WO1996008578A2

WO1996008578A2 - Improved procedure for the solid phase synthesis of oligonucleotides

Info

Publication number: WO1996008578A2
Application number: PCT/US1995/011302
Authority: WO
Inventors: Radhakrishnan P. Iyer; Sudhir Agrawal
Original assignee: Hybridon, Inc.
Priority date: 1994-09-06
Filing date: 1995-09-05
Publication date: 1996-03-21
Also published as: CA2199246A1; AU3546995A; WO1996008578A3; EP0779892A2

Abstract

An improved method of solid phase oligonucleotide synthesis is presented. The method results in lesser contamination of the final product with N-1, N-2, and other failure sequences. The method comprises employing a second capping step after initial loading and capping of the solid support and before addition of the second nucleoside. Further diminution in the N-1 oligonucleotide content can be realized when this second capping stepped is used in combination with loading the support to higher densities than have previously been deemed acceptable and by cleaving the full length oligonucleotide from the solid support using mildly basic conditions.

Description

IMPROVED PROCEDURE FOR THE SOLID PHASE SYNTHESIS OF

OLIGONUCLEOΗDES

BACKGROUND OF THE INVENTION Field of the Invention

This invention relates to the field of solid phase synthesis of oligonucleotides. Description of the Related Art

The synthesis of oligonucleotides, in most instances, is carried out on a solid-phase matrix, such as controlled pore glass (CPG), using phosphoramidite chemistry. Beaucage and Iyer, Tetrahedron Report 48(12), 2223-2311 (1992). One of the first steps in oligonucleotide synthesis is the attachment of the first nucleoside to the solid support. A variety of protocols exist for the preparation of support-bound nucleosides. Protocols for Oligonucleotides and Analogs, S. Agrawal, Ed., Humana Press, 1993. Once the first nucleoside has been attached to the support, it is important to "cap" the unreacted sites on the support to prevent subsequently added nucleosides from binding to the support rather than to the growing oligonucleotide chain. Beaucage and Iyer, Tetrahedron Report 48(12), 2223-2311 (1992) and Protocols for Oligonucleotides and Analogs, S. Agrawal, Ed., Humana Press, 1993. Incomplete capping of CPG during initial preparation of support- bound nucleoside or uncapping of its nucleophilic sites during extended storage are factors which could conceivably affect both the quality and yield of an oligonucleotide. Thus, for example, incompletely capped CPG can give rise to an oligonucleotide with a greater proportion of N-1 (i.e., oligonucleotide of one fewer nucleotide in length than the desired, or "N," oligonucleotide) and other "failure sequences". Most failure sequences, except N-1 and N-2, can be separated easily from the desired N oligonucleotide by reverse-phase HPLC. At the end of the synthesis, the crude oligonucleotide is obtained in two steps: the first step involves the cleavage of the oligonucleotide from the support and the second step involves complete deprotection of the phosphate (e.g. cyanoethyl) and base protecting groups (e.g. benzoyl and isobutyroyl) by treatment with 28-30% ammonium hydroxide. Two different protocols are commonly used: (a) one in which the support- bound oligonucleotide is treated with ammonia at ambient temperature for 1-2 hours, followed by heating the supernatant ammonia solution in a closed tube for 8-16 hours at 55°C ("SUP" protocol) (Protocols for Oligonucleotides and Analogs, S. Agrawal, Ed., Humana Press 1993) or (b) the CPG is removed from the synthesis column and directly heated with ammonia in a closed tube for 8-16 hours at 55°C ("CPG" protocol) (Protocols for Oligonucleotides and Analogs, S. Agrawal, Ed., Humana Press 1993). The latter procedure is more convenient, especially in large-scale synthesis of oligonucleotides (10 μmol - 1 mmol scale).

