CA2103000A1 - Method for promoting specific alignment of short oligonucleotides on nucleic acids - Google Patents

Abstract

METHOD FOR PROMOTING SPECIFIC ASSIGNMENT OF SHORT
OLIGONUCLEOTIDES ON NUCLEIC ACIDS
Abstract of the Disclosure Disclosed is a method for promoting specific alignment of short oligonucleotides on a nucleic acid polymer. The nucleic acid polymer is incubated in a solution containing a single-stranded DNA-binding protein and a plurality of oligonucleotides which are perfectly complementary to distinct but adjacent regions of a predetermined contiguous nucleotide sequence in the nucleic acid polymer. The plurality of oligonucleotides anneal to the nucleic acid polymer to form a contiguous region of double stranded nucleic acid. Specific application of the methods disclosed include priming DNA
synthesis and template-directed ligation.

Description

.` ' '. .', ' ,~" 2la3~

METHOD FOR PROMOTING SPECIFIC ~LIGNMENT OF_SHORT
OLIGONUCLEOTIDES ON NVCLEIC ACIDS

ackgrpund Substan~ial improvement in ~he efPiciency of nucleotide sequencing i~ needed if the goals of the human genome ~equencing project are to be realized.
Improvements in sequencing technology would also provide substantial benefit to molecular genetics by liberating creative ~cientists from the repetitive, but highly informative task of equencing newly isolated DNAs of interest.
A potentially efficient method of se~uencing by current technology i5 by primer walking. By this technique, priming an enz~matic ~equencing reaction within a ~egment of known sequence (such as vector sequence) is used to extend the sequence into the unknown region, The newly d~termined sequence in turn is used to select a primer to extend the sequence further, and this process is repeated until the sequence o~ the entire molecule has been determined. Advantages of primer walking are that.
the entire sequence can be determined on a single preparation of template DNA without ~ubcloning, and the sequence can be determined in the minimum number of sequencing reactions.
A disadvantage of primer walking has been the inconvenience and expense of having to synthesize a primer for each sequ~ncing reaction. ~n improvement in pri~ing methods which would eliminate this disadvantage would represent an important advance in the art.
,' ~.

'~

2~03a;~0 Summary of the Invention The subject invention relates to a method for promoting specific alignment o~ short oligonucleotides on a nucleic acid polymer. The nucleic acid polymer ie incubated in a solution containing a single-stranded DNA-binding protein and a plurality o~ oligonucleotides which are preferably perfectly complementary to distinct but adjacent regions oP a predetermined contiguous nucleotide sequence in the nucleic acid polymer. The plurality of oligonucleotides anneal to the nucleic acid polymer to Porm a contiguous region of double stranded nucleic acid.
Among the important applications of the æubject invention is the use o~ the method ~or priming nucleic acid polymerization. Prior to contacting the template molecule with an appropriate polymerase en2yme, the nucleic acid template is incubated with a sin~le-stranded DNA-binding protein (SSB) and a plurality (two or more) of nucleotide primers which are perfectly complementary to distinct but adjacent regions of a predetermined contiguous nucleotide sequence in the template molecule.
The mixture i5 incubated under conditions appropriate for annealing of the primers to the predetermined contiguous nucleotide sequence to form a contiguous region of double stranded nucleic acid, and binding of the SSB to the nucleic acid template. The nucleic acid template can be either DNA or ~NA.
The priming method described in the preceding paragraph can be employed to prime DNA sequencing reactions by the dideoxynucleotide chain termination method. This novel approach represents a particularly signi~icant advance in the ~ield of DNA sequencing by directed primer se~uencing. The conventional approach to directed primer sequencing requires the synthesis of a new oligonucleotide primer designed to anneal near the downstream terminus of a newly determined DNA sequence.

2 1 ~ 0

-3 This processive approach to DNA sequencing is also known as primer walking.
In the method of this invention, a preexisting oligonucleotid primer library (~or example, a hexamer oligonucleotide library) i8 used as the source of primer to initiate DNA synthesis from the downstream region o~ a newly determined DNA sequence. ~ complet~ library of all possible hexa~ers would contain ~096 unique hexamer sequences. Sets of these unique primer sequences ~for example, a set ~ay contain 3 or 4 unique hexamer sequences) are selected to anneal at a predetermined downstream region of a newly determined ~NA sequence to form a contiguous duplex region which will prime specific DNA polymerization. Thus, a single hexamer library can be used to prime litarally millions of DNA sequencing reactions thereby obviating the need to generate custom oligonucleotide primers. This improvement will substantially reduce the cost and time associated with large scale DNA sequencing pro~ects.

~ r o..~ ~ on of the Dr~awing Figure 1 is a diagram representing the M13mpl8 DNA
template sequence complementary to hexamer set D, as well as the relative annealing positions of primer set D
prlmers.

Detailed Description of th~ Invention The subject invention is based on the discovery that the inclusion of a single-stranded DNA-binding protein in an incubation mixture comprising a nucleic acid polymer, and a plurality (i.e., two or more) of ~hort oligonucleotides complementary to a predetermined contiguous series of nucleotides in the nucleic acid polymer, functions to promote the specific binding of the short oligonucleotides to the predetermined contiyuous 2103~33 series of nucleotides in the nucleic acid polymer. The short oligonucleotides are preferably perfectly complementary to distinct but adjacent regions of the predetermined contiguous series of nucleotides.
The methods of this invention are applicable primarily for use in connection with oligonucleotides having a monomer number of between about 5 and 10 (re~erred to herein as short oligonuclaotides) which are prepared by conventional methods. Such oligonucleotides can be labeled with a detectable group, or modified to generate a particular functional group, depending upon the specific application. The methods are applicable for any application in which it is desirable to specifically anneal two or more short oligonucleotides to a contiguous ~tretch of nucleotides of known sequence in a nucleic acid polymer. As will be discussed in greater detail below, there are a wide variety of such applications.
The expression "nucleic acid polymer" as used herein refers to any nucleic acid molecule having a monomer number which is greater than or equal to the total number of nucleotides in the contiguously annealing set of oligonucleotides~ If the polymer is double stranded DNA, the monomer number count is determined by counting the number o~ nucleotide monomers in a single strandO In practice, the methods of this invention will most often be employed to specifically align a plurality of oligonucleotides at a predetermined contiquous series of nucleotides in a nucleic acid polymer such as genomic DNA
isolated from cells or from a genomic DNA library, cDNA
from a cDNA library, mRNA, rRNA or restriction fra~ments of same. The methods are particularly applicable to the analysis of cloned DNA (e.g., M13, cosmid, P~ and YAC
clones).
In the past, a problem associated with efforts to promote the specific alignment of two or more short 2 ~ o oligonucleotides to a contiguous stretch of nucleotides of known sequence in a nuclei¢ acid polymer has been the occurrence of undesirable oli~onucleotide binding events.
For example, individual oligonucleotides, or multiples of oligonucleotides can anneal at locations other than the desired location (the conti~uous stretch of nucleotides of known sequence in the nucleic acid polymer) in the polymer. In addition, th~ bindiny of less than the desired number of contiguously annealing oligonucleotides to the predetermined conkiguous nucleotide sequence i8 an undesirable bindin~ event. Th~ undesirable binding events mentioned are only a few examples of a wide variety of possibilities. These undesirable binding events interfere with the application or assay (e.g., DNA sequencing or template-directed ligation~.
It has been determined that the inclusion of SSB in the annealing mixture promotes the specific alignment of short oligonucleotides at the predetermined contiguous nucleotide sequence in the nucleic acid polymer while, at the same time, inhibiting undesirable binding events.
This discovery was completely unexpected and is likely to revolutionize DNA sequencing methods.
SSB (single-stranded DNA-binding protein) is a descriptive label which has been applied to a class of proteins which bind more strongly to single-stranded DNA
than to double stranded DNA ~see e.g., Meyer and Laine, Nicrobiological Review~ 5~, 342 (1990); Chase and Williams, Ann. Rev. Biochsm. 5$: 103 (1986)). They tend to be proteins which are involved in DNA metabolism.
Specific examples include the E. coli SSB, the bacteriopha~e T4 gene 32 protein and the T7 gene 2.5 protein. The methods of this invention are not limited, however, to the use of the specific SSB examples recited.

