Pentopyranosyl oligonucleotide systems. Part 11: Systems with shortened backbones: (D)-β-ribopyranosyl-(4′→3′)- and (L)-α-lyxopyranosyl-(4′→3′)-oligonucleotides

Bioorganic & Medicinal Chemistry 9 (2001) 2411–2428

Pentopyranosyl Oligonucleotide Systems. Part 11: Systems with

Shortened Backbones: (^D)-ꢀ-Ribopyranosyl-(4⁰!3⁰)- and (L)-ꢁ-

Lyxopyranosyl-(4⁰!3⁰)-oligonucleotides^y

Harald Wippo,^aFolkert Reck,^aRene Kudick,^a,bMahesh Ramaseshan,^b

Griet Ceulemans,^aMartin Bolli,^a,bRamanarayanan Krishnamurthy^a,*

and Albert Eschenmoser^a,b,

*

^aThe Skaggs Institute for Chemical Biology at The Scripps Research Institute (TSRI),

10550 North Torrey Pines Road, La Jolla, CA 92037, USA

^bLaboratory of Organic Chemistry, Swiss Federal Institute of Technology (ETH),

Universitatstrasse 16, CH-8092 Zurich, Switzerland

¨

This paper is dedicated to Peter Dervan who, through his pioneering work, has taught organic chemists to look at DNA as ‘an organic molecule’

Received 28 February 2001; accepted 14 May 2001

Abstract—The (l)-a-lyxopyranosyl-(4⁰!3⁰)-oligonucleotide system—a member of a pentopyranosyl oligonucleotide family con-

taining a shortened backbone—is capable of cooperative base-pairing and of cross-pairing with DNA and RNA. In contrast, cor-

responding (d)-b-ribopyransoyl-(4⁰!3⁰)-oligonucleotides do not show base-pairing under similar conditions. We conclude that

oligonucleotide systems can violate the ‘six-bonds-per-backbone-unit’ rule by having ﬁve bonds instead, if their vicinally bound

phosphodiester bridges can assume an antiperiplanar conformation. An additional structural feature that seems relevant to the

cross-pairing capability of the (l)-a-lyxopyranosyl-(4⁰!3⁰)-oligonucleotide system is its (small) backbone/basepair axes inclination.

An inclination which is similar to that in B-DNA seems to be a prerequisite for an oligonucleotide system’s capability to cross-pair

Introduction

an informational base-pairing system, a property con-

sidered to be a requisite for any type of oligomer that is

supposed to have the functional potential of a genetic

system. For nucleic acid alternatives recruited from the

close structural neighborhood of RNA, the ﬁrst criter-

ion demands a candidate structure to be derivable from

a (CH₂O)_nsugar (n=6, 5, 4) by the same type of

potentially natural chemistry that allows the structure

type of RNA to be derived from ribose. Applying the

second criterion, on the other hand, amounts to pre-

dicting a candidate system’s capacity of base-pairing.

Whereas this can, in principle, be attempted at various

levels of theory-assisted molecular modeling, we have

developed, and systematically applied, a qualitative

conformational analysis of oligonucleotide single strand

units as a simple and, in favorable cases, remarkably

reliable tool for assessing a given oligonucleotide sys-

tem’s chance of being a base-pairing system.^2,3Those

favorable cases are oligonucleotide backbones contain-

ing their sugar units in the pyranose form.

Central to the strategy of our studies toward a chemical

etiology of nucleic acid structure¹are the chemical cri-

teria by which we decide which of the formally con-

ceivable alternative nucleic acid structures should, with

priority, be selected for experimental study. The criteria

are of two kinds: they are generational and functional in

nature, they refer to the type of chemistry to be con-

sidered for the generation of a candidate structure under

potentially natural conditions, and they relate to the

question whether such a structure will likely prove to be

*Corresponding authors. Fax: +41-1-632-1043; e-mail: rkrishna@

scripps.edu; eschenmoser@org.chem.ethz.ch

^yFor part 10, see ref 18. The paper is also communication No. 33 in the

series ‘Chemistry of a-aminonitriles’. For No. 32 of this latter series,

see ref 18, and for No. 30, see ref 17. The paper that appeared in ref 34

counts as communication No. 31 in this series. For a survey of the

numbering within this series, see footnote 1 in ref 17.

PII: S0968-0896(01)00220-6

2412

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

In the task of deﬁning the constitution of candidate

systems in the ﬁrst place, we adhered to an empirical

rule concerning the constitutional length of an oligonu-

cleotide system’s repetitive backbone unit. That rule⁴

requires the number of covalent bonds per repeating

backbone unit to be six (number of atomic centers, six)

emulating the backbone of the natural nucleic acids.

This constitutional rule has been crucial for our choice

of the chemical structures of nucleic acid alternatives in

both the hexopyranosyl-(6⁰!4⁰)- and the pentopyr-

anosyl-(4⁰!2⁰)-oligonucleotide series (Scheme 1).^3,5The

ﬁnding of strong (orthogonal!) base-pairing in these

oligonucleotide families^3,6À8did greatly strengthen the

belief in the relevance of the rule for predicting the base-

pairing potential of new oligonucleotide systems. In

addition, massive support came from the quarters of

anti-sense oriented oligonucleotide chemistry, where a

large number of base-pairing systems that cross-pair

with RNA and DNA align to the rule with almost no

exception.⁹So far, exceptions from the ‘six-bonds-per-

backbone-unit’ rule comprise, at least in phosphodie-

ster-based oligonucleotide systems, lengthened but not

shortened oligonucleotide backbones.^10{The best known

(and etiologically important) case is the (5⁰!2⁰)- isomer

of RNA, in which the phosphodiester bridge links the

sugar units between the 5⁰- and the 2⁰-positions, corre-

sponding to seven (instead of six) bonds per backbone

unit. The system shows both Watson–Crick self pairing

and cross-pairing with RNA, though distinctly weaker

than the natural (5⁰!3⁰)-isomer.¹¹Recent investigation

from two laboratories have demonstrated (5⁰!2⁰)-

DNA-oligonucleotides to behave correspondingly.¹²

(l)-a-lyxopyranosyl-(4⁰!3⁰)-oligonucleotides (Scheme 2)

not only undergo base-pairing by themselves, but also

have the capability of (weakly) cross-pairing with DNA

and RNA. The access to that discovery had been by no

means straightforward. It came about the following

way. Template-controlled ligation of (d)-b-ribopyr-

anosyl-(4⁰!2⁰)-tetramer cyclophosphates¹⁴(Scheme 3)

had been predicted, as well as observed, to proceed with

(4⁰!2⁰) regioselectivity. In the course of that work, it

had become desirable to strengthen the experimental

evidence on this matter by synthesizing, at least for one

example, the alternatively possible (4⁰!3⁰)-ligation pro-

duct and determine its chromatographic behavior in

order to be able to decide whether the (4⁰!2⁰) ligation

product can be actually₀chromatographically dis-

tinguished from its (4⁰!3 ) regiosiomer.^15,16This, in

fact, turned out to be the case.¹⁵However, the corre-

sponding octamer with a (4⁰!3⁰) connection at the cen-

tral ligation site, when mixed with the template strand,

surprisingly showed about the same T_mvalue (74 ^ꢀC, c

3 mM, 0.15 M NaCl) as the duplex derived from the

(4⁰!2⁰) ligation product (T_m=76 ^ꢀC). Therefore, it was

decided to conﬁrm experimentally what the ‘six-bonds-

per-backbone-unit’ rule was supposed to predict in this

series, namely, that b-ribopyranosyl-(4⁰!3⁰)-oligonu-

cleotides (Scheme 2) should be devoid of the capacity of

base-pairing. To this end, the two (d)-b-ribopyranosyl-

(4⁰!3⁰)-octamers pr(4⁰!3⁰)(A₈) and pr(4⁰!3⁰)(T₈) con-

taining exclusively (4⁰!3⁰)-phosphodiester links were

synthesized. They, in fact, did not undergo duplex for-

mation under conditions where their pr(4⁰!2⁰) analo-

gues show distinct base pairing (T_m=40 ^ꢀC, c 10 mM,

0.15 M NaCl).¹³Much later, when our investigations on

As we reported in a preliminary communication,¹³an

example of a phosphodiester-based base-pairing system

with a shortened oligonucleotide backbone was encoun-

tered in our systematic studies on the properties of

pentopyranosyl oligonucleotides, where it was found that

Scheme 1. Constitution and conﬁguration of the repeating units of

RNA, homo-DNA, and the family of the four (4⁰!2⁰)-pentopyranosyl

oligonucleotides.

^{Note added in proof: After the present paper was submitted, it came

to our attention (Herdewijn, P. Agnew. Chem. 2001, 113, 2309) that an

example of such a system had in fact been described by Koga, M.; Abe,

K.; Ozaki, S.; Schneller, S. W. Nucl. Acids. Symp. 1994, 31, 65.

Scheme 2. (4⁰!2⁰)-(d)-b-ribo- and (4⁰!2⁰)-(l)-a-lyxo-pyranosyl oli-

gonucleotides and their corresponding (4⁰!3⁰)-pentopyranosyl coun-

terparts in their idealized pairing conformations.

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

2413

pyranosyl–RNA were extended to the isomeric mem-

bers of the pentopyranosyl-(4⁰!2⁰)-oligonucleotides

family and when it was found, to our surprise, that (l)-

again, this time, however with a new perspective: In the

lyxopyranosyl series, a (4⁰!3⁰)-oligonucleotide would

have its phosphodiester bridges attached to the neigh-

boring pyranosyl chairs in diaxial (instead of axial-

equatorial) conformation (Scheme 2). This would imply

a conformational ﬁxation of the phosphodiester bridge,

and enforce a maximal spatial separation for the two

vicinally bound phosphate centers from each other, and

so, perhaps, allow for a violation of the ‘six-bond-per-

backbone-rule’. On the basis of this analysis, work on

the synthesis of lyxopyranosyl-(4⁰!3⁰)-oligonucleotides

was pursued and it was found that in this series the rule

is indeed violated.¹³In the present paper, we describe

the experimental details of that investigation and also

complement the previously published preliminary com-

munication.¹³

a-lyxopyranosyl-(4⁰!2⁰)-oligonucleotides represent

a

tendentially even stronger base-pairing system than

pyranosyl–RNA,⁷the question of whether the isomeric

(4⁰!3⁰)-oligonucleotide system with the shortened

backbone would be capable of base-pairing came up

Preparation of Nucleoside Building Blocks

The building blocks required for the synthesis of

(4⁰!3⁰)-oligonucleotides in the (d)-b-ribopyranosyl-

and (l)-a-lyxopyranosyl series were prepared from

intermediates which we had previously used in the

synthesis of (4⁰!2⁰)-oligonucleotides in the (d)-b-ribo-

pyranosyl^3,17and the (l)-a-lyxopyranosyl series.^7,18

Schemes 4 and 5 summarize the adaptations and exten-

sions of these procedures for the preparation of com-

pounds 4a–d (Scheme 4) and 8a–d (Scheme 5) required

for the synthesis of (4⁰!3⁰)-oligonucleotide-isomers in

the two series. In both the controlled pore glass (CPG)

supported nucleosides 2a–d and 9a–d were the same as

used previously in the synthesis of the corresponding

(4⁰!2⁰)-oligonucleotides.

Since in the (4⁰!3⁰)-ribopyranosyl series (Scheme 4), the

synthesis of oligomers was to be carried out using the

allylphosphoramidite chemi₀stry¹⁹similar to the synthesis

of the ribopyranosyl-(4⁰!2 )-oligomers,^3,17the axial 3⁰-

hydroxyl groups of the intermediates 3a–d were phosphi-

tylated by treatment with (allyloxy)(diisopropylamino)-

chlorophosphine²⁰in the presence of Huning’s base. The

reaction proceeded, not surprisingly, slower than the

corresponding phosphitylation of the equatorial 2⁰-

hydroxy group in the synthesis of (4⁰!2⁰) oligomers, and

an excess of the phosphine reagent was needed in order

to drive the reaction to completion. Avoiding overly long

reaction times was essential, because of the danger that

Scheme 3. Template-controlled ligation of (4⁰!2⁰)-(d)-b-ribopyr-

anosyl tetramer cyclophosphates leading to regioselective (4⁰!2⁰as

against their 4⁰!3⁰)-ligation product (from ref 15).

Scheme 4. Preparation of the A, T, G and C building blocks in the (4⁰!3⁰)-(d)-b-ribopyranosyl series. 3a(3b, 3c, 3d)!4a (4b, 4c, 4d): (a) (allylox-

y)(diisopropylamino)chlorophosphine, diisopropyl ethyl amine, CH₂Cl₂, rt, 94% (78%, 63%, 63%). DMT=4,4⁰-dimethoxy trityl. CPG=long-

chain-alkylamino-controlled pore glass.

2414

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

(2⁰!3⁰) benzoate migration might interfere during the

than in the synthesis of the corresponding (4⁰!2⁰)-oligo-

mers. Two diﬀerent phosphitylation reagents were used

in the two series (allyloxy in the ribopyranosyl- and 2-

cyanoethoxy in the lyxopyranosyl series), the deprotec-

tion protocols of the syntheses diﬀering correspondingly.

Scheme 6 summarizes the deprotection, puriﬁcation and

isolation protocols followed in the two series and Table

1 gives an overall survey of the sequences synthesized,

together with relevant analytical information.

reaction (for that migration see refs 3 and 17).

The preparation of the building blocks 8a–d (Scheme 5)

required for studying the (l)-a-lyxopyranosyl-(4⁰!3⁰)-

series made use of components 6a–d which had served

as intermediates in the synthesis of (l)-a-lyxopyranosyl-

(4⁰!2⁰)-oligonucleotides.^7,18Removal of the acyl pro-

tection of the axial 3⁰-hydroxyl group by alkaline

hydrolysis followed by regioselective benzoylation of

the equatorial 2⁰-hydroxyl group lead to the mono-

benzoylated intermediates 7a–d²²which, in turn, were

phosphitylated with (2-cyanoethoxy)(diisopropylamino)

chlorophosphine in dichloromethane in the presence of

2,4,6-collidine and N-methyl imidazole²³to give the

required phosphitylated derivatives 8a–d in good yields.