Only with the advent of capillary-gel electrophoresis has the technology advanced to the point where N-1 sequences could be distinguished from N sequences. Our earlier work regarding the analysis of a purified oligonucleotide revealed the presence of significant amounts of N-1 content in the oligonucleotide of a given length N. (unpublished observations) Sequence analysis of the N-1 oligonucleotide revealed a complex heterogeneous population of oligomers (unpublished observations). Quite clearly, several factors, including efficiency of synthesis, nature of the solid support, work- up and purification protocols etc., contribute towards the heterogenous nature of the N-1 oligonucleotide. To date, there has been no systematic study of the effect of these factors on the purity of oligonucleotide end-product. Improved methods of obtaining desired synthetic oligonucleotides, free from N-1 and N-2 sequences, are desirable. All patents and other prior art cited within this specification are hereby incorporated by reference in their entirety.

SUMMARY OF THE INVENTION

The present invention provides an improved method of solid phase synthesis of oligonucleotides. In one aspect of the invention, a second capping step is introduced after the first nucleoside has been bound to the solid support. We have unexpectedly found that this second capping step decreases the amount of N-1 oligonucleotide detected after synthesis is complete.

In a second aspect of the present invention, we have unexpectedly found that implementation of the "SUP" protocol (as defined below) in combination with the second capping results in a further decrease in the amounts of N-1 and other failure sequences. Both the SUP protocol and the "CPG" protocol (also defined below) are methods of removing the synthesized oligonucleotide from the solid support and deprotecting the bases and phosphate linkages. Because the CPG protocol is easier, there has been a tendency in the prior art to use it (and similar methods) for the cleavage step. We have discovered, however, that such protocols result in higher N-1 content. The SUP protocol uses milder conditions to cleave the synthesized oligonucleotide from the solid support. It is theorized that the milder conditions preferentially cleave the desired N oligonucleotide from the support.

In a third aspect of the present invention, it has been unexpectedly found that the initial loading density of the first nucleoside can be substantially greater than previously thought. The prior art theorizes that growing oligonucleotides on highly loaded supports prevents diffusion of reactants to the growing chain ends, resulting in a greater number of failure sequences as compared to supports with lower loads. We have found, however, that high loading densities, when used in combination with the second capping step alone or in addition to the SUP protocol, results in a final product having a smaller N-1 content than when lower loading levels are used.

The foregoing merely summarizes some aspects of the present invention and is not intended, nor should it be construed, as limiting the invention in any way.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 displays a schematic representation of the generalized method of solid phase oligonucleotide synthesis.

Figure 2 displays a schematic representation of solid support functionalization, loading, and capping. Figures 3A-B display ion-exchange profiles of dT₅ synthesized on a T-CPG column

(53 μmol/g) worked up using the SUP protocol, with a second capping (3A) and without one (3B).

Figure 4 displays a histogram of the percentage N-1 content in a dT₅ homopolymer purified with the SUP protocol as a function of initial nucleoside loading and second capping of the CPG.

Figure 5 displays a histogram of the percentage N-1 content in a dT₅ homopolymer purified with the CAP protocol as a function of initial nucleoside loading and second capping of the CPG.

Figure 6 displays the capillary electrophoretic profile of a crude phosphorothioate oligonucleotide (25-mer) synthesized on a 1 mmol scale and worked up using the SUP protocol.

Figure 7 displays the chemistry of adding the first nucleoside to the CPG.

Figure 8 displays the chemistry of trityl exchange.

Figures 9A-B display the capillary electrophoretic profile of two 25-mer phosphorothioate oligonucleotides using the SUP protocol with (top) and without

