Innovir Laboratories, Inc., 510 East 73rd Street, New York, NY 10021, USA

Received June 10, 1997; Revised and Accepted July 28, 1997

ABSTRACT

3-ethoxy-1,2,4-dithiazoline-5-one (EDITH) was recently introduced as an efficient sulfurizing reagent for solid-phase oligonucleotide synthesis. The successful syntheses were performed using standard base protecting groups (i.e. benzoyl for A and C, isobutyryl for G), which required deprotection in concentrated ammonium hydroxide at 55^oC for 15-18 h. We have explored the possibility of using EDITH in combination with fast deprotection chemistry (e.g. Expedite Chemistry using tert-butylphenoxy acetyl as a base protecting group). Surprisingly, poor synthesis performance was observed when syntheses were conducted with EDITH, Expedite Chemistry and standard synthesis cycle (i.e. Coupling-Thio-Cap). Potential G modification seemed to be the source of incompatibility since sequences containing no G or carrying isobutyryl- protected G residues could be synthesized with high efficiency. However, the deleterious G modification can be readily eliminated by inserting a capping step before the sulfurization reaction. Oligomers prepared with the Coupling-Cap-Thio-Cap cycle contained few phosphodiester contaminants as measured by ³¹P-NMR, anion-exchange HPLC and MALDI-TOF mass spectrometry. In addition to reducing deprotection time, this new combination also provides a mild method for the preparation of certain phosphorothioate oligomers that may be sensitive to prolonged ammonia treatment (e.g. thioated RNAs).

Phosphorothioate (PS) oligonucleotides are an important class of modified analogues (1 ,2 ). Introduction of the non-bridging sulfur atom into phosphodiester linkages is typically achieved by chemical synthesis via either the phosphoramidite approach (3 ,4 ) or the H-phosphonate approach (5 -7 ). Among various sulfurizing reagents that are commercially available, 3H-1,2-benzodithiol-3-one 1,1-dioxide (Beaucage reagent) has been the most widely used (8 ). However, this reagent suffers from problems associated with sustained solubility and/or stability in acetonitrile, which often cause precipitation in the instrument even when silanized bottles are used. Recently, a new sulfurizing reagent, 3-ethoxy-1,2,4-dithiazoline-5-one (EDITH), was introduced by Xu et al. (9 ,10 ). In addition to reported rapid and efficient sulfurization, this reagent should alleviate the precipitation problem encountered with the Beaucage reagent. However, the reported syntheses were performed with standard chemistry using benzoyl for A and C, and isobutyryl for G as protecting groups for the exocyclic amino functions. As a result, a time-consuming deprotection step (e.g. concentrated ammonium hydroxide at 55^oC for 15-18 h) was required for post-synthesis processing.

During the course of development of modified oligoribonucleotides as external guide sequences (EGSs) (Ma, M.Y.-X. et al., manuscript in preparation), we have been using fast deprotection chemistry [i.e. t-butylphenoxyacetyl (tBPA) Expedite Chemistry] (11 ), which not only reduces the lengthy ammonia treatment to <30 min, but can also prevent premature chain cleavage and 2' <-> 3' chain migration in RNA synthesis (12 ). In this communication, we describe the initial complications experienced with EDITH in combination with Expedite Chemistry using standard synthesis protocols (2 ,8 ,9 ), and a simple solution which results in fully-thioated oligonucleotides with high efficiency and little phosphodiester (PO) contaminants.

Oligomer 1 (Table 1 ) was the first sequence synthesized, at 1 [mu]mol scale, using EDITH as sulfurizing reagent and tBPA as base protecting group for A, C and G. Because of the presence of 2'-O-methyl residues, the sulfurization conditions for oligoribonucleotides were adapted from previous protocols (10 ). Furthermore, the capping reaction was carried out after sulfurization as is typically practiced for DNA thioates (2 ,8 ,9 ). Upon completion of synthesis and 30 min deprotection in concentrated ammonium hydroxide, the crude deprotection mixture was analyzed by capillary gel electrophoresis (CGE) (Fig. 1 A). Although the amount of the crude material was apparently not reduced, only 43% of full-length oligomer was generated from this synthesis. Extension or reduction in sulfurization time as well as variation in EDITH concentrations did not change the final result (data not shown). To ensure that the phosphoramidites and all other reagents employed were of high quality, synthesis was repeated with 0.1 M Beaucage reagent as the sulfurizing reagent for 5 min. As shown in Figure 1 B and Table 1 (oligomer 2), 77% of full-length material was detected in the crude mixture, indicating the effectiveness of each synthesis step.