Base-Pairing Properties

The base pairing properties of (d)-b-ribopyranosyl-

(4⁰!3⁰)- and (l)-a-lyxopyranosyl-(4⁰!3⁰)-oligonucleo-

tides were characterized by methods routinely used in

oligonucleotide chemistry²⁵and used previously in the

homo-DNA⁶and p-RNA³series, namely, (a) tempera-

ture-dependent UV spectroscopy for determination of

T_mvalues, (b) concentration-dependent T_mmeasure-

ments by UV spectroscopy for determination of ther-

modynamic data of duplex formation, (c) molar-ratio-

dependent UV spectroscopy for determination of com-

plexation stoichiometry and (d) temperature-dependent

CD spectroscopy. Measurements were normally made

in 10 mM aqueous NaH₂PO₄buﬀer containing 0.1 mM

Na₂EDTA, 150 mM NaCl at pH 7.0, with a total oli-

gonucleotide concentration of approximately 10 mM,

unless otherwise stated. Table 2 gives the T_mvalues and

thermodynamic data of the sequences prepared in this

study.

Synthesis, Puriﬁcation and Characterization of Oligo-

nucleotides

The protocols for the automated synthesis of (4⁰!3⁰)-

linked oligonucleotides in the ribopyranosyl- and lyx-

opyranosyl- series as well as their deprotection, pur-

iﬁcation and characterization, basically followed the

procedures developed for the preparation of ribopyr-

anosyl-(4⁰!2⁰)-^3,17and lyxopyranosyl-(4⁰!2⁰)-^7,18oli-

gonucleotides respectively. The syntheses were carried

out on 10 mmol scale on automated DNA synthesizers

(Pharmacia’s Gene Assembler Plus, and Perseptive’s

Expedite), coupling, deprotection, and puriﬁcation pro-

cedures adapted where necessary (for details, see

Experimental). Individual coupling yields in the

(4⁰!3⁰)-ribopyranosyl series were, on the average, a

mere 88%, in sharp contrast to the yields obtained in

the respective (4⁰!2⁰)-oligonucleotide series (ꢁ96%).

As illustrated for the sequences pr(4⁰!3⁰)-(A₄T₄) and

pr(4⁰!3⁰)-(T₄A₄) in the Experimental, the lower-than-

usual coupling yields led to rather complex HPLC tra-

ces of the crude oligonucleotide material, rendering

puriﬁcation rather tedious, and yields of isolated and

puriﬁed oligonucleotide sequences (>95% pure accord-

ing to HPLC and authenticated by mass spectroscopy)

correspondingly low. Average coupling yields in the

lyxopyranosyl-(4⁰!3⁰) series were not markedly lower

The (d)-b-ribopyranosyl-(4⁰!3⁰) octamer strands

pr(4⁰!3⁰)(A₈) and pr(4⁰!3⁰)(T₈) do not form a duplex,

as probed by the above UV and CD spectroscopy

methods, neither do the formally self-complementary

sequences pr(4⁰!3⁰)(A₄T₄) and pr(4⁰!3⁰)(T₄A₄) as Fig-

ure 1a demonstrates. The absence of duplex formation

in the ribopyranosyl-(4⁰!3⁰) series [under conditions

where corresponding ribopyranosyl-(4⁰!2⁰)-strands

show T_mvalues in a range up to 45 ^ꢀC] demonstrates

that shortening the backbone unit in pyranosyl–RNA

from the (4⁰!2⁰) to the (4⁰!3⁰) junction leads to loss

or marked weakening of the base-pairing capacity.

Since, however, in our ligation studies in the pyranosyl–

Scheme 5. Preparation of the A, T, G and C building blocks in the (4⁰!3⁰)-(l)-a-lyxopyranosyl series. 6a(6c, 6d)!7a (7c, 7d): (a) aq NaOH in THF,

H₂O (2:1), 0 ^ꢀC; (b) benzoyl chloride, pyridine, 0 ^ꢀC, 90% (87%, 95%). 6b!7b: (c) K₂CO₃in MeOH, 0 ^ꢀC; (d) benzoyl chloride, pyridine, 0 ^ꢀC, 91%.

7a(7b, 7c, 7d)!8a (8b, 8c, 8d): (e) (2-cyanoethoxy)(diisopropylamino)chlorophosphine, 2,4,6-collidine, 0.5 equiv N-methylimidazole, CH₂Cl₂, rt,

73% (82%, 70%, 72%). DMT=4,4⁰-dimethoxy trityl; CPG=long-chain-alkylamino-controlled pore glass.

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

2415

Scheme 6. The deprotection, puriﬁcation and isolation protocols followed in the (4⁰!3⁰)-(d)-b-ribo- and (4⁰!3⁰)-(l)-a-lyxo-pyranosyl oligonucleo-

tide series.

RNA series¹⁵we had observed of between a pr(4⁰!2⁰)

octamer containing a single (4⁰!3⁰) connection at the

central ligation site with the corresponding (d)-b-ribo-

pyranosyl (4⁰!2⁰) template strand [the observation that

had induced us to study the pr(4⁰!3⁰) systems to begin

with], it seemed advisable also to check the capability of

a com-

plementary all-(4⁰!2⁰)-strand. Weak duplex formation

between pr(4⁰!3⁰)(A₈) and pr(4⁰!2⁰)(T₈) was in fact

observed [T_mꢂ10 ^ꢀC, c 10 mM, 1 M NaCl, as compared

0

ꢃ

to the corresponding (4 !2 )A₈(4 !2 )T₈duplex,

T_mꢂ45.5 ^ꢀC, Nos 12 and 1 in Table 2]. In contrast, the

reverse combination of pr(4⁰!3⁰)(T₈) with pr(4⁰!2⁰)(A₈)

does not show duplex formation. That pr-(4⁰!3⁰) strands

have a weak aﬃnity to complementary pr-(4⁰!2⁰)

strands is also indicated by the T_mvalue of 23 ^ꢀC (c

5 mM, 150 mM NaCl, No. 14 in Table 2) of the complex

formed from pr-(4⁰!3⁰)(C₈) and pr-(4⁰!2⁰)(G₈).²⁶

an all-(4⁰!3⁰)-strand to cross-pair with

2416

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

The data presented in Table 2 concerning the lyxopyr-

anosyl-(4⁰!3⁰) series demonstrate that pl-(4⁰!3⁰)-oligo-

nucleotides are a fairly well functioning base-pairing

system (Fig. 1b and d).²⁷Yet, base-pairing strength, as

judged by T_mvalues, is consistently lower than that of

corresponding oligonucleotides in the lyxopyranosyl-

(4⁰!2⁰)-series. Another diﬀerence seems to be a marked

sequence-dependence of pairing strength; the range of

sequences studied is, however, too limited²⁸for a reli-

able analysis of this eﬀect to be feasible. All our obser-

vations indicate that duplex formation with antiparallel

strand orientation is preferred. The sequence pl-

(4⁰!3⁰)(TTAAAATA) and its antiparallel complement

pl-(4⁰!3⁰)(TATTTTAA) form a duplex clearly melting

at 12.6 ^ꢀC (c 10 mM, 1 M NaCl, No. 9 in Table 2, Fig. 2a),

whereas the T_mvalue of the corresponding combination

with its parallel complement pl-(4⁰!3⁰)(AATTTTAT) is

below the range of observation (No. 10 in Table 2). The

very existence of some of the duplexes listed in Table 2

(Nos 3, 11, 26–31) supports the preference for anti-

parallel strand orientation. Stoichiometry of duplex

formation, tested in three cases, was found to be 1:1 by

UV mixing curve determination (Fig. 2b).²⁷As judged

from pairing tests with the homobasic (A₈) and (T₈)

octamers, pl-(4⁰!3⁰)-oligonucleotides do not crosspair

with either pl-(4⁰!2⁰)- nor pr-(4⁰!2⁰)-partners.

moderate and somewhat capricious in the sense that it

can strongly depend on which of the two com-

plementary base-sequences belong to which of the two

participating pairing systems. This is most clearly

exempliﬁed by the behavior of the homobasic sequences

pl-(4⁰!3⁰)(A_n) and pl-(4⁰!3⁰)(T_n) (n=8, 12) towards the

corresponding complementary DNA and RNA sequen-

ces (Fig. 1c). Whereas the combination of pl-(4⁰!3⁰)(A₈)

with d(T₈) forms a duplex with a T_mvalue of 42 ^ꢀC (No.

17), as well as a triplex pl(4⁰!3⁰)(A₈)^ꢃ2d(T₈) (see repro-

duction of mixing curve in ref 13), no cross-pairing is

detected in the inverse combinations pl-(4⁰!3⁰)(T₈) with

d(A₈) and r(A₈), respectively (Nos 18 and 22). T_mdif-

ferences of approx 40 and 60 ^ꢀC are observed with the

corresponding dodecamer sequences (Nos 19 and 20)

(for a discussion of this kind of phenomenon, see ref 8).

Intersystem cross-pairing among sequences in which all

four canonical nucleobases are irregularly mixed does

not show this kind of backbone dependence (Nos 26–

31), however. With regard to the pairing mode, our

observations are compatible with the presumption that

the self- as well as the cross-pairing of lyxopyranosyl-

(4⁰!3⁰)-oligonucleotides proceeds by the Watson–Crick

mode; alternative modes in duplexes with homobasic

sequences are not excluded, however.

Table 2 contains also the information we have collected

about the capability of lyxopyranosyl-(4⁰!3⁰)-oligonu-

cleotides to cross-pair with antiparallel-complementary

DNA and RNA sequences. Cross-pairing strength is

Discussion

Although the study of the two members of the pento-

pyranosyl oligonucleotide family with shortened phos-

Table 1. HPLC and MS data of (d)-b-ribopyranosyl-(4⁰!3⁰) and (l)-a-lyxopyranosyl-(4⁰!3⁰)-oligonucleotides

Oligonucleotides

all (4⁰!3⁰)

Deprotection^a

method

Amount made

(OD260) (yield after

puriﬁcation)

Analytical HPLC

MonoQ-Ion Exchange^b

gradient, t_r(min)

MALDI-TOF MS^c

[M+H]⁺(obsd)

[M+H]⁺(calcd)

Ribopyranosyl

pr(A₈)

pr(T₈)

pr(A₄T₄)

pr(T₄A₄)

pr(C₈)

A

22.0 (2%)^d

30.0 (3%)^d

4.0 (4%)^d

21.0 (2%)^d

0–100% in 30 min, t_r15.5

0–100% in 30 min, t_r25.5

0–100% in 30 min, t_r18.5

0–100% in 30 min, t_r18.7

0–100% in 30 min, t_r16.1

2571

2500

2537

2376

2572

2500

2536

2379

Lyxopyranosyl

pl(A₈)

pl(T₈)

B

C

B

23.2 (12%)

24.0 (15%)

12.9 (5%)

24.6 (28%)

18.4 (24%)

22.1 (14%)

13.8 (18%)

9.6 (6%)

35.3 (24%)

39.4 (24%)

11.8 (35%)

7.8 (18%)

5.1 (7%)

0–50% in 30 min, t_r17.4

20–80% in 30 min, t_r26.1

0–100% in 30 min, t_r17.9

0–100% in 30 min, t_r27.4

0–100% in 30 min, t_r20.2

0–100% in 30 min, t_r20.4

0–100% in 30 min, t_r20.6

0–100% in 30 min, t_r20.8

0–100% in 30 min, t_r21.4

0–100% in 30 min, t_r18.6

0–100% in 30 min, t_r21.8

0–100% in 30 min, t_r21.9

0–100% in 30 min, t_r21.7

0–100% in 30 min, t_r22.1

2572

2502

3888

3780

2542

2539

2540

2528

2546

1890

1889

2539

2538

2572

2500

3889

3780

2536

2527

2545

1889

2538

pl(A₁₂

)

pl(T₁₂)

pl(A₄T₄)

pl(T₄A₄)

pl(AT)₄

pl(TA)₄

pl(TATTTAA)

pl(TTAAATA)

pl(CG)₃

pl(GC)₃

pl(ATTCAGCG)

pl(CGCTGAAT)

4.0 (6%)

^aMethod A: 25% aq NH₂NH₂, 4 ^ꢀC; ca. 20 h; method B: CH₃ONH₂^ꢃHCl in concd aq NH₃and EtOH (3:1) at 4 ^ꢀC for approx 6.5 h (ca. 70 h for G,C

containing sequences); method C: 40% aq CH₃NH₂in concd aq NH₃(1:1) at rt for 6.5 h (ca. 70 h for G,C containing sequences). All oligonucleo-

tides were puriﬁed by ion exchange chromatography on Mono Q HR 5/5 column (58Â6.0 mm, Pharmacia); elution with 10 mM Na₂HPO₄in H₂O

and a linear gradient of 1 M NaCl with a ﬂow of 1 mL/min; followed by desalting on Sep-Pak cartridges.

^bMonoQ HR 5/5 column; elution with 10 mM Na₂HPO₄in H₂O and a linear gradient of NaCl, pHꢂ10.5, ﬂow 1 mL/min; peak purity (260 nm) 95–

99%.

^cMatrix assisted laser-desorption ionisation time-of-ﬂight mass spectroscopy; matrix: 2,4,6-trihydroxyacetophenone and ammonium citrate buﬀer.

^dOD yields were measured at 270 nm.

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

2417

Table 2. T_mvalues and thermodynamic data

Experiment no.

Duplex

T_m^ꢀC (ꢂ10 mM, 1.0 M NaCl

pl(4⁰!3⁰)(0.15 M NaCl)

pr 4⁰!2⁰

pr 4⁰!3⁰

pl 4⁰!2⁰

pl 4⁰!3⁰

ÁG 25 ^ꢀC

ÁH

TÁS 25 ^ꢀC

Intra-system pairing

1.

2.

