated at 94° C. for 5 minutes in 10 μl of formamide loading buffer (80% formamide, 10 mM Tris-HCl, pH 8, 1 mM EDTA, 0.02% bromphenol blue) to denature the DNA and release the labelled primer from the primer:template complex. Samples were analyzed by electrophoresis on 8 or 12.5% polyacrylamide/8 M urea sequencing gels (19:1 acrylamide:bis-acrylamide ratio; 100 mM Tris-HCl, 100 mM borate, 2 mM EDTA, pH 8.3, running buffer; 60 watts constant power). After electrophoresis, gels were either dried down onto filter paper or frozen at −80° C. to prevent diffusion, covered with plastic wrap, and exposed to X-ray film to visualize the labelled DNA by autoradiography (FIG. 4).

2. Direct autoradiographic analysis of labelled DNA. For the analysis of total radioactivity bound to the beads, 10 μl aliquots of the bead suspensions in TNET were spotted directly onto filter paper or nylon membranes. Filters or membranes were dried under an incandescent lamp, covered with plastic wrap, and exposed to X-ray film (FIG. 5).[0108]

Example 5

[0109]


	TGL240:	^5′AGATGATGCTTTTGTGCAAAACAC^3′

	TGL239:	^5′TCAATACCTGAGTCCCGACACCCTG^3′

	TGL308:	^5′AGCCTCAGACCGCGTGGTGCCTGGT^3′

Oligonucleotide TGL240 was synthesized with a primary amino group attached to its 5′ terminus and coupled with biotin as described above. TGL240 (biotinylated) and TGL239 (unbiotinylated) were used to amplify, via the polymerase chain reaction procedure (see “A. General Methods”), a region of DNA comprising a particular genetic locus in samples of mammalian genomic DNA. DNAs from two different individuals, each homozygous for a particular set of linked sequence polymorphisms (the “A” allele and the “B” allele—see FIG. 6), were examined. After the PCR reaction, 2-20 pmoles of duplex PCR DNA was incubated with 100 u[0110]¹of streptavidin-conjugated M-280 Dynabeads (7×10⁸beads/ml) in TNET buffer in order to bind the biotinylated strand to the beads. After binding, the beads were magnetically immobilized and washed three times with 200 μl of TNET, then resuspended in 100 μl of TNET. To remove the non-biotinylated strand, 500 μl of 0.15 N NaOH was added and the suspension incubated for 30 minutes at 20° C. The beads were then magnetically immobilized and washed once with 250 μl of 0.15 N NaOH, three times with 500 μl TNET, and resuspended in 100 μl of TNET.

The detection primer, oligonucleotide TGL308 (FIG. 6), was annealled to the bead-bound PCR-generated template as described above in Example 4. Further washes, extension reactions, and detection assays were also carried out as described in Example 4. A gel autoradiographic analysis of the labelled primer extension products for the two homozygous individuals, ESB164 (“AA” genotype) and EA2014 (“BB” genotype), is shown in FIG. 7. Autoradiographic analyses of total bead-bound radioactivity, or primer-associated radioactivity after NaOH elution, are shown for these same individuals using the filter spotting assay (FIG. 8). For the analysis of primer only, 10 μl of 0.4 N NaOH was added to 10 μl of the bead suspension. After incubation for 10 minutes at room temperature, the beads were immobilized magnetically and the supernatant withdrawn and spotted onto nylon blotting membrane.[0111]

C. PREFERRED EMBODIMENT

A particularly advantageous way to practice the present invention involves obtaining from a convenient source, such as blood, epithelium, hair, or other tissue, samples of DNA or RNA, then amplifying in vitro specific regions of the nucleic acid using the polymerase chain reaction, transcription-based amplification (see Kwoh, et al., Proc. Natl. Acad. Sci. 80:1173 (1989)), etc. Amplification is accomplished using specific primers flanking the region of interest, with one or more of the primers being modified by having an attached affinity group (although in any given reaction only one such primer is modified at a time). A preferred modification is attachment of biotin moieties to the 5′-termini of the primers. A sample (typically, 0.5-5 pmoles) of the amplified DNA is then bound to streptavidin-conjugated magnetic microspheres (e.g., Dynal M-280 “Dynabeads”) via the attached biotin moiety on the amplification primer. The DNA is denatured by adjusting the aqueous suspension containing the microspheres to a sufficiently alkaline pH, and the strand bound to the microspheres via the biotin-streptavidin link is separated from the complementary strand by washing under similar alkaline conditions. To accomplish this, the microspheres are centrifuged or immobilized by the application of a magnetic field. The microsphere-bound strand is then used as a template in the remaining manipulations.[0112]

