3'-(0-2-cyanoethyl-N,N-diisopropyl)phosphoramidites (1), the corresponding ... which is analogous to nucleoside reagents bound to solid support, used as the ...
Nucleic Acids Research, Vol. 18, No. 8 2065
Backbone-modified oligonucleotides containing a butanediol-1,3 moiety as a 'vicarious segment' for the deoxyribosyl moiety synthesis and enzyme studies Andrzej Wilk, Maria Koziolkiewicz, Andrzej Grajkowski, Bogdan Uznanski and Wojciech J.Stec Polish Academy of Sciences, Centre of Molecular and Macromolecular Studies, Department of
Received December 21, 1989; Revised and Accepted March 19, 1990
ABSTRACT Sequential single replacement of nucleosides within the decanucleotide d[GGGAATTCCC] (7) by means of a butanediol-1 ,3 residue allowed us to obtain a set of ten decanucleotides containing 'vicarious' (V) carbonphosphate fragments. These analogues were further used as substrates for svPDE, nuclease PI and EcoRl endonuclease. Interestingly, replacement of any of the nucleosides within the canonical sequence ...GAATTC... by the butanediol-1 ,3 'vicarious' segment discriminates such constructs 8c-h as substrates for Eco-Ri enzyme.
INTRODUCTION The search for convenient routes of the synthesis of biopolymers bearing a carbon-phosphate backbone analogous to that in nucleic acids prompted Penczek et al.I to study the process of ringopening polymerization of 2-hydro-2-oxo-1,3,2-dioxaphosphorinanes. The resulting poly(trimethylene H-phosphonates) were subsequently oxidized leading to poly(trimethylene phosphates) of molecular weight 5 x 103. In the course of studies on the synthesis of biopolymers bearing a carbon-phosphate skeleton vicarious to the natural deoxyribose-phosphate fragment we2, and others3, have used different approaches, which rely upon the phosphite-triester method of oligonucleotide synthesis originally designed by Letsinger4 and elegantly developed by Caruthers5. Instead of the routinely used 5'-DMT-nucleoside 3'-(0-2-cyanoethyl-N,N-diisopropyl)phosphoramidites (1), the corresponding 1-0-DMT-butane-3-0-(-2-cyanoethyl-N,Ndiisopropyl)phosphoramidite (2) was employed. Reagent 2 was prepared by dimethoxytritylation of (i) butanediol-1,3 and subsequent phosphitylation of 1-O-dimethoxytrityl-butane-3-ol with [(N,N,N',N'-tetraisopropyl)2-cyanoethyl] phosphordiamidite. Its spectral and analytical characteristics are given in the Experimental Section. Use of 2 under conditions of routine manual oligonucleotide synthesis allowed us to introduce the butanediol-1,3 unit at any preselected position(s) in an oligonucleotide and thus to obtain oligonucleotides lacking the deoxyribonucleoside moiety at these preselected positions. In this way a series of modified oligonucleotides were obtained and characterized by means of enzymatic and spectroscopic methods. Since it was essential to establish solid support synthesis of
oligo(butanediol-1,3-phosphates) competitive with that based on the ring-opening polymerization, we have also obtained 1-0dimethoxytrityl-butanediol unit bound to the solid support (3), which is analogous to nucleoside reagents bound to solid support, used as the 'starter' reagent for oligodeoxyribonucleotide synthesis. Combination of reagents 1, 2 and 3 allowed us to synthesize oligomers which are listed in Tables 1 and 2. A drawback of our approach to the preparation of oligo(butanediol-1,3-phosphates), as compared with ring-opening polymerization', is the scale of synthesis, which is limited to /.tmole amounts and does not allow us to synthesize the desired material in bulk quantities. On the other hand, our approach has the advantages of flexibility, i.e. introduction of the vicarious segment at any desired, preselected position, reliability of getting oligomers with a precisely defined number of butanediol units (predeterminated by the number of steps of the synthesis), and the possibility of 'labeling' of our oligomers with nucleoside components, which allows the purification of synthesized material using an HPLC equipped with a UV detector.