3.

4.

5.

6.

7.

8.

A₈+T₈

A₁₂+T₁₂

T₄A₄

A₄T₄

(TA)₄

(AT)₄

(CG)₃

45.5

67.6

40^a

27^a

40^a

38^a

65^a

62^a

45.9

<0

51.0

74.0

49.6

41.1

43.2

44.7

41.2

67.0

17.5

<5

30.1

26.6

45.1

35.0

12.6

—

À10.0^b

À14.0^b

À42.1^b

À51.8^b

À32.1^b

À37.8^b

<0

À6.3

À6.1

À9.5

À8.0

À33.9

À32.5

À42.5

À34.3

À27.6

À26.4

À33.0

À26.3

(GC)₃

9.

10.

11.

ÀTTAAAATA+AATTTTAT—

ÀTTAAAATA+–AATTTTAT

ÀATTCAGCG+TAAGTCGCÀ

46.3

61.9

67.6

39.0

À8.4

À24.9

À16.5

Inter-system cross-pairing

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

4⁰,3⁰-pr(A₈)+4⁰,2⁰-pr(T₈)

10

—

23.5

—

42

—

63.4

17.2

21

4⁰,3⁰-pr(T₈)+4⁰,2⁰-pr(A₈)

4⁰,3⁰-pr(C₈)+4⁰,2⁰-pr(G₈)

4⁰,3⁰-pr(A₈)+d(T₈)

d(A₈)+4⁰,3⁰-pr(T₈)

4⁰,3⁰-pl(A₈)+d(T₈)

d(A₈)+4⁰,3⁰-pl(T₈)

4⁰,3⁰-pl(A₁₂)+d(T₁₂

d(A₁₂)+4⁰,3⁰-pl(T₁₂

4⁰,3⁰-pl(A₈)+r(U₈)

r(A₈)+4⁰,3⁰-pl(T₈)

)

—

4⁰,3⁰-pl(A₁₂)+r(T₁₂

)

62.1

<5

—

28.6

26.4

36.3

42.0

36.5

52.0

r(A₁₂)+4⁰,3⁰-pl(T₁₂

)

4⁰,3⁰-pl(TTAAAATA)+d(AATTTTAT)

d(TTAAAATA)+4⁰,3⁰-pl(AATTTTAT)

4⁰,3⁰-pl(ATTCAGCG)+d(CGCTGAAT)

d(ATTCAGCG)+d(CGCTGAAT)

4⁰,3⁰-pl(ATTCAGCG)+r(CGCTGAAT)

r(ATTCAGCG)+4⁰,3⁰-pl(CGCTGAAT)

r(ATTCAGCG)+r(CGCTGAAT)

À7.9

À9.2

À52.6

À55.1

À44.7

À45.9

À11.2

À52.8

À41.5

Measurements were made in 10 mM NaH₂PO₄buﬀer, 0.1 mM EDTA, 150 mM NaCl, pH 7.0:

^aIn 0.15 M NaCl.

^bIn 1.0 M NaCl. pr=ribopyranosyl, pl=lyxopyranosyl, r=ribofurnaosyl, d=2⁰-deoxyribofuranosyl.

phodiester backbones remains rather limited and the

observations on these systems incomplete, the study,

nevertheless, taught us two signiﬁcant lessons. The ﬁrst

concerns the range of ﬂexibility of the ‘six-bonds-per-

backbone-unit’ rule toward backbone shortening. The

diﬀerence in pairing behavior of the two diastereomeric

pentopyranosyl-(4⁰!3⁰)-oligonucleotide systems clearly

indicates that it is most probably the ﬁxed anti-

periplanarity in the arrangement of the two vicinal

phosphodiester groups at the pyranosyl chair²⁹that is

the essential conformational feature responsible for the

lyxopyranosyl-(4⁰!3⁰)-system to be a base-pairing sys-

tem. The requirement for an antiperiplanar conforma-

tion of vicinal phosphodiester groups in a ﬁve-bonds-

per-backbone-unit can be expected to be a valuable cri-

terion in the design of new oligonucleotide systems

possessing base-pairing capability (for one explicit

example, see below). It also raises the interesting ques-

tion whether base-pairing capability would be even

compatible with aliphatic backbones containing phos-

phodiester groups that would assume an antiperiplanar

conformation as a result of a tendency to minimize

electrostatic repulsion (Scheme 7). At any rate, the

ﬁndings about the lyxopyranosyl-(4⁰!3⁰)-oligonucleo-

tide system widens the constitutional range for the

design of new oligonucleotide base-pairing systems.

The second lesson refers to the perhaps most remark-

able property of the lyxopyranosyl-(4⁰!3⁰)-system,

namely its capability to undergo cross-pairing with the

natural nucleic acids. After all, none of the four mem-

bers of the family of pentopyranosyl-(4⁰!2⁰)-oligonu-

cleotides⁷—although all of them showing strong

Watson–Crick base-pairing by themselves and among

themselves- has been observed to possess the capability

of cross-pairing with RNA or DNA.

General constitutional criteria that correlate the con-

stitution (including conﬁguration) of an oligonucleotide

backbone with a system’s capability of cross-pairing

with the natural nucleic acids would be valuable tools

for oligonucleotide chemistry, but they are notoriously

hard to deﬁne. It seems to us that the lyxopyranosyl-

(4⁰!3⁰)-system, besides extending the ‘six-bonds-per-

backbone-unit’ rule, points to one such constitutional

criterion, namely, the oligonucleotide system’s back-

bone/basepair axes inclination.^30,31Among oligonu-

cleotide families, there are at present three groups of

base-pairing systems known which are capable of eﬃ-

ciently communicating by base-pairing among them-

selves, but diﬀer in their ‘language’ in the sense that they

are unable to communicate by cross-pairing with each

other. In other words, the base-pairing of each of these

2418

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

Figure 1. UV–T_mcurves documenting the pairing behavior in (A) the (4⁰!3⁰)-(d)-b-ribopyranosyl oligonucleotide series; curves of the non-self-

complementary single strands are also shown, (B, D) (4⁰!3⁰)-(l)-a-lyxoyranosyl oligonucleotide series; curves of the non-self-complementary single

strands are also shown, and (C) the intersystem cross-pairing of (4⁰!3⁰)-(l)-a-lyxoyranosyl oligonucleotides with RNA and DNA. All measure-

ments were made with total oligonucleotide concentration ꢂ10 mM in 10 mM aq NaH₂PO₄buﬀer containing 0.1 mM Na₂EDTA, 150 mM NaCl at

pH 7.0.

Figure 2. (A) UV–T_mcurves documenting the preference of the anti-parallel pairing behavior in the (4⁰!3⁰)-(l)-a-lyxoyranosyl oligonucleotide

series: pl-(4⁰!3⁰)-(TTAAAATA) with its antiparallel complement pl-(4⁰!3⁰)(TATTTTAA) strand and its parallel complement pl-

(4⁰!3⁰)(AATTTTAT) strand. For conditions see caption of Figure 1. (B) Molar-ratio-dependent UV spectroscopy–mixing curve indicating the 1:1

complexation stoichiometry between the sequence pl-(4⁰!3⁰)(TATTTTAA) and its antiparallel complement pl-(4⁰!3⁰)(TTAAAATA) (260 nm,

0 ^ꢀC).

groups is orthogonal to that of the other two (Fig. 3).

These three groups are, ﬁrst, the ‘DNA–RNA group’ to

which belong those systems which are able to cross-pair

in the Watson–Crick mode with the natural nucleic

acids, second, the ‘homo-DNA group’ of which the 2⁰,3⁰

-dideoxy-glucopyranosyl-(6⁰!4⁰)-oligonucleotide system

(‘homo-DNA’)⁶is the prototype of the family of poten-

tial hexopyranosyl-(6⁰!4⁰)-oligonucleotide pairing sys-

tems and, third, the ‘pyranosyl–RNA group’ in which

the pyranosyl isomer of RNA is the most representative

member of the family of the four diastereomeric pento-

pyranosyl-(4⁰!2⁰)-oligonucleotide systems, all of which

are stronger pairing systems than RNA itself.⁷We had

implied that the orthogonality of the base-pairing in

homo-DNA versus that of the pentopyranosyl-(4⁰!2⁰)-

systems is the consequence of the opposite sense of

the backbone/basepair axes inclination;³⁰Figure 4,

showing idealized pairing conformations, qualitatively

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

2419

illuminates the meaning of this statement, and, at the

same time, indicates that a shift from a pentopyranosyl-

(4⁰!2⁰)- to a pentopyranosyl-(4⁰!3⁰)-system can be

expected to result in a pronounced reduction of that

inclination.

It had been taken for granted that no candidates have to

be considered that contain a tetrose sugar as a backbone

unit since no oligonucleotide system can be constructed

from a tetrose that would satisfy the ‘six-bonds-per-

backbone-unit’ rule. The pairing properties of the lyx-

opyranosyl-(4⁰!3⁰) system as against those of the ribo-

pyranosyl-(4⁰!3⁰) isomer led us to extend our screening

for potentially natural nucleic acid alternatives to the

threofuranosyl-(3⁰!2⁰) series, counting on the possibi-

lity that their 2⁰,3⁰-trans-phosphodiester groups at the

threofuranose rings might conformationally emulate the

diaxial 3⁰,4⁰-trans-phosphodiester groups of the lyx-

opyranosyl-(4⁰!3⁰)-oligonucleotide system (Scheme 2).

As described in a recently published report³⁴this con-

jecture turned out to be fruitful in that it led us to

investigate the pairing properties of the threo-(3⁰!2⁰)-

oligonucleotide system which are, so we think, of con-

siderable interest in the context of nucleic acid etiology.

In fact, model considerations of (idealized) pl-(4⁰!2⁰)-

and pl-(4⁰!3⁰)-duplexes show that the strong

inclination^30À32of pl-(4⁰!2⁰) duplexes largely dis-

appears in shifting to corresponding (4⁰!3⁰) duplexes.

This shift is accompanied by a change in the constella-

tional nature of base-stacking: What in lyxopyranosyl-

(4⁰!2⁰) duplexes (in their idealized conformation) is

essentially pure interstrand-stacking, becomes intra-

strand stacking in (4⁰!3⁰) duplexes. The two alternative

stacking modes correlate directly with the degree of

(positive or negative) backbone/basepair axes inclina-

tion. Systems with only small (or no) inclination are

expected to display intrastrand base-stacking. The pro-

totype of this type of system is B-DNA and, in fact, an

analysis of inclinations in known B-DNA X-ray struc-

tures³³conﬁrms that B-DNA-backbones on the average

show essentially no backbone/basepair axes inclination.

Experimental

General

Such qualitative reasoning encourages us to postulate

that one of the criteria which must be fulﬁlled by an

oligonucleotide system (or any other type of base-pair-

ing system) in order to possess the capability of cross-

pairing with the natural nucleic acids is that they have

access to a pairing conformation that has no, or only a

small, backbone/basepair axes inclination. This postu-

late certainly deserves in future work a more quantita-

tive deﬁnition as well as a systematic testing of

oligonucleotide and other pairing systems of known

structure and known to cross-pair with DNA and

RNA.³²

Solvents for extraction: technical grade, distilled. Sol-

vents for reaction: reagent grade. Reagents: unless other-

wise noted, from Acros, Fluka or Aldrich, highest quality

Finally, our studies on the lyxopyranosyl-(4⁰!3⁰)-oligo-

nucleotide system led us to conclude—and this may

perhaps be the most consequential lesson we received

from it—that we should question our originally held

view that tetrofuranosyl-(3⁰!2⁰)-oligonucleotides are

irrelevant to a search for potentially natural nucleic acid

alternatives from the structural neighborhood of RNA.

Figure 3. The three groups of base-pairing systems which exhibit

strong intrasystem base-pairing but do not cross communicate with

each other; they represent orthogonal base-pairing systems. This

orthogonality is proposed to be related to the diﬀerent backbone/base-

pair-axes inclinations of these systems.

Scheme 7. A hypothetical aliphatic oligonucleotide backbone system

in which the vicinal phosphodiester groups might—as a result of a

tendency to minimize electrostatic repulsion—assume an anti-

periplanar conformation.

2420

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

Figure 4. Qualitative vertical projections of formulas of the pairing conformations (idealized, nucleosidic torsion angles À120^ꢀ) of the four oligo-

nucleotide systems: (A) homo-DNA (g/t phosphodiester conformation); (B) pyranosyl—RNA; (C) (4⁰!2⁰)-(l)-a-lyxopyranosyl-, and D: (4⁰!3⁰)-(l)-

a-lyxopyranosyl-oligonucleotides. Projections allow a rough estimate of the sense and the degree of the four systems’ backbone/base-pair-axes

inclination.³²Note that sense and degree of the inclination depends sensitively on the nucleosidic torsion angle.

available. (2-cyanoethoxy)(diisopropylamino)phosphine

(97%) was purchased from Chem-Impex Inc., Wood

Dale, IL, USA. TLC: silica gel 60 F₂₅₄aluminum plates,

(Whatman, Type Al Sil G/UV, 250 mm layer); visuali-

zation by UV absorption and/or (A) by dipping in a

soln of H₂SO₄/H₂O/EtOH 14:4:1 or (B) cerium (IV)

sulfate (3 mM)/ammonium molybdate (250 mM) in aq

H₂SO₄(10%), followed by heating. Flash column chro-

matography (CC) was performed on silica gel 60 (40–

63 mm, 230–440 mesh, EM Science) at low pressure (max

2 bar). The columns were ﬁlled with a suspension of

silica gel in MeOH. The MeOH was then replaced by

ﬂushing the column with the appropriate solvent mix-

ture. In case of acid sensitive compounds, the silica gel

was pre-treated with MeOH containing about 0.5%

Et₃N. Melting points (uncorrected) were measured with

MEL-TEMP II (Laboratory Devices Inc., USA). NMR:

¹H: d values in ppm (TMS as internal standard), J (Hz),

Na₂HPO₄in H₂O, 1 M NaCl, pH=10.5. UV spectro-

scopy was measured on a Cary 1 C spectrophotometer

(Varian) and a Perkin–Elmer Lambda 2. Melting tem-

perature (T_m) measurements of oligonucleotides were

determined with Cary 1 Bio spectrophotometer (Varian)

or with Perkin–Elmer Lambda 2 equipped with Perkin–

Elmer Digital Controller/Temperature Programmer

C570. CD spectrum was measured on (A) JASCO J-710

or (B) AVIV 61 DS circular dichroism spectro-

polarimeter. Concentrations of oligonucleotide solu-

tions were calculated from the UV absorbance of the

solutions at 260 nm (pH 7) at ca. 80 ^ꢀC using the fol-

lowing molar extinction coeﬃcients: epr(A)=-

pl(A)=15,000, epr(T)=pl(T)=10,000, epr(C)=8400,

epr(G)=11,900. Abbreviations: CPG: ‘controlled pore

glass’, DMAP: 4-dimethylaminopyridine, DMT: 4,4⁰-

dimethoxytrityl, LCAA-CPG: long-chain aminoalkyl-

CPG (500 A).

