WO1993014224A1 - Size-calibration dna fragment mixture and method - Google Patents

Abstract

A size-calibration mixture for use in achieving single basepair resolution of polynucleotide fragments fractionated by high-resolution electrophoresis. The mixture includes two or more polynucleotide strand pairs, each pair having one strand which has a selected number of polynucleotide bases and a detectable reporter in its 5'-end region and a complementary strand which lacks the reporter. Also disclosed are methods of producing and using the mixture.

Description

SIZE-CALIBRATION DNA FRAGMENT MIXTJJRE AND METHOD

1. Field of the Invention

The present invention relates to a size calibration mixture for use in calibrating the sizes, i.e., number of bases, of polynucleotides fractionated by high-resolution electrophoresis, and to methods for making and using the mixture.

2. References

Cantor, C. , et al., "Biophysical Chemistry Parts I, II and III," .H. Freeman and Co., N.Y. (1980).

Carrano, A.V. , et al, Genomics, 4:129 (1989). Connell, C. , et al., BioTechnigues, 5:342-348 (1987). Fung, S., et al., U.S. Patent 4,855,225 (1989).

Litt, M. , et al., Am. J. Human Genetics, 44:397-401 (1989) .

Maniatis, T. , et. al., "Molecular Cloning: A Labora¬ tory Manual," Cold Spring Harbor Labs (1982). Maxam, A.M., et al., Meth Enzymol, 65:499 (1980).

Sanger, F. , et al., Proc Nat Acad Sci, USA, 74:5363 (1977) .

Smith, L.M., et al., Nature, 321:674 (1986). Weber, J.L. , et al., Am. J. Human Genetics, 44:388-396 (1989) .

3. Background of the Invention

Methods for size fractionating mixtures of single- strand or double-strand (duplex) nucleic acids are crucial to a variety of analytical and preparative techniques in

SUBSTITUTESHEET biochemistry. Such methods generally employ a solid or semi-solid gel matrix for fragment separation. In the case of larger molecular weight fragments, typically greater than about 700 bases, the preferred gel material is aga- rose, where the concentration of the agarose may vary from about 0.3%, for separating fragments in the 5-60 kilobase size range, up to about 2%, for separating fragments in the 100-3,000 basepair range (Maniatis) . Smaller size frag¬ ments, less than about 1,200 basepairs, are usually separa- ted in polyacrylamide gel. The concentration of acrylamide polymer can range from about 3.5%, for separating fragments in the larger basepair range, up to about 20%, for achie¬ ving separation in the size range 10-100 basepairs.

One important application of DNA size fractionation is restriction-site analysis, in which duplex DNA is digested with selected restriction enzyme(s) , fractionated according to digest fragment size, then analyzed for fragment posi¬ tions. This method is widely used in molecular cloning to determine the number and arrangement of restriction sites in a cloning vector, and to confirm insert location and/or orientation in the vector.

Restriction analysis is an important tool for genetic mapping as well. In one method which has been proposed for human genomic mapping, genomic fragments from a cosmid lib- rary are digested with a selected group of restriction en¬ zymes to generate duplex subfragments predominantly less than about 700 basepairs in size. The digest subfragments are then fractionated by polyacrylamide gel, and the sub- fragment sizes are compared with those of subfragments derived from other library inserts, to identify fragment overlap on the basis of common subfragment sizes. This mapping method can be adapted for subfragment detection by

-SUBSTITUTESHEET fluorescent labels, allowing rapid gel fractionation analy¬ sis (e.g., Carrano) .

The success of the above mapping method depends in part on accurate size determination of the fractionated subfragments. In particular, it is desirable to determine fragment sizes to a resolution of one nucleotide base.

In a related application, the discovery of linkages between a number of genetic diseases and restriction frag¬ ment length polymorphisms in humans has provided a tool for screening individuals for these genetic diseases. More recently, genomic fragments containing a variable number of tandem repeat (VNTR) sequences have been discovered, and it is likely that individual-specific VΝTR regions may be used as linkage markers for certain genetic diseases. Several VΝTR loci which have been reported to date contain a 2- nucleotide repeat sequence (Weber; Litt) . Thus, distin¬ guishing VΝTRs which differ by a single repeat sequence may require one- or two-base size resolution by eleσtrophore- sis. Rapid DΝA sequencing methods currently in use rely on size fractionation of single-stranded DΝA fragments on polyacrylamide gel. Both enzymatic sequencing techniques, in which random-termination fragments are generated enzyma- tically in the presence of dideoxynucleotides (Sanger) , and chemical methods in which random-termination fragments are generated chemically (Maxam) , rely on discrimination of fragment size differences of one base. However, at larger fragment sizes, it becomes progressively more difficult to match and compare fragment peaks, and the precise base nu - ber of any large peak becomes more uncertain.

In the various high-resolution DΝA fractionation me¬ thods described above, it is useful to include known-size reference fragments in the fractionated material, for cali-

SUBSTITUTESHEET brating fragment sizes on the gel. Typically, the size standards are generated by restriction digestion of plas- mids with known restriction-fragment sizes, where the frag¬ ments are pre-labeled with a selected reporter, such as a radioactive or a fluorescent reporter.

One limitation of this type of size standard is that both strands of the digest fragment are labeled and there¬ fore detectable in the gel. Typically, the opposite la¬ beled strands have slightly different migration rates on the gel, giving a broadened or even double band at the reference fragment which makes accurate size determination more difficult. In addition, the sizes of the fragments which can be produced by this approach are limited to the relative positions of the restriction sequences in the digested fragment.

4. Summary of the Invention

It is one general object of the invention to provide a size-calibration fragment mixture for use in high-resolu- tion electrophoretic fractionation of polynucleotide frag¬ ments.

It is another object of the invention to provide a method of preparing and using such a fragment mixture.

The invention includes, in one aspect, a polynucleo- tide strand-pair mixture containing two or more polynucle¬ otide strand pairs, where each pair has one strand having a selected number of polynucleotide bases and a detectable reporter in its 5'-end region, and a complementary strand which lacks this reporter. The different strand pairs in the mixture have different selected sizes (numbers of basepairs) .

For use in size calibration over a wide range of analyte fragment sizes, the number of bases in the repor-

SUBSTITUTESHEET ter-containing strands in the strand pairs preferably dif¬ fer from one another by a multiple of a selected number of bases, typically 10-100 bases. Where the analyte fragments are fractionated by high-resolution polyacrylamide gel electrophoresis in the presence of a denaturant, such as urea, the fragments preferably have sizes less than about 1,200 bases.

In one preferred embodiment, for use in quantitating the amount of analyte fragments, the concentration of each strand pair fragment in the mixture is known, and one strand of each fragment is labeled with a fluorescence reporter. The concentration of analyte polynucleotides can be estimated, after electrophoretic separation, by compar¬ ing the total fluorescence emission signal of analyte peaks with those of adjacent-size calibration strands.

The invention also includes a method of producing a polynucleotide fragment mixture for use in size-calibrating polynucleotide analytes fractionated by electrophoresis. In practicing the method, a segment of template DNA having a selected number of basepairs is identified, and the seg¬ ment-containing template is reacted with reporter-labeled and unlabeled single-stranded primers which are complemen¬ tary to 3'-end regions of opposite strands of the segment, when the segment is in double-stranded form. The reaction is carried out under conditions which produce multiple rounds of primer-mediated duplex DNA replication, producing a duplex DNA fragment having the basepair composition of the segment, and one reporter-labeled strand only.

The steps of identifying a selected-size segment, and reacting the segment-containing DNA with reporter-labeled and unlabeled single-stranded primers are repeated with different, selected-size segments until a desired mixture of different-size fragments in the mixture is produced.

SUBSTITUTESHEET In one embodiment, each subsequently-identified seg¬ ment contains the same basepair sequence as the previously identified segment, plus an end sequence of at least about 10 basepairs. In this embodiment, the reporter-labeled primer may be homologous to the common-sequence end of the sequences, and the unlabeled primer, homologous to the 5'- end sequence of the opposite strand of each successively larger sequence. The fragments may be produced simul¬ taneously in a reaction mixture containing the common- sequence primer and each of the individual-sequence pri¬ mers. Alternatively, the fragments may be produced indi¬ vidually and combined to produce the desired mixture. In the latter method, each different-size fragment can be added in the mixture to a known final concentration. In another embodiment, an excess of one or more labeled strands is produced, relative to the corresponding unlabeled complementary strands, through the use of a limiting amount of the unlabeled primer, relative to the amount of the corresponding labeled primer, in primer- mediated replication of the selected size segments.