DMTO
DMTOQ
/
0
0.,PCN
3
| CA CPG1
1
2
3
EXPERIMENTAL General. Eco RI endonuclease was isolated from over-producing strain, obtained from Prof. F. Eckstein, according to the procedure described by Bickle et.al. 1. Other enzymes were obtained from Boehringer. All solvents and chemicals were either DNA synthesis grade or of quality necessary in enzyme studies. HPLC was performed using LDC Milton Roy, NMR spectra were recorded on Bruker MSL 300, Tm values were measured on Specord M40 (Carl Zeiss Jena).
2066 Nucleic Acids Research, Vol. 18, No. 8 Table 1. Chromatographic and spectroscopic characteristics of di- and trinucleotide analogues dNpV, VpdN and dNpVpdN
dNpV 4
Elution time (mm)
VpdN 5
Elution time (min)
dNpVpdN 6
Elution time
N A
a
8.75
a
a
C
b
4.40
b
G
c
7.00
c
8.10 8.70 5.70 6.50 6.90 7.25 8.00 8.30
10.20 10.60 6.20 6.75 7.75 8.10 9.60 10.15
T
d
8.10
d
b c
d
31P-NMR of dNpVpdN*
(min)
6 (ppm)
0.522 0.494;
0.388 0.413 0.322
0.484; 0.450
0.618;
0.464; 0.492;
0.455 0.187 0.173
* Spectroscopic analysis of the diastereomers of compounds 6 was performed with a Bruker MSL 300 Spectrometer
Synthesis of 1-0-Dimethoxytrityl-butane-3-ol. Into the solution of (i)butanediol-1,3 (530 mg, 5.9 mmol) in pyridine (2.5 ml) was added the solution of dimethoxytrityl chloride (2 g, 5.9 mmol) in methylene chloride (5 ml). This reaction was maintained below -20°C to achieve desired regioselectivity. After 2h solvents were evaporated and the residue was purified on a silica gel column using diethyl ether-hexane (4:1) as an eluting system. The product was isolated as colorless resinous material (yield 2g, 90%; tic (diethyl ether-hexane 1:1), Rf=0.40; 'H NMR 6 0.93, 2H, dt (2-methylene protons), 3JH-H =7Hz, 3JH-H =1.7 Hz; 6 1.08. 3H, dd (3-methyl protons), 3JH-H=8 Hz, 4JH-H=1.5 Hz; 6 3.28, 6H, s (two methyl groups of DMT); 6 3.30, 2H, m (1-methylene protons); 6 3.29, 1H, s (methine proton); protons of the aromatic rings:6 6.73, 4H, d, 3JH-H=9 Hz; 6 7.05, 1H, t, 3JH-H=7 Hz; 6 7.18, d, 3JH-H=7 Hz; 6 7.45, d, 4H, 3JH-H=9 Hz; 6 7.63, 2H, d, 3JH-H=7 Hz) Synthesis of 1-0-Dimethoxytrityl Butane-3-0-(2-cyanoethylN,N-diisopropyl)-phosphoramidite (2). The mixture of 1-0dimethoxytrityl butane-3-ol (42 mg, 0.108 mmol) and tetrazole (7.6 mg, 0.108 mmol) was dried under high vacuum over P205. Anhydrous acetonitrile (400 1d) was added and the resulting solution was cooled to -20°C. To this solution [(N,N,N',N'-tetraisopropyl)2-cyanoethyl]phosphordiamidite (30 mg, 0.1 mmol) was added dropwise. The reaction mixture was allowed to warm to room temperature and was used as an 'in situ' prepared substrate for solid support synthesis [50 1.l of the solution added to 60 il of the acetonitrile solution of tetrazole (2.8 mg, 0.04 mmol)] [31P NMR: 6 147.97 ppm and 147.13 ppm (84:100), the purity estimated by NMR >90%.