1

assignments of H resonances were in some cases based

on 2-D experiments (¹H–¹H-COSY); ¹³C: d values in

ppm (TMS as internal standard), J (Hz); assignments

and multiplicities were based on 2-D experiments

(¹H–¹³C-COSY); ³¹P: d values in ppm (85% H₃PO₄as

external standard). FAB⁺-MS (matrix-sol): m/z (inten-

sity in%), performed in the positive-ion mode on a VG

ZAB-VSE double focusing high resolution mass spec-

trometer equipped with a cesium ion gun. Matrix-assis-

ted laser-desorption-ionization time-of-ﬂight mass

spectrometry (MALDI-TOF-MS) was performed on a

Voyager-Elite mass spectrometer (Perseptive Biosys-

tems) with delayed extraction with THAP as the matrix

with ammonium citrate added to the sample. Elemental

analysis were performed using Perkin–Elmer PE2400

CHN analyzer. Oligonucleotides were synthesized on an

Expedite 8909 Nucleic Acid Synthesis system (Perseptive

Biosystems) or on a Pharmacia Gene Synthesizer Plus.

HPLC: Anion exchange (IA)-HPLC was performed on

(A) Pharmacia GP-250 Gradient Programmer equipped

with two Pharmacia P-500 pumps, ABI-Kratos Spec-

traﬂow 757 UV/Vis detector and a Hewlett Packard HP

3396A analogue integrator or (B) Pharmacia Akta

puriﬁer (900) controlled by UNICORN system. Col-

umns: Mono Q HR 5/5 (Pharmacia); buﬀer A: 10 mM

Na₂HPO₄in H₂O, pH=10.5; Buﬀer B: 10 mM

Procedure

(Allyloxy)dichlorophosphine. To a mechanically stirred

solution of 88 mL (137 g, 1 mol) of PCl₃in 250 mL of

dry ether, 81 mL (79 g, 1 mol) of dry pyridine was slowly

added under argon. The slightly turbid solution was

chilled with dry ice in acetone. Over a period of 6 h, a

soln 61 mL of allyl alcohol in 100 mL of dry diethyl

ether was added. On the addition of the alcohol a white

precipitate formed and eventually, the mixture became

thick and vigorous stirring was needed. On completion

of addition, the mixture was allowed to warm up to rt

and stirring continued for 16 h. Finally, the mixture was

diluted with 100 mL dry diethyl ether and ﬁltered under

nitrogen. The precipitate was carefully washed with

3Â100 mL portions of dry diethyl ether. The washings

were combined with the ﬁltrate and the bulk of the

ether was removed by distillation under atmospheric

pressure. The residue was subject to fractional distilla-

tion under reduced pressure (1 torr). A ﬁrst fraction

with a boiling point of 28–30 ^ꢀC was collected to give

74 g (47%) of the desired product. A second fraction

with a boiling point of 48–50 ^ꢀC, was collected to give

15 g (8%) of what is considered to be the bis(allyloxy)-

chlorophosphine.

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

2421

Data of (allyloxy)dichlorophosphine. ¹H NMR

(600 MHz, CDCl₃): 4.72 (ddt, J=8.4, 5.7, 1.4, 2H,

OCH₂); 5.31 (dq, J=10.3, 1.2, 1H, H₂C¼); 5.39 (dq,

J=18.5, 1.4, 1H, H₂C¼); 5.95 (ddt, J=16.1, 10.6, 6.5,

1H, HC¼).

H-C(4⁰), H-C(5⁰)); 4.09–4.13 (m, 0.6 H, CH₂O(allyl));

0

4.19–4.23 (m, 0.6 H, CH₂O(allyl)); 4.49 (d, J_{3 ,P}=10.8,

0

0.4 H, H-C(3⁰)); 4.72 (d, J_{3 ,P}=10.8, 0.6 H, H-C(3⁰));

4.85 (d, J=10.3, 0.4 H, H₂C=(allyl)); 4.91 (dd, J=1.4,

17.2, 0.4 H, H₂C=(allyl)); 5.07 (d, J=10.2, 0.6 H,

H₂C=(allyl)); 5.19 (dd, J=2.4, 9.5, 0.4 H, H-C(2⁰));

(Allyloxy)(diisopropylamino)chlorophosphine. To a vig-

orously stirred solution of 74 g (0.47 mol) of (allyloxy)-

dichlorophosphine in 400 mL of dry diethyl ether a soln

of 130 mL (94.3 g, 0.93 mol) of diisopropylamine in

100 mL of dry diethyl ether was added within 3 h at

À78 ^ꢀC under nitrogen. After complete addition of the

amine, stirring of the thick suspension was continued

for 16 h at rt and was diluted with 200 mL of dry ether.

The white precipitate was ﬁltered and washed with

3Â100 mL portions of dry diethyl ether under nitrogen.

The ﬁltrate and washings were combined and the diethyl

ether was removed on the rotavapor under nitrogen

atmosphere. The residue was carefully distilled under

reduced pressure (nitrogen atmosphere, 50 mtorr) and a

fraction with a boiling point of 55–60 ^ꢀC (50 mtorr) was

collected to give 69.3 g (66%) of the desired product.

The purity of the material was 94% (¹H NMR), the

impurity being (allyloxy)(diisopropylamino)-H-phos-

phonate.

5.25 (dd, J=1.5, 17.1, 0.6 H, H₂C=(allyl)); 5.34 (d,

0

J_{1 ,2}=9.7, 0.6 H, H-C(2 )); 5.43 (m, J=5.3, 10.6, 17.1,

0.4 H, HC-(allyl)); 5.92 (m, J=5.3, 10.5, ₀17.1, 0.6 H,

0

HC-(allyl)); 6.26, 6.27 (d, J_{1 ,2}=9.5, H-C(1 )); 6.87–8.59

(m, 28 H, H-C(ar)); 8.11, 8.13 (m, H-C(2)), 8.59, 8.64 (s,

H-C(8)). ¹³C NMR (150 MHz, CDCl₃): 24.77, 24.81,

25.25, 25.31, 25.4, 25.5 (6q, CH₃(isoprop)); 43.1, 43.2,

43.2 (3d, CH(isoprop)); 55.5 (q, CH₃O(DMT)); 64.1 (t,

CH₂O(allyl)); 64.4 (dt, J_CP=19.3, CH₂O(allyl)); 65.8,

66.0 (2t, C(5⁰)); 69.0, 69.3, 70.8, 71.7, 72.15, 72.23, 72.28

(7d, C(2⁰), C(3⁰), C(4⁰)); 78.8, 79.0 (2d, C(1⁰)); 87.6, 87.9

(2s, CAr); 113.66, 113.71, 113.74, 113.76 (4d, C(ar));

115.7, 116.4 (2t, H₂C¼(allyl)); 127.4, 127.5, 128.3, 128.4,

128.6, 128.77, 128.82, 129.0, 129.2, 129.8, 130.2, 130.5,

130.88, 130.91, 133.3, 133.5, 133.8, 135.75, 135.78,

136.02, 136.07 (21d, C(ar), CH(allyl)); 127.61, 127.65

(2s, C(5)); 129.2, 134.3, 136.5, 136.61, 136.64, 136.9,

145.7, 145.9, 159.2, 159.3 (10s, C(ar)); 143.2, 143.5

(2d, C(8)); 152.1, 152.1, 153.7, 153.8 (4s, C(4),

C(6)); 152.9 (d, C(2)); 165.1, 165.5, 172.61, 172.63

(4s, CO); ³¹P NMR (242 MHz, CDCl₃): 150.0,

151.6, HR-FAB-MS (pos, NBA/CsI): 1201.3295 (100,

[M+Cs]⁺).

Data of allyloxy(diisopropylamino)chlorophosphine. ¹H

NMR (600 MHz, CDCl₃): 1.25–1.29 (m, 12H, CH₃);

3.79–3.85 (m, 2H, H-CN); 4.38 (brs, 2H, OCH₂); 5.22

(d, J=10.5, 1H, H₂C¼); 5.34 (d, J=16.9, 1H, H₂C¼);

5.98 (ddt, J=16.4, 10.7, 5.4, 1H, HC¼). ¹³C NMR

(150 MHz, CDCl₃): 24.2, 23.5 (q, CH₃); 46.2 (d,

J_CP=12.9, CH–N); 66.6 (t, J_CP=18.2, CH₂–O); 118.3

(t, CH₂¼); 134.3 (d, J_CP=7.9, CHC=). ³¹P NMR

(242 MHz, CDCl₃): 183.0.

1-[3⁰-O-((Allyloxy)(diisopropylamino)phosphino)-2⁰-O-

benzoyl-4⁰-O-((dimethoxy-triphenyl)methyl)-ꢀ-D-ribopyr-

anosyl-(1⁰)]-thymine (4b). A solution of 1.99 g (3.0 mmol)

of 3b in 9 mL CH₂Cl₂was cooled to 0 ^ꢀC under argon.

1.57 mL (9.0 mmol) diisopropylethylamine was added,

immediately followed by 0.95 mL (4.6 mmol) (allyl-

oxy)(diisoproylamino)chlorophosphine. The reaction

was treated with extra equivalents of the phosphityl-

ating reagent and base as described for the preparation

of 4a. Column chromatography over silica gel (acetone/

hexanes 2:8 with 1% triethylamine) yielded 2.0 g (78%)

of 4b as a white foam.

N⁶- Benzoyl - 1 - [3⁰- O - ((allyloxy)(diisopropylamino)pho-

sphino)-2⁰-O-benzoyl-4⁰-O-((dimethoxytriphenyl)methyl)-

ꢀ-^D-ribopyranosyl-(1⁰)]adenine (4a). A solution of 2.52 g

(2.9 mmol) of 3a in 9 mL CH₂Cl₂was cooled to 0 ^ꢀC

under argon. 1.5 mL (8.7 mmol) Diisopropylethylamine

was added, immediately followed by 0.95 mL (4.6 mmol)

(allyloxy)(diisoproylamino)phosphine. The reaction

mixture was allowed to warm to rt and stirred at rt for

1 h. The reaction was mixture was cooled to 4 ^ꢀC,

0.95 mL (4.6 mmol) (allyloxy)(diisoproylamino)chloro-

phosphine, 0.79 mL (4.6 mmol) of diisopropylethyl-

amine were added and stirred at rt After 2 h, 0.36 mL

(1.7 mmol) of (allyloxy)(diisoproylamino)phosphine was

added (without base). The mixture was stirred for

another 1.5 h and 1 mL of methanol was added. After

dilution with CH₂Cl₂and washing with satd aq

NaHCO₃soln, the organic phase was dried over

Na₂SO₄and evaporated. Column chromatography over

silica gel (acetone/hexanes 2:8 with 1% triethylamine)

yielded 2.88 g (94%) of 4a as a white foam.