Table 1 . Use of EDITH as a sulfurizing reagent in combination with Expedite Chemistry

Oligomer	Sequence	Thio conditions	Crude (mg)	% FLa
1	GsAsGs GsAsAs AsCsGs CsCsGs C	0.05 M/2 min	3.87	43
2	GsAsGs GsAsAs AsCsGs CsCsGs C	0.10 M/5 minb	4.05	77
3	d(GsAsGs GsAsAs AsCsGs CsCsGs C)	0.05 M/30 s	3.45	27
4c	d(GsAsGs GsAsAs AsCsGs CsCsGs C)	0.05 M/30 s	3.85	70
5	(Us)14U	0.05 M/2 min	4.30	73
6	(Cs)14C	0.05 M/2 min	3.46	81
7	(As)15(dT)	0.05 M/2 min	4.26	80
8	UsAsUs UsAsAs AsCsUs CsCsUs C	0.05 M/2 min	4.10	91
9	GsUsGs GsUsUs UsCsGs CsCsGs C	0.05 M/2 min	3.74	54
10d	GsAsGs GsAsAs AsCsGs CsCsGs C	0.05 M/2 min	3.88	82
11d	d(GsAsGs GsAsAs AsCsGs CsCsGs C)	0.05 M/30 s	4.28	74
12	d(GsAsGs GsAsAs AsCsGs CsCsGs C)	0.05 M/30 s	3.28	92
13	GsAsGs GsAsAs AsCsGs CsCsGs C	0.05 M/2 min	3.82	91
14e	UsGsAs GsGsAs AsAsCs GsCsCs GsUsUs C	0.05 M/2 min	31.8	85

Oligonucleotides (shown 5' to 3', uppercase letters represent 2'-O-methyl ribonucleotides and boldfaced letters represent 2'-deoxyribonucleotides) were prepared on an Applied Biosystems (ABI) (Foster City, CA) model 394 DNA/RNA synthesizer according to previously published procedures (2 ,11 ,18 ), using controlled pore glass (CPG) as the solid support matrix. 2'-deoxy and 2'-O-methyl RNA phosphoramidites with tBPA protection and standard base protecting groups were purchased from PerSeptive Biosystems (Framingham, MA) and ABI, respectively. Standard synthesis reagents were purchased from ABI with the exception of the capping reagents (t-butylphenoxyacetic anhydride/tetrahydrofuran as Cap A) which were purchased from PerSeptive Biosystems. 3H-1,2-benzodithiol-3-one 1,1-dioxide was obtained from Glen Research (Sterling, VA) and 3-ethoxy-1,2,4-dithiazoline-5-one (EDITH), from PerSeptive Biosystems.
s indicates phosphorothioate linkages.
^aPercentage of full-length (FL) oligomer in crude deprotection mixture was determined by CGE.
^b3H-1,2-benzodithiol-3-one 1,1-dioxide was used as the sulfurizing reagent.
^cThis synthesis was performed using identical conditions as oligomer 3, except standard base protecting groups were employed.
^dThese syntheses were performed using tBPA as base protecting group for A and C and isobutyryl for G.
^eThis synthesis was performed at 10 [mu]mol scale.

To investigate whether these synthesis failures were unique to sequences containing 2'-O-methyl modification, a DNA version of oligomer1 (Table 1 , oligomer 3) was synthesized. Again, a low percentage of full-length oligomer was produced. However, when phosphoramidites carrying standard base protecting groups (i.e. benzoyl for A and C, isobutyryl for G) were used, high quality synthesis was achieved for DNA oligomer 4(Table 1 ),which is in accordance with the previous success reported by Xu et al. (9 ). These data clearly indicate that EDITH is an effective sulfurizing reagent when syntheses are conducted with standard chemistry, but not Expedite Chemistry using the standard synthesis cycle (Coupling-Thio-Cap).

In order to identify the source(s) of incompatibility, a series of all 2'-O-methyl homo-oligomers were synthesized. As illustrated in Table 1 , high percentages of full-length oligomers were obtained for syntheses derived from 2'-O-methyl U, C and A. Since G-rich oligomers tend to form aggregates which can complicate analysis (13 ), we prepared a variant of oligomer 1 in which all G residues were replaced with U residues (e.g. oligomer 8, Table 1 ). Interestingly, this G replacement led to a dramatic improvement in synthesis efficiency. In contrast, poor synthesis performance was observed when G residues were re-installed and A residues were substituted with U residues (oligomer 9, Table 1 ). G modification was further confirmed when synthesis of oligomer 10 (replacing G^tBPA with G^iBu deprotection in concentrated NH₄OH at 55^oC for 16 h) was carried out successfully. It is noteworthy to mention that prolonged ammonia treatment is not the reason for the observed changes in the synthesis of oligomer 10 because similar treatment of oligomer1 did not increase the amount of full-length oligomer produced (data not shown). G modification is most likely the source of incompatibility for the DNA series as well, as demonstrated by efficient synthesis of oligomer 11 (A and C were protected by tBPA, and G by the isobutyryl group, Table 1 ).