Synthesis of J-0-Dimethoxytrityl-Butane-3-ol Bound to the Solid Support (3). The synthesis was performed according to procedure described by Gait et.al.6 via active p-nitrophenyl ester of succinylated 1-0-DMT-butane-3-ol. Trityl cation assay of the support showed a loading of 20 Amol/g. Synthesis of Di- and Trinucleotide Analogues dNpV (4), VpdN (5) and dNpVpdN (6) Containing the Butanediol Unit (Segment V) Instead of One of the Nucleosides. Manual synthesis was carried out according to usually employed procedure6 using described above reagents 2 and 3, respectively. The purification of compounds 4, 5 and 6 was carried out by means of HPLC on ODS-Hypersil column (30 cmx4.6 mm) with the linear gradient 5-20% CH3CN-0. 1 M triethylammonium bicarbonate (TEAB) (pH 7.4) 0.75 %/min at a flow rate of 1.5 ml/min. During the purification of compounds 5 and 6 by means of HPLC it appeared that, under the conditions used, it was
possible to separate two isomers of each of these compounds (see Table 1).
Synthesis of the Decanucleotide d[GGGAATTCCC](7) and Its Analogues Containing the Butanediol Unit at a Preselected Position (8a-j). The synthesis of oligomer 7 was performed as described previously7. The synthesis of oligomers 8 was carried out under the conditions of the standard protocol6 using 5'-ODMT-2'-deoxynucleoside 3'-O-(2-cyanoethyl-N,N-diisopropyl) phosphoramidites. 1-0-DMT-Butane-3-0-(2-cyanoethyl-N,Ndiisopropyl) phosphoramidite was used for the introduction of the butanediol unit into the growing oligomer chain as described for dNpVpdN (6). After standard work-up6 5'-DMT derivatives of decanucleotides 8a-j were obtained and their purification was performed by means of HPLC on an ODS-Hypersil column with an exponential gradient (exp. 0.25) starting from 5% CH3CN in 0.1 TEAB (pH 7.4) to 30% CH3CN in the same buffer for 20 min, then isocratic at a flow rate of 1.5 ml/min. Under these conditions satisfactory purification of 5'-DMT derivatives of 8a-j was achieved. Isolated decanucleotides were detritylated by treatment with a 20% aqueous solution of CH3COOH (25°C, 20 min) and then purified on an ODS-Hypersil column under conditions described above for the isolation of compounds 4-6. Tm Measurements. About 0.5-0.7 A260 units of each decanucleotide 7 and 8a-j were dissolved in buffer containing 10 mM Tris-Cl (pH 7.6), 80 mM NaCl, and 20 mM MgCl2 (1 ml). The melting temperature (Tm) was measured spectrophotometrically at max =258 nm. The Tm values for these decanucleotides are presented in Table 2.
Enzymatic Digestion of Decanucleotides 7and 8a-j. About 0.25 A260 units of each decanucleotide 7 and 8a-j were separately dissolved in 200 /d of buffer containing 0.1 M Tris-Cl (pH 8.5) and 15 mM MgCl2 and incubated with snake venom phosphodiesterase (svPDE) (10,.g, 1.5 units/4Lg, Boehringer) for 12 h and then with alkaline phosphatase (AP) (1,ug, 0.47 units/ug, Boehringer) for 1 h at 37°C. After heat-denaturation (100°C, 1 min) of protein the digestion mixtures were analyzed by means of HPLC on an ODS-Hypersil column with the linear gradient 5-20% CH3CN-0.1 M TEAB, 0.75%/min at a flow-rate of 1.5 nl/min. Undigested dinucleotide analogues containing butanedol-1,3 unit were compared by co-injection with genuine samples of these compounds (see Table 1). Independently, compounds 8a-j were digested with nuclease P1. About 0.25 A260 units of the oligomers 8a-j dissolved in 200 yl of the buffer containing 0.1 M Tris-Cl (pH 7.2) and 1 mM ZnCl2 was incubated with nuclease P1 (1 jAg, 0.37 units/Atg, Sigma) for 12 h
Nucleic Acids Research, Vol. 18, No. 8 2067 Table 2. The analogues of decamer d(GGGAATTCCC) containing a butanediol-1,3 unit (V): RP-HPLC isolation and their digestion with Eco RI endonuclease. Retention time [min] Tm
Identified product of Eco RI digestion
Nucleoside ratio observed after svPDE/AP digestion
Comp. no.
5' - 3' Sequence
5'-DMT
5'-HO
[1C]
7.