Data of 4b. TLC: (hexane/ethyl acetate 1:1): R_f0.71. ¹H

NMR (600 MHz, CDCl₃): 1.16, 1.22, 1.29 (3d, J=6.8,

CH₃(isoprop)); 1.841 (s, 2.1H, CH₃–C(5)); 1.84 (s, 0.9H,

0

CH₃–C(5)); 2.62 (dd, J_{4 ,5}=3.9, J_gem=10.0, 0.7H, H-

C(5⁰)); 2.83–2.88 (m, 0.3 H, H-C(5⁰)); 3.61–3.66 (m,

0.3H, CH₂–O(allyl)); 3.75–3.95 (m, 10.3H, CH₂–

O(allyl), CH₃–O(DMT) HC(isoprop), H-C(4⁰), H-

C(5⁰)); 4.06–4.11 (m, 0.7H, CH₂–O(allyl)), 4.16–4.21 (m,

0

0.7H, CH₂–O(allyl)); 4.41 (d, J_{3 ,P}=11.1, 0.3H, H-

0

C(3⁰)); 4.63 (d, J_{3 ,P}=10.6, 0.7H, H–C(3⁰)); 4.67 (dd,

J=2.2, 9.5, 0.3H, CH₂¼(allyl)); 4.82–4.85 (m, 1H, H-

C(2⁰)); 4.90 (dq, J=1.6, 17.2, 0.3H, CH₂¼(allyl)); 5.05

(dd, J=0.9, 10.4, 0.7H, CH₂¼(allyl)); 5.23 (dq, J=1.6,

17.1, 0.7H, CH₂¼(allyl)); 5.44 (m, J=5.3, 10.4, 17.1,

1

Data of 4a. TLC: (hexane/ethyl acetate 3:2) R_f0.58. H

NMR (600 MHz, CDCl₃): 1.18, 1.23, 1.29 (3d, 12 H,

0.3H, HC=(allyl)); 5.90 (m, J=5.3, 10.4, 17.1, 0.7H,

0

HC¼(allyl)); 6.21 (d, J_{1 ,2}=9.5, 1H, H–C(1 )); 6.86–6.88

(m, 4 H, H–C(ar)); 7.02 (d, J=1.1, 1H, H–C(6)); 7.22–

7.64 (m, 12H, H–C(ar)); 7.97–8.02 (dd, J=1.2, 8.2, 2H,

H–C(ar)). ¹³C NMR (150 MHz, CDCl₃): 12.6 (q, CH₃–

C(5)); 24.7, 25.2 (2q, CH₃(isoprop)); 43.1, 43.2 (2d,

0

J=6.8, CH₃(isoprop)); 2.72 (dd, J_{4 5}=4.9, J_gem=10.9,

0.6H, H-C(5⁰)); 2.95 (dd, J_{4 ,5}=4.9, J_gem=10.9, 0.4 H,

H-C(5⁰)); 3.62–3.70 (m, 0.4 H, H₂CO(allyl)); 3.82–4.00

(m, 10.4 H, CH₃O(DMT), CH(isoprop), H₂CO(allyl),

0

2422

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

CH(isoprop)); 55.5 (q, CH₃O(DMT)); 63.9, 64.3 (2dt,

J_CP=19.3, CH₂O(allyl)); 65.5, 65.6 (2t, C(5⁰)); 69.2,

69.4, 70.7, 71.0, 71.5, 72.4 (6d, C(2⁰), C(3⁰), C(4⁰)); 78.1,

78.2 (2d, C(1⁰)); 87.5, 87.8 (2s, CAr); 111.6, 111.7 (2s,

C(5)); 113.6, 113.7 (2d, C(ar)); 115.6, 116.3 (2t,

CH₂¼(allyl)); 127.4, 127.5, 128.3, 128.6, 128.8, 130.4,

130.7, 130.9, 133.5, 134.0, 135.3, 135.6, 135.9, 136.0,

136.1, 136.2 (16d, C(ar), CH(allyl), C(6)); 129.4, 130.0,

136.5, 136.7, 136.9, 145.7, 145.9 (7s, C(ar)); 150.7, 150.8

(2s, C(2)); 159.2, 159.3 (2s, C(ar)); 163.9 (s, C(4)); 165.6,

165.9 (2s, CO)); ³¹P NMR (242 MHz, CDCl₃): 149.7,

151.6. HR-FAB-MS (pos, NBA/CsI): 984.2641 (100,

[M+Cs]⁺).

(40, [M+1]⁺), 726 (35), 637 (71), 303 (100), 188 (10),

105 (29).

9-[3⁰-O-((Allyloxy)(diisopropylamino)phosphino)-2⁰-O-

benzoyl-4⁰-O-((dimethoxy-triphenyl)methyl)-ꢀ-D-ribopyr-

anosyl-(1⁰)]-N²-isobutyrylguanine (4d). To a stirred solu-

tion of 613 mg (0.77 mmol) of 3d in 10 mL dry CH₂Cl₂

0.25 mL (1.16 mmol) of (allyloxy)(diisopropylamino)-

chlorophosphine and 0.34 mL (1.93 mmol) of diiso-

propylethylamine were added. The reaction was treated

with extra equivalents of the phosphitylating reagent

and base as described for the preparation of 4c. The

solvent was evaporated, and the residue was puriﬁed by

chromatography over silica gel (hexane/ethyl acetate 5:1

to 1:1). The product fractions were combined, evapo-

rated and the residue was twice precipitated from

150 mL of ice cold pentane to yield 492 mg (63%) of 4d

as a 1:2 mixture of diastereoisomers.

N⁴-Benzoyl-1-[3⁰-O-((allyloxy)(diisopropylamino)phos-

phino)-2⁰-O-benzoyl-4⁰-O-((dimethoxy-triphenyl)methyl)-

ꢀ-^D-ribopyranosyl-(1⁰)]cytosine (4c). To a stirred solution

of 540 mg (0.72 mmol) of 3c in 10 mL of dry CH₂Cl₂,

0.31 mL (1.8 mmol) of diisopropyl ethylamine and

0.23 mL (1.08 mmol) of (allyloxy)(diisoproylamino)-

chlorophosphine were added. After 3 h stirring at rt,

an additional 0.06 mL (0.36 mmol) of diisopropyl-

ethylamine and 0.07 mL (0.36 mmol) of (allyloxy)(di-

isoproylamino)phosphine were added. After 5 h, the

reaction was quenched by the addition of 1 mL of

methanol. The solvent was evaporated and the residue

was puriﬁed by chromatography on silica gel column

(hexane/ethyl acetate 5:1 to 2:1). The product fractions

were combined, evaporated and the residue was further

puriﬁed by precipitation from 150 mL of ice cold pen-

tane. This yielded 426 mg (63%) of 4c as a 1:2 mixture

of diastereoisomers.

1

Data of 4d. TLC (hexane/ethyl acetate 3:2): R_f0.67. H

NMR (400 MHz, CDCl₃): 1.15, 1.20, 1.21, 1.23, 1.28

(5d, J=6.8, 12H, CH₃(isoprop)); 2.62 (dd, J=4.7, 10.7,

0.7 H, H-C(5⁰)); 2.86 (dd, J_gem=4.8, 10.5, 0.3 H, H–

C(5⁰)); 3.35–3.50 (m, 1H, H–C(isobut)); 3.62–3.70 (m,

0.3H, H₂C–O(allyl)); 3.75–3.98 (m, 10.3H, CH₃–

O(DMT), CH₂–O(allyl), CH(isoprop), H-C(4⁰), H-

C(5⁰)); 4.05–4.12 (m, 0.7 H, CH₂–O(allyl)); 4.16–4.22

0

(m, 0.7H, CH₂–O(allyl)); 4.54, 4.74 (d, J_{3 ,P}=10.7, 0.7

0

H, H-C(3⁰)); 4.83 (d, J_{1 ,2}=10.4, 0.3 H, H-C(2 )); 4.90

0

(dd, J=1.6, 17.2, 0.3H, H₂C¼(allyl)); 4.96–4.99 (m, 2H,

0

H₂C–O(allyl)); 5.05 (d, J_{1 ,2}=10.3, 0.7 H, H–C(2 ));

5.19–5.49 (m, 4 H, H₂C¼(allyl), HC¼(allyl)), 5.89 (m,

J=5.4, 10.5, 17.1, 0.7H, HC¼(allyl)); 5.99–6.12 (m, 2H,

HC¼(allyl), H-C(1⁰)); 6.83–7.82 (m, 25H, H–C(ar), H–

C(8)). ¹³C NMR (100 MHz, CDCl₃): 19.11, 19.16, 19.22,

19.42 (4q, CH₃(isobut)); 24.39, 24.45, 24.48, 24.54,

25.00, 25.09, 25.12, 25.2 (8q, CH₃(isoprop)); 34.4 (d,

CH(isobut)); 42.8, 42.9 (2dd, J_CP=12.8, CH(isoprop));

55.2 (q, CH₃O(DMT)); 63.7 (dt, J_CP=22.1, CH₂O(al-

lyl)); 64.1 (dt, J_CP=19.8, CH₂O(allyl)); 65.1, 65.2 (2t,

C(5⁰)); 67.8 (t, CH₂O(allyl)); 68.9, 69.1, 70.5, 70.6, 71.3,

71.8, 71.9 (7d, C(2⁰), C(3⁰), C(4⁰)); 78.7, 78.9 (2d, C(1⁰));

87.3, 87.5 (2s, CAr); 113.23, 113.27, 133.31 (3d, C(ar));

115.3, 115.9 (2t, CH₂¼(allyl)); 117.9, 118.0 (2s, C(5));

118.6 (t, CH₂¼(allyl)); 126.9, 127.0, 127.86, 127.93,

128.1, 128.3, 129.7, 129.9, 130.0, 130.37, 130.40, 132.0,

132.9, 133.3, 135.3, 135.3, 135.5, 135.6 (18d, C(ar), CH–

(allyl)); 128.9, 129.5, 136.0, 136.15, 136.15, 136.4, 145.2,

145.5, 158.7, 158.8 (10s, C(ar)); 139.8, 139.9 (2d, C(8));

151.99, 152.03, 152.9, 153.0 (4s, C(2), C(4)); 160.51,

160.54 (2s, C(6)); 164.6, 164.9, 177.3 (3s, CO); ³¹P NMR

(162 MHz, CDCl₃): 138.5, 139.9. FAB-MS (pos, NBA):

1974 (15, [2M+1]⁺), 987 (48, [M+1]⁺), 683 (30), 303

(100), 105 (85).

1

Data of 4c. TLC (hexane/ethyl acetate 1:1): R_f0.58. H

NMR (400 MHz, CDCl₃): 1.13 (d, J=6.8, 0.9H,

H₃C(isoprop)); 1.18, 1.27, 1.43 (3d, J=6.7, 11.1H,

CH₃(isoprop)); 2.64 (dd, J_gem=4.9, 10.3, 0.7H, H-

C(5⁰)); 2.83 (dd, J_{4 ,5}=4.9, J_gem=10.3, 0.3 H, H-C(5⁰));

3.76–3.88 (m, 10 H, CH₃O(DMT), H–C(isoprop), H-

C(4⁰), H-C(5⁰)); 4.06–4.17 (m, 2H, H₂C–O(allyl)); 4.46

(d(br), J=10.0, 0.3H, H–C(3⁰)); 4.62–4.68 (m, 1 H, H-

C(3⁰), H₂C¼(allyl)); 4.82 (m, 1H, H–C(2⁰)); 4.90 (dd,

J=1.7, 17.2, 0.3H, H₂C¼(allyl)); 5.03 (d(br), J=10.3,

0.7H, H₂C¼(allyl)); 5.18 (dd, J=1.6, 17.1, 0.7H,

H₂C¼(allyl)); 5.43 (m, J=5.4, 10.3, 17.1, 0.3H,

HC¼(allyl)); 5.82 (m, J=5.4, 10.6, 17.1, 0.7H,

HC¼(allyl)); 6.40–6.46 (m, 1H, H-(C1⁰)); 6.83 (d,

J=9.0, 4H, H-C(ar)); 7.19–8.42 (m, 21H, H–C(ar), H–

C(5), H–C(6), NH). ¹³C NMR (100 MHz, CDCl₃):

24.47, 24.53, 24.59, 24.95, 25.04, 25.09 (6q, CH₃(iso-

prop)); 42.8 (dd, J_CP=12.6, CH(isoprop)); 43.02 (dd,

J_CP=12.8, CH(isoprop)); 55.22 (q, CH₃O); 63.8 (dt,

J_CP=19.3, CH₂–O(allyl)); 65.3, 65.5 (2t, C(5⁰)); 68.9,

69.2, 70.5, 72.05, 72.13 (5d, C(2⁰), C(3⁰), C(4⁰)); 80.0

(d, C(1⁰)); 87.2 (d, C(5)); 87.5 (s, CAr); 113.21,

113.26, 113.28, 126.9, 127.0, 127.8, 128.1, 128.26,

128.32, 128.36, 129.0, 129.62, 130.1, 130.34, 130.36,

130.43, 130.46 (17d, C(ar)); 115.11, 115.76 (2t,

CH₂¼(allyl)); 132.9, 133.1, 133.4, 135.8, 136.1, 136.25,

136.28, 136.5, 145.5 (8s, C(ar)); 135.6 (d, CH¼

(allyl)); 145.3 (d, C(6)); 158.7 (s, C(2)); 165.1, 165.5

(2s, C(4), CO). ³¹P NMR (162 MHz, CDCl₃): 138.1;

140.2; FAB-MS (pos, NBA): 1882 (17, [2M+1]⁺), 941

0

N⁶-Benzoyl-9-[2⁰-O-benzoyl-4⁰-O-((4⁰⁰,4⁰⁰⁰-dimethoxytri-

phenyl)methyl)-ꢁ-^L-lyxopyranosyl-(1⁰)]adenine (7a). To

a stirred soln of 3.21 g (4.13 mmol) of 6a in 196 mL

THF/H₂O (2:1) at 0 ^ꢀC was added dropwise 6.42 mL of

15% aq NaOH soln. After stirring for 2 h, 40 mL of an

1 M E t NHHCO₃soln was added and THF was

3

removed by distillation at rt. The soln was diluted with

CH₂Cl₂and the resulting mixture was washed with satd

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

2423

aq NaHCO₃soln. The org phase was dried (Na₂SO₄)

and evaporated. The residue was subjected to column

chromatography on silica gel (petroleum ether/acetone

2:1) to aﬀord an amorphous colorless solid. A soln of

this compound in 50 mL of dry CH₂Cl₂and 1.7 mL

pyridine was cooled to 0 ^ꢀC. After dropwise addition of

0.96 mL (8.28 mmol) benzoyl chloride over a period of

30 min, 50 mL of CH₂Cl₂and 20 mL of satd aq

NaHCO₂soln were added. The org phase was washed

with 20 mL of satd aq NaHCO₃soln, 20 mL of H₂O,

dried (Na₂SO₄) and evaporated. The residue was pur-

iﬁed by column chromatography on silica gel (petro-

leum ether/acetone 2:1) and the combined product

fractions were dried (Na₂SO₄) and evaporated. The

colorless amorphous product was dried for 12 h in

vacuo (ꢄ0.1 torr) to furnish 2.60 g (81.1%) of 7a.

C(3⁰)); 70.9 (d, C(2⁰)); 71.3 (d, C(4⁰)); 78.1 (d, C(1⁰));

87.9 (s, C(Ph)₃); 112.1 (s, C(5)); 113.8 (d, arom C); 127.6

(s, arom C); 128.5, 128.9, 129.2, 130.2, 130.67, 130.71,

133.9, 135.9 (8d, arom C); 136.4, 136.5 (2s, arom C);

145.6, 151.2 (2s); 159.2 (s, arom C); 164.1, 165.6 (2s,

CO). HR-FAB-MS (pos, NBA/CsI): 797.1498 (15.5

[M+H⁺Cs]⁺). Anal. calcd for C₃₈H₃₆N₂O₉: C 68.66, H

5.46, N 4.21, found: C 68.45, H 5.48, N 4.74.