(bottom) second capping.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

The present invention provides improved methods for solid phase synthesis of oligonucleotides. The methods yield desired products having less contamination from N-1 and other failure sequences. As used herein, an "N" oligonucleotide is the synthetic oligonucleotide of the desired length. An "N-1" oligonucleotide is 1 less nucleotide in length as compared to the N oligonucleotide. Similarly, an "N-2" oligonucleotide is 2 less nucleotides in length as compared to the N oligonucleotide. The term "failure sequence" generally refers to oligonucleotides synthesized on the solid support but having fewer nucleosides in length than the oligonucleotides of desired length. The term "load" as used herein refers to the amount of first nucleoside attached to the solid support, expressed in units of (μmol nucleoside)/(g support material), or μmol/g. Figure 7 displays two illustrative methods by which a solid support may be loaded. We have systematically studied the effects on N oligonucleotide recovery of different protocols for oligonucleotide synthesis, including the amount of initial nucleoside loaded, the effect of a second capping step after loading the CPG column but before oligonucleoside synthesis, and the effect of the duration and temperature of ammonium hydroxide treatment to cleave the oligonucleotide from the solid support. The results presented herein demonstrate for the first time that a second capping step decreases the amount of N-1 nucleoside produced. In addition, combination of a second capping step with milder cleavage conditions, as well as higher nucleoside loading densities, all lessen the contamination of the desired oligonucleotide with N-1, N-2, and other failure sequences. The following is a description of the generalized protocol for synthesizing an oligonucleotide of N nucleotides in length (also displayed in Fig. 1), wherein N is an integer between 3 and 150: a) a first nucleoside is chemically linked to a solid support having reactive sites, forming a support-bound mononucleoside and a mononucleoside- bound support with reactive sites, b) the mononucleoside-bound support with reactive sites is contacted with a capping solution, forming a capped mononucleoside-bound support, c) a second nucleoside is chemically linked to the support-bound mononucleoside, forming a nascent oligonucleotide-bound support and a support-bound nascent oligonucleotide, d) the nascent oligonucleotide-bound support is contacted with a capping solution, e) another nucleoside is chemically linked to the support-bound nascent oligonucleotide, forming a nascent oligonucleotide-bound support and a support-bound nascent oligonucleotide being one nucleotide greater in length, f) d) and e) are repeated until the nascent oligonucleotide is N nucleotides in length, g) the N nucleotide long nascent oligonucleotide is cleaved from the solid support.

As used herein, the term "reactive site" refers to terminal amino and hydroxyl groups capable of forming a chemical bond with incoming nucleosides during oligonucleotide synthesis. The term "nascent oligonucleotide" refers to any oligonucleotide of 2 or more nucleotides in length that must be subjected to further treatment to obtain the desired oligonucleotide. The "capping solution" is one or more solutions that combined comprise a reactive mixture capable of linking a blocking/protecting group to unreacted reactive sites. Contacting the capping solution with the support results in some or all of the unreacted reactive sites being chemically linked to a blocking/protecting group, thereby becoming "capped."

The general protocol for loading nucleoside and capping the non-functionalized hydroxyl groups is schematically displayed in Figure 2. The solid support initially has terminal hydroxyl groups that may be functionalized with a long chain amino-alkyl- triethoxy silane. Not all terminal hydroxyl groups become functionalized, however. The DMT protected nucleoside is attached to the solid support via the amino moiety of a long chain alkyl amine, resulting in a support-bound nucleoside and, conversely, a nucleoside-bound support. This is followed by capping any reactive sites on the solid support. Capping prevents subsequent nucleosides from attaching directly to the support rather than the growing oligonucleotide chain. Often the capped, nucleoside-bound support is stored for future use.

The prior art has assumed that this first capping step is complete. We have found, however, that the addition of a second capping step (a process herein called "second capping") decreases the amount of N-1 oligonucleotide. While applicants do not wish to be bound by any theory, we hypothesize that the initial capping step is not complete or that "uncapping" may occur during handling and storage, leaving unreacted hydroxyl and amino groups that may react with incoming nucleosides. This is depicted in Figure 2. Presumably, the second capping step results in capping of some or all of the remaining uncapped hydroxyl and amino moieties, also as shown in Figure 2. Thus, in a first embodiment of the present invention, we provide an improved method of capping a mononucleoside-bound solid support. In one aspect of this embodiment, therefore, we provide a method of capping a mononucleoside-bound support comprising consecutively contacting the mononucleoside- bound support with same or different capping solutions two or more times. In the most preferred embodiment, two cappings (i.e., contacting the mononucleoside-bound support with a capping solution) are conducted. The term "consecutively" means that the mononucleoside-bound support is subjected to temporally separated capping procedures. In a second aspect of this embodiment, we provide an improved method of synthesizing an oligonucleotide, the improvement comprising contacting a capped mononucleoside-bound support with a capping solution before a second nucleoside is added.