GGGAATTCCC
15.50
10.0
47.0
GGG (6.4 min) +pAATTCCC(8.7 min)
8a.
VGGAATTCCC
15.7OFast
16.50Slow
10.5
36.0
VGG(Slow) 7.3 min GVG(Fast) 7.0 min GVG(Slow) 7.3 min +pAATTCCC 8.7 min
2
2
2
3
1
2
2
3
1 2 3 3 3 3 3 3
2 2 3 1 2 3 0 2 3 1 1 3 2 0 3 2 1 2 2 2 1 2 2 1
G: A: T: C 3 2 2 3
VGG(Fast)7.0 min
8b.
GVGAATTCCC
16.0
10.4
29.0
8c. 8d. 8e. 8f. 8g. 8h. 8i. 8j.
GGVAATTCCC GGGVATTCCC GGGAVTTCCC GGGAAVTCCC GGGAATVCCC GGGAATTVCC GGGAATTCVC GGGAATTCCV
15.50 16.0 16.0 15.5 15.5 16.0 16.0 16.0
10.5 10.5 10.5 10.5 10.5 10.5 10.4 10.5
21.0
-
-
-
26.0 28.0 39.0
-
1 h at 37°C. HPLC analysis of the digestion mixture was performed as it was described for svPDE/AP digestion.
at 37°C and then with alkaline phosphatase for
Digestion of Decanucleotides 7 and 8a-j with Eco RI Endonuclease. About 0.5 A260 units (3 nmole) of each oligomer 7 and 8a-j were dissolved in 400 Al of buffer containing 10 mM Tris-Cl (pH 7.6), 80 mM NaCl and 20 mM MgCl2. To this solution a 2 itl aliquot of Eco RI endonuclease stock solution (5.5 106 units/mg) was added (100 units, 0.18 jg of the protein, 3 pmol of the protein in its dimer form). The digestion mixture was incubated at 18°C for 24 h. Independently, the digestion with Eco RI endonuclease was performed at 12°C (for 24 h) with the use of increased quantity of the enzyme, namely 200 units/0.5 A260 unit of substrate. The 50 1I aliquots of the digestion mixture were removed periodically, heat-denaturated (100°C, 1 min) and analyzed on an ODS-Hypersil column under conditions described above for the analysis of the oligonucleotide analogues 4-6. In all analyzed incubation mixtures one or two possible products were identified (see Table 2). x
RESULTS AND DISCUSSION Synthesis of Oligodeoxyribonucleotides. Routine solid-support synthesis of oligodeoxyribonucleotides based on the phosphoramidite approach as described elsewhere5 allowed us to synthesize the decamer d[GGGAATTCCC] (7). This compound contains the canonical hexanucleotide ..GAATTC.. which is the recognition site for endonuclease Eco RI8. Starting in each case with 5'-DMT-cytidine bound to CPG (via a longchain alkylamine-succinic ester bond), and replacing at a preselected step the routinely used 5'-DMT-nucleoside 3'-O(2-cyanoethyl-N,N-diisopropyl)phosphoramidite I with reagent 2, compounds 8a-i were obtained. Effectiveness of coupling of 2 with the growing polymer was measured by the yield of released dimethoxytrityl cation from the attached 1-O-dimethoxytritylbutane-3-(O-(2-cyanoethyl phosphate) moiety. This yield was never lower than 98.7%. Except for compound 8a, the efficiency of introduction of butanediol-1,3 moiety has been confirmed by couplings of subsequent nucleotide units. In the case of compound
GGG (6.4 min) GGG (6.4 min)
8j the synthesis of nonanucleotide d[GGGAATTCC] attached via 3'-phosphate to 1-hydroxyl function of butanediol-1,3 was performed using reagent 3 as the 'starter'. All compounds 7 and 8a-j, when the synthesis was completed, were released from support and liberated from phosphate- and exoaminofunctionblocking groups by treatment with a 25 % ammonia solution for 24 hours at 25°C, and purified on a reverse-phase HPLC column as 5'-dimethoxytritylated species. Detritylation of each compound was achieved under standard conditions and 5 '-hydroxyl oligonucleotides bearing a 'vicarious' butanediol-1,3 unit were isolated as pure specimens by means of RP-HPLC. Digestions with Snake Venom Phosphodiesterase (svPDE) and Nuclease Pl. The structure of compounds 8a-j was confirmed by means of their hydrolysis in the presence of nucleases such as snake venom phosphodiesterase (svPDE) and nuclease P1. If the svPDE-digest of 7 was subsequently treated with alkaline phosphatase, HPLC analysis allowed assignment of the expected