N⁴-Benzoy1-[2⁰-O-benzoyl-4⁰-O-((4⁰⁰,4⁰⁰⁰-dimethoxytriphe-

nyl)methyl)-ꢁ-^L-lyxopyranosyl-(1⁰)]cytosine (7c). To a

soln of 500 mg (0.72 mmol) of 6c in 30 mL 1,4-dioxane/

water (2:1) cooled to 0 ^ꢀC, was added dropwise, 1.5 mL

of 15% aq NaOH soln. The soln was allowed to stir for

2 h, at 0 ^ꢀC, until the reaction was complete (TLC). The

soln was diluted with CH₂Cl₂, at 0 ^ꢀC, and extracted

ﬁrst with 50 mL of 1 M aq NaH₂PO₄(pH 6.5), and then

with satd aq NaHCO₃soln. The organic layer was dried

(MgSO₄), ﬁltered and evaporated under vacuum. The

residue was dissolved in 10 mL of dry CH₂Cl₂, cooled to

0 ^ꢀC, and 291 mL (3.6 mmol) of pyridine was added. To

the stirred soln, 162 mL (1.4 mmol) benzoylchloride dis-

solved in 5 mL dry CH₂Cl₂was added dropwise. The

soln was allowed to stir for an additional hour at 0 ^ꢀC. It

was diluted with CH₂Cl₂and extracted with satd aq

NaHCO₃soln. The aqueous layer was extracted twice with

CH₂Cl₂, the combined organic phases dried (MgSO₄), ﬁl-

tered, and was evaporated under vacuum. The residue was

puriﬁed by ﬂash chromatography on a silica gel (acetone/

hexane 1:2 to 2:3) to aﬀord 516mg (95%) of 7c.

Data of 7a. TLC (Et₂O, 100%): R_f0.52. ¹H NMR

(600 MHz, CDCl₃): 3.45 (d, J_gem=12.2, H–C(5⁰)); 3.79

(m, 6H, OCH₃); 3.88–3.90 (m, 2H, H–C(3⁰), H–C(4⁰),

overlapping signals); 4.09 (d, J_gem=12.2, H–C(5⁰)); 6.06

0

(dd, J_{2 ,3}=2.4, J_{1 ,2}=9.6, H–C(2 )); 6.44 (d, J_{1 ,2}=9.6,

H–C(1⁰)); 6.87–7.98 (m, 23 arom H); 8.37, 8.67 (2s, H–

C(2), H–C(8)); 9.14 (s, NH). ¹³C NMR (150.9 MHz,

CDCl₃): 55.6 (q, OCH₃); 67.3 (t, C(5⁰)); 69.2, 71.4 (2d,

C(3⁰), C(4⁰)); 71.6 (d, C(2⁰)); 78.5 (d, C(1⁰)); 88.0 (s,

C(Ph)₃); 113.8 (d, arom C); 122.8 (s, C(5)); 127.6, 128.2,

128.5, 128.7, 129.2, 129.8, 130.7, 133.2, 133.9 (9d, arom

C); 134.0, 136.4, 136.5 (3s, arom C); 141.5 (d, C(8));

145.5, 149.9, 152.1 (3s, arom C); 153.1 (d, C(2)); 159.2

(s, arom C); 164.8, 165.0 (2s, CO). HR-FAB-MS (pos,

NBA/CsI): 1042 (21, [M+H+2Cs]⁺), 910.1822 (100,

[M+H+Cs]⁺). Anal. calcd for C₄₅H₃₉N₅O₈: C 69.49,

H 5.05, N 9.00, found: C 69.24, H 5.34, N 9.10.

1

Data of 7c. TLC (acetone/hexane 1:1): 0.48. H NMR

(200 MHz, CDCl₃): 3.37 (d, br, J_gem=12.4, H–C(5⁰));

3.82 (s, 6H, OCH₃); 3.82–3.90 (m, 2H, H–C(3⁰), H–

C(4⁰)); 3.99 (d, br, J_gem=12.4, H–C(5⁰)); 5.62 (dd,

1-[2⁰-O-Benzoyl-4⁰-O-((4⁰⁰,4⁰⁰⁰-dimethoxytriphenyl)-methyl)-

ꢁ-^L-lyxopyranosyl-(1⁰)]thymine (7b). To a stirred soln of

5.92 g (8.91 mmol) of 6b in 400 mL dry MeOH was

added 400 mg of K₂CO₃and kept for 4 h at 4 ^ꢀC. After

the addition of 50 mL of 1 M Et₃NHHCO₃soln, the

mixture was evaporated and the residue dissolved in

CH₂Cl₂. The org phase was washed with satd aq

NaHCO₃soln, dried (Na₂SO₄) and evaporated. The

remaining violet solid was dried overnight in vacuo

(ꢄ0.2 torr). After dissolving this solid in 27 mL of

CH₂Cl₂and cooling to 0 ^ꢀC, 3.8 mL (47.1 mmol) of dry

pyridine was added. To the stirred soln 1.32 mL

(11.4 mmol) of benzoyl chloride was added dropwise

over a period of 20 min. The soln was washed with satd

aq NaHCO₃soln and extracted with CH₂Cl₂. The org

phase was dried (Na₂SO₄) and evaporated. Puriﬁcation

of the residue by column chromatography on silica gel

(petroleum ether/acetone 2:1) aﬀorded 4.35 g (73.5%)

7b as an amorphous colorless solid..

J_{2 ,3}=2.4, J_{1 ,2}=9.8, H–C(2 )); 6.47 (d, J_{1 ,2}=9.8, H–

C(1⁰)); 7.25–8.12 (m, 23 arom–H, H–C(5), H–C(6)).

FAB-MS (pos, NBA): 303 (74, [DMTr]⁺); 755 (100,

[M+H]⁺); 1508 (23, [2M+H]⁺).

0

9-[2⁰-O-Benzoyl-4⁰-O-((4⁰⁰,4⁰⁰⁰-dimethoxytriphenyl)methyl)-

ꢁ-^L-lyxopyranosyl-(1⁰)]-N²-isobutyrylguanine (7d). To a

soln of 300 mg (0.43 mmol) 6d in 30 mL 1,4-dioxane/

water (2:1) cooled to 0 ^ꢀC, was added dropwise, 1 mL of

15% aq NaOH soln and stirred at 0 ^ꢀC for 3 h. The soln

was diluted with CH₂Cl₂and extracted with 30 mL 1 M

aq NaH₂PO₄soln (pH 6.5) followed by satd aq

NaHCO₃soln. The organic phase was dried (MgSO₄)

and evaporated under vacuum to dryness. The resulting

syrup was puriﬁed by ﬂash chromatography on silica gel

(acetone/hexane 3:2) to aﬀord 263 mg of the debenzoy-

lated intermediate. This was dissolved in 7.5 mL dry

CH₂Cl₂, cooled to 0 ^ꢀC under an argon atmosphere and

added dropwise a soln of 92 mL (0.8 mmol) ben-

zoylchloride dissolved in 3 mL CH₂Cl₂. The soln was

allowed to stir for 1 h at 0 ^ꢀC before it was diluted with a

mixture of CH₂Cl₂and satd aq NaHCO₃soln. The

aqueous phase was extracted twice with CH₂Cl₂. The

combined organic phases were dried (MgSO₄) and

evaporated under vacuum. The resulting residue was

puriﬁed by chromatography on a silica gel column

(acetone/hexane 1:2 to 3:2) to give 280 mg (87%) of

7d.

Data of 7b. TLC (petroleum ether/acetone 2:1): R_f0.40.

¹H NMR (600 MHz, CDCl₃): 1.97 (s, 3H, CH₃–C(5));

3.29 (d, J_gem=12.1, H–C(5⁰)); 3.79 (2s, 6H, OCH₃); 3.85

(m, H-C(4⁰)); 3.89 (m, H–C(3⁰)); 3.95 (d, J_gem=12.1, H–

0

C(5⁰)); 5.59 (dd, J₀_{2 ,3}=2.6, J_{1 ,2}=9.6, H–C(2 )); 6.21 (d,

0

J_{1 ,2}=9.6, H–C(1 )); 6.87–7.94 (m, 18 arom H, H–C(6));

9.36 (s, br, NH). ¹³C NMR (150.9 MHz, CDCl₃): 13.2

(q, CH₃–C(5)); 55.6 (q, OCH₃); 67.0 (t, C(5⁰)); 69.3 (d,

2424

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

1

Data of 7d. TLC (acetone/hexane 2:1): 0.62. H NMR

(300 MHz, CDCl₃): 1.14, 1.16 (d, J=6.9, (CH₃)₂CH));

2.63 (m, 1H, (CH₃)₂CH); 3.29 (d, br, J_gem=12.3, H–

1-[2⁰-O-Benzoyl-3⁰-O-((2-cyanoethoxy)(diisopropylami-

no)phosphino)-4⁰-O-((4⁰⁰,4⁰⁰⁰-dimethoxytriphenyl)methyl)-

ꢁ-^L-lyxopyranosyl-(1⁰)]thymine (8b). To a soln of 620mg

(0.93 mmol) of 7b in 7 mL of dry CH₂Cl₂, 1.0 mL

(7.6 mmol) 2,4,6-collidine and 40 mL (0.5 mmol) N-methy-

limidazole were added followed by 520 mL (2.0mmol) of

97% (2-cyanoethoxy)(diisopropylamino)chlorophosphine

and stirred for 1 h at rt. The soln was diluted with 5 mL

AcOEt and washed with satd aq NaHCO₃soln. The org

phase was dried (MgSO₄) and concentrated. Puriﬁca-

tion of the resulting oil by column chromatography on

silica gel (petroleum ether/AcOEt 3:2) aﬀorded 663 mg

C(5⁰)); 3.75 (s, 6H, OCH₃); 3.81 (d, br, J_{3 ,4}=3.9, H–

C(4⁰)); 3.89 (d, br, J_gem=12.3, H-C(5⁰)); 4.02 (dd,

0

J_{2 ,3}=2.1, J_{3 ,4}=3.9, H–C(3 )); 5.96 (d, J_{1 ,2}=9.6, H–

0

C(1⁰)); 6.12 (dd, J_{3 ,4}=2.1, J_{1 ,2}=9.6, H–C(2 )); 6.84–

7.74 (m, 21H, arom); 7.98 (s, H–C(8)); 8.10 (d,

J=8.1, 1H, arom); 8.58 (d, J=3.6, 1H, arom); 9.11,

11.97 (2s, br, 2H, H–N). FAB-MS (pos, NBA): 303

(100, [DMTr]⁺); 760 (31, [M+H]⁺); 1520 (7,

[2M+H]⁺).

0

1

(82%, mixture of diastereomers ꢄ1:1 by H NMR) of

N⁶-Benzoyl-9-[2⁰-O-benzoyl-3⁰-O-((2-cyanoethoxy)(diiso-

propylamino)phosphino)-4⁰-O-((4⁰⁰,4⁰⁰⁰-dimethoxytriphe-

nyl)methyl)-ꢁ-^L-lyxopyranosyl-(1⁰)]adenine (8a). To a

soln of 787 mg (1.0 mmol) 7a in 7 mL of dry CH₂Cl₂,

0.56 mL (4.3 mmol) 2,4,6-collidine, 40 mL (0.5 mmol) N-

methylimidazole were added followed by 564 mL

(2.2 mmol) of 97% (2-cyanoethoxy)(diisopropyl amino)-

chlorophosphine and stirred for 1 h at rt The soln was

diluted with 5 mL of AcOEt and washed with satd aq

NaHCO₃soln. The org phase was dried (MgSO₄) and

concentrated. The residual oil was puriﬁed by column

chromatography on silica gel (petroleum ether/AcOEt

1:1) to give 727 mg (73%, mixture of diastereomers

ꢂ3:2 by ¹H NMR) of 8a as a slightly yellowish

foam.

8b as a colorless foam.

Data of 8b. TLC (petroleum ether/AcOEt 1:1): R_f0.48.

¹H NMR (600 MHz, CDCl₃): 0.95, 1.07 (2d, J=6.8,

(CH₃)₂CH(a)); 1.00, 1.13 (2d, J=6.8, (CH₃)₂CH(b));

2.01 (s, CH₃–C(5a,b)); 2.33 (t, 2H, J=6.6, –OCH₂CH₂–

CN); 2.78 (d, J_gem=12.4, H–C(5⁰b)); 3.07 (d,

J_gem=12.2, H–C(5⁰a)); 3.42–3.77 (m, H–C(4⁰a), 2H–

OCH₂CH₂CN, H–C(5⁰b), (CH₃)₂CH, overlapping sig-

nals); 3.82 (s, 6H, OCH₃); 3.86 (d, J_gem=12.2, H–

C(5⁰a)); 3.94 (m, H–C(4⁰b)); 4.39 (dt, J_{2 ,3}=2.4,

0

J_{3 ,P}=11.1, H–C(3⁰a)); 4.55 (dt, J_{2 ,3}=2.4, J_{3 ,P}=9.3,

0

H–C(3⁰b)); 5.55 (dd, J_{2 ,3}=2.4, J_{1 ,2}=9.₀6, H–C(2 a));

0

5.64 (dd, J_{2 ,3}=2.4, J_{1 ,2}=9.62, H–C(2 b)); 6.16₀(d,

0

J_{1 2}=9.6, H–C(1 b)); 6.21 (d, J_{1 ,2}=9.6, H–C(1 a));

6.88–7.57 (18 arom H, H–C(6)); 7.97, 8.00 (2d, arom.