In another embodiment, the two primers required for primer-mediated replication of the selected size segments may be labeled with different, distinguishable reporter groups, such as different fluorescent reporters (Fung et al.)_r giving strand pairs in which the two complementary strands are labeled with the different, distinguishable reporters, permitting study of their relative electropho¬ retic mobilities.

Also disclosed are improvements in genomic mapping and sequencing methods employing the strand-pair mixture of the invention.

These and other objects and features of the invention will become more fully apparent when the following detailed

SUBSTITUTESHEET description of the invention is read in conjunction with the accompanying drawings.

Brief Description of the Drawings Figure 1 illustrates how the method of the invention may be practiced to generate a series of strand pair fragments having a common downstream end delimited by a labeled forward primer, and upstream end regions whose lengths increase in increments of 50 basepairs, delimited by reverse primers;

Figure 2 shows the polynucleotide sequence of the sense (+) strand of M13mpl8, where the sequence comple¬ mentary to the 5'-end labeled primer is double underlined and the sequences of the reverse primers are underlined; Figure 3 illustrates the primer-initiated reactions which occur in forming the strand-pair fragments of the invention;

Figure 4 is an electrop^'herogram of a strand-pair mixture having labeled strands which differ from one another by 50 bases, fractionated on a polyacrylamide gel under denaturing conditions;

Figure 5 shows a portion of a plot of migration time as a function of size for the fragment separation in Figure

4; Figure 6 shows an enlarged portion of an electrophero- gra , illustrating how the concentration of analyte peaks (indicated by dashed lines) can be quantitated by reference to adjacent labeled strands of the strand-pair mixture (solid lines) ; Figure 7 shows an electropherogram of digest subfrag¬ ments produced by digesting phage lambda DNA with each of three combinations of restriction enzymes (upper panels) ,

SUBSTITUTESHEET with strand-pair fragments (lower panel) , fractionated on a polyacrylamide gel under denaturing conditions;

Figures 8A-8D show electropherograms of ddATP, ddTTP, ddGTP and ddCTP sequencing fragment mixtures, respectively, fractionated on polyacrylamide gels under denaturing condi¬ tions, with size calibration fragments indicated by the arrows;

Figure 9 shows an electropherogram of double-stranded fluorescence-labeled restriction fragments and double- stranded strand-pair fragments labeled with a second fluo¬ rescence label, and fractionated by agarose gel electro¬ phoresis under non-denaturing conditions; and

Figure 10 shows electropherograms of the same digest subfragments (upper panels) and strand pair fragments (lower panel) as were used in Figure 7, separated in this case on a 2% agarose gel, under non-denaturing conditions.

Detailed Description of the Invention I. Strand-Pair Mixture The strand pair mixture of the invention includes a plurality of selected-size polynucleotide strand pair frag¬ ments, such as fragments 20, 22, and 24, shown at the top in Figure 1. These fragments are also designated F_J0, F₁₀₀, and F_1S0, respectively, indicating fragment lengths of 50, 100, and 150 basepairs, respectively. Fragment 30, which is representative, includes a polynucleotide strand 34 which has a detectable reporter 36 (indicated at R) in its 5'-end region, and an opposite, complementary polynucle¬ otide strand 32 which lacks this reporter. The reporter-labeled strand has a known, selected number of bases, and the opposite strand preferably has the same number of bases, i.e., the strand-pair fragment is base-paired along its entire length. It is known, however.

SUBSTITUTESHEET that Taq DNA polymerase may add one unpaired base (dA) at the 3' end of a DNA chain.

A. Size-Calibration Mixture In one embodiment, the strand-pair mixture is designed for use in size calibration for DNA (or RNA) fractionation by electrophoresis, over a selected nucleic acid size range. The size range typically extends from about 50 bases up to as high as 1,200 bases for polyacrylamide elec- trophoresis under denaturing conditions, and from about 50 bases up to as high as 2,000 or more for agarose electro¬ phoresis of duplex nucleic acid fragments under non-denatu¬ ring conditions.

Preferably the strand-pair fragments in this embodi- ment differ from one another in size by a selected multiple of N bases, where N is typically between about 10-100 bases, e.g., 50 bases. The strand-pair fragments shown at the top in Figure 1 have sizes between 50-300 basepairs, in increments of 50 basepairs. Figures 1-3 illustrate how the size-calibration strand-pair mixture can be prepared in accordance with the method of the invention. Shown at the bottom in Figure 1 is a template 38 used in the method. For illustrative pur¬ poses, the template shown here is the sense (+) strand of an M13mpl8 plasmid. The strand serves as the template strand in the method. This strand has a region 40 of known sequence, which is shown in Figure 2, extending from a 5' end to a 3' end in the sense indicated by the arrows.

Region 40, in turn, is divided into a plurality of partially overlapping segments, such as segments 42, 44, 46, which have a selected number of bases (or basepairs if both strands of a duplex template are considered) . Each successively larger segment includes the bases in the pre-

SUBSTITUTESHEET vious segment plus an additional 50 bases in the upstream direction, i.e., toward the 5' end of region 40 in Figure 1. Thus, each segment includes a common 3' end sequence and a unique 5' end sequence. A common-sequence primer P₀ is complementary to the 5' end region of each template strand sequence. Primer P₀ terminates, at its 5' end, in a detectable reporter 36, such as the ROX (6-carboxy rhodamine X) fluorescence reporter which has been described (Fung) . A variety of detectable fluorescent, colorimetric, or radioactive repor¬ ters may be employed as the detectable reporter in the primer. Alternatively, the reporter may be a ligand, such as biotin, which is capable of binding by specific, high- affinity binding to an antiligand, such as streptavidin. The particular P₀ primer for the M13 sense sequence shown in Figure 2 has the sequence: d(5'-R0X-TGT AAA ACG ACG GCC AGT-3') , where ROX is coupled via a linker to the 5'-phosphate on the terminal thymidine as shown.

A second, reverse primer has the sequence of the 5'- end region of each template-strand segment, i.e., in the template illustrated in Figures 1 and 2, the first segment 42 is bounded by P₀ and a second primer P_J0, which has the sequence shown by underlining for the P₅₀ primer. The specific P₅₀ primer shown in the figure has the form: d(5'- CTA GAG TCG ACC TGC AGG CAT GCA AGC TTG-3') . The two primers P₀ and P₅₀ are used in producing fragment F₅₀, a 50- basepair fragment, according to procedures described below. Segment 44 in the template is a 100-base segment ex¬ tending upstream from the 3'-end of template region 40 in Figure 1. This segment is bounded by the two primers P₀ and P₁₀₀, the latter of which has the sequence shown by under¬ lining in Figure 2. As above, P₁₀₀ is unlabeled, and the two primers P₀ and P₁₀₀ are complementary to the 3'-end

SUBSTITUTESHEET regions of opposite strands of segment 22, when the segment is in duplex form. The two primers are used in producing strand-pair fragment F₁₀₀, a 100-basepair fragment, as detailed below. Similarly, each larger segment in the template, such as segment 46, is bounded by the primer P₀ and an upstream primer, such as primer P_1J0, whose sequence bounds the up¬ stream end of that segment. The sequences of these addi¬ tional primers -?_1S0-^'P_lOQO are underlined in Figure 2. As shown in Figure 1, each of these additional primers is used in conjunction with the common, reporter-labeled primer P₀ to produce the strand-pair fragments F₁₅₀ to F₁₀₀₀.

Figure 3 illustrates the reaction steps used in gene¬ rating selected-size strand-pair fragments, in accordance with the method of the invention. Shown at A in this fi¬ gure is a portion of the single-stranded M13mpl8 template 38, containing a selected-size segment, such as segment 48 (a 200 basepair segment) , from the template. The 3' and 5' ends of the size-selected segment are shown, along with the P₀ primer, which is complementary to the 3'-end of the selected segment. The template and selected pair of primers are employed in a fragment amplification method based on the PCR (polymerase chain reaction) amplification method disclosed in U.S. Patent No. 4,683,202. Details of an exemplary reaction method are given in Example 1. The general reaction steps involve first mixing the template with the reporter-labeled primer and unlabeled primer which bound the selected segment, heating the DNA components to denature the template strands (necessary in the case of a duplex template) , and cooling the components under condi¬ tions which favor primer annealing to the template strand(s) , as shown at A in Figure 3.