In this embodiment of the present invention, a second capping step is performed after the initial loading of nucleoside, but before addition of the second nucleoside. This second capping step can be by any means that renders the free hydroxyl and amino groups unreactive to incoming nucleoside. A number of such means are known in the art and may be used in the present invention. Kg., Methods in Molecular Biology, Vol 20: Protocols for Oligonucleotides and Analogs (S. Agrawal, Ed., Humana Press, 1993). Any non-acid labile capping group can be used in the present invention. In the most preferred embodiment, the capping is accomplished by adding to the loaded support equal volumes of Cap A and Cap B reagents (acetic anhydride and N-methylimidazole, respectively, or acetic anhydride and DMAP, respectively), shaking or stirring for about two hours followed by filtering and washing with dichloromethane. The second capping step may use the same materials and protocol as the first step or use different materials and a different protocol. The second capping step may immediately follow the first capping step, or it may be employed some time later, such as after storage of the loaded support. In the most preferred embodiment, the second capping step is carried out just before the remaining nucleotides are to be added.

In a second embodiment of the present invention, a milder cleavage step is employed in combination with the second capping step to remove the full length oligonucleotide from the solid support. Any method of cleavage that detaches the full length oligonucleotide from the solid support without significantly or substantially altering one's ability to obtain the desired full length oligonucleotide can be employed. The cleavage step will preferably preferentially cleave the full length N oligonucleotide from the solid support without cleaving from the support N-1 oligonucleotides and other failure sequences. Although we do not wish to be bound by any theory, we believe that some N-1 sequences arise due to linkage of the second nucleoside to uncapped support- bound amino groups to form phosphoramidate linkages. These linkages are more difficult to cleave than the ester linkage that connects the support-bound succinic acid moiety to the ribose moiety of the first nucleoside at the 3' position. Thus, the cleavage step of the present invention will comprise any suitable method that is able to cleave preferentially the desired N oligonucleotide from the support-bound succinic acid without substantially effecting cleavage of N-1 oligonucleotides linked to the support via phosphoramidate linkages.

Another possible explanation for how N-1 failure sequences arise is that a trityl exchange reaction occurs. Figure 8. It is conceivable that the acetylated groups in the capped CPG matrix could participate in trityl exchange reaction through neighboring group participation, resulting in a tritylated amino group and a capped nucleoside. Detritylation of the amino group could then result in the synthesis of N-1 sequences.

More stringent cleavage conditions result in higher yields of desired N oligonucleotide, but also in higher levels of contamination by N-1 and other failure sequences. Milder conditions result in a lesser contamination by N-1 and other failure sequences. Using the methods described below, it is a routine matter to vary the cleavage conditions to obtain an acceptable combination of yield and purity.

This aspect of the present invention comprises using what has herein been called the "SUP" protocol for cleavage of oligonucleotides from the solid support and deprotection of the base and sugar phosphate moieties. The SUP protocol comprises contacting the support-bound oligonucleotides with 5-30% ammomum hydroxide at ambient temperature for 30 minutes to 2 hours, followed by heating the supernatant in a closed tube for 8-16 hours at 55°C. Heating the supernatant is for the purpose of deprotecting the base and sugar moieties and does not affect the concentration of N-1 oligonucleotides obtained. Cleavage is accomplished by contacting the nascent oligonucleotide-bound support with a weak base. In preferred embodiments, the oligonucleotide is cleaved from the support with from about 10% to about 30% (saturated) ammonium hydroxide for about 30 minutes to 2 hours. In the most preferred embodiment, about 28-30% ammonium hydroxide is used for about 1 to 2 hours. It is a routine matter to vary the conditions under which the N oligonucleotide is cleaved from the solid support to adjust yield and purity of the desired N oligonucleotide to an acceptable level. Harsher conditions and treatment for longer times will increase yield and decrease purity; less harsh conditions will decrease yield, but increase purity. The higher the concentration of ammonium hydroxide and the higher the temperature at which the support-bound oligonucleotides are subject to the ammomum hydroxide, the harsher the conditions. Thus, by increasing the temperature above ambient temperature and/or using higher concentrations of ammonium hydroxide, the apparent yield of N oligonucleotide is increased. But, as mentioned before, this will also increase contamination from N-1 and other failure sequences. Conversely, using lower temperatures and lower concentrations of ammonium hydroxide will result in lesser contamination by N-1 and other failure sequences as well as lesser yield. As noted before, the period of heating the supernatant is for the purpose of cleaving the base and sugar phosphate protecting groups. Thus, only varying the conditions to which the support-bound oligonucleotides are subject affects the N-1 oligonucleotide level. It is a routine matter to adjust the ammonium hydroxide concentration and temperature consistent with the objectives of the particular synthesis.