H); 8.73, 9.22 (2s, br, NH(a,b)). ¹³C NMR (150.9 MHz,

CDCl₃): 13.13 (q, CH₃–C(5a,b)); 20.05, 20.09 (2t, –

OCH₂CH₂CN(a,b)); 24.65, 24.78 (2q, (CH₃)₂CH(a,b));

43.30, 43.38, 43.52, 43.60 (4d, (CH₃)₂CH(a,b)); 55.58 (q,

OCH₃); 59.08, 59.20 (2t, –OCH₂CH₂CN)); 66.91 (t,

C(5⁰b)); 67.01 (t, C(5⁰a)); 69.82 (d, C(3⁰a)); 69.92 (d,

C(2⁰b)); 70.85 (d, C(2⁰a)); 70.95 (2d, C(4⁰a,b), over-

lapping signals); 72.14 (d, C(3⁰b)); 78.10 (d, C(1⁰b));

78.20 (d, C(1⁰a)); 87.84 (s, C(Ph)₃); 112.21 (s, C(5));

113.82, 113.89 (2d, C(ar)); 118.10 (s, CN); 127.63,

128.53, 128.66, 128.86 (4d, arom C); 129.63 (s, arom C);

130.11, 130.39 (2d, arom C); 130.67 (s, arom C); 130.75,

130.82 (2d, arom C); 130.87 (s, arom C); 134.01, 135.53,

135.84 (3d, arom C); 136.18, 136.29, 145.61 (3s, arom

C); 151.32 (s, C(2)); 163.89, 164.05 (2s, CO); 165.76 (s,

CO (Bz)). ³¹P NMR (242.9 MHz, CDCl₃): 150.20,

152.51 (2s, (a,b)). HR-FAB-MS (pos, NBA/CsI):

997.2518 (100 [M+Cs]⁺), 1129 (9, [M+2Cs]⁺). Anal.

calcd for C₄₇H₅₃N₄O₁₀P: C 65.27, H 6.18, N 6.48,

found: C 65.28, H 5.99, N 6.39.

Data of 8a. TLC (petroleum ether/AcOEt 1:1): R_f0.32.

¹H NMR (600 MHz, CDCl₃): 1.02, 1.09 (2d, J=6.8,

(CH₃)₂CH(a)); 1.12, 1.18 (2d, J=6.8, (CH₃)₂CH(b));

2.10–2.31 (m, 2H, –OCH₂CH₂–CN); 2.87 (d,

J_gem=12.4, H–C(5⁰b)); 3.25 (d, J_gem=12.1, H–C(5⁰a));

3.45 (m, H–C(4⁰a)); 3.56 (2m, (CH₃)₂CH; 1H–CH₂O,

overlapping signals); 3.68 (m, 1H–CH₂–O); 3.78–3.87

(m, H–C(5⁰b); 6H, –OCH₃(a,b), overlapping signals);

3.97 (d, J_gem=12.1, H–C(5⁰a)); 4.02 (d, J_{3 ,4}=4.0, H–

0

C(4⁰b)); 4.43 (m, J_{3 ,P}=10.9, H–C(3⁰a)); 4.65 (m,

0

J_{3 ,P}=9.0, H–C(3⁰b)); 6.03 (dd, J_{2 ,3}=2.5, J_{1 ,2}=9.6, H–

0

C(2⁰a)); 6.15 (dd, J_{2 ,3}=2.5, J_{1 ,2}=9.6, H–C(2 b)); 6.25

0

(d, J_{1 ,2}=9.6, H–C(1 b)); 6.34 (d, J_{1 ,2}=9.6, H–C(1 a));

6.89–8.0 (m, 23 arom H); 8.37, 8.40, 8.82, 8.87 (4s, H–

C(2a,b), H–C(8a,b)); 9.35 (s, NH(a,b)). ¹³C NMR

(150.9 MHz, CDCl₃): 19.8, 19.9 (t,–CH₂–CN); 24.7, 24.9

(q, CH₃(a)); 24.95, 24.97 (q, CH₃(b)); 43.6 (d, CH_i-pr);

57.9 (d, C(4⁰b)); 59.0 (t, O–CH₂); 67.0 (t, C(5⁰b)); 67.2 (t,

C(5⁰a)); 70.2 (d, C(3⁰a)); 70.3, 70.94, 70.99 (3d, C(2⁰b),

C(4⁰a), C(4⁰b)); 71.8 (d, C(2⁰a)); 72.0 (d, C(3⁰b)); 78.7 (d,

C(1⁰b)); 78.9 (d, C(1⁰a)); 113.9 (s, arom C); 122.9 (s,

CN); 127.6, 128.25, 128.27, 128.56, 128.66, 128.67,

128.8, 128.9, 129.2, 129.3, 129.5, 130.0, 130.2, 130.7,

130.81, 130.84, 131.0, 133.12, 133.16, 133.92, 133.98,

134.1, 134.20, 136.3, 136.33, 136.37, 136.6, 145.5, 145.6,

149.9, 150.0, 152.5, 152.5, 159.3, 159.4 (arom C,

C(4a,b), C(5a,b), C(6a,b)); 141.54, 141.63 (2d,

C(8a,b)); 153.5 (2d, C(2a,b), overlapping signals);

165.1 (2s, CO, overlapping signals); 165.3, 165.6 (2s,

CO). ³¹P NMR (242.9 MHz, CDCl₃): 151.54, 152.35

(2s, (a,b)). FAB-MS (pos, NBA/CsI): 1110.2967

(100.0, [M+Cs]⁺), 1242 (9.8, [M+2Cs]⁺). Anal.

calcd for C₅₄H₅₆N₇O₉P: C 67.42, H 5.87, N 10.19,

found: C 66.88, H 6.18, N 9.74.

N⁴-benzoyl-1-[2⁰-O-benzyl-3⁰-O-((2-cyanoethoxy)(diiso-

propylamino)phosphino)-4⁰-O-((4⁰⁰,4⁰⁰⁰-dimethoxytriphe-

nyl)methyl)-ꢁ-^L-lyxopyranosyl-(1⁰)]cytosine

(8c).

A

solution of 252 mg (334 mmol) 7c, 14 mL (0.17 mmol) N-

methylimidazole and 177 mL (1.4 mmol) 2,4,6-collidine

in 3 mL dry CH₂Cl₂was stirred under argon at rt. To

this soln, 203 mL (0.9 mmol) of (2-cyanoethoxy)

(diisopropylamino)chlorophosphine was added drop-

wise and the soln was stirred over night at rt. The

soln was concentrated in vacuum and puriﬁed by chro-

matography on a silicagel column (acetone/hexane 1:2

to 3:2) to yield 242 mg (72%) of 8c as a 4:5 mixture of

diastereomers.

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

2425

Data of 8c. TLC (acetone/hexane 1:1): R_f0.8. ¹H

NMR(300 MHz, CDCl₃): 0.9–1.14 (m, 12H, (CH₃)C);

2.30–2.37 (m, 2H, H–CN(CH₃)₂); 2.78 (d, J_gem=12.0,

H–C(5⁰b)); 3.03 (d, br, J_gem=12.0, H–C(5⁰a)); 3.39–3.61

(m, 4H, NC(CH₂)₂O); 3.73–3.96 (m, H–C(4⁰a,b)); 3.80–

2 min with 0.01 M iodine in acetonitrile/sym-collidine/

water 100:9.2:46. On the Expedite 8909 synthesizer:

Coupling was performed by using a 0.05–0.075 M solu-

tion of the 3⁰-phosphoramidite (4–6 equiv with respect

to nucleotide bound to CPG) and a solution of 0.15 M

4-nitrophenyl tetrazole and 0.35 M tetrazole. As the

machine does not allow cyclic pumping of the amidite/

tetrazole mixture over the CPG-support, the coupling

was divided into four steps. In each step the reaction

mixture, which contained about 1–1.5 equiv of phos-

phoramidite with respect to nucleotide bound to the

CPG support, was allowed to stand on the support for

900 s. Then it was slowly (within 300 s) washed out by

pressing tetrazole solution through the support. Detri-

tylation, capping and oxidation procedures were analo-

gous to the Pharmacia protocol. All oligonucleotides

were synthesized in the ‘Trityl-on’ mode.

3.81 (s, 6H, OCH₃); 4.44–4.62 (m, H–C(3⁰a,b)); 5.55, 5.63

0

(dd, J_{2 ,3}=2.4, J_{1 ,2}=9.3, H–C(2 a,b)); 6.41, 6.48 (d,

0

J_{1 ,2}=9.3, H–C(1 a,b)); 6.86–8.04 (m, 26H, arom, H–C(5),

H–C(6)); 11.35 (s, br, 1H, H–N). ³¹P NMR (300MHz,

CDCl₃): 152.03, 151.14. FAB-MS (pos, NBA): 303 (100,

[DMTr]⁺); 954 (3, [M+H]⁺); 1909 (4, [2M+H]⁺).

9-[2⁰-O-Benzoyl-3⁰-O-((2-cyanoethoxy)(diisopropylami-

no)phosphino)-4⁰-O-((4⁰⁰,4⁰⁰⁰-dimethoxytriphenyl)methyl)-

ꢁ-^L-lyxopyranosyl-(1⁰)]-N²-isobutyrylguanine (8d).

A

solution of 120 mg (158 mmol) 7d, 8 mL (0.1 mmol) N-

methylimidazole and 93 mL (0.7 mmol) 2,4,6-collidine in

1 mL dry CH₂Cl₂was stirred under argon at rt. To this

soln, 90 mL (0.4 mmol) of (2-cyanoethoxy) (diisopropyl-

amino)chlorophosphine was added dropwise and the

soln stirred overnight at rt. The soln was concentrated

in vacuum and puriﬁed by chromatography on a silica-

gel column (AcOEt/hexane 2:3 to 1:0) aﬀording 110 mg

(70%) of 8d as a mixture of diastereomers.

pl(4⁰!3⁰)-Oligonucelotides. Oligonucleotide synthesis

were carried out on a 1 mM scale. The Expedite 8909

synthesizer column was ﬁlled with the CPG solid sup-

port loaded with the appropriate nucleobase. The sub-

strates and reagents required were prepared as follows.

Phosphoroamidites: The amount of phosphoramidite

solution was determined as follows: For adenine, thymine

or the four-base containing sequences: (n+1)Â22 mg of

phosphoramidite dissolved in (n+1)Â312 mL of dry ace-

tonitrile. For guanine or cytosine containing sequences:

(n+2)Â18 mg of phosphoramidite dissolved in

(n+2)Â230 mL of dry acetonitrile (n=number of cou-

plings). The phosphoramidite solution (ca. 0.08 M) was

dried over 3 or 4 A molecular sieves (8–12 mesh, freshly

activated by heating at ca. 300 ^ꢀC under HV overnight)

overnight at rt prior to use. The excess of phoshpor-

amidites, depending on the sequence synthesized, ran-

ged from 160 to 323 equiv Activator solution: 5-

Ethylthio-1H-tetrazole in dry acetonitrile (0.25 M for

adenine or thymine containing sequences and 0.35 M

for guanine or cytosine containing sequences). Capping

A: A solution of 3.0 g of 4-(dimethylamino)-pyridine in

50 mL of dry acetonitrile and ﬁltered to remove any

undissolved solid particles. Capping B. A solution of

10 mL of acetic acid anhydride and 15 mL of 2,4,6-col-

lidine in 25 mL of dry acetonitrile. Oxidizing solution: A

solution of 220 mg of iodine and 4.6 mL of 2,4,6-colli-

dine in 23 mL of water and 50 mL of acetonitrile and

ﬁltered to remove any undissolved residue. Detritylation

reagent: A solution of 6% dichloroacetic acid in 1,2-

dichloroethane. The synthesis of oligonucleotides using

the Perseptive Expedite Gene Synthesizer required the

following modiﬁcations to the protocol provided by

Perseptive for the DNA/RNA synthesis: (1) The dura-

tion of the coupling time of phosphoramidite was about

15–16.7 min and (2) the detritylation was accomplished

by 6% dichloroacetic acid in 1,2-dichloroethane over a

3 min period. All oligonucleotides were synthesized in

the ‘Trityl-on’ mode.

Data of 8d. TLC (AcOEt/hexane 2:1): R_f0.21. ¹H NMR

(300 MHz, CDCl₃): 0.99–1.21 (m, 6H, (CH₃)₂CH); 1.25

(d, J=6.9, 6H, (CH₃)₂CHN); 1.27 (d, J=6.6, 6H,

(CH₃)₂CHN); 2.31–2.75 (m, 6H, NC(CH₂)₂O, H–

CN(CH₃)₂); 2.92–2.99 (2d, J_gem=12.3, H–C(5⁰a,b));

3.30–3.70 (m, 3H, H–C(4⁰a,b), H–C(5⁰a,b), H–

C(CH₃)₂); 3.79 (s, 6H, OCH₃); 4.70, 4.76 (m, H–

C(3⁰a,b)); 5.81–5.92 (m, H–C(2⁰b), H–C(1⁰a)); 6.02 (d,

0

J_{1 ,2}=9.6, H–C(1 b)); 6.36 (dd, J_{2 ,3}=2.1, J_{1 ,2}=9.6,

H–C(2⁰a)); 6.82–7.95 (m, 18H, arom); 8.01, 8.04 (s, 1H,

H–C(8a,b)); 9.20 (s, br, 1H, H–N). ³¹P NMR (121 MHz,

CDCl₃): 150.4, 151.3. FAB-MS (pos, NBA): 303 (100,

[DMTr]⁺); 960 (13, [M+H]⁺); 1920 (52, 2M+H]⁺).

Automated solid-phase synthesis on the gene synthesizer

General. The phosphoramidite solutions and the acti-

vator solutions were dried over 3 or 4 A molecular sieves

(8–12 mesh, freshly activated by heating at ca. 300 ^ꢀC

under HV overnight) overnight, in a dessicator (con-

taining KOH), at rt prior to use.

pr(4⁰!3⁰)-oligonucelotides.

p-RNA-Oligonucleotide

syntheses were performed in a 10 mmol scale on either a

Gene Assembler Plus^TM(Pharmacia) or on an Persep-

tive Expedite 8909 Nucleic Acid Synthesis System (Per-

Septive Biosystems). On the Pharmacia synthesizer, the

protocol included the following steps: Detritylation for

4.5 min with 6% dichloroacetic acid in dichloroethane,

Coupling for 60 min with a 10% (w/v) solution of

phosphoramidite in acetonitrile (0.1 M, 500 mL per cou-

pling; nÂ40 mg of phosphoramidite dissolved in

nÂ500 mL of dry acetonitrile, n=number of couplings).