21030~D

SSBs can be used in~ividually, or mixturas of different SSBs can be used. The optimal con~entration of SSB can vary depending upon the particular SSB employed, and the nature of the template molecule. This concentration i5 easily determined empirically by the methods described in detail in the Exemplification section which follows. For example, i~ has been determined thak the optimal ratio for E. col i SSB with M13 template is a mass ratio of about 2.5 or greater.
lo SSB is thought to bind single-stranded DNA by wrapping the DNA around an octamer of SSB, protecting a~out 145 nucleotides from digestion by DNase but leaving an average of about 30 unbound nucleokides between DNAT
octamer beads (Chrysogelos and Griffith, Proc. Natl. Acad.
15 Sci. ~SA 79, 5803 (1982); Griffith et al., Cold Spring Narbor Symp. Quant. Biol. 49, 553 (1984)). Although not wishing to be bound by a mechanism, it is poss~ble that these unbound nucleotides are the sites of initial binding of oligonucleotides to the DNA. Random movement of the octamer beads along the DNA strand might well expose all potential binding sites in the DNA and also displace weakly bound oligonucleotides.
A saturating amount of SS~ both suppresses binding by individual short oligonucleotides at locations other than the predetermined nucleotide ~equenc~ in the nucleic acid polymer and stimulates contiguous annealing of primers at the predetermined nucleotide sequence. Masking of individual binding sites is an important factor in the success of many applications, and presumably increases the effective concentration of oligonucleotides available for binding at the desired location on the polymer.
Clearly, molecular cooperativity plays a role in the formation of a stable complex between contiguously annealing short oligonucleotides and SSB-coated single-stranded DNA. This cooperativity requires that the 2l030~a binding sites be adjacent in the DNA, without any gaps.
There~ore, base-stacking interactions between adjacent oligonucleotides must be responsible for stabilizing binding to SSB-coated DNA.
As w~s mentioned above, methods for the specific alignment of two or more short oligonucleotides to a co~tiguous stretch of nucleotide~ o~ known sequence in a nucleic acid polymer have utility in a wide variety of applications. A few examples are di~cussed below.

Priminq_Nucleic_Acid Synthesis The methods of this invention can be used to prime nucleic acid synthesis ~rom a nucleic acid template.
Prior to discu6sing the speci~ics of the methodology, it is important ~o highlight the importance of the invention.
The methods relate to the use of a plurality of short oligonucleotide primers to prime nucleic acid synthesis.
Prior art methods have employed a single oligonucleotide primer to prime such polymerization reactions. The importance of the u~e of multiple contiguously annealing primers is that specific members o~ a library o~ such oligonucleotides can be combined to generate a contiguous region of duplex DNA. It is no longer necessary to custom synthesize primers in order to complement a known DNA
sequence~ For example, rather than custom synthesizing a nucleic acid primer of 24 nucleotides in length which is complementary to a known template sequence, four hexamers are selected from a preexisting hexamer library. The four hexamers are perfectly complementary to the known template ~equence and anneal to form a duplex region o~ 24 base pairs in length. The implications of this will be discussPd in greater detail below in connection with DNA
sequencing.
In the methods for priming nucleic acid synthesis, the nucleic acid template can be either DNA or RNA.

"~ 2~033~0 Although the experiments discussed in the Exemplification section below are limited to studies of DNA, it is known that SSB binds to RNA as well. In fact, SSB has been used to bind to RNA in order to remove secondary structure for electron microscopy studies ~see e.g., Chase and Williams, Ann. RevO Biochem. ~ 03 (1986); ~angel et al. Proc.
Natl . Acad . Sci. USA 71: 4541 (1974)).
In general, for purposes of priming polymerization, the use~ul lower limit for the number o~ base pairs in a duplex region formed by the contiguou~ annealinq of a plurality of primers to a single stranded nucleic acid template is 12. In ~uch a case, the number of contiguously annealing primers would be 2, each being a hexamer. Any combination of short oligonucleotides (oligonucleotides having a monomer number of between about 5-10, inclusive) which anneal contiguously to form a duplex region of greater than 12 base pairs, can be used in connection with the methods described herein.
Preferably, the duplex region which is formed by contiguous annealing is from about 18-24 base pairs in length.
Following the incubation o~ the template molecule with SSB and the selected oligonucleotide primers, the primed template is incubated with an appropriate polymerase enzyme. For example, if the template is RNA, the polymerase can be an RNA-dependent DNA polymerase. If the template is DNA, the polymerase can be a DNA-dependent DNA polymerase.
The preferred size of the oligonucleotide primer is 6 monomer units. This size is preferred due to consider~tions o~ the binding stability and statistical factors relating to library size. A library (i.e., a collection of unique members, with each unique member being specifically accessible) o~ all possible hexamers would consist of 4096 unique members. As discussed in - 2l~3a3~
- 9~
greater detail below, a practical minimum number for a useful hexamer library is about 1500. It is also possible to optimize an oligonucleotide library to include or exclude primers known to be particularly useful, or particularly troublesome, xespectively. For example, base composition ¢an be taken into account when optimizing the composition of a library.

DNA ~ a~in~
A particularly important application of the methods of this invention is DN~ sequencingO Conventional approaches to directed primer sequencing (primer walking) require the synthesis of new oligonucleotide primers to extend a known sequence into an unknown sequence. Primer walking, particularly with contiguously annealing sets of three or four hexamers, should be an efficient way to sequence DNAs of at least 40 kbp directly without subcloning. Almost all hexamers appear likely to be useable, and a library of all 4096 possible hexamers would be manageable. Smaller libraries could also be effective, but a practical minimum would probably be around 1500 hexamers, which would have a 99% chance of providing at least one contiguous hexamer set of three haxamers within a stretch of 100 nucleotides of template DNA, and an 84%
chance of providing a contiguous set of four. Synthesis on even the 0.2 ~mole scale provides enough primer for thousands of sequencing reactions at an average cost of only pennies per reaction. Once suitable primer libraries are available, they should improve the efficiency of sequencing in individual laboratories as well as in large-scale ~equencing centers.
The size limit of template DNA that can be ~equencedby direct priming will ultimatPly be determined by the sensitivity o~ detection of sequence ladders~ because the concentration of priming sites at a given mass ~` 2l~3~

concentration decreases with DNA length. Another limitation may be the c~ance occurrence of sequences that provide secondary priming sites in the DNA, which should increase with DNA length. It has be2n clearly demonstrated that the methods of this invention are use~ul for sequencing DNA which has been cloned in a cosmid vector (~40 kb). It seems likely that the method should also be applicable to the ~equencing of DNA which is cloned in higher capacity vectors such as Pl (~100 kb) and YAK (~500 kb) cloning vectors.
Currently, abou~ 60-90% of newly selected contiguously annealing hexamer sets provide useful sequence information, and this percentage seems likely to increase as more is learned about how to select the hexamer sets most likely to primP well. Regions of secondary structure that would interfere with priming should be easy to identify in the template DN~ and avoid.
The success rate might also be increased by using primer sets consisting of a single heptamer flanked by two hexamers. Addition of only a few hundred heptamers to a hexamer library could provide a useful density of such priming sites.
~ primer walking strategy allows the complete sequence of both DNA strands to be determined ~rom the minimum possible amount of primary sequence information, and provides complete freedom to choose additional priming sites for resolving ambiguities. The huge burden of subcloning, template preparation, excess sequencing, and sequence assembly imposed by the currently favored shotgun sequencing is eliminated. Since each template is used repeatedly, a battery of templates plus a primer library would allow sequencing reactions to be assembled rapidly enough to saturate any current or easily foreseeable means of analysis. The entire process is susceptible to computer control and automation, which should increase the -3 3 ~ 3 efficiency of large-scale DNA sequencing at least an order of magnituds over current practice. Ssquencing ~achines based on these principles could operate with little requirement for skilled human intervention, and could provide the capacity and ef~iciency needed ~or thQ success of the Human Genome Project.