In another embodiment of the invention, decreased contamination from N-1, N-2, and other failure sequences is obtained by increasing the initial nucleoside loading density. The term "loading" as used herein refers to chemically linking the first nucleoside to the solid support. The prior art teaches that too high levels of loading are undesirable because the growing oligonucleotide chains prevent diffusion of reactants, thereby resulting in more failure sequences. Agrawal, Protocol in Molecular Biology, supra. We have unexpectedly found, however, that when used in combination with the second capping step and with or without mild cleavage conditions, higher loading densities actually result in lesser contamination by N-1 failure sequences.

Once again, it is the case that decreasing the N-1 concentration results in a decrease in yield. Thus, it is up to the artisan to determine the appropriate nucleoside loading density that gives a satisfactory balance of yield and purity to suit his purposes. It is a routine matter to do so, however, by simply carrying out the desired oligonucleotide synthesis on several samples of loaded support having different loading densities. Methods for determining the loading density of a sample are known in the art. Methods of Molecular Biology, Vol 20, supra.

Initial loading densities are typically in the range of about 50 to 70 μmol/g. We have unexpectedly found that an initial loading density of 100 μmol/g resulted in a lesser amount of N-1 oligonucleotide content than oligonucleotides synthesized on supports having loading densities in the range of about 50 μmol/g.

The methods of the present invention are suitable for synthesizing any type of oligonucleotide capable of being synthesized on a solid support. Examples of oligonucleotides for which the present methods are suitable are RNA, 2'-0-methyl RNA, DNA or RNA/DNA hybrids, each of which can be unmodified or modified in any number of positions, including the sugar phosphate backbound and/or the nucleoside base. Examples of the types of oligonucleotides that may be synthesized and isolated with the methods of the present invention include, but are not limited to, phosphodiesters, phosphorothioates, phosphorodithioates, phosphoroamidates, methyl- or other alkyl-phosphonates, carbonates, carbamates and oligonucleotides having modified bases. Modifications may occur in any number of nucleotides and may occur alone or in any combination.

Oligonucleotide synthesis may be accomplished in any manner consistent with application of the presently disclosed methods. E.g., Methods in Molecular Biology, Vol. 20, supra; Uhlmann and Peyman, Chem Rev. 90, 543 (1990); Oligonucleotides and Analogues: A Practical Approach (REckstein, Ed., 1991). Although the Examples presented below disclose use of DMT-dT-succinic acid, any unmodified base (e.g., A, G, C, T, and U) may be used as can any modified base suitable for incorporation by solid phase synthesis. In addition, any suitable linking group that links the first nucleoside to the denvatized solid support via an ester bond may be used in the present invention. A number of linkers are known in the art. Beaucage and Iyer, Tetrahedron Report 48(12), 2223-2311 (1992). These include 1,10-decanediol bissuccinate and N-methyl sarcosine. For other examples, ≤s& Beaucage and Iyer. Tetrahedron 48, 2223-2311 (1992) Succinic acid is most preferred. Any solid support that can be or is derivatized for solid phase oligonucleotide synthesis can be used in the present invention. As used herein, the terms "derivatized" and "functionalized" are used interchangeably and, when used in relation to a solid support, mean that the support has reactive hydroxyl and/or amino moieties suitable for oligonucleotide synthesis. A number of such supports are known in the art. Among these are low cross-linking polystyrene, polyamide, polyamide bonded silica gel, cellulose, silica gel, controlled pore glass, polystyrene/PEG "tentacle" copolymer, and high cross- linking polystyrene. See Methods of Molecular Biology, Vol 20, supra, and references cited therein. In the preferred embodiments, the solid support is capable of being used in phosphoramidate synthesis of oligonucleotides. In the most preferred embodiment, the solid support is controlled pore glass (CPG).