To the template strand, generated as described above, a specific primer oligonucleotide is bound under high stringency annealing conditions, the sequence of the primer being consistent with unique binding to a site on the template strand immediately adjacent to a known DNA sequence polymorphism. A preferred sequence and mode of binding for the primer ensures that the primer forms a duplex with the template such that the 3′-terminal nucleotide of the primer forms a Watson-Crick basepair with the template nucleotide immediately adjacent to the site of the first nucleotide in the sequence polymorphism, without the duplex overlapping any of the polymorphic sequence to be analyzed. This arrangement causes the nucleotides added via template-directed, DNA polymerase-catalyzed, extension of the primer to be determined unambiguously by the polymorphic nucleotide sequence in the template.[0113]

The above-described primer:template complex is contacted, under conditions of salt, pH, and temperature compatible with template-directed DNA synthesis, with a suitable DNA polymerase and four different chain-terminating nucleotide analogues known to form specific base pairs with the bases in the template. Most likely, but not necessarily, the bases in the template as well as the chain-terminating analogues are based on the common nucleosides: adenosine, cytosine, guanine or inosine, thymidine or uridine. A preferred set of chain-terminating analogues are the four dideoxynucleoside triphosphates, ddATP, ddCTP, ddGTP, and ddTTP, where each of the four ddNTPs has been modified by attachment of a different fluorescent reporter group. These fluorescent tags would have the property of having spectroscopically distinguishable emission spectra, and in no case would the dideoxynucleoside triphosphate modification render the chain-terminating analogue unsuitable for DNA polymerase-catalyzed incorporation onto[0114]primer 3′-termini. The result of DNA polymerase-catalyzed chain extension in such a mixture with such a primer:template complex is the quantitative, specific and unambiguous incorporation of a fluorescent chain-terminating analogue onto the 3′-terminus of the primer, the particular fluorescent nucleotide added being solely dictated by the sequence of the polymorphic nucleotides in the template.

The fluorescently-tagged primer:template complex, still attached to the magnetic microspheres, is then separated from the reaction mix containing the unincorporated nucleotides by, for example, washing the magnetically immobilized beads in a suitable buffer. Additionally, it is desirable in some circumstances to then elute the primer from the immobilized template strand with NaOH, transfer the eluted primer to a separate medium or container, and subsequently determine the identity of the incorporated terminator. The identity of the attached fluorescent group is then assessed by illuminating the modified DNA strand with light, preferably provided by a laser, of a suitable wavelength and intensity and spectrophotometrically analyzing the emission spectrum produced. In general, for a two allele (diploid) system at any given site in the DNA sequence, there are ten possible canonical emission spectra produced, corresponding to the sixteen possible homozygotic and heterozygotic pairings. By suitable matching of the measured spectra to this library of canonical spectra it is possible to identify which chain-terminating nucleotide(s) have been added to the 3′-terminus of the primer and thereby identify the nature of the sequence polymorphism in the template. Spectra produced by multiple allele systems or by alleles present in a ratio other than 1:1 can also be deconvolved by suitable mathematical treatments to identify and estimate the relative ratios of each contributing nucleotide.[0115]

All of the above steps involve chemistries, manipulations, and protocols that have been, or are amenable to being, automated. Thereby, incorporation of the preferred mode of practice of this invention into the operation of a suitably programmed robotic workstation should result in significant cost savings and increases in productivity for virtually any diagnostic procedure that depends on the detection of specific nucleotide sequences or sequence differences in nucleic acids derived from biological samples.[0116]