Sequencin~ ~ Hybrid~ation The methods o~ this in~ention are also likely to be useful for ~equencing by hybridizat:ion (~ee e.g., Strezoska et al. Proc. Natl . Acad . Sci . USA 88: 10,089 (1991); KhrapXo et al., DNk Sequence 1: 375 (199~) ) . In this method, arrays of oligonucleotides are hybridized to larger DNAs, or arrays o~ larger DNAs are hybridized to oligonucleotides. A current difficulty in applying this method is efficiently discriminating perfectly paired from imperfectly pair~d hybrids. By using SSB coated DNA the reliability o~ sequencing by hybridization is likely to be improved. Khrapko et al. had proposed using continuous stacking hybridization for extending the length of sequence that can be read by sequencing by hybridization.
The method of this invention would greatly facilitate the oligonucleotide pairing which is essential for this method.

Template-Directed Li~ation Template-directed ligation is a method wherein a plurality of contiguously annealiny oliqonucleotides (typically two~, modified if neces~ary to provide appropriate ~unctional groups, are incubated with a template molecule which contains a nucleotide sequence which is perfectly complementary to the oligonucleotides.
The oligonucleotides anneal to form a contiguous duplex stru¢ture. The complex is then contacted with a liga~e .. . .

r~

~,J 2l~)3~3a3 enzyme which joins the adjacent oligonucleotides through a phosphodiester linkage.
Such a method can be used, ~or example, in a diagnostic method ~or the detection o~ point mutations in DNA (see e.g., Landegren et al., Science 241: 1077 (1988)). The inclusion o~ S~B in an incubation mixture of this type has b~en clearly demonstrated to improve binding specificities. Thus, the ~ethods herein are applicable to improving the results of template-directed ligation experiments~

EXEMPLIFICATION

Conditions for primin~ by hexamer oliqonucleotides Standard conditions. The template DNA in initial experiments was single-stranded M~3 DNA (6407 nucleotides) or the M13mpl8 derivative (7250 nucleotides) (Van Wezenbeek et al., Gene 11, 129 (1980); Ebright et al., G~ne 114, 81 (1992)). In an early experiment it was determined that a group of four hexamers, which were per~ectly complementary to distinct but adjacent regions of a predetermined contiguous nucleotide sequence of the template molecule (A4~A1, Table 1), primed well. These primers were used to test the range of conditions suitable for specific pri~ing by hexamers. The standard reaction conditions were derived from the protocols for sequencing with Sequenase 2.0 using 35S label (US Biochemicals) which employs a modified T7 DNA polymerase.
An equilibration reaction contained 0.7 ~g of M13 DN~, 3 ~g of E. coli SSB and 50 pmole of each hexamer (added last) to give a final volume of 10 ~1 in 40 mM
Tris-Cl, pH 7.5, 50 mM NaCl, 10 mM MgC12 ~nd concentrations of 33 nM M13 DNA, 16 ~M SSB monomer, and 5 ~M of each hexamer. Reaction mixtur~s were assembled at ~103~0 room temperature (assembly at 0 C gave the same results) and e~uilibrated for at least 5 min at 0 C ~but usually 30 to 60 min for con~enience). ~abeling was for 5 min at 0 C, adding 6 ~l of an ice-cold solution containing 2.5 units of Sequenase, 313 nM each of dCTP, dGTP and dTTP, about 3.5 ~Ci r~-35S]dATP (a 61ight molar exc~ss over the unlabeled dNTPs) in 10 m~ dithiothr~itol, 10 mM Tris-Cl, pH 7.5, 0.1 mM E~TA. The termination reaction was for 5 min at 37 C in the standard Sequenase protocol, and 0.1%
sodium dodecyl sulfate was added to the stop solution to prevent SSB ~rom int~rfering with electrophoresis o~ the DNA on sequencing gels.

Reaction solvent. Although not needed during the equilibration reaction, MgCl2 is needed for Sequenase activity and was added to the equilibration mixture for convenience. Decreasing the MgC12 concentration to 5 mM
reduced labeling of the sequence ladder substantially;
increasing it to 20 mM also decreased labeling but less significantly. Reactions seemed rather insensitive to NaCl concentration between 40 and 100 mM. The minimum practical NaCl concentration was 40 mM because of contribution from the stock solutions of SSB. The stock solution of SSB also contributed glycerol, typically a final concentration of 5% in the equilibration reaction.

SSB concentration. Similar results were obtained with SSB purified after axpression from the cloned gene, and with SSB obtained from two commercial sources (US
Biochemicals, Promega). Titration showed that a certain level of SSB is required to stimulate priming at the desired polymerization start point, and to suppress priming ~t sites other than the desired polymerization start point. In an experiment using 0.6 ~g M13 DNA, 1.2 ~g SSB was suff icient to stimulate a readable se~uence -~:` 21~3~3~

ladder that nevertheless showed significant priming at secondary sites, 1.5 ~g SSB almost eliminated this secondary priming, and ~.8-2.5 ~g gaYe equivalent patterns with no apparent secondary priming. Thus, a mass ratio of SSB to DNA of slightly greater than 2.5 appears to provide maximum stimulation of priming by the contiguously annealed hexamer set and essentially complete suppression o~ secondary priming. This saturating ratio corresponds to about 22 nucleotides of DNA per SSB monomer, 88 per tetramer, or 176 per octamer, consi~tent with the value of 175 nucleotides per octamer estimated for khe beaded form of the SSB-DNA complex (Chrysogelos and Griffith, Proc.
Natl. Acad. Sci. USA 79, 5803 (1982~; Griffith et al., Cold Spring Harbor Symp. Quant. Biol . ~9, 553 ~,1984) ) .
Labeling intensity was maximal ~rom the minimum saturating amount of SSB to a level of at least 3.5 ~g per reaction, but decreased somewhat at 5 ~g and higher levels. Labeling was almost completely suppressed by 7.5 10 ~g of one preparation o~ SSB but remained substantial with two others. This complete suppression appeared to be reversed upon dilution. The decrease in priming ef~iciency at high SSB concentrations is not simply due to a higher ratio of SSB to DNA, since diluting the template DNA 4 fold in 5 ~g SSB had little effect on labeling (see below). Most of the early experiments used 5 ~g SSB, which appears to be slightly higher than the optimal level for stimulating priming by a contiguously annealing hexamer set, but which may be somewhat more e~fectivP in suppressing secondary priming.

Hexamer concentration. Oligonucleotides were synthesized (1.0 ~mole scale) on a MilliGen 8750 DNA
synthesizer. Initially, hexamers were purified using Poly-Pac purification cartridges (Glen Research Corp., Sterling, VA~ according to the manu~acturar's 2~03a~0 specifications. However, tests showed that simply removing the dimethoxytrityl group as part o~ the synthesis procedure, releasing the oligonucleo~ide from the support with 30% ammonium hydroxide, incubating at 55 C to remove protecting groups, lyophilizing, and dissolving in water produced hexamer~ that gave eguivalent results and this simplified procedure has been employed.
Specific primin~ by hexamers ~rom the desired polymerization start point depends upon having a high enough concentration o~ hexamers to displ~ce SSB and pair contiguously to the template at the desired location.
About 5 ~M of each hexamer appeared to be sufficient to promote maximum intensiky and uniformity of labeling of the sequence ladders under the reaction conditions employed. Increasing the concentrations to 10-50 ~M
showed only marginal improvement. Reducing the hexamer concentration to 2.5 ~M reduced the labeling of shorter DNAs in the sequence ladder, indicating a lower frequency of priming during the labeling reaction. Labeling was much reduced at 1 ~M and almost lmdetectable at 0.5 ~M, even though this concentration was still a 15-fold molar excess over template.