EXAMPLES Example 1

Effect on N-1 Content of Pre-capping and Load Level Using the SUP Protocol Phosphoric diester homopolymer dT₅ was synthesized on a Milligen/Biosearch 8700 DNA synthesizer using standard 1 μmol synthesis cycles as recommended by the manufacturer, but with or without the incorporation of a second capping step just prior to the start of the synthesis. Several syntheses were conducted, each with a different CPG loading of 47, 48, 53, and 100 μmol/g, respectively. After the synthesis, each of the CPG-bound oligonucleotides was dried and subjected to the SUP protocol. In the SUP protocol, the CPG was treated with about 28-30% ammonia (1.5 ml for 1 μmol and 5 ml for 10 μmol scale) at ambient temperature for 30 to 120 minutes. The supernatant was removed and heated at 50-55°C for 8-10 hours.

After the SUP protocol, the isolated crude homopolymer was subjected to ion- exchange chromatography. Ion exchange chromatography was carried out on a Waters 660 E high pressure liquid chromatograph (HPLC) equipped with a Waters 996 photodiode Array Detector. The sample was dissolved in 0.1 M ammonium acetate and analyzed on a GEN-PAK FAX column (4.6 X 100 mm) at 45°C using a linear gradient of 100% -> 50% buffer A over 90 minutes at a flow rate of 0.5 ml/minute. The following buffers were used:

Buffer A: 25 mM Tris HC1; pH 8.5, 10% CH₃CN,

Buffer B: 25 mM Tris HC1, 1 M LiCl; pH 8.5, 10% CH₃CN.

Figure 3 shows the typical HPLC profile of a crude homopolymer dT₅ as analyzed on a

GEN-PAK FAX column. The retention times of the 5-mer (N), 4-mer (N-1) and 3-mer (N-2) were 31, 23 and 14 minutes respectively.

Figure 4 displays the N-1 content of the 5-mer (depicted as a percentage of the area of the N peak) expressed as a function of initial loading of the nucleoside on the

CPG and second capping versus single capping of the CPG just prior to synthesis. As can be seen, the N-1 content is less in oligonucleotides which were prepared using the SUP protocol with second capping (left column of each loading pair) as compared to those prepared using the SUP protocol without second capping (right column of each loading pair).

Example 2

Affect on N-1 Content of Pre-capping and Load Level Using the CPG Protocol The same experiment as presented in Example 1 was repeated using the CPG protocol. In the CPG protocol the CPG is directly heated with 28-30% ammonia (1.5 ml for 1 μmol at 50-55°C for 8-10 hours). The results are presented in Figure 5. The left column of each loading pair in Figure 5 represents the percentage N-1 when the CPG was subjected to a second capping and the right column represents the percentage of N-1 without second capping the CPG. Overall, it is seen that capping the capped, loaded support substantially reduces the amount of N-1. Combined with the results from

Example 1, it is seen that second capping decreases the N-1 content irrespective of whether the CPG or SUP protocol is used. Furthermore, on comparison of these results with those from Example 1, it is seen that the combination of second capping and the

SUP protocol yields less N-1 than does second capping with the CPG protocol.

Example 3 Affect on N-2 Content of Pre-capping and

Load Level Using the SUP and CPG Protocols

The same experiments were performed as described in Examples 1 and 2 and the

N-2 content was measured. The same results were obtained.