SUBSTITUTESHEET Additional reaction components, which may be added before or after primer annealing to the template strands, include all four deoxynucleotide,5'-triphosphates (dNTPs) and a thermal-stable DNA polymerase, such as the polymerase obtained from Thermus aguaticus. The polymerase operates, in the presence of the dNTPs, to extend the primer in the 5' to 3' direction by the addition of bases complementary to the template strand, forming duplex fragments which include the selected template fragment plus a strand 54 extending in a 5'-3' direction from the P₀ sequence as shown at B in the figure. As seen, the newly formed strand 54 is reporter labeled.

Heating to denature the newly formed duplex frees the labeled product strand 54, which can then anneal with the unlabeled primer P₂₀₀, as shown at C in Figure 3. Polymer¬ ization gives the new duplex shown at D, with an unlabeled strand 56 corresponding to the size-selected segment 48 of template 38. Denaturation of this duplex and annealing the unlabeled strand with labeled primer P₀, as shown at E in Figure 3, then permits polymerization to give the duplex shown at F, consisting of a newly synthesized, labeled product strand 58 and the unlabeled template strand 56 corresponding to the selected segment 48 of template 38, both strands having the selected number of basepairs. The above denaturing, annealing, and polymerizing steps are repeated, with each of the six strands shown at B, D and F being annealed with one of the two primers in the reaction mixture. It will be appreciated that the reporter-labeled strand at B will be annealed at its 3' end with the unlabeled primer, forming, on polymerization, the duplex shown at D, while the unlabeled strand at B will anneal to the labeled primer to repeat the process shown at A and B. Similarly, the primer-induced, unlabeled strand

SUBSTITUTESHEET at D will be annealed at its 3' end with the reporter- labeled primer, again forming on polymerization the duplex shown at F, while the labeled strand at D will be annealed at its 3' end by the unlabeled primer, again forming on polymerization the duplex shown at D. The duplex shown at F contains labeled and unlabeled strands of equal length; on denaturation, the unlabeled strand will anneal to the labeled primer (as at E) , to give on polymerization the same duplex shown at F; the labeled strand will anneal to the unlabeled primer, also giving, after polymerization, another duplex identical to that shown at F.

The denaturing, annealing, and polymerization cycle is repeated N times, forming approximately 2^N strand-pair fragments of the type just described having the basepair sequence of the selected template segment, and a single reporter-labeled strand. After terminating the reaction, the selected-size fragments can be freed of primer and other unwanted reaction components by centrifugal filtra¬ tion using commercial membrane devices such as Centricori™ (Amicon) or Centri/Por (Spectrum Medical Industries) , wash¬ ing with Tris-EDTA buffer (pH 8) . The purified strand-pair components are precipitated with cold ethanol and resus- pended in a suitable buffer to a desired DNA concentration. Each selected-size strand-pair fragment, such as the F^ fragment, is formed by the above method, using the template and appropriate set of primers, such as the labeled P₀ primer and unlabeled P₁₀₀ primer described above. The final strand-pair mixture is formed by combining the individual strand-pair fragments. Thus, for example, a strand-pair mixture for size calibration in the 50-600 basepair range would be prepared by making each of the individual fragments F₅₀, F₁₀₀, ....F^, and combining these into a single mixture.

SUBSTITUTESHEET In practice, F₅₀ is more conveniently prepared by auto¬ mated synthesis of the two strands, e.g. using an Applied Biosystems Model 38OB DNA synthesizer. The strand to be labeled with the reporter is synthesized with a linker group at its 5' end, to which the reporter group is subse¬ quently attached. After purification by conventional me¬ thods, the two strands are combined in a one-to-one ratio to give a duplex identical to that produced by primer- mediated synthesis using a size-selected template. Automa- ted synthesis of F₁₀₀ is also practical. These synthetic fragments are resuspended at the desired concentration and combined with the other fragments made as described above.

A size-calibration standard was prepared as described above, with strand-pair fragments having selected size between 50-1,000 basepairs, at 50 basepair intervals. The fragments were denatured and fractionated by polyacrylamide gel under denaturing conditions, as detailed in Example 2. Fluorescence emission of the bands was measured at a posi¬ tion about 25 cm from the initial sample position, with the results shown in Figure 4. As seen, all of the reporter- labeled strands except two migrated as single bands with well-defined peak positions. The two exceptions — at 350 and 400 base sizes — gave double peaks evidencing frag¬ ments with one basepair size differences. The single peaks allows the migration time of any given size fragment to be unambiguously determined. When the peak migration times in Figure 4 are plotted against peak size, the plot shown in Figure 5 is obtained. The non-linear behavior is due to variation in migration rate as a function of strand size. To determine the size of any unknown analyte fragment which is co-electrophoresed with the size-calibration frag¬ ments, the migration time of the analyte peak is measured and this peak time is plotted on the Figure 5 time/size

SUBSTITUTE SHEET plot. Thus, for example, an analyte peak with measured migration time of 364.5 minutes is determined to have a 288 base size. It will be appreciated that the frequent and regularly spaced calibration peaks, in conjunction with the sharp peaks achieved with a single-labeled strand, allows unambiguous determination of single-strand analyte peaks based on their migration rates in a denaturing gel. Appli¬ cations of the strand-pair mixture of the invention for fragment size determinations are described in Section II below.

As an alternative method for producing the above de¬ scribed size-calibration mixture, a single template, such as template 38, is mixed initially with all of the primers used in generating the desired selected-size fragments. For example, in the case above, the M13 template is mixed with the P₀, P₅₀, P₁₀₀, •••P₁₀₀o primers in the presence of the dNTPS and thermal-stable polymerase, and the mixture taken through several cycles of denaturation, annealing, and polymerization. All of the strand-pair fragments are thus formed simultaneously in one primer-initiated amplification reaction. The method has the advantage of simplicity; however, the relative concentrations of the different-size fragments formed in the reaction may be difficult to con¬ trol, and this may preclude the use of fragment mixture for quantitating electrophoresis peaks, as described below.

It will also be appreciated that the different-size strand-pair fragments in the mixture of the invention may be formed using a different template for each fragment. This method would require an independent pair of a labeled and unlabeled primers for each different template sequence.

SUBSTITUTESHEET B. Size Calibration Mixture for DNA Quantitation

In one preferred embodiment of the invention, for use in quantitating the amounts of individual analyte peaks on electrophoresis, each strand-pair fragment in the mixture is present at a known concentration. Typically in such a mixture, all of the strand-pair fragments are present at the same concentration. The mixture is prepared by combin¬ ing known amounts of each of the individual strand-pair fragments prepared as above. In one preferred embodiment, each of the strand-pair fragments in the mixture has the same concentration, in a concentration range between about 1 and 10 _/yg/ml, corresponding to about 10 pmol/mL.

The use of the above strand-pair mixture for quanti¬ tating the amount of an unknown analyte by gel electro- phoretic fractionation is illustrated in Figure 6, which shows an idealized portion of an electropherogram contain¬ ing size calibration peaks for 250 and 300 bases, and an intermediate analyte peak, indicated at 70. The amount of each of the two size-calibration peaks applied to the gel is known from the known concentration of each fragment in the strand-pair mixture.

The total fluorescence emission associated with each band may be more accurately determined by sampling fluores¬ cence emission across the length and width of the band, and using the sampled values to construct a three-dimensional fluorescence emission surface. Total fluorescence is then calculated from the volume defined by the emission surface, and the calculated volume is plotted as a function of elec¬ trophoretic migration rate. A variation in measured fluo- rescence emission with migration time will be observed, due to the change in peak shape as a function of fragment size, as illustrated in Figure 4, and this variation can be stan-

SUBSTITUTESHEET dardized from the known amount of material in each band. It will be appreciated that the frequent, regularly spaced and sharp strand peaks provided by the strand-pair mixture of the invention, combined with the known DNA amount associated with each peak, allows the relationship between peak fluorescence emission, DNA amount and migration distance on a denaturing gel to be accurately calibrated.