ratio of nucleosides G:A:T:C = 3:2:2:3. However, the analysis of the digests of compounds 8b-j obtained under analogous conditions did not give the ratio of nucleosides lacking the expected nucleoside, and the HPLC trace contained an additional peak corresponding to an unidentified product. One could conclude that svPDE which is known to hydrolyse oligodeoxyribonucleotide from the 3'- end does not recognize the butanediol-1,3 moiety as an appropriate substrate and 'jumps over' a dNpV segment. This endonucleolytic activity of svPDE has been recognized in earlier studies on nucleotide phosphorothioates9. Indeed, we have synthesized compounds dNpV (4) (N=G,A,T,C) (see Table 1) and have found, that compounds 4 are not substrates for svPDE, and their HPLC mobilities are identical with respective, but formerly unidentified, products of the svPDE digest of compounds 8b-j. This picture was consistent with the interpretation of the results of svPDE/AP catalyzed hydrolysis of compound 8a where the expected ratio G:A:T:C = 2:2:2:3 was obtained. When these studies were in progress, a paper by Seela and Kaiser3 was published, wherein the authors described their results on the synthesis of oligonucleotides bearing 'vicarious' propandiol-1,3 residues; the results of svPDE-catalyzed hydrolysis of their products are consistent with these presented in this report. In the light of the
2068 Nucleic Acids Research, Vol. 18, No. 8 above findings, less unexpected were the results of the digestion of compounds 8a-j by nuclease Pl/alkaline phosphatase. HPLC analysis of products suggested that this enzyme does not catalyze the hydrolysis of a phosphodiester bond in nucleoside-5'-(l-hydroxy-butan-3-oxy)phosphate. Indeed, the substances corresponding to unidentified products of hydrolysis of compounds 8a-i, according to HPLC, have appeared to be identical with VpdN (5) synthesized independently. This result is also consistent with our former studies on nuclease P1-catalyzed hydrolysis of oligonucleotides bearing phosphorothioate moieties at two adjacent positions"I and indicates that nuclease P1 'digests' oligonucleotides in the direction 5' to 3', possesses endonucleolytic activity (jumps over VpdN unit) and does not recognize a butanediol-1,3 residue attached to phosphate via 3-hydroxyl function as the substrate. Tm Measurements. The analogues of decanucleotide d[GGGAATTCCC], which contain at preselected positions a butanediol-1,3 unit instead of a nucleoside moiety, do have a selfcomplementary sequence, and under appropriate conditions can form a duplex which has been thought to be a prerequisite for Eco RI endonuclease digestion. However, it has been recently reported that Eco RI enzyme recognizes and cleaves single stranded substrates'2 13. Therefore the duplex form may not be necessary for recognition and cleavage by this enzyme. Nevertheless we performed measurements of the melting temperature (Tm) for all compounds 8a-j. As assumed, that the different position of butanediol-1,3 unit in each of the ten oligonucleotides 8a-j strongly influences the Tm value. Results of these measurements are presented in Table 2. The oligonucleotides in which this unit is located at the positions 1,2,3,8,9 and 10 (numbered from the 5' to 3' end) form duplex molecules, although the Tm values are substantially lower than that of the unmodified decanucleotide. Oligonucleotides in which the butanediol-1,3 unit is located within the central tetranucleotide do not form a stable duplex at the temperatures and concentrations of these measurements. That the duplex is increasingly destabilized as the two vicarious units on opposite strands are brought closer together suggests destabilization by distance-related conformational distortions.