Activation by adding a mixture of 0.15 M 4-nitrophe-

nyl-tetrazole and 0.35 M tetrazole in acetonitrile (600 mL

per coupling). Capping for 1.5 min with 0.45 M dime-

thylaminopyridine in acetonitrile and a mixture of col-

lidine/acetic anhydride/acetonitrile 3:2:5. Oxidation for

Post-automation procedures

pr(4⁰!3⁰)-Oligonucleotides. The post-automation work

up procedures were identical with those developed for

the 4⁰,2⁰-p-RNA.³

2426

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

Removal of the phosphotriester-allyl protecting group.

The CPG solid support, containing the oligonucleotide

(‘trityl-on’) from the synthesizer, was dried in HV for

30 min at rt, and treated with 90 mg (78.0 mmol) of

Pd(PPh₃)₄, 90 mg (344 mmol) of PPh₃, 90 mg (666 mmol)

of Et₂NH^ꢃH₂CO₃in 5 mL of CH₂Cl₂at rt, following the

procedure of Noyori et al.²¹The CPG was ﬁltered,

washed sequentially with 0.1 M aq sodium diethyl

dithiocarbamate, CH₂Cl₂, acetone and water, then

dried.

tection conditions (25% aq hydrazine hydrate). There-

fore a modiﬁed approach was necessary to minimize

strand scission during the cleavage of the oligo from the

CPG solid support and the removal of the protecting

groups on the base and sugar moieties.

Removal of phosphotriester-ꢀ-cyanoethyl protecting

group. After the automated synthesis was completed,

the CPG-solid support containing the oligonucleotide

(‘Trityl-on’) was dried in vacuo for 30 min at rt, trans-

ferred to a pear shaped 10 mL ﬂask and treated with

2.4 mL of pyridine/triethylamine (5:1) for 6.5 h at rt.

Evaporation of pyridine and triethylamine in vacuo

followed by coevaporation with toluene (or DMF) to

dryness—avoiding temperatures over 35 ^ꢀC—resulted in

dry CPG solid support material.

Removal of oligonucleotide from solid support and depro-

tection of acyl-protecting groups. The dried CPG was

treated with 25% aq NH₂NH₂hydrate in water (pre-

pared from 1 mL NH₂NH₂hydrate and 4 mL water) at

4 ^ꢀC for 24 h. Deprotection was monitored by ion

exchange HPLC. After complete deprotection, the CPG

was ﬁltered and washed sequentially with water. The

washings were combined with ﬁltrate and were desalted

(refer to desalting of oligonucelotides) over a Waters

Sepak-C18 cartridge [eluted with 10–15 mL acetonitrile/

water (1:1)] to aﬀord the salt free, crude oligonucleo-

tides (‘Trityl-on’) in solution.

Detachment of oligonucleotide from CPG solid support

with concomitant deprotection of sugar and nucleobase

protecting groups. One of the following two procedures

were used depending on the sequence of the oligonu-

cleotides (Table 1 lists the speciﬁc deprotection method

for the speciﬁc sequence).

Detritylation. The crude oligonucleotide solution

obtained by above was concentrated in vacuo to dry-

ness, the residue treated with about 10 mL of 80% aq

formic acid (a red color appears within seconds indicat-

ing detritylation) at rt for 15–30 min and concentrated

in vacuo to dryness. The residue was dissolved in about

2 mL of water, ﬁltered (Nalgene syringe ﬁlter, 0.2 mM)

and taken to the next step of HPLC puriﬁcation (Fig.

5).

Method A. To the ﬂask containing the dry CPG solid

support was added 2 mL of a mixture of 40% aq

CH₃NH₂in concd aq NH₃(1:1) and shaken at rt for

6.5 h (ca. 70 h for G,C containing sequences). The sus-

pension was co-evaporated with water, carefully (with

the temperature always less than 35 ^ꢀC) to remove the

volatiles CH₃NH₂and NH₃, and ﬁltered. The ﬁltrate

was diluted with about 5–10 mL of 0.5 M aq

Et₃NH₂CO₃buﬀer and desalted (refer to desalting of

oligonucleotides) over a Waters Sepak-C18 cartridge

[eluted with 10–15 mL acetonitrile/water (1:1)] to aﬀord

the salt free, crude oligonucleotides (‘Trityl-on’) in

solution.

pl(4⁰!3⁰)-oligonucleotides. The oligonucleotides with the

4⁰,3⁰-connection were less stable to the regular depro-

Method B. To the ﬂask containing the dry CPG solid

solution of

support was added 2 mL of

a

ꢃ

CH₃ONH₂HCl dissolved in concd aq NH₃and EtOH

(3:1) and shaken at rt for approx. 6.5 h (ca. 70 h for G,C

containing sequences). The suspension was ﬁltered, the

ﬁltrate diluted with about 5–10 mL of 0.5 M aq

Et₃NH₂CO₃buﬀer and desalted (refer to desalting of

oligonucleotides) over a Waters Sepak-C18 cartridge

[eluted with 10–15 mL acetonitrile/water (1:1)] to aﬀord

the salt free, crude oligonucleotides (‘Trityl-on’) in

solution.

All of the above deprotections were monitored by anion

exchange HPLC for optimum deprotection time.

Detritylation of ‘Trityl-on’ oligonucleotides. This was

performed as described for pr(4⁰!3⁰)-oligonucleotides.

HPLC puriﬁcation of oligonucleotides

The crude oligonucleotides were puriﬁed by anion

exchange (IA)-HPLC system, over a Mono Q HR 5/5

(Pharmacia) column, performed on (A) Pharmacia GP-

250 Gradient Programmer equipped with two Pharma-

cia P-500 pumps, ABI-Kratos Spectraﬂow 757 UV/Vis

Figure 5. HPLC trace of the crude, fully deprotected, (d)-b-

and

ribopyranosyl(4⁰!3⁰)-oligonucleotides,

pr(4⁰!3⁰)(A₄T₄)

pr(4⁰!3⁰)(T₄A₄). For conditions of chromatography, see Table 1.

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

2427

detector and a Hewlett Packard HP 3396A analogue

integrator or (B) Pharmacia Akta puriﬁer (900) con-

trolled by UNICORN. The oligonucleotides were eluted

from the column using a linear gradient of 1 M NaCl in

10 mM aq Na₂HPO₄, pH=10.5 over a period of 20–

30 min with the following buﬀer systems: buﬀer A:

10 mM Na₂HPO₄in H₂O, pH=10.5; buﬀer B: 10 mM

Na₂HPO₄in H₂O, 1 M NaCl, pH=10.5. Exact gradient

conditions are given in Table 1. The pure product con-

taining fractions were collected in Eppendorf vials con-

taining 20 mL of 1M aq acetic acid (to neutralize the

high pH), combined and desalted to remove excess salt.

In the cases where the ﬁrst HPLC puriﬁcation gave still

impure oligonucleotides, the combined product frac-

tions were desalted and re-puriﬁed by HPLC.

9. (a) For recent reviews of anti-sense oriented oligonucleotide

chemistry, see: Hunziker, J.; Leumann, C. In Modern Syn-

thetic Methods; Ernst, B., Leumann, C., Eds.; Helv. Chim.

Acta: Basel, Switzerland, 1995; Vol, 7, p 331: P. Herdewijn,

Liebigs Ann. Chem. 1996, 1337. (b) Hyrup, B.; Nielsen, P.

Bioorg. Chem. Med. Chem. Lett. 1994, 4, 5.

10. (a) For exceptions to the rule in nucleic acid analogues

that are not based on phosphodiester inter-nucleotide lin-

kages, see: Stork, G.; Zhang, C.; Gryaznov, S.; Schultz, R.

Tetrahedron Lett. 1995, 36, 6387. (b) Maesmaker, A.; Wald-

ner, A.; Wendeborn, S.; Wolf, R. M. Pure Appl. Chem. 1997,

69, 437. (c) Diederichsen, U. Angew. Chem. Int. Ed. Engl.

1996, 35, 445.

11. Kierzek, R.; He, L.; Turner, D. H. Nucl. Acids Res. 1992,

20, 1685.

12. (a) Dougherty, J. P.; Rizo, C. J.; Breslow, R. J. Am.

Chem. Soc. 1992, 114, 6254. (b) Switzer, C.; Hashimoto, H. J.

Am. Chem. Soc. 1992, 114, 6255. (c) Breslow, R.; Sheppard,

T. L. J. Am. Chem. Soc. 1996, 118, 9810. (d) Prakash, T. P.;

Jung, K.-E.; Switzer, C. Chem. Commmun. 1996, 1793. (e) See

also: Usher, D. A. Nature New Biol. 1972, 235, 207. (f) Usher,

D. A.; McHale, A. H. Proc. Natl. Acad. Sci. U.S.A. 1976, 73,

1149.

13. Reck, F.; Wippo, H.; Kudick, R.; Bolli, M.; Ceulemans,

G.; Krishnamurthy, R.; Eschenmoser, A. Org. Lett. 1999, 1,

1531.

14. Pitsch, S.; Krishnamurthy, R.; Bolli, M.; Wendeborn, S.;

Holzner, A.; Minton, M.; Lesueur, C.; Schlonvogt, I.; Jaun,

B.; Eschenmoser, A. Helv. Chim. Acta 1995, 78, 1621.

15. Bolli, M.; Micura, R.; Pitsch, S.; Eschenmoser, A. Helv.

Chim. Acta 1997, 80, 1901.

Desalting of oligonucleotides

For desalting, a Waters Sepak-C18 cartridge was equi-

librated with 10 mL of acetonitrile/water (1:1) followed

by 10 mL of water and ﬁnally 10 mL of 0.2 M aq

Et₃NH₂CO₃. The oligonucleotide solution was diluted

with 10 mL of 0.1–0.2 M aq Et₃NH₂CO₃solution and

loaded onto the Waters Sepak-C18 cartridge. The car-

tridge was eluted with 10 mL of 0.2 M (or 0.5 M)

Et₃NH₂CO₃, followed by 15–30 mL of acetonitrile/

water (1:1). The eluted acetonitrile/water fractions con-

taining the oligonucleotide (monitored by UV at

260 nm) were combined and concentrated in vacuo to

dryness. The residue was co-evaporated three times with

10 mL water to remove excess buﬀer. The residue was

dissolved in the desired amount of water to give a salt-

free stock solution of the oligonucleotide and stored at

À20 ^ꢀC.

16. Bolli, M.; Micura, R.; Eschenmoser, A. Chem. Biol. 1997,

4, 309.

17. The full paper on the work described in the preliminary

communications of refs 3, 14, 15, 16, and 30 will be published

in Helv. Chim. Acta (S. Pitsch, S. Wendeborn, R. Krishna-

murthy, A. Holzner, M. Minton, M. Bolli, R. Micura, C.

Miculka, N. Windhab, M. Stanek, B. Jaun, A. Eschenmoser).

18. Reck, F.; Wippo, H.; Kudick, R.; Krishnamurthy, R.;

Eschenmoser, A. Helv. Chim. Acta In press.

19. (a) Hayakawa, Y.; Wakabayashi, S.; Kato, H.; Noyori, R. J.

Am. Chem. Soc. 1990, 112, 1691. (b) Carruthers, M. H.; Matteuci,

M. D. J. Am. Chem. Soc. 1981, 103, 3185. (c) Carruthers, M. H.;

Beaucage, S. L. Tetrahedron Lett. 1981, 22, 1859.

Acknowledgements

The work was supported by the Skaggs Foundation

(TSRI) and Novartis AG, Basel.

20. As the preparation of (allyloxy)(diisopropylamino)-

chlorophosphine has recently been reported to be trouble-

some,²¹a detailed procedure for its preparation is given in the

Experimental.

References and Notes

1. Eschenmoser, A. Science 1999, 284, 2118.

2. Eschenmoser, A.; Dobler, M. Helv. Chim. Acta 1992, 75,

218.

21. (a) Pirrung, M. C.; Fallon, L.; Lever, D. C.; Shuey, S. W.

J. Org. Chem. 1996, 61, 2129. (b) Kamber, M.; George, J. Can.

J. Chem. 1985, 63, 823.

3. Pitsch, S.; Wendeborn, S.; Jaun, B.; Eschenmoser, A. Helv.

Chim. Acta 1993, 76, 2161.

4. See e.g. Moser, H.; Fliri, A.; Steiger, A.; Costello, G.;

Schreiber, J.; Eschenmoser, A. Helv. Chim. Acta 1986, 69,

1224.

22. Reaction with 1.2 equiv of benzoyl chloride in dichloro-

methane at 0 ^ꢀC in the presence of 5 mol equiv Pyridine. 2⁰-

and 3⁰-benzoyl derivatives 6a,b and 7a,b, respectively, are

clearly distinguishable by their NMR spectra.

23. Scaringe, S. A.; Francklyn, C.; Usman, N. Nucl. Acids.

Res. 1991, 18, 5433 The choice of allyl protection of the

phosphate in the (d)-b-ribopyranosyl-(4⁰!3⁰)-oligomers was

an extension of the protocol used in the pr-(4⁰!2⁰)-series while

the change to 2-cyanoethyl protection of the phosphate in (l)-a-

lyxopyranosyl-(4⁰!3⁰)-oligomers was a matter of convenience

and preference to use the commercially available phosphorylat-

ing agent, (2-cyanoethoxy)(diisopropylamino)phosphine.

24. Sood, A. K.; Narang, S. A. Nucl. Acids. Res. 1977, 4,

2757.

5. Pitsch, S.; Pombo-Villar, E.; Eschenmoser, A. Helv. Chim.

Acta 1994, 77, 2251.

6. (a) Hunziker, J.; Roth, H.-J.; Bohringer, M.; Giger, A.;

Diederichsen, U.; Gobel, M.; Krishnan, R.; Jaun, B.; Leu-

mann, C.; Eschenmoser, A. Helv. Chim. Acta 1993, 76, 259.

(b) Groebke, K.; Hunziker, J.; Fraser, W.; Peng