Example 4

Affect on N-1 and N-2 Content of Duration of Ammonia Treatment in the SUP Protocol

The same procedures as used in Examples 2 and 3 were employed to determine the affect on N-1 and N-2 content of the duration of ammonia treatment in the SUP protocol. The dT₅ homopolymer was synthesized with and without second capping treatment. CPG was removed from the column and divided into aliquots. Each aliquot was exposed to 28-30% ammonia treatment for 30, 60, and 120 minutes. The supernatant was removed and heated at about 50-55°C for 8 hours. The crude oligonucleotide was then analyzed by ion-exchange chromatography.

Figs. 3 - 5 are typical examples of the experimental data used to evaluate the Example 4. It was observed that the difference in N-1 content with and without second capping is 26% for 30 minutes of ammonia treatment and 20% for 2 hours of ammoma treatment. By comparison, the difference in N-2 content with and without second capping is 23% for 30 minutes of ammonia treatment and 14% for 2 hours of ammonia treatment. Thus, longer treatment with ammoma releases more N-1 and other failure sequences.

Example 5

Large-Scale (10 μmol) Synthesis of Oligonucleotides Following the procedures outlined in Examples 1 and 2, a 25-mer phosphodiester oligonucleotide was synthesized on a 10 μmol scale with and without second capping and with the SUP and CPG protocols. 5 ml of 28-30% ammonia was used for both the SUP and CPG protocols. Following synthesis, the oligonucleotide was purified by reverse- phase HPLC, detritylated, dialyzed, and analyzed by capillary electrophoresis. Standard procedures were used. .See. Agrawal, Protocols in Molecular Biology, supra. Capillary electrophorestic analysis (CE) was done on a Beckmann P/ace 2200 instrument. Before CE, the samples were desalted on a "SEP-PAK" cartridge. A decrease in N-1 and N-2 content was noted in the oligonucleotide subjected to the SUP and second capping protocol. The results are consistent with those previously obtained.

Example 6

Large-Scale (1 mmol) Synthesis of Oligonucleotides The same procedure as described in Example 5 was repeated on a 1 mmol scale using 28%-30% ammonia for the synthesis of a 25-mer oligonucleotide phosphorothioate. Three batches were prepared and aliquots of CPG divided and subjected to both SUP and CPG protocols. Figure 6 shows the capillary electrophoretic profile of a typical crude phosphorothioate oligonucleotide. The electrophoretic mobility of the N peak was 27.14 minutes and that of the N-1 peak was 26.7 minutes. A reduction in N-1 content of up to 30% was observed in aliquots subjected to the SUP protocol.

The reduction in the amount of failure sequences was also tested on the 1 mmol scale with two 25-mer phosphorothioate oligonucleotides using the SUP protocol with and without second capping. The same experimental protocols as previously described were used. The capillaiy electrophoresis profiles are displayed in Figure 9A-B. The top chromatograms in Figures 9A and 9B display the capillary electrophoresis analysis of the each 25-mer phosphorothioate synthesized with second capping. The bottom chromatograms show the capillary electrophoresis analysis of the same 25-mers synthesized without second capping. In both cases the relative amount of failure sequences (the smaller peaks) was less when the oligonucleotide was synthesized with second capping. In Figure 9A the amount of N-1 failure sequence relative to the amount of N sequence was 15% and 20% when synthesis was conducted with and without second capping, respectively. In Figure 9B the relative amount of N-1 failure sequence was 4% and 7% when synthesis was conducted with and without second capping, respectively.

Claims

What is claimed is:

1. An improved method of synthesizing an oligonucleotide on a solid support, the improvement comprising contacting a capped mononucleoside-bound support with a capping solution before a second nucleoside is added.

2. The method of Claim 1 wherein the improvement further comprises cleaving a support-bound oligonucleotide from the solid support by contacting support- bound oligonucleotide with a weak base at ambient temperature.

3. The method of claim 2 wherein the support-bound oligonucleotide is contacted with the weak base for about 30 minutes to about 2 hours.

4. The method of claim 2 wherein the support-bound oligonucleotide is contacted with the weak base for about 1 to about 2 hours.