Digestion of Oligonucleotides 8a-j with Eco RI. Diminished temperature stability of the duplex form of compounds 8a-j prompted us to incubate them with Eco RI endonuclease at 12°C and 18°C. However, the results of these two experiments were very similar. Therefore, we performed the Eco RI digestion only at 12°C (with the use of 400 units of the enzyme/l A260 unit of the substrate). Under these conditions complete digestion of the unmodified decamer 7 was observed after 2 h of incubation. The same rate of cleavage was observed for decanucleotides 8i and 8j, i.e oligomers which contained the butanediol-1,3 unit instead of deoxycytidine at the position 9 and 10 (Table 2). Also, decanucleotides 8a and 8b appeared to be substrates for Eco RI endonuclease, although the relative rate of their cleavage was lower than that one observed for unmodified decamer. None of the decanucleotides 8c-8h was hydrolyzed by Eco RI endonuclease (no traces of products were detected). Thus, the presence of the butanediol-1,3 unit in the canonical hexamer GAATTC caused the resistance of decanucleotides 8c-h to Eco RI action. Therefore it is rather surprising that the substitution of deoxyguanosine and deoxycytidine at positions closely neighbouring in this hexamer (2 and 9, respectively), by
butanediol-1,3, did not cause the resistance of oligonucleotides 8b and 8i to the enzyme action. Moreover, the relative ratio of cleavage appeared to be rather high: 60% for decamer 8b and 100% for decamer 8i during the period of the time sufficient for complete digestion of decamer 7. These results allow us to assume that substitution of nucleosides 1 and 10 as well as 2 and 9 by butanediol-1,3 unit does not influence the duplex conformation what is required for the recognition and cleavage by Eco RI restrictase. In our earlier studies on mono-O-ethyl ester analogues of decamer 7 it was found that the area of interactions between Eco RI protein and phosphate groups is greater than hexamer GAATTC and it is extended to the nonamer d[GGGAATTCC]. The modification of the phosphate moiety within this nonamer by means of O-ethyl group caused the protection from the nucleolytic action of Eco RI enzyme7. The comparison between the results obtained for mono-O-ethyl ester analogues of 7 and those presented here, allows to suggest that Eco RI requires for the digestion the presence of unmodified phosphate groups flanking the recognition site.
ACKNOWLEDGEMENT We wish to thank Prof. F.Eckstein for a gift of Eco RIendonuclease over-producing E. coli strain. This paper is dedicated to Professor C.B.Reese on the occasion of his 60th birthday.
REFERENCES 1. S.Penczek, G.Lapienis and P.Klosinski, Phosphorus and Sulfur, 27, 153 (1986). 2. B.Uznanski, A.Sibinska, W.J.Stec, poster presented at 2nd International Symposium on Phosphorus Chemistry Towards Biology, Lodz, Poland, 8-12 September 1986. 3. F.Seela and K.Kaiser, Nucleosides and Nucleotides, 6, 447 (1987). 4. R.L.Letsinger, W.B.Lunsford, J.Am.Chem.Soc., 98, 3655 (1976). 5. S.L.Beaucage and M.H.Caruthers, Tetrahedron Lett., 22, 1859 (1981); L.J.McBride and M.H.Caruthers, ibid., 24, 245 (1983). 6. T.Atinson, M.Smith in Oligonucleotide Synthesis a Partical Approach, Edited by M.J.Gait IRL Press, 1984, p. 35. 7. M.Koziolkewicz, B.Uznanski, W.J.Stec, Nucleosides and Nucleotides, 8(2), 185 (1989). 8. W.J.Stec, G.Zon, B.Uznanski, J.Chromatography, 326, 263 (1985). 9. B.Uznanski, W.Niewiarowski, W.J.Stec, J.Biol.Chem., 261, 529 (1986). 10. T.A.Bickle, V.Pivvotta, R.Imber, Nucleic Acid Res. 4, 2561 (1977). 11. W.J.Stec, G.Zon, W.Egan, B.Stec, J.Am.Chem.Soc., 106, 6077 (1984). 12. N.Bischofberger, P.G.Ng, T.R.Webb, M.D.Matteucci, Nucleic Acid Research, 15, 709 (1987). 13. K.Mishigaki, Y.Kaneko, H.Wakuda, Y.Husimi, T.Tanaka, Nucleic Acid Research, 13, 5747 (1985).