From the fluorescence emission plot just described, the amount of DNA in any analyte peak can be determined from the measured fluorescence emission surface and migra¬ tion rate of that peak. If the fluorescence reporter attached to the analyte strand is different from that in the size-calibration strands, an additional correction for fluorescence efficiencies will also be required. The correction factor can be determined readily from sample peaks containing known amounts of each reporter.

C. Small-Increment Fragment Mixture

In another embodiment, the strand-pair mixture of the invention includes two or more strand-pair fragments having defined sizes which differ from one another by a small number of bases, typically 1-10 bases. The mixtures are designed for use, for example, in detecting small base- number mutations in a genomic fragments having defined- sequence ends, or in detecting polymorphisms in genomic sequences containing variable numbers of tandem repeat sequences (VNTRs) . In each of these applications, the strand-pair fragments are tailored in size to provide strand sizes which span the size of the analyte(s) of interest.

For example, a strand-pair mixture for use in detec¬ ting a one- or two-base mutation in a genomic segment would contain strand pairs with sizes corresponding to the normal

SUBSTITUTESHEET (non-mutated) genomic segment + one or two bases, respec¬ tively. As another example, a strand pair mixture for use in detecting polymorphisms of variable numbers of tandem repeats (VNTRs) would contain size markers corresponding to upper and lower expected sizes of genomic segments having minimum and maximum expected numbers of VNTRs, respective¬ ly.

Such strand-pair mixtures are prepared as described above, preferably using a single template with substanti- ally overlapping segments which differ in base number from one another by the selected number(s) of bases, e.g., 1-30 bases.

II. Electrophoretic Methods Involving Single-Strand DNA The strand-pair mixture described above is designed for use in a variety of electrophoretic methods in which DNA analyte fragments are fractionated in single-stranded form by electrophoresis. The mixture of the invention allows analyte sizes to be determined exactly (to a one- base accuracy) , and optionally allows the amount of DNA in the bands to be quantitated.

The electrophoretic methods are carried out in any electrophoretic system capable of resolving analyte frag¬ ments to a resolution of one base, i.e., of separating fragments which differ from one another in size by a single base. One preferred system is a polyacrylamide slab gel in a denaturant buffer, as detailed below. Single base resolution has also been achieved by capillary electro¬ phoresis using either a polyacrylamide matrix, or a solu- tion of a hydroxylated polymer as the separation medium, as described in co-owned U.S. Patent application for "Nucleic Acid Fractionation by Counter-Migration Capillary Electro¬ phoresis," Ser. No. 390,631 filed August 7, 1989.

SUBSTITUTESHEET A. Genomic Mapping

As a general strategy for gene mapping a large genomic region, such as an entire chromosome, the region is first fragmented, and the fragments are cloned to form a genomic library. The library clones are then examined to identify regions of sequence overlap between cloned inserts, with the aim of identifying and ordering clones with overlapping sequences. Once several clones with consecutively over¬ lapping regions are identified, the basepair sequences of the clones can be determined, and the sequences can then be matched to produce extended basepair sequences.

One method which has been used for identifying regions of sequence overlap in cloned inserts uses restriction ana¬ lysis to characterize the restriction fragment sizes in a digested insert, then matches the digest fragment sizes a- mong insert fragments to identify the fragments with pos¬ sible sequence, overlap. Carrano et al have described such a method in which each cloned insert is digested with dif¬ ferent pairs of restriction enzymes, generating two or more sets of restriction fragments for each insert. Each frag¬ ment set is labeled with a distinctive fluorescence repor¬ ter, and the fragment sets are run together in a common electrophoresis lane, as described in Example 3, in a poly¬ acrylamide gel under denaturing conditions. The resulting peaks are monitored by fluorescence emission and signal processing techniques which allow each set of fragments to be individually monitored. A typical electropherogram generated by this method is shown in Fi¬ gure 7, where the three upper lines represent electrophero- grams of Hindlll/Haelll, BamHI/Haelll, and EcoRI/Haelll fragments of phage lambda DNA as indicated, and the lower trace represents size-calibration fragment peaks.

SUBSTITUTESHEET The success of the fragment matching method for geno¬ mic mapping depends on the ability to accurately determine the sizes of the digest fragments by electrophoretic frac¬ tionation. In particular, it is important to determine the exact size (number of bases) of each fragment, for accurate size matching among digest fragments from different clonal inserts.

In the present invention, the ability to determine fragment size to an accuracy of one base is accomplished by co-electrophoresing the digest fragments with the size- calibration strand-pair mixture described in Section I, where the reporter in the labeled size-calibration strands is preferably distinguishable from the reporter(s) used for labeling the digest fragments. In the Figure 7 electrophe- rogram, the size-calibration strands and each of the three restriction digests are labeled with a separate fluores¬ cence reporter, which allows the four samples to be sepa¬ rately detected in a single electropherogram.

The sizes of the digest fragments in the electrophero- gram are determined, as described above, by plotting the known-size strands as a function of migration time, and using this plot to determine analyte fragment size.

B. DNA Sequencing Accurate size determination of single-stranded DNA is also required in DNA sequencing methods. In a standard dideoxy sequencing method, chain elongation in the presence of a selected dideoxynucleoside triphosphate (ddNTP) is used to generate random-size fragments which terminate at the selected ddNTP, where the primer employed in the elongation reaction is fluorescence labeled, to label the fragments. Each of the four ddNTP reaction mixtures is fractionated by electrophoresis and the fragment lengths

SUBSTITUTESHEET are determined to construct the unknown sequence (Sanger) . The four reaction mixtures may be co-fractionated in a single lane, where each of the four reaction mixtures are labeled with a distinguishable fluorescence probe (Smith) . Alternatively, the four reaction mixtures may be fractiona¬ ted in parallel, in side-by-side lanes.

Figures 8A-8D illustrate fragment electropherograms for each of the four ddNTPs in a typical dideoxy sequencing method. The sequence is "read" conventionally by identify- ing and assigning a fragment size to each peak in the four reactions mixtures, and counting the peak fragment sizes in numerical succession. One limitation of this method heret¬ ofore has been the uncertainty in fragment size, particu¬ larly as the fragment size extends beyond about 400 bases, leading to loss or misidentification of bases in the sequence.

In the present method, the fragments produced by ran¬ dom termination in the above sequencing method are fractio¬ nated by gel electrophoresis in the presence of the size- calibration, strand-pair mixture of the invention. In one embodiment, illustrated in Figures 8A-8D, the four ddNTP reaction mixtures are separately fractionated, each in the presence of the strand-pair mixture. In this embodiment, the random-size termination fragments and the strand-pair fragments are labeled with distinguishable fluorescence reporters, as described in Example 4. In Figures 8A-8D, the traces represent the measured fluorescence emission from the random-termination strands and the reporter- labeled size-calibration strands (arrows) . For each of the four electropherograms, the size-cali¬ bration peaks are plotted as a function of strand size, and this plot is then used to determine fragment size for each of the random-termination strands in that electropherogram.

SUBSTITUTE SHEET As described above, this method provides precise size determination up to 700 bases or more, allowing accurate sequence determination for fragments up to 700 bases or greater in size.

III. Applications to Double-Strand DNA Electrophoresis

In another general method, the strand-pair mixture of the invention is employed as a size-calibration standard for double-stranded DNA analytes fractionated by electro- phoresis. The electrophoresis is typically carried out by agarose gel electrophoresis in a slab gel, although separa¬ tion may be carried out in other electrophoretic systems capable of resolving double-stranded DNA fragments with high resolution, generally within one or two basepairs. One application of the duplex-fractionation method is for detecting relatively small size changes in restric¬ tion-enzyme digest fragment, where a duplex fragment is digested partially or completely with a selected restric¬ tion enzyme or set of enzymes, and the resulting digest fragments are characterized on the basis of size. This application may involve, for example, detecting small bases number mutations in a restriction fragment, or small size changes related to restriction fragment length polymorphisms (RFLPs) , or polymorphisms involving variable numbers of tandem repeats (VNTRs) , as discussed above.

Another application is in genomic mapping, where a number of genomic inserts are characterized on the basis of restriction-fragment sizes, to map regions of overlap among the inserts. This approach is similar to the one discussed in Section II, except that the digest fragments are fractionated in double-stranded form. Figure 10 shows a portion of an electropherogram of the same phage lambda restriction digest fragments as in Figure 7, here fractio-

SUBSTITUTESHEET nated in double-stranded form on a 2% agarose horizontal slab gel. The fragments were digested and labeled with a fluorescence reporter, and mixed with a double-strand size calibration mixture prepared as in Section II and labeled with a second fluorescence reporter, following the proce¬ dures detailed in Example 5. The size-calibration frag¬ ments with sizes between 100 and 1000 basepairs are indi¬ cated in the figure.