Figure 1. (A) Analysis by CGE of the crude mixture upon deprotection of oligomer 1. Synthesis was performed at 1 [mu]mol scale using Expedite Chemistry, standard synthesis cycle (Coupling-Thio-Cap) and EDITH (0.05 M in CH₃CN/2 min) as the sulfurizing reagent (Table 1). Fast deprotection was conducted in concentrated ammonium hydroxide at 55^oC for 30 min. CGE analysis was carried out on a Beckman P/ACE 5000 system according to manufacturer's instructions. (B) Electropherogram of crude oligomer 2 which has the same sequence as oligomer 1. All conditions were identical except 3H-1,2-benzodithiol-3-one 1,1-dioxide (Beaucage reagent) was used for the sulfurization reaction (0.1 M in CH₃CN/5 min).

G modification is not a new problem in oligonucleotide synthesis. For example, Pon et al. (13 ,14 ) described explicitly the addition of tetrazole-activated phosphoramidites to the O⁶ position in G residues. This side-reaction can lead to many serious problems such as base modification, chain cleavage, chain branching, etc. Although significant efforts have been devoted to the development of an ideal O⁶ protecting group (13 -16 ), a simple solution is to insert an additional capping step before the oxidation reaction (13 ). Weak nucleophiles such as N-methylimidazole can break down the highly unstable addition species, and regenerate unmodified guanine residues.

As a result of these studies, a capping step of 15 s, followed by an acetonitrile wash (3 * 10 s), was added to the synthesis cycle before the EDITH-mediated sulfurization reaction. When DNA oligomer 3 was re-synthesized using this modified cycle (oligomer 12, Table 1 ), the percentage of full-length product in the crude mixture was increased from 27 to 92%. Similarly, repeated synthesis of the 2'-O-methyl oligomer1 (oligomer 13) was also very effective. To make sure that the Coupling-Cap-Thio-Cap cycle is applicable to sequences containing all four bases, oligomer 14 (Table 1 ) was synthesized at 10 [mu]mol scale with a sulfurization time of 2 min. Analysis of the crude oligomer by CGE (Fig. 2 ) indicates that EDITH is a highly effective sulfur-transfer reagent, producing 85% of full-length oligomer which implies an overall efficiency in the range of 99% per synthesis cycle. Analysis conducted on MALDI-TOF mass spectrometry also displayed a sharp peak (5584.45) which matches the calculated molecular weight (5584.80) of the intended product, suggesting the authenticity of the oligomer generated (data not shown).

Figure 2. CGE analysis of crude oligomer 14. Synthesis was performed at 10 [mu]mol scale using Expedite Chemistry, EDITH (0.05 M in CH₃CN/2 min) and the modified Coupling-Cap-Thio-Cap cycle. Deprotection was conducted in concentrated ammonium hydroxide at 55^oC for 1 h.

For the modified Coupling-Cap-Thio-Cap cycle to be widely accepted as a routine method, it is of primary importance to determine the efficiency of sulfur incorporation, given the fact that premature oxidation may occur during the capping reaction (8 ). Three analytical methods were employed to determine the PO/PS ratios. First, crude mixture of oligomer 14 was analyzed by ³¹P-NMR. To ensure a good signal to noise ratio, 25 mg crude oligomer was used in the experiment. The spectrum was recorded on a JOEL GX-400 spectrometer using 85% phosphoric acid as external reference. Multiple PS peaks were observed near 56 p.p.m. due to the presence of numerous diastereomers. Integration of this region and the region near 0 p.p.m. (phosphodiesters) gave an area to area ratio of 302:1, indicating the presence of 0.33% PO contaminants (data not shown). Secondly, the same crude mixture was analyzed by anion-exchange HPLC. Systematic optimization of a previously published protocol (17 ) generated a much improved anion-exchange HPLC method which can separate the all-PS oligomer 14 and three standards with high resolution (Fong, G.W. et al., unpublished results). The three standards had the same length and base composition as oligomer 14 but containing one phosphodiester linkage at three different positions (3'-end, 5'-end and the middle position). The anion-exchange data (76.55% of the full-length all-PS oligomer 14 in the crude mixture) revealed that a stepwise sulfurization efficiency of >99% was achieved under these conditions (assuming the stepwise coupling yield was 99% as estimated by trityl measurements and CGE) (9 ,17 ). Thirdly, the absence of peaks corresponding to lower molecular weights (due to desulfurization, S -> O switching) in the MALDI-TOF mass spectrometry analysis (data not shown) is also a good indication of efficient sulfurization although this technique is not quantitative at this time.