DNA concentration. Reducing the M13 DNA to 150 ng (7 nM) had little ef~ect on the labeling intensity of the saquence ladder, but further reduction to 50 ng (2.35 nM) significantly reduced labeling, and reduction to 15 ng (0.7 nM) reduced it still further. The patterns of labeling upon dilution of the template DNA were similar whether the SSB concentration remained constant or was diluted in parallel with the DNA.
Equivalent labeling patterns were also obtained when the M13 DNA was diluted in the presence of denatured T7 DNA, keeping a total of 0.6 ~g of DNA in each reaction mixture. It is not surprising that competing DNA with a 2103~

12-fold higher complexity had little effect on priming ef~iciency at the desired pol~merization start point: ~13 DNA itself contains 1.5 times as many hexamers as the 4096 that are possible, so further increases in complexity should have little effe~t on the density of potential interaction sites for individual hexamers.

Reaction~mperature. ~pecific priming at the desired polymerization start point clecreased markedly as reaction temperature increased: considerabl~ priming remained at 5 C, much less at 10 C, and very little at 15 C. The average length o~ the DNA chains in the sequence ladder also increased with temperature, consistent with a reduced frequency of priming. This decrease in priming apparently reflects competition between the contiguously annealing hexamer set and SSB for binding the template DNA, as priming decreased little if at all over this temperature range in the absence of SSB.
In the absence of SS~, the dominant sequence at 0 oc reflects priming by the A2 hexamer at nucleotide 4193 of M13 DNA. This pattern is apparent up to 15 C but not at 25 or 30 C. The changes in priming pattern with increasing kemperature in the absence of SSB presumably reflect differences in pairing stabilities of the four different hexamers, and differences in local context or conformation of the template DNA at the different priming sites. A ~hi~t to longer DNA chains is evident at temperatures of 15 C and higher, presumably reflecting a decreasing overall priming efficiency.

Order of addition. Equilibration of hexamer binding at the desired position in the template appeared to take place rapidly in the presence of SSB at 0 C: labeling of sequence ladders was similar when the 5~min labeling reaction was initiated 2.5, 5, 10, 20 or 30 min after 2~ 03a3~

hexamers were added to a pre-equilibrated mixture of DNA
and SSB, and labeling was only slightly less when initiated immediately after adding the hexamers.
Essentially identical patterns and kinetics of lab~ling were obtain~d wh~n SSB wa8 add~d to a pre-equilibr~ted mixture of DNA and hexamer~, indicating that e~uilibrium is established rapidly with either order of addition.
In one experiment, weak secondary priming was detected when labeling was initiated 5 or 10 min a~ter SSB
was added to a pre-equilibrated mixture of DNA and hexamers, but not 20 min after, whereas no secondary priming was observed when the hexamers were added last.
Although any dif~erences in sequence ladders due to order of addition appear to be slight, the procedure routinely employed is to add hexamers after SSB to minimize the potential for s~condary priming.

Gener~lity of priming reaction ~ examer priminq in M13 DN~. To test the generality of priming by sets of contiguously annealing hexamers in the presence of SSB, 20 different sets of contiguously annealing hexamers, containing from 4 to 18 contiguous hexamers complementary to 15 different regions of M13 or M13mpl8 DNA, were synthesized. The haxamers in the first three sets of contiguously annealing hexamers (sets A, B
and C) are listed in Table 1. Most subsequent setæ of contiguously annealing hexamers were built around an already available hexamer whose complement is found at more than one site in M13 DNA. Thus, hexamer B1 is the same as A4, and hexamer Cl is the same as A2. Altogether, these 20 sets contained a total of 119 different hexamers of widely different composition.

: 21~3~0 , ~

~ - , __ . ,, _ _ . _ I
Numb0r Sequence Position in Other ~5'-3') M13 DNAt Sites~
. . . _ I
~6 ~CCCCC ~194 0 l __ ~ . -.. I

_ .. . _ __ ._ ._. _~ ., _ . .
A3 AI~GCGC 117 6 2 _ ~,.",_ . , _ .
A2 GAAACA 117 0 6 ¦
_ _ _ . .__ 11 , A1 AAGTAC 1164 3 ¦
-._ ~ - _~ ~ - ~ __ ' ~
B6 TACCTT 830 4 :~
. _ ~ . _ _._ .................... .
B4 TTTTAA 818 9 ~ :q.~
~ -. ._.. _ _~ ._ ._....... - ,, ~ _ . .
Bl TATACC 800 1 .; __ _ ~. = _ _ ~ ~ . -- . _ .. . . .. _ ._ _ ., :`
C5 GTAACA ~383 5 ..
, . . ~

~ . .. _ - ~.__w_ .

. _ _ , , _ . =----_ __. .-- ,,, _ ~ _ -- ~ , _ , =
Table 1. Sequences of hexamers comprising ontiguously annealing hexamer sets A, B and C in Ml3 DNA.
* Numbers decrease in the 5' to 3' dixection, so that :;:
the hexamer with the lowest number is at the 3' end o~ a contiguously annealing hexamer set.

``~ 21~3~0 t The nucleotide in M13 DNA that is complementary to the 3' nucleotide of the hexam~r at the desired polymPrization start site. Priming proceeds toward lower numbers.
~ Number of sites in M13 DNA complementary to the hexamer at positions other than the desired annealing location.
In contiguously annealing hexamer s0ts A, B and C, each single hexamer and every ~ubset of contiguously annealing hexamers of two to ~ix hexa~er units in length was tested for ability to prime ~equencing reaction~ in the presence and absence of SSB. In the remaining 17 contiguously ann~aling hexamer sets, most subsets of contiguously annealing hexamers of two, three and four hexamer units in length were tested in the presence of SSB. A total of 63 con~iguously annealing hexamer sets or subsets of two, 7~ contiguously annealing hexamer sets or subsets of three, and 55 contiguously annealing hexamer sets or subsets of four wer2 tested in the 19 contiguously annealing hexamer sets excluding set D. Sequence ladders were analyzed to determine the specificity of priming.
In the absence of SSB, sequence ladders were generally weak and ambiguous, whether primed by individual hexamers or by any o~ the contiguously annealing set~ or subsets of hexamers. Exceptions include hexamer A6, which primed moderately well at :its single priming site in M13 DNAI and the contiguous pair B3-B2, which primed selectively as a pair even though neither hexamer by itself primed significantly at this site (and each is complementary to three additional sites in M13 DNA).
In the presence of SSB, priming by individual hexamers was almost always strongly suppressed~ Priming by some contiguous pairs o~ hexamers was also suppressed, but about 40% of those tested primed to greater or l~sser extents specifically as a pair in the presence of SSB.
Examples include the A6~-A5 pair, the B3-B2 pair and the -:` 2:~a3~:3~