To determine the sizes of the analyte fragments, the peak positions of the size-calibration fragments are plot¬ ted as a function of fragment size, and the analyte sizes are determined from their peak migration rates, using this plot. As in the analogous application involving single- stranded fragments, the size-calibration mixture of the in- vention increases the accuracy of fragment size determina¬ tion by providing a series of uniformly spaced size markers whose frequent spacing throughout a size region of interest allows an accurate plot of peak migration time as a func¬ tion of fragment size. It is recognized that some accuracy in size determination will be sacrificed in fragment sepa¬ ration by agarose electrophoresis under non-denaturing conditions, due to peak broadening, seen, for example, by comparing peak widths in Figure 10 with those in Figure 7. However, duplex-fractionation method is generally applic- able to fragment separation of larger fragments, e.g., above about 700-1,000 bases, which may not be well resolved by polyacrylamide gel under denaturing conditions.

From the foregoing, it can be seen how the various objects and features of the present invention are met. The strand-pair mixture provides, in one aspect, a. series of uniformly spaced size markers which can be used for con¬ structing a plot of migration distance vs fragment size, allowing any analyte to be sized accurately with reference

SUBSTITUTESHEET to two closely spaced standard markers. Moreover, the spacing between the standard markers can be easily selected for greater accuracy, if needed. That is, size-dependent migration effects can refined with increasing numbers of strand-pair sizes.

The strand-pair fragments of the invention contain a single reporter-labeled strand, and thus give sharp, single-peak bands when used in a denaturing electrophoresis system. In one preferred embodiment, the individual-size fragments are present in known amounts, allowing for con¬ struction of a plot of peak reporter signal as a function of migration distance. From this plot, quantitative deter¬ mination of analyte fragments, based on total reporter sig¬ nal and migration distance, can be made. The strand-pair fragment mixture can be conveniently prepared, in accordance with the method of the invention, from a single template, a single reporter labeled primer, and a group of second (reverse) primers whose sequences are selected to produce desired fragment sizes. The following examples illustrate methods for prepa¬ ring and using the strand pair fragment mixture of the present invention. The examples are intended to illus¬ trate, but in no way to limit the invention.

Example 1

Preparing a Size Ladder Strand-Pair Mixture A. Preparation of Single-Strand Primers

Size-fragment primers were prepared by standard auto¬ mated DNA synthesis methods, using an Applied Biosystems Model 38OB DNA synthesizer. The unlabeled primers were purified using OPC™ cartridges (Applied Biosystems) , fol-

SUBSTITUTESHEET lowed by drying and resuspension at the desired concentra¬ tion (20_/M) .

The P₀ primer was synthesized with a linker group (Ami- nolink 2™) coupled to the 5' T residue as the final step of automated synthesis. After deprotection and evaporation of the ammonia solution, the ROX fluorescent reporter was added by reaction of the linker-modified primer with ROX- NHS ester (Applied Biosystems) in sodium carbonate/sodium bicarbonate buffer (250 mM, pH 9) overnight at room temperature. Excess fluorescent dye was removed by Sephadex™ G-25 (Pharmacia) gel filtration, and preparative high-performance liquid chromatography (prep HPLC) was used to remove other impurities, including any primer that failed to couple with the fluorescent reporter. Prep HPLC was carried out on an Applied Biosystems 15OA DNA separa¬ tion system, using an Aquapore RP-300 column (10 x 250 mm) , eluting with 0.1 M triethylammonium acetate and a gradient of 5 to 35% acetonitrile over 30 minutes. The flow rate was 4 ml/min. The desired fraction was collected, and its purity confirmed by HPLC. The purified, dye-labeled primer was concentrated and resuspended at 10 μϋ final concentra¬ tion.

The labeled and unlabeled strands of F_J0 and F₁₀₀ were prepared analogously, as explained above. The P₀ and P₅₀- P₁₀₀₀ primer sequences are shown in Table 1 below.

SUBSTITUTESHEET Table 1

Rev

Primers Sequence (5 ^f - 3 CDU/unA

ROX-TGT AAA ACG ACG GCC AGT 3 ^r50 CTA GAG TCG ACC TGC AGG CAT GCA AGC TTG 33

-^100 AAA CAG CTA GAG TCG ACC TCG AGG CAT GCA AGC TTG 4)

^?150 CCG GCT CGT ATG TTG TGT GGA ATT GTG AGC 33

^?200 ATG TGA GTT AGC TCA CTC ATT AGG CAC CCC 35

?250 GCA CGA CAG GTT TCC CGA CTG GAA AGC GGG 35

^OO ATA CGC AAA CCG CCT CTC CCC GCG CGT TGG ϋ

"^SO TCA GCT GTT GCC CGT CTC ACT GGT GAA AAG 33

^•'-too AGC GTG GAC CGC TTG CTG CAA CTC TCT CAG B

^■^SO CCG ATT TCG GAA CCA CCA TCA AAC AGG ATT 3 ^?500 CAA CAC TCA ACC CTA TCT CGG GCT ATT CTT _Q ^?550 GAC GTT GGA GTC CAC GTT CTT TAA TAG TCG 3. ?600 GAT GGT TCA CGR AGT GGG CCA TCG CCC TGA 39 -^>650 TCC GAT TTA GTG CTT TAC GGC ACC TCG ACC 3L ^■*700 GTT CGC CGG CTT TCC CCG TCA AGC TCT AAA 21 P750 AGC GCC CTA GCG CCC GCT CCT TTC GCT TTC 25 ^•*800 TAA GCG CGG CGG GTG TGG TGG TTA CGC GCA 35 ^?850 TAT ACG TGC TCG TCA AAG CAA CCA TAG TAC 3 ^*900 TAA TCG GCC TCC TGT TTA GCT CCC GCT CGT Z ^?950 TAA AAA CAC TTC TCA AGA TTC TGG CGT ACC 3D ^•"lOOO GTT AAT TTG CGT GAT GGA CAG ACT CTT TTA 3J

B. Preparation of Strand-Pair Mixture M13mpl8 (+) single-stranded DNA (1 mg/mL) was obtained from Promega Corporation (Madison, WI) . 10x PCR amplifica¬ tion buffer (Perkin Elmer Cetus) contained 100 mM Tris-HCl pH 8.3, 500 mM KC1, and 0.01% gelatin. A dNTP mix was prepared by combining dATP, dCTP, dGTP, and dTTP at equal concentrations (1.25 mM each).

In a 0.5 mL microcentrifuge were combined water (139 μL) , lOx amplification buffer (20 μl) , 25 mM magnesium chloride solution (16 μL, 2 mM final reaction concentra¬ tion) , forward primer P₀ (5 μL, 50 pmol)., reverse primer P₃₀₀ (2 μL, 40 pmol) , dNTP mix (16 μL, to give 100 μM of each dNTP in the reaction) , M13mpl8 solution (1 μL, 1 μg) , and 5 units (1 μL) of Thermus aguaticus DNA polymerase (Tag

SUBSTITUTESHEET polymerase, Perkin Elmer Cetus) . The mixture was covered with two drops of mineral oil, and the closed tube placed in a preheated thermocycler (Perkin Elmer Cetus DNA Thermal Cycler) at 95°C for 5 minutes (DNA duplex denaturation) . The mixture was then subjected to cycles of heating to 95°C for 1 minute, cooling to 60°C for 30 seconds (primer annealing), and heating to 72°C for 1 minute (primer extension by Tag polymerase) . The replication reaction, each cycle involving successive denaturation, primer annealing, and DNA polymerization steps, was repeated a total of 35 times.

The 300-base duplex fragments produced in the reaction were freed of single stranded primers by dilution into 1.0 mL water and filtration on a Centricon-100 microconcentra- tor (Amicon) , washing 3 times with 2.0 mL of water accord¬ ing to the manufacturer's instructions. The purified DNA was recovered by inverting the microconcentrator and collecting the product in the concentrator cap. The product (recovered in about 60 μL water) was assayed by capillary electrophoresis (Applied Biosystems 270A with

SepraGene™ 500 buffer, monitored by absorption at 260 nm) .