In summary, results presented in this study have demonstrated that EDITH can be a highly effective sulfurizing reagent in combination with Expedite Chemistry if a capping step is inserted before the sulfurization reaction. Without this additional capping step, significant modification may occur at least in the case where the G residues are protected by the tBPA group. Insertion of this capping step prior to sulfurization does not cause the formation of additional PO contaminants. This simple change in the synthesis cycle provides a rapid and mild method for the preparation of certain PS-containing oligonucleotides which may be sensitive to the harsh deprotection conditions when overnight ammonia treatment is used (e.g. thioated RNAs, PS oligomers carrying various tags) (10 ).

ACKNOWLEDGEMENTS

We would like to thank Mr Jeffery Zonderman and Dr Subhasish Purkayastha (PerSeptive Biosystems) for a generous gift of EDITH and Dr Michael Blumenstein (Hunter College, New York) for ³¹P-NMR analysis.

REFERENCES

1 Eckstein, F. (1985) Annu. Rev. Biochem., 54, 367-402. MEDLINE Abstract

2 Zon, G. (1993) InAgrawal, S. (ed.), Protocols for Oligonucleotides and Analogs: Synthesis and Properties. Humana Press, Totowa, New Jersey,pp. 165-189.

3 Stec, W.J., Zon, G., Egan, W. and Stec. B. (1984) J. Am. Chem. Soc., 106, 6077-6079.

4 Connolly, B.A., Potter, B.V.L., Eckstein, F., Pingoud, A. and Grotjahn, L. (1984) Biochemistry, 23, 3443-3453. MEDLINE Abstract

5 Garegg, P.J., Regberg, T., Stawinski, J. and Strömberg, R. (1985) Chemica Scripta,25, 280-282.

6 Froehler, B.C. (1986) Tetrahedron Lett., 27, 5575-5578.

7 Agrawal, S. and Tang, J.-Y. (1990) Tetrahedron Lett., 31, 7541-7544.

8 Iyer, R.P., Phillips, L.R., Egan, W., Regan, J.B. and Beaucage, S.L. (1990) J. Org. Chem., 55, 4693-4701.

9 Xu, Q., Musier-Forsyth, K., Hammer, R.P. and Barany, G. (1996) Nucleic Acids Res., 24, 1602-1607. MEDLINE Abstract

10 Xu, Q., Barany. G., Hammer, R.P. and Musier-Forsyth, K. (1996) Nucleic Acids Res., 24, 3643-3644. MEDLINE Abstract

11 Sinha, N.D., Davis, P., Usman, N., Perez, J., Hodge, R., Kremsky, J. and Casale, R. (1993) Biochimie,75, 13-23. MEDLINE Abstract

12 Chaix, C., Molko, D. and Teoule, R. (1989) Tetrahedron Lett.,30, 71-73.

13 Pon, R.T., Usman, N., Damha, M.J. and Ogilvie, K.K. (1986) Nucleic Acids Res., 14, 6453-6470. MEDLINE Abstract

14 Pon, R.T., Damha, M.J. and Ogilvie, K.K. (1985) Nucleic Acids Res., 13, 6447-6465. MEDLINE Abstract

15 Kamimura, T., Tsuchiya, M., Urakami, K., Koura, K., Sekine, M., Shinozaki, K., Miura, K. and Hata, T. (1984) J. Am. Chem. Soc.,106, 4552-4557.

16 Gaffney, B.L., Marky, L. and Jones, R.A. (1984) Tetrahedron,40, 3-13.

17 Bergot, B.J. and Egan, W. (1992) J. Chromatogr.,599, 34-42.

18 Beaucage, S.L. and Caruthers, M.H. (1981) Tetrahedron Lett.,22, 1859-1862.

---

*To whom correspondence should be addressed. Tel: +1 212 327 9127; Fax: +1 212 249 4881
CiteULike    Connotea    Del.icio.us    What's this?

This Article Abstract Print PDF (50K) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (5) Commercial Re-use Guidelines for Open Access NAR Content Google Scholar Articles by Ma, M. Articles by George, S. T. Search for Related Content PubMed PubMed Citation Articles by Ma, M. Articles by George, S. T. Social Bookmarking          What's this?