C5-C4 pair. Most sets or subsets of contiguously annealing hexamers having a length of three or four h~xamer units were stimulated by SSB to prime intensely and specifically at the desired poly~erization start point. Sets of more than four hexamers did not ssem to offer any advantage.
The seguence ladders obtained in the presence of SSB
were usually primed almost ~xclusively by the hexamer at the 3' end of the contiguously annealing hexamer set, as shown by a shift of the sequence ladder by six nucleotides with the addition or subtraction o~ a hexamer at the 3' but not the 5' end of the contiguously annealing hexamer set. However, in many cases priming could also be observed by one or two internal hexamers, producing superimposed se~uence ladders six nucleotides apart.
Substantial amounts of such double priming are evident in the patterns generated by B4-B1, B3-Bl, C6-C3, C5-C3 and C4-C1. In a few cases, the 3' hexamer of the contiguously annealing hexamer set primed weakly if at all, and priming was predominantly or almost exclusively at the next hexamer. Thirteen of the 70 contiguously annealing hexamer sets or subsets of three (19%), and 16 of the 55 contiguously annealing hexamer sets or subsets of four (29%) had enough double priming to make reading of the sequence ladder difficult.
Two other problems interfered with determining sequence primed by contiguously annealing hexamer sets;
weak priming and priming at secondary sites other than the desired polymerization start site. A few sets of contiguously annealing hexamers primed so weakly that 2-5 day exposures of the autoradiogram were required to read the saquen~e. Relatively weak priming by several sets o~
contiguously annealing hexamers of three hexamers increased substantially when a fourth contiguously annealing hexamer was added. Significant interference by 2~3~30 priming at secondary sites was observed in three cases affecting ll contiguously annealing hexamer sets or ~ubsets of three or four hexamer units in length.
Ladd~rs ~rum which sQquence could be read unambiguously without di~ficulty were obtained from 49 of the 70 contiguously annealing hexamPr sets or subsets of three hexamer units (70%) and 33 of the 55 contiguously annealing hexamer sets or ~ubset~ o~ four hexamers (~0%).
At least some sequence infor~ation could be obtained ~rom many of the other ladder~ a well. The most frequent problem, overlapping ladders primad by two or more hexamers in a striny, might well be resolvable by computer analysis to generate reliable sequence information.

Inter~erence by base-pairing in template DNA. The l~ set of contiguously annealing hexamer primers referrsd to as hexamer set D is a special case not included in the above analysis. This set was built from a site complementary to hexamer A3 at nucleotide 6446-6451 in Ml3mpl8 DNA and extended initially to comprise a contiguously annealing set o~ six hexamers. Unlike the other ~ets of` contiguously annealing hexamers, none of the combinations of these six hexamers primed a sequence ladder from the desired polymerization start site.
Examination of the template sequence revealed a perfect 11-base palindrome plus considerable potential for additional base pairing that might compete directly against pairing of the~e hexamers with template DNA. The M13mpl8 DNA template sequence complementary to hexamer set D, as well as th~ relative annealing positions of primer set D primers, is shown in Fig. 1.
Hexamer set D was extended in both directions by the addition of contiguously annealing hexamers in an e~fort to extend the duplex region away from the in~luence of the intramolecular base-paired structure in the template DNA.

.: :-.. .: :~.. ... : . ~ . .... .. : .. -; . .... .. . ..

- - ;

~: ` 2:~33~0 It was determined that specific priming was obtained with contiguously annealing hexamer set D14-D11 on the upstream side and set Dl'-D4' on the downstream side, whereas set DlO-D7, which substantially overlaps the region of potential pairing, did not prime specifically from the desired polymerization start siteO Thus, competition from base~pairing in th~ template DN~ seems to pr~vent priming by hexamer setsO
SSB is thought to remove most base-paired structures from single-stranded DNA (Meyer and Laine, Microbiological ~eviews 5~, 342 (1990)), but the potential structure at the position of contiguously annealing hexamer set D may be too stable to be removed by ~SB under the conditions used for priming sequencing reactionæ. In an effort to remove the secondary structure from the template molecule, the template DNA was heated in the presence of SSB. SSB
is known to be very thermostable and the rationale was that heating might allow SSB to stabilize the unfolded structure and stimulate priming. However, heating the mixture of primer~, DNA an~ SSB to temperatures as h.igh as 90 C before attempting sequencing reactions at 0 C did not promote specific priming by hexamer string D4-Dl.
Perhaps the structure in the template DNA can form again after cooling in the presence of SSB, or perhaps displacement of SSB in the process of forming a contiguous hexamer string allows the structure to form and displace the hexamers.
Since structure in the template DNA forms by intramolecular association but contiguously annealed primer sets form by intermolecular associations, it was reasoned that priming miqht be ~avored by increasing the concentration of hexamers. However, increasing the hexamer concentration tenfold, from the usual 5 ~M to 50 ~M was not sufficient to promote specific priming by the primer subset D4~D1.

-- 21~3~

Although strong local base pairing in the template DNA seems to prevent priming by hexamer sets, inspection of the sequence of template DNA in th~ region where priminq i~ de~ired should allow most ~uch problem areas to be identified and avoided.

Priming by hexamer setæ ln denatured double-stranded DNAs. Conditions ~or priming by contiguously annealing hexamer sets were developed using the naturally single-stranded M13 viral DNA, but the goal was to prime directly on the sinqle strands from double-s~randed DNAs o~ at least cosmid length (40,000 bp or larger). The contiguously annealing hexamer set A4-Al primed speci~ically on heat- or alkali-denatured linear or supercoiled forms of double-stranded M13 DNA, demonstrating that the presence of the complementary strand in the reaction mixture does not prevent speci~ic priming by contiguously annealing hexamer sets. Good sequence ladders were also obtained from a heat-denatured 2.1-kbp PCR product from T7 gene 5.
To test a DNA in the size range of cosmid DNAs, priming was attempted with contiguously annealing hexamer sets at three different regions in T7 DNA. T7 DNA is a linear dou~le-stranded DNA of 39,937 base pairs whose eequence is known. Standard reaction conditions contained 0.6-1 ~g of denatured T7 DNA, which provided 2.3-3.8 nM of unique priming sites, a concentration range where the intensity of sequence ladders primed on M13 DNA had decreased but was still substantial. In each region of T7 DNA, contiguously annealing hexamer sets of three or four hexamers primed specific sequence ladders that were usually readable after overnight exposure of the autoradiograms. Some sets of contiguously annealing hexamers primed very weakly or primed double or triple ladders, which represent problems similar to those . 2~3~3 observed in M13 DNA. Increasing the hexamer concentration as much as 10 fold slightly changed the distribution of priming in a triply primed ladder but did not reduce priming to a single site.
Different procedures were tested for converting T7 DNA to single strands for sequencing reactions.
Equivalent sequence ladders were obtained after the following tr~atments o~ khe DNA. 2 min at 100 C before adding lOx reaction buffer; 2 min at 100 C in th~
presence of SSB (which is highly thermostable); 2 min at 100 C or 5 min at room temperature in 50 mM NaOH, 50 mM
NaCl followed by neutralization at O ~C. In almost all of our experiments, DNA was denatured by heating, either in the presence or absence of SSB.
~ince a single strand of T7 DN~ is over six times the length of M13 DNA, an experiment was designed to test whether reducing the size of the DNA containing the specific priming site ~or a contiguously annealing hexamer set would have any effect on priming. Reducing the size to 20, 14, 7 or 4 kbp by cutting the DNA with different restriction enzymes before denaturation had no detectable effect on the sequence ladders obtained.
As an initial test of primer walking with contiguously annealing hexamer sets on a cosmid-sized DNA
of unknown sequence, the DNA of LPP-l was used. LPP-1 is a T7-like cyanobacteriophage (Sherman and Haselkorn, J.
Virol . 6, 841 (1970)) having a sequence which we had partially determined. To prime synthesis, 34 contiguously annealing hexamer sets, each of four hexamer units in length, were designed to prime within blocks of known sequence and to prime synthesis into unknown reqions.
These contiguously annealing hexamer sets were chosen at a relatively early stage in the analysis of primary sequence information, and 7 of them were later found to be unsuitable because of sequence errors at the position to :- 2~33~a which the primers were designed to anneal, or because the sequence chos~n was present at more than one site in LPP-1 DNA. Of the remaining 27 sets, 24 have given readable sequence ladders whose guality ranged from Pair to excellent. 5SR itself was suc~e~sful in about half of these cases, and addition o~ T7 gene 2~5 protein produced readable ladders in the others. The longest read from one of these sequencing reactions ~o far is 461 nucleotides, but the sequencing reactions were not optimized ~or long reads, and only ~ome reaction~ have been analyzed under conditions that allow reading as far as possible. We continue to optimize priming by hexamer strings on LPP-l DNA and expect to complete the sequence entirely by primer walking with contiguously annealing hsxamer sets.