The primer-initiated amplification described above was repeated individually with each of the primers -?_1∞-^'Pι_000'

The final reaction components for each primer were freed of primer material, and concentrated as described above. Each fragment solution was analyzed for purity and relative concentration by capillary electrophoresis as described above. A fragment mixture was prepared by combining aliquots of each fragment solution in proportion to their relative concentrations, to a final concentration of approximately 1-10 μg/mL for each size fragment. Each fragment in the mixture was characterized by (a) a known number of bases between 50-1000, in increments of 50 bases.

SUBSTITUTE SHEET and (b) one strand which has a 5'end ROX fluorescence reporter, and an opposite, unlabeled strand.

Example 2 Fractionating a Size Ladder Mixture by Polyacrylamide Electrophoresis A slab polyacrylamide gel containing 6% polyacrylamide was prepared by standard methods (User Bulletin Number 16 for the Model 373A DNA Sequencing System, Applied Biosyste- ms, Inc., Foster City, CA, January 1991) between 5 mm high- optical quality glass plates, to slab dimensions of 0.4 mm thick, 22 cm width, and 40 cm length. The gel was prepared in a standard TBE buffer (0.089 M Tris base, 0.089 M boric acid, 0.002 M EDTA) containing 8 M urea. An aliquot of the strand-pair mixture prepared in Example 1 was dried down and resuspended in 5 μL of loading buffer (containing 5 parts v/v deionized formamide to one part 50 mM EDTA solution, pH 8.0); the mixture was heated for 2 minutes at 90°C to denature the strand-pair fragments, and 5 μl of the denatured strand-pair mixture was added to a sample well at the top of the gel.

Electrophoresis was carried out on a Model 373A DNA Sequencing System (Applied Biosystems, Inc, Foster City, CA) , using a standard Tris-borate buffer as the running buffer. Electrophoretic voltage was 1300-1400 V, at approximately 20 ma with the power regulated at 30 W and the temperature regulated at 40°C. The peaks were moni¬ tored by a scanning laser fluorometer positioned to detect fragments exactly 25 cm from the sample origin in the gel. Details of the fluorometer and fluorescence-emission signal processing system have been published (Connell) . The electropherogram generated from the sample is shown in Figure 4. The double peaks at 350- and 400-base fragment

SUBSTITUTE SHEET sizes are presumably due to the action of the Tag polymer¬ ase, which may add one additional base (dA) beyond the end of the template strand.

Example 3

Fingerprinting for Genomic Mapping Lambda DNA was digested with either Hind III, Bam HI, or Eco RI restriction endonuclease, and the restriction fragments were 5'-end labeled by ligation to labeled oligodeoxynucleotides according to published methods (Carrano) . Each restriction/ligation reaction contained lambda DNA (6.4 μg, 0.2 pmol), the labeling oligodeoxynucl- eotide and its complement or "splint" (30-40 pmol of each, for a 10-fold excess relative to the number of restriction cut ends expected), restriction enzyme (6 units per μg of DNA) , T4 DNA ligase (1 unit per μg of DNA) , and ATP at 1 mM, in a total -volume of 40 μL, buffered with lx restric¬ tion digest buffer supplied by the enzyme manufacturer (Promega Corporation, Madison, WI) . The 18-mer labeling oligodeoxynucleotides had the same sequence as the primer P₀ shown in Example 1, above, except that the fluorescent label used with Hind III was FAM, the label used with Bam HI was JOE, and that used with Eco RI was TAMRA (Connell) . The three complementary "splint" oligodeoxynucleotides were 22-mers, with four extra bases at the 5'-end chosen to complement the 5'-overhang created by the corresponding restriction enzyme:

Hind III: 5'-A GCT ACT GGC CGT CGT TTT ACA-3'

Bam HI: 5'-G ATC ACT GGC CGT CGT TTT ACA-3'

ECO RI: 5'-A ATT ACT GGC CGT CGT TTT ACA-3'

SUBSTITUTESHEET Each reaction was incubated overnight at 37°C. The labeled fragments were then further digested by incubation for an additional two hours at 37°C with restriction endonuclease Hae III (6 units per μg of DNA) . The reactions were then stopped by addition of 0.8μL of 500 mM EDTA solution (pH 8.0) .

Aliquots of the three fragment mixtures were combined in a DNA concentration ratio of 1:1:2 for the FAM-, JOE- and TAMRA-labeled fragments. The ROX-labeled strand-pair fragments from Example 1 were added to this mixture in an approximately 1:1 DNA concentration ratio. The pooled sample was precipitated with 0.1 volume of 3 M sodium acetate (pH 5.5) plus 2.5 volumes of absolute ethanol. After centrifugation at 12,000 rpm for 20 minutes at 4°C, the supernatant was decanted. The precipitated sample was washed by addition of 250 μL of cold 70% ethanol followed by centrifugation for 5 minutes. The supernatant was again decanted, and the residue dried briefly in a vacuum evaporator. It was resuspended in gel loading buffer (see Example 2) and denatured by heating for 2 minutes at 90°C. The resuspended sample was applied to a lane of an 8 M urea polyacrylamide slab gel prepared as in Example 2. Electro¬ phoresis and fluorescence-peak monitoring were carried out as in Example 2, with the fluorescence emission from each of the four different fluorophors being monitored by four different band-pass filters as described by Connell et- al. Data acquisition was carried out using DNA sequencing software provided with the Model 373A Sequencer (Applied Biosystems, Inc., Foster City, CA) . Data analysis to resolve the peaks associated with each of the flourophores was carried out using the Model 672 GENE SCANNER™ Analysis Software for the DNA Sequencer {Applied Biosystems, Inc., Foster City, CA) .

SUBSTITUTESHEET For comparison, samples of the restriction digests described above were analyzed on the same polyacrylamide gel, using a ROX-labeled size standard prepared by restric¬ tion endonuclease PstI digestion of phage lambda DNA (GENESCAN-2500 ROX, Applied Biosystems, Inc., Foster City, CA) .

The electropherogram obtained is shown in Figure 7, where the FAM (Hindlll/Haelll- JOE (BamHI/Haelll) , TAMRA (EcoRI/Haelll) , and ROX (size fragments) fluorescent sig- nals are as indicated. The positions of the ROX-labeled size-calibration strands was plotted as a function of strand size, and this plot was used to establish the basepair numbers of each of the FAM-, TAMRA, and JOE- labeled fragment strands. The observed and expected base- pair number of each of the latter fragments is given in Table 2 below. For almost every expected restriction digest band, the DNA standard (Selected Size Ladder) prepared according to the method of the present invention gives more accurate fragment sizing than does the conven- tional method (GENESCAN-2500) . The difference in accuracy is summarized by the chi squared values, which measure the relative discrepancies between observed and expected values (the larger the chi squared value, the greater the devia¬ tions from the expected values) . Note that the 697 base- pair Hind III/Hae III fragment migrates anomalously; while its actual size cannot be determined correctly with either standard, its characteristic mobility can nonetheless be used to confirm its identity.

SUBSTITUTE SHEET

SUBSTITUTE SHEET Example 4 DNA Sequencing The M13mpl8(+) plasmid, in single-stranded form, was used as the template for DNA polymerization, in combination with the P₀ primer noted above. Chain elongation and random-base termination in the presence of a selected-base 2',3'-dideoxynucleoside 5'-triphosphate was carried out according to standard procedures (Sanger) .

Briefly, the primer used for each reaction mixture was a FAM-labeled primer. The template and primer in each of the four reaction mixtures were heated to 60°C for 10 minutes, then cooled slowly to room temperature. The chain elongation reactions were each carried out in the presence of all four deoxynucleoside 5'-triphosphates (dNTPs) , Tag DNA polymerase, (Promega Corporation, Madison, WI) and the selected dideoxynucleoside-triphosphate (ddNTP) for 30 minutes at 60°C. The reaction volumes were precipitated by addition of sodium acetate (to 0.3 M) and 2.5 volumes of ethanol, followed by centrifugation at 12,000 rpm for 20 minutes at 4°C. The supernatant was decanted, and 250 μL of cold 70% ethanol was added to each tube. The tubes were centrifuged an additional 5 minutes, the supernatant again decanted, and the products dried briefly.