5. The method of claim 4 wherein the weak base is about 5% to about 30% ammomum hydroxide.

6. The method of claim 4 wherein the weak base is about 28% to about 30% ammomum hydroxide.

7. The method of claim 1 wherein the improvement further comprises the support-bound mononucleoside being present at a density of greater than about 70

8. The method of claim 1 wherein the improvement further comprises the support-bound mononucleoside being present at a density of greater than about 90

μmol/g.

9. The method of claim 6 wherein the improvement further comprises the support-bound mononucleoside being present at a density of greater than about 70

10. The method of claim 6 wherein the improvement further comprises the support-bound mononucleoside being present at a density of greater than about 90 μmol/g.

11. The method of claim 9 wherein the support-bound oligonucleotide is contacted with the weak base for about 30 minutes to about 2 hours.

12. The method of claim 9 wherein the support-bound oligonucleotide is contacted with the weak base for about 1 to about 2 hours.

13. The method of claim 12 wherein the weak base is about 5% to about 30% ammomum hydroxide.

14. The method of claim 12 wherein the weak base is about 28% to about 30% ammomum hydroxide.

15. A method of synthesizing an oligonucleotide of N nucleotides in length on a solid support having reactive sites, wherein N is from about 3 to about 150, comprising: a) chemically linking a first nucleoside to a plurality of the reactive sites, forming a mononucleoside-bound support, b) contacting the mononucleoside-bound support with a capping solution, forming a capped mononucleoside-bound support, c) contacting the capped mononucleoside-bound support with a capping solution, forming a twice-capped, mononucleoside-bound support, d) chemically linking another nucleoside to the support-bound mononucleoside of the twice-capped, mononucleoside-bound support, forming a nascent oligonucleotide-bound support and a support-bound nascent oligonucleotide, e) contacting the nascent oligonucleotide-bound support with a capping solution, f) chemically linking another nucleoside to the support-bound nascent oligonucleotide, forming a nascent oligonucleotide-bound support and a support-bound nascent oligonucleotide being one nucleotide greater in length than the support-bound nascent oligonucleotide to which the nucleoside was linked, g) sequentially repeating e) and f) until the support-bound nascent oligonucleotide is N nucleotides in length, h) cleaving the support-bound nascent oligonucleotide from the solid support, yielding an oligonucleotide of N nucleotides in length.

16. A method according to claim 15 wherein the support-bound nascent oligonucleotide is cleaved from the support by contacting the nascent oligonucleotide- bound support with a weak base at ambient temperatures.

17. The method of claim 16 wherein the nascent oligonucleotide-bound support is contacted with the weak base for about 30 minutes to about 2 hours.

18. The method of claim 16 wherein the nascent oligonucleotide-bound support is contacted with the weak base for about 1 to about 2 hours.

19. The method of claim 18 wherein the weak base is about 5% to about 30% ammonium hydroxide.

20. The method of claim 18 wherein the weak base is about 28% to about 30% ammomum hydroxide.

21. The method of claim 15 wherein the improvement further comprises the support-bound mononucleoside being present at a density of greater than about 70 μmol/g.

22. The method of claim 15 wherein the improvement further comprises the support-bound mononucleoside being present at a density of greater than about 90 μmol/g.

23. The method of claim 20 wherein the improvement further comprises the support-bound mononucleoside being present at a density of greater than about 70 μmol/g.

24. The method of claim 16 wherein the improvement further comprises the support-bound mononucleoside being present at a density of greater than about 90 μmol/g.

25. The method of claim 24 wherein the nascent oligonucleotide-bound support is contacted with the weak base for about 30 minutes to about 2 hours.

26. The method of claim 24 wherein the nascent oligonucleotide-bound support is contacted with the weak base for about 1 to about 2 hours.

27. The method of claim 26 wherein the weak base is about 5% to about 30%

ammonium hydroxide.

28. The method of claim 26 wherein the weak base is about 28% to about 30% ammonium hydroxide.

29. A method of capping a mononucleoside-bound support comprising consecutively contacting the mononucleoside-bound support with same or different capping solutions two or more times.

30. The method of claim 29 wherein the mononucleoside-bound support is consecutively contacted with the same or different capping solutions twice.