SSB inhibition of primers o~ different lengths Priming by contiguously annealing hexamer sets is effective because SSB both stimulates priming by the hexamer ~ets and suppresses priming by individual hexamers at other sites in the DNA. To explore the limits of the effectiveness of SSB in suppressing priming, a set of oligonucleotides of increasing length was synthesized, which are complementary to M13 DNA ak the position of contiguously annealing hexamer set B and have the same 3' nucleotide as hexamer B2. The B2 hexamer has ~our complementary binding sites in M13 DNA, but the heptamer and longer oligonucleotides have only one perfectly complementary site. Priming by 5 ~M oligonucleotide on 0.6 ~g of Ml3 DNA was tested in the presence of 0, 2 or 5 ~g SSB under standard reaction conditionsO
In the absence of SSB, maximum priming efficiency was reached at primer lengths of 9 or greater; weaker specific priming was seen by the octamer or heptamer, and only very weak priming was apparent ~or the hexamer. Adding 2 ~g SSB only slightly suppressed priming by the heptamer and 2 1 ~

~eemed to enhance priming by the oligonucleotides of lengths 8 to 11. Adding 5 ~g SSB rather skrongly suppressed priming by oligonucleotides up to length 8, moderately suppressed priming by those of length 9 and 10, and appeared not to suppress priming by those of length 11 and longer. Increasing the temperature to 22 or 37 C had relatively mild effects, increasing the length of primer needed for maximum efficiency by only one nucleotide or so, and only moderately increasing the suppression by SSB.
I0 ~lthough only one ne~ted set oE primers was tested, these result~ suggest that interaction between contiguous hexamers in a contiguously annealing hexamer set need not be very great to drive the astablishment of priming complexes in the presence of SSB.

Nucleotide sequence of_hexamers in contiquously annealing .
hexamer~s Initially, contiguously annealing hexamer sets were selected without applying specific criteria of base composition or sequence. As more information b~came available, some sets were built to test the priming behavior of individual hexamers in other sets. The contiguously annealing hexamer sets for LPP-1 DNA were selected to contain hexamers predicted to have a relatively high affinity for template DNA (Breslauer et 25 al., Proc. Natl. Acad. Sci. USA 8~, 3746 (1986); Quartin and Wetmur, Biochemis~ry 28, 1040 (1989)). In total, more than 200 hexamers have been used in contiguously annealing hexamer sets that primed success~ully in more than 45 regions in three different template DNAs. These hexamers ~0 have a wide range of sequence and composition.
In analyzing priming behavior in a particular contiguously annealing set of seven hexamers, it was observed that TAATAA did not prime effectively as the 3' hexamer in a contiguously annealing set of three or four ~` 21~3~0 hexamers. This hexamer also failed to stimulate priming as the 5' hexamer in a contiguously annealing set of three hexamers, but functioned internally in contiguously annealing sets of three or four. On the basis of this observation, ~our additional contiguously annealing hexamer sets were built which conkained TAATAA or ATTATT.
It was determined that both hexamers behaved similarly in each contiguously annealing cet. Another contiguously annealing hexamer set happened to contain TTAATT, and this hexamer likewi~e did not function at the 3' end of contiyuously annealing hexamer sets of three or four, but did function internally. Five other hexamers that contained only A and T were used successfully internally in contiguously annealing hexamer sets. The only one of them tested in the 3' position of a set was B4 (Table 1), which performed effectively there and internally. (A
determination of whether B4 stimulates priming in the 5' position could not be made because the B3-B2 pair primes effectively by itself.) It is possible that excluding hexamers containing only A and T from the end positions in selecting hexamer sets would improve the probability of successful priming.

;:- 2~3~0 E~fects of ~atc~es About a dozen instances of secondary priming outsid~
of contiguously annealing hexamex sets in the presence of SSB were identified in which a determination of the specific site of secondary priming could be made. In almost every case, the site of secondary priming was a perfect complement to one of the hexamers in the set of contiguously annealing hexamers t which was ~lanked by one or more contiguous but mismatched pairiny sites for the same hexamer or another hexamer in ~he mixture. Two cases were also observed where a second sequence ladder was primed six nucleotides past the 3' end of a contiguously annealing hexamer set, and two cases where a second ladder was displaced five nucleotides. When the ladder was shifted by six, one of the hexamers present in the mixture could pair with the six bases immediately past the 3' end of the contiguously annealed hexamer set with a single mismatch; when shi~ted by five, the ~ive nucleotides at the 3' end of the hexamer could pair with the five bases immediately past the 3' end of the annealed hexamer set.
These observations made it seem likely that a range of contiguous but partially mismatched hexamer sets would be able to prime sequence ladders to some extent in the presence of SSB. To test the effects of mismatches more systematically, all possible single-base mismatches in each of the three hexamers of contiguously annealing hexamer set A3 to A1 were tested for their ef~ect on priming under standard conditions (3 ~g SSB). No hexamer with any mismatch in Al, the 3' hexamer of the annealed hexamer set, stimulated priming significantly. On the other hand, eight of the 18 possible mi~matches in A2 and 12 of 18 in A3 primed correct sequence, the intensity o~ :
the sequence ladder ranging from very weak to moderate.
Extending the analysis to other contiguously annealing hexamer sets revealed that all 18 mismatches in B4 primed :" 21~3~

correct sequ~nce in combination with hexamers B5 and B3, again ranging from very weak to moderate levels. ~ven more striking, all 18 mismatches in the middle hexamer of yet another contiguously annealing hexamer set primed sequence ladders almost as int~nsely as the perfectly matched hexamer.

Primin~ w th contiguously~ a_alina ~ets of pPntamers or he~tamers If sets o~ sontiguously annealing pentamers would also prime sequencing reactions specifically, the size of the library needed for ePficient sequencing would be

4-fold smaller than that needed for hexamers. To te~t this possibili-ty, a contiguously annealing hexamer set of seven contiguous pentamers complementary to M13 DNA at the same position as the hexamer set A was synthesized. Under standard conditions, where contiguously annealing hexamer sets primed intensely, priming by contiguously annealing sets of pentamers could be detected but was relatively weak and ambiguous. Decreasing the reaction temperaturs to -2.5 C or -5 C, or increasing the primer concentration 10-fold to 50 ~M did not provide much improvement. Pentamer sets alone appear unlikely to be useful in the priming methods described.
Priming with a set of five contiguously anneling heptamers was also tested. T7 DNA was used as template, and the primers were tested individually and in all ~-contiguous combinations, in the presence and absence of SSB. Priming by the individual heptamers was generally weak in khe absence of SSB and similar or slightly enhanced in its presence. Contiguous sets of two or more heptamsrs all stimulated priming from the expected polymerization start point in the ab ence or presence of SSB. It was observed that the sequence ladders seemed somewhat better in the presence of SSB. A heptamer ,,.. ,i~ 2l03~a~

flanked by two hexamers, and a hexamer flanked by two heptamers primed clean ~equence ladders in the presence of SSB.