Each of the four dideoxy termination reaction mixtures from above was combined with 1 μL (approximately 5 fmol) of the ROX-labeled strand pair mixture from Example 1. Each of these mixtures was then denatured by addition of 5 volumes of deionized formamide, with heating for 2 minutes 90°C. A polyacrylamide slab gel in Tris-borate buffer, and containing 8M urea was prepared as in Example 2. To the top of the gel, at four different wells, were added 5 μl of the denatured DNA mixtures. Electrophoresis was carried out under the conditions described in Example 2. The fluo- rescent-labeled DNA peaks were monitored, for each lane, at

SUBSTITUTE SHEET fluorescence emission wavelengths effective to resolve fluorescence emission from the FAM and ROX reporters. Data collection was performed on a Macintosh® Ilex computer (Apple Computer, Cupertino, CA) using the instrument's DNA sequencing software. The results, plotted for each lane, are shown in Figures 8A-8D, for ddATP, ddTTP, ddGTP, and ddCTP terminations, respectively.

For each lane, the migration times (to detection) of the ROX peaks were plotted as a function of fragment size. The number of bases in each of the FAM peaks was then determined from this plot, using the Model 680A Contig Mapper software (Applied Biosystems, Foster City, CA) running on a VAX computer (Sun Microsystems, Mountain View, CA) . From the results, the fragment sequence was deter- mined. The sequence could be read with reasonable accuracy (97.5%) for 600 bases from the starting point twelve bases downstream of the primer binding site. In 8 instances, the correct base could not be read from the data, and in 7 instances the correct order of two bases could not be determined (in one of the cases, the data appeared to require an extra base, or insertion error) . These results compare favorably to manual or automated sequencing, as carried out with no internal-lane size standard.

Example 5

Restriction Digest Analysis on Agarose Gels Purified pBR322, obtained from Promega Corporation (Madison, WI) , was digested to completion with Alul. The 18-base, single-stranded oligonucleotide from Example 3 (labeled at its 5' end with a FAM fluorescence reporter) was annealed with a complementary oligonucleotide. The Alu I fragments were ligated with FAM-labeled duplex fragments, as in Example 3, to form end-labeled pBR322 digest frag¬ ments (GENESCAN-1000™ FAM) . The fragments were diluted in

SUBSTITUTESHEET 10 mM EDTA solution (pH 8.0) to a working concentration of 1 nM.

A calibration mixture containing ROX-labeled strand pair fragments having sizes between 50-600 basepairs, in 50-basepair increments, was prepared as in Example 1. This mixture was added to the FAM-labeled pBR322 fragments in a

DNA ratio of about 4:1.

A slab agarose gel containing 2% agarose was prepared in the above Tris-borate buffer by standard methods (Maniatis) to slab dimensions of 4 mm thick, 20 cm width, and 16 cm length in the submarine gel tray provided with the 362 GENE SCANNER (Applied Biosystems, Foster City, CA) .

To a gel well was added 2.5 μL of the combined pBR322/str- and-pair fragment mixture, diluted 1:1 with 2 Agarose Loading Buffer (containing 50 mg/mL Ficoll 400-DL, 8.3 mg/mL blue dextran, and 1.67 mg/mL dextran sulfate, all obtained from Sigma Chemical Corporation, St. Louis, MO) . Electrophoresis was carried out in a Model 362 GENE SCANNER (Applied Biosystems, Inc, Foster City, CA) , using a standard Tris-borate buffer as the running buffer. Electrophoretic voltage was 100 V, 80 ma, 8 W, at a temperature of about 25°C. The peaks were monitored by a laser fluorometer, as described in Example 3. The electro- pherogram generated from the sample is shown in Figure 9, where the pBR322 fragments (FAM label) are shown in the upper trace, and the strand-pair-fragments (ROX label) are shown in the lower trace. The peak positions (electropho¬ resis times) of the pBR322 were used to establish a plot of basepair number as a function of electrophoresis time. This plot was then used to determine the sizes of the ROX- labeled fragments. The sizes determined for the ROX- labeled fragment peaks are shown in Table 3 below. The calculated size is within 5 basepairs of the correct value

SUBSTITUTESHEET for every fragment except the smallest one listed (250 basepairs) , which lies outside of the size range accurately defined by the pBR322 fragments.

Example 6 Restriction Digest Analysis on Agarose Gels

The FAM JOE, and TAMRA-labeled phage lambda restric¬ tion fragments from Example 3 were combined with ROX- labeled fragments of 50-1000 basepairs prepared according to the present invention (Example 1) . The mixture was diluted 1:1 with 2x Agarose Loading Buffer and 5 μL aliquot was separated on a 2% agarose gel as described in Example 5. Figure 10 shows the electropherograms for the lambda restriction fragments (upper panels) and the ROX-labeled standard fragments (bottom panel) . The Model 362 GENE SCANNER Analysis Software (Applied Biosystems, Foster City, CA) used the ROX-labeled fragment positions to establish the relationship between electrophoresis time and fragment size, and then calculated sizes for each of the observed lambda restriction fragments. In another gel lane, the

SUBSTITUTE SHEET lambda fragments were analyzed by comparison to ROX-labeled restriction fragments (GENESCAN-2500™ ROX, Applied Biosystems, Foster City, CA) .

SUBSTITUTESHEET Example 7

Large-scale Preparation of a Size Ladder

Strand-Pair Mixture

The same amplification primers were used as in Example 1. The 10x amplification buffer was similar, except that it contained 15 mM magnesium chloride. The M13mpl8(+) template (Promega Corporation, Madison, WI) was diluted to

25 μg/mL with sterile water. A dNTP mix was prepared containing dCTP, dGTP, and dTTP at 2.5 mM, plus dATP at 1.5 mM. Each fragment was prepared by combining water (835 μL) , 10x buffer (100 μL) , template (10 μL, 250 ng) , ROX- labeled forward primer (20 μL, 200 pmol) , reverse primer (10 μL, 200 pmol) , dNTP mix (20 μL, reaction concentration 50 μM dCTP, dGTP, dTTP and 30 μM dATP) , and AmpliTaq® (Perkin Elmer Cetus, Norwalk, CT; 5 μL, 25 units) . After mixing, each solution was divided into five 200 μL aliquots in five separate 0.5 mL microcentrifuge tubes. Each aliquot was covered with mineral oil (75 μL) . The tubes were placed in a preheated (95°C) DNA Thermal Cycler (Perkin Elmer Cetus, Norwalk, CT) and 30 cycles consisting of double strand denaturation (95°C/30 sec) , primer annealing (55°C/30 sec) , and polymerase-mediated primer extension (70°C/30 sec) were executed.

Following thermal cycling, the five aliquots were recombined with 1 mL of lx TE buffer (10 Tris-HCl 1 mM EDTA, pH 8.0) in a Centricon-100 filter (Amicon Division of W.R. Grace & CO., Beverly, MA). The mixture was filtered by centrifugation at 5000 RPM (approximately 3000x g) for 10 minutes at 4°C. The products were recovered in approxi- mately 50 μL of TE buffer by inverting the filter and spinning at 5000 RPM for 2 minutes. After dilution to 1.0 mL with TE buffer, the DNA yield was determined by measure¬ ment of _jgo in a 1.0 cm cuvette. The yields of all frag-

SUBSTITUTESHEET ments from 150-1000 basepairs are shown in Table 5. It is evident that fragments smaller than 300 basepairs were recovered with reduced efficiency on the Centricon-100 filter.

Strand pair mixtures were prepared by combining aliquots of all the fragment sizes listed in Table 5, as well as synthetic double stranded 50 and 100 basepair

SUBSTITUTESHEET fragments, in the proportions indicated in the last column of the table, to give equal molar ratios of the fragments. The fragments were then precipitated by addition of sodium acetate to 0.3 M, followed by 2.5 volumes of ethanol, as described in Examples 1 and 3. After centrifugation and removal of the supernatant, the strand-pair mixture was resuspended in the desired amount of TE buffer, from which individual aliquots containing 5-10 fmol of each fragment could be removed for analysis. Although the invention has been described with respect to preferred compositions and methods, it will be apparent that various modifications can be made without departing from the invention.

SUBSTITUTESHEET

Claims

IT IS CLAIMED:

1. A polynucleotide strand-pair mixture for use in calibrating the basepair numbers of one or more analytes fractionated by electrophoresis, comprising a plurality of polynucleotide strand pairs, each pair having one strand which has a selected number of polynucle¬ otide bases and a detectable reporter in its 5'-end region and a complementary strand which lacks said reporter, where the concentration of said complementary strand may be different from the concentration of the corresponding reporter-containing strand, and where the selected number of bases in each pair's reporter-containing strand is different for each pair in the mixture.

2. The strand-pair mixture of claim 1, wherein the reporter is a radiolabeled atom in a nucleotide in the 5'- end region of the reporter-containing strand in each strand pair.

3. The strand-pair mixture of claim 1, wherein the reporter is a fluorescence reporter carried on the 5'-end nucleotide of the reporter-containing strand in each strand pair.

4. The strand-pair mixture of claim 1, wherein the selected number of bases in the reporter-containing strands in the strand pairs differ from one another by a multiple of a selected number of bases.

5. The fragment mixture of claim 4, wherein the strand pairs differ in size by a multiple of about 40-60 bases.

SUBSTITUTESHEET

6. The fragment mixture of claim 4, for use in high-resolution polyacrylamide gel electrophoresis in the presence of a denaturant agent, wherein the strand pairs have selected sizes less than about 1,200 bases.

7. The fragment mixture of claim 1, wherein each successively larger reporter-containing strand in the mixture contains the same basepair sequence as the next- smaller reporter-containing strand, plus a 3'-end sequence of at least about 10 bases.

8. The fragment mixture of claim 7, wherein each successively larger reporter-containing strand differs from the next-smaller strand by substantially the same number of bases.

9. The fragment mixture of claim 1, for use in detecting a small base number mutation in a genomic segment having a given number of basepairs, wherein the mixture contains first and second fragments whose reporter-contain¬ ing strands differ from one another by a selected number of bases, and are shorter and longer, respectively, than the genomic segment in the absence of a mutation.

10. A method of producing a polynucleotide fragment mixture for use in calibrating the basepair numbers of one or more polynucleotide analytes^' fractionated by electropho¬ resis, comprising identifying a segment of a DNA template having a selected number of basepairs between about 25-2,000 base¬ pairs,

SUBSTITUTESHEET reacting the segment-containing template with repor¬ ter-labeled and unlabeled single-stranded primers which are complementary to 3'-end regions of opposite strands of the identified segment, when the segment is in a double- stranded form, under reaction conditions which produce multiple rounds of primer-mediated duplex DNA replication, by said reacting, producing a duplex DNA fragment having the basepair composition of said segment, and one strand only which includes the reporter-labeled primer, and repeating said identifying and reacting steps by (i) identifying a segment of a DNA template containing a selected number of basepairs which is different from that of an already-selected segment, (ii) reacting the DNA template containing the just-identified segment with reporter-labeled and unlabeled single-stranded primers which are complementary to 3'-end regions of opposite strands of the just-identified segment, when the segment is in double-stranded form, under reaction conditions which produce multiple rounds of primer-mediated duplex DNA replication, thereby producing a duplex DNA fragment having the basepair composition of the just-identified segment, and one strand only which includes the reporter label, until a desired number of different-length fragments are produced.

11. The method of claim 10, wherein the templates containing the selected-length segments are single-stranded DNA templates.

12. The method of claim 10, wherein the different selected-length segments are contained in a single tem¬ plate, and have an overlapping segment end region which is common to each of the segments.

SUBSTITUTESHEET

13. The method of claim 12, wherein the labeled pr:Lmer is complementary to said common-sequence segment 3'- end region.

14. The method of claim 13, wherein the fragments are produced in a single reaction mixture containing said template, the labeled primer, and a series of unlabeled primers which are complementary to the 3' ends of the selected-length segment strands which are opposite to the common-sequence end region.

15. The method of claim 10, for use in determining the amounts of one or more of such polynucleotide analytes fractionated by electrophoresis, wherein the fragments are produced individually and combined in known concentrations to produce the mixture.

16. The method of claim 15, wherein the different selected-length segments are contained in a single tem- plate, and have an overlapping segment end region which is common to each of the segments, and the labeled primer is homologous to said common-sequence segment end region.

17. The method of claim 10, for use in calibrating the basepair numbers of one or more polynucleotide analytes fractionated by polyacrylamide gel electrophoresis in the presence of a denaturant agent, wherein the selected segments contain less than about 1,200 bases, and differ from one another in number of bases by a multiple of about 40-60 bases.

SUBSTITUTESHEET

18. The method of claim 10, wherein the reporter is a fluorescent reporter carried on the 5'-end nucleotide of the reporter-containing primer.

19. In a method for genomic mapping in which indi¬ vidual library genomic fragments are digested by restric¬ tion endonucleases, to produce digest subfragments having sizes in the size range between about 50-2,000 basepairs, and the subfragments are (i) labeled in at least one strand with a fluorescent reporter, (ii) fractionated by high- resolution electrophoresis in a polyacrylamide matrix, and (iii) compared according to size with subfragments derived from other library genomic fragments, to determine frag¬ ments which have overlapping regions, an improvement for enhancing the resolution of subfragment size comprising fractionating the subfragments in the presence of a polynucleotide mixture composed of a plurality of polynu¬ cleotide strand pairs, each pair having one strand which has a selected number of polynucleotide bases and a detectable fluorescence reporter in its 5'-end region and a complementary strand which lacks said reporter, where the selected number of bases in each fragment's reporter- containing strand is between about 30-2,000 and is differ¬ ent for each fragment in the mixture, and determining the size of the subfragments, to a resolu¬ tion of 1-2 bases, from their relative migration rates in the acrylamide matrix with the known sizes of the reporter- containing strand-pair polynucleotide strands.

SUBSTITUTESHEET

20. The improvement of claim 19, wherein the subfragments and polynucleotide mixture are fractionated in single-stranded form on a polyacrylamide gel in the presence of a denaturing agent, and the size of the subfragments is determined to a 1-base resolution.

21. The improvement of claim 19, wherein the subfragments and polynucleotide mixture are fractionated in double-stranded form on an agarose, polyacrylamide, or other gel.

22. The improvement of claim 19, wherein the fluorescent probe carried on the digest subfragments is distinguishable from the fluorescent probe carried on the fragment polynucleotide strand pairs.

23. In a dideoxy polynucleotide sequencing method in which various-length fragments which terminate at one of four dideoxynucleotide bases are fractionated by high- resolution electrophoresis in the presence of a denaturing agent, and the sequence of the polynucleotide is determined from the basepair number determined for each fragment, an improvement for increasing the resolution of basepair number determination comprising fractionating the various-length fragments in the presence of a polynucleotide mixture composed of a plura¬ lity of polynucleotide strand pairs, each pair having one strand which has a selected number of polynucleotide bases and a detectable fluorescence reporter in its 5'-end region and a complementary strand which lacks said reporter, where the selected number of bases in each fragment's reporter- containing strand is between about 30-2,000 and is dif¬ ferent for each fragment in the mixture, and determining the size of the various-length fragments, to a resolution of 1 base, from their relative migration rates in the acrylamide matrix with the known sizes of the reporter-containing strand-pair polynucleotide strands.

24. The improved method of claim 23, wherein the various-length fragments have a distinctive fluorescence reporter for each dideoxynucleotide, and the various length fragments with the different reporters and the reporter- labeled strand pairs are co-electrophoresed.

25. The improved method of claim 23, wherein the various-length fragments which terminate at a common dideoxynucleotide and the reporter-labeled pair strands are co-electrophoresed.

26. A method of quantitating the amount of one or more nucleic acid fragments fractionated by electrophoresis comprising, mixing the fragments with a polynucleotide mixture composed of a plurality of polynucleotide strand pairs, each pair having one strand which has a selected number of polynucleotide bases and a detectable fluorescence reporter in its 5'-end region and a complementary strand which lacks said reporter, where the selected number of bases is different for each fragment in the mixture, and the concentration of each strand pair in the mixture is known, fractionating the fragments and polynucleotide mixture, determining the relationship between fragment concentration and total fluorescence emission, as a function of number of bases, for strand pairs polynucleo¬ tides whose concentrations are known, and

S using said relationship to determine the concentration of analyte fragments, according to total fluorescence emission of the fragments and number of bases in the frag¬ ments.

27. The method of claim 26, wherein the fragments and polynucleotides in the mixture are fractionated on polyacrylamide gel in the presence of a denaturant agent.

SUBSTITU