Other obl~a~E~inn~
Modified T7 nNA pol~merase (5equenase) initiates DNA
chains preferentially and perhaps exclusively from the 3' hexamer of a contiguously annealing hexamer set. The double and occasionally triple ~equence ladders observed invariably arose from priming by the hexamer(s) adjacent to the 3' hexamer in the contiguously annealing hexamer set. The factors influencing the frequency of double priming are not yet well understood but presumably involve relatively weak binding of the 3' hexamer or weak -~
interaction with the rest of the contiguously annealing hexamer set. Three different hexamers containing only A
and T residues appeared to prime very poorly as the 3' hexamer in different contiguously annealing hexamer sets.
Attempting to eliminate double priming by increasing the concentration of the 3' hexamer has been only marginally effective in the few cases it has been tested.
Although thè conditions for priming with conti~uously annealed hexamers were worked out with modified T7 DNA
polymerase and ~. coli SSB, other polymerases and other single-stranded DNA-binding proteins might be found to work as well or better. Preliminary experiments indicate that the ba~teriophage T4 gene 32 protein and the T7 gene 2.5 protein are both capable of stimulating specific priming by hexamer strings, at least under some conditions, and might offer some advantages in combination with the E. coli SSB. If a DNA polym~rase and S5B that are both thermostable could prime from contiguously annealed primers, rapeated cycles of synthesis and d~naturation might be used to obtain sequence ladders from much smaller concentrations of template DNA. Since ~ 2~3~0 , hexamers sets provide great specifieity (there are almost 70 billion possible 18-mers), ~uch amplification might allow sequencing directly on DNAs much larger than cosmids.
Preliminary experiments also indicate that SSB can stimulate template-directed ligation of contiguously annealed short oligonucleotides. Although perhaps offering no advantages over direct priminq for D~A
equencing, the ability to ligate short oligonucleotides might be very useful in other ~pplications.

Equivalents Those skilled in the art will know, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. These and all other equivalents are intended to be encompassed by the following claims.

2l03~ao SEQUENCE LISTING
(1) GENERAL INFORMATION:
(i) APPLICANT:
(A) ADDRESSEE: A~ociated Univer~iti~ c.
(B) STREET: 1400 16th Street, N~W., Sulte 730 (C) CITY: W~hington (D) 8TATE: Di~trict of Columbia (E~ COUNTRY: USA
~F~ POSTAL CODE (ZIP): 20036 (ii) TITLE OF INVENTION: Method or Pr~moting Speclfic Alignment of Short Oligonu~leotide~ on Nucl~ic Acid~
(iii) NUMBER OF SEQUENCES: 1 (iv) CORRESPONDENCE ADDRESS:
~A3 ADDRESSEE: Broo~haven National Laboratory (B) STREET~ Building 902C
(C) CITY: Upton (D) STATE: NY
(E) COUNTRY: USA
(F) ZIP: 11~73 ~v) CO~PUTER READABLE FORM:
~A) MEDIUM TYPE: Floppy dl~k (B) COMPUTER: I8M PC compatibl~
(C) OPEPATING SYSTEM: PC-DOS/MS-DOS
(D) SOFTWARE: PatentIn Releaqe ~1.0, Ver~ion #1.25 (vi) CURRENT APPLICATION DATA~
(A) APPLICATION NUMBER: 07/916,062 (B) FILING DATE: July 17, 1992 ~viii) ATTORN~Y/AGENT INFORMATION:
tA) NAME: Margaret ~ogo~ian (B) REGISTRATION NUNBER: 25,324 ~C) REFERENC~/DOCRET NUMBER: AUI92-11 .
~ix) TELECOMMUNICATION INFORMATION:
~A) TELEPHONE: 516 282-3341 ~B) TELEFAX: 516 282-3000 :
~2) INFORMATION FOR SEQ ID NO:l:
~i) SEQUENCE CHARACTERISTICS:
~A) LENGTH: 108 ba~2 pair~
~B) TYPE: nuclei~ a~id ~C) STRANDEDNESS: doubl~
(D) TOPOLOGY: linear ~li) MOLECULE TYPE: DNA
~xl) S~QUENCE D~SCRIPTION: SEQ ID NO:l:
GTTGCGCAGC CTGAATGGCG AATGGCGCTT TGCCTGGTTT CCGGCACCAG A~GCGGTGCC 60

Claims (47)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE PROPERTY OR PRIVILEGE
IS CLAIMED ARE DEFINED AS FOLLOWS:

1. A method for promoting the specific alignment of a plurality of short oligonucleotides on a nucleic acid polymer, comprising:
a) providing a nucleic acid polymer having a predeter-mined contiguous series of nucleotides;
b) forming an incubation mixture comprising:
i) the nucleic acid polymer;
ii) a saturating amount single-stranded DNA bind-ing protein;
iii) a plurality of short oligonucleotides which are perfectly complementary to distinct but adjacent regions of the predetermined con-tiguous series of nucleotides;
c) incubating the incubation mixture at a temperature of from about 0° C to about 15° C, under conditions appropriate for:
i) annealing of the oligonucleotides to the pre-determined contiguous series of nucleotides;
and ii) binding of the single-stranded DNA-binding protein to the nucleic acid polymer.

2. A method of Claim 1 wherein the short oligonucleotides have a length of between about 5 to 10 nucleotides.

3. A method of Claim 1 wherein the single-stranded DNA-binding protein is selected from the group consisting of the E. Coli single-stranded DNA-binding protein, the gene 32 protein of bacteriophage T4, the gene 2.5 pro-tein of bacteriophage T7 or combinations of same.

4 A method for priming nucleic acid polymerization, com-prising:
a) providing a single-stranded nucleic acid template having a predetermined contiguous nucleotide se-quence of at least about 12 nucleotides in length, the order and identity of each of the 12 nucleo-tides being known;
b) forming an incubation mixture comprising:
i) the nucleic acid template;
ii) a saturating amount single-stranded DNA bind-ing protein;
iii) a plurality of short oligonucleotides which are perfectly complementary to distinct but adjacent regions of the predetermined con-tiguous nucleotide sequence; and iv) a polymerase enzyme;
c) incubating the incubation mixture at a temperature of from about 0° C to about 15° C, under conditions appropriate for:
i) annealing of the oligonucleotides to the pre-determined contiguous series of nucleotides;
ii) binding of the single-stranded DNA-binding protein to the nucleic acid polymer; and iii) primer directed polymerization initiated at the 3' end of the contiguously annealed pri-mers.

5. A method of Claim 4 wherein the nucleic acid template is deoxyribonucleic acid.

6. A method of Claim 5 wherein the polymerizing enzyme is a DNA-directed DNA polymerase.

7. A method of Claim 4 wherein the nucleic acid template is ribonucleic acid.

8. A method of Claim 7 wherein the polymerizing enzyme is a RNA-directed DNA polymerase.

9. A method of Claim 4 wherein the nucleotide primers are selected from the group consisting of pentamers, hexa-mers and heptamers.

10. A method of Claim 4 wherein the plurality of short oligonucleotides consist essentially of three hexamers which anneal to the single stranded nucleic acid tem-plate to form a contiguous region of double stranded nucleic acid of 18 base pairs in length.

11. A method of Claim 9 wherein the plurality of short oligonucleotides consist essentially of four hexamers which anneal to the single stranded nucleic acid tem-plate to form a contiguous region of double stranded nucleic acid of 24 base pairs in length.

12. A method of Claim 9 wherein the plurality of short oligonucleotides consist essentially of two hexamers and one heptamer which anneal to the single stranded nucleic acid template to form a contiguous region of double stranded nucleic acid of 19 base pairs in length.

13. A method of Claim 4 wherein the single-stranded DNA-binding protein is selected from the group consisting of the E. coli single-stranded DNA-binding protein, the gene 32 protein of hacteriophage T4, the gene 2.5 pro-tein of bacteriophage T7 or combinations of same.

14. A method of Claim 13 wherein the single-stranded DNA-binding protein is E. coli single-stranded DNA-binding protein, and the mass ratio of E. coli single-stranded DNA-binding protein to DNA is from about 2.5 to about 10.

15. A method for determining the identity and order of nucleotides in a single-stranded deoxyribonucleic acid template the deoxyribonucleic acid template having a predetermined contiguous nucleotide sequence of at least about 12 nucleotides in length, comprising:
a) providing a single-stranded deoxyribonucleic acid template having a predetermined contiguous nucleo-tide sequence of at least about 12 nucleotides in length, the order and identity of each of the 12 nucleotides being known;
b) forming an incubation mixture comprising:
i) the deoxyribonucleic acid template;
ii) a saturating amount single-stranded DNA bind-ing protein;
iii) a plurality of short oligonucleotides which are perfectly complementary to distinct but adjacent regions of the predetermined con-tiguous nucleotide sequence; and iv) a polymerase enzyme;
c) incubating the incubation mixture at a first tem-perature of from about 0° C to about 15° C, under conditions appropriate for:
i) annealing of the oligonucleotides to the pre-determined contiguous series of nucleotides;

ii) binding of the single-stranded DNA-binding protein to the deoxyribonucleic acid template;
and iii) primer directed polymerization initiated at the 3' end of the contiguously annealed prim-ers with periodic chain termination;
d) following the initial low temperature polymeri-zation of step c) which stabilizes the hybrid formed between the priming strand and the template of interest, raising the temperature of the incu-bation mixture to a second temperature which more closely approximates the optimum temperature for the polymerizing enzyme; and e) analyzing the products of step d) to determine the identity and order of nucleotides in the single stranded deoxyribonucleic acid template.

16. A method of Claim 15 wherein the polymerizing enzyme is selected from the group consisting of T7 DNA polymerase and modified T7 DNA polymerase.

17. A method of Claim 15 wherein the short oligonucleotides are selected from the group consisting of pentamers, hexamers and heptamers.

18. A method of Claim 17 wherein the plurality of short oligonucleotides consist essentially of three hexamers.

19. A method of Claim 17 wherein the plurality of short oligonucleotides consist essentially of four different hexamers.

A method of Claim 17 wherein the plurality of short oligonucleotides consist essentially of two hexamers and one heptamer.

21. A method Claim 15 wherein the single-stranded DNA-bind ing protein is selected from the group consisting of the E. coli single-stranded DNA-binding protein, the gene 32 protein of bacteriophage T4, the gene 2.5 protein of bacteriophage T7 or combinations of same.

22. A method of Claim 21 wherein the single-stranded DNA-binding protein is E. coli single-stranded DNA-binding protein and the mass ratio of single-stranded DNA-bind-ing protein to DNA is from about 2.5 to about 10.

23. A method for synthesizing cDNA from MRNA, the MRNA
template having a predetermined contiguous nucleotide sequence of at least about 12 nucleotides in length, the method comprising:
a) providing a MRNA template having a predetermined contiguous nucleotide sequence of at least about 12 nucleotides in length, the order and identity of each of the 12 nucleotides being known;
b) forming an incubation mixture comprising:
i) the MRNA template;
ii) a saturating amount single-stranded DNA bind-ing protein;
iii) a plurality of short oligonucleotides which are perfectly complementary to distinct but adjacent regions of the predetermined con-tiguous nucleotide sequence; and iv) a polymerase enzyme;

c) incubating the incubation mixture at a first tem-perature of from about 0° C to about 15° C, under conditions appropriate for:
i) annealing of the oligonucleotides to the pre-determined contiguous series of nucleotides;
ii) binding of the single-stranded DNA-binding protein to the MRNA template; and iii) primer directed polymerization initiated at the 3' end of the contiguously annealed prim-ers;
d) following the initial low temperature polymeri-zation of step c) which stabilizes the hybrid formed between the priming strand and the template of interest, raising the temperature of the incu-bation mixture to a second temperature which more closely approximates the optimum temperature for the polymerizing enzyme.

24. A method of Claim 23 wherein the nucleotide primers are selected from the group consisting of pentamers, hexa-mers and heptamers.

25. A method of Claim 24 wherein the plurality of nucleotide primers consist essentially of three hexamer nucleotide primers.

26. A method of Claim 24 wherein the plurality of nucleotide primers consist essentially of four hexamer nucleotide primers.

27. A method Claim 24 wherein the plurality of nucleotide primers consist essentially of two hexamer nucleotide primers and one heptamer nucleotide primer.

28. A method of Claim 23 wherein the single-stranded DNA-binding protein is selected from the group consisting of the E. coli single-stranded DNA-binding protein, the gene 32 protein of bacteriophage T4, the gene 2.5 pro-tein of bacteriophage T7 or combinations of same.

29. A method of Claim 28 wherein the mass ratio of single-stranded DNA-binding protein to DNA is from about 2.5 to about 10.

30. A method for template directed ligation of oligonucleo-tides, comprising:
a) providing a single-stranded nucleic acid polymer having a predetermined contiguous series of nucleo-tides, the order and identity of the nucleotides in the contiguous series of nucleotides being known;
b) forming an incubation mixture comprising:
i) the nucleic acid polymer;
ii) a saturating amount single-stranded DNA bind-ing protein;
iii) a plurality of short oligonucleotides which are perfectly complementary to distinct but adjacent regions of the predetermined con-tiguous series of nucleotides;
iv) a DNA ligase; and c) incubating the incubation mixture at a temperature of from about 0° C to about 15° C, under conditions appropriate for:
i) annealing of the oligonucleotides to the pre-determined contiguous series of nucleotides;
ii) binding of the single-stranded DNA-binding protein to the nucleic acid polymer; and iii) formation of phosphodiester bonds between adjacent oligonucleotides by the DNA ligase.

31. A method of Claim 30 wherein the nucleotide primers are selected from the group consisting of pentamers, hexa-mers and heptamers.

32. A method of Claim 31 wherein the plurality of nucleotide primers consist essentially of three hexamer nucleotide primers.

33. A method of Claim 31 wherein the plurality of nucleotide primers consist essentially of four hexamer nucleotide primers.

34. A method of Claim 31 wherein the plurality of nucleotide primers consist essentially of two hexamers nucleotide primers and one heptamer nucleotide primer.

35. A method Claim 30 wherein the single-stranded DNA-bind-ing protein is selected from the group consisting of the E. coli single-stranded DNA-binding protein, the gene 32 protein of bacteriophage T4, the gene 2.5 protein of bacteriophage T7 or combinations of same.

36. A method of Claim 35 wherein the single-stranded DNA
binding protein is E. coli single-stranded DNA binding protein and the mass ratio of single-stranded DNA-bind-ing protein to DNA is from about 2.5 to about 10.

37. A hexamer oligonucleotide library comprising at least about 1500 unique hexamer sequences.

38. A hexamer oligonucleotide library comprising about 4096 unique hexamer sequences.

39. A reagent kit comprising single-stranded DNA-binding protein, an oligonucleotide primer library and a poly-merizing enzyme.

40. A reagent kit of Claim 39 wherein the polymerizing enzyme is a DNA-directed DNA polymerase.

41. A reagent kit of Claim 40 wherein the DNA-directed DNA-polymerase is a modified T7 DNA polymerase which lacks exonuclease activity.

42. A reagent kit of Claim 41 wherein the single-stranded DNA-binding protein is selected from the group consist-ing of the E. coli single-stranded DNA-binding protein, the gene 32 protein of bacteriophage T4, the gene 2.5 protein of bacteriophage T7 or combinations of same.

43. A reagent kit of Claim 42 wherein the oligonucleotide primer library is a hexamer oligonucleotide library comprising about 4096 unique hexamer sequences.

44. A method of Claim 13 wherein the single-stranded DNA-binding protein is E. coli single-stranded DNA-binding protein, and the mass ratio of E. coli single-stranded DNA-binding protein to DNA is from about 2 to about 200.

45. A method of Claim 21 wherein the single-stranded DNA-binding protein is E. coli single-stranded DNA-binding protein and the mass ratio of single-stranded DNA-bind-ing protein to DNA is from about 2 to about 200.

46. A method of Claim 28 wherein the mass ratio of single-stranded DNA-binding protein to DNA is from about 2 to about 200.

47. A method of Claim 35 wherein the single-stranded DNA
binding protein is E. coli single-stranded DNA binding protein and the mass ratio of single-stranded DNA-bind-ing protein to DNA is from about 2.5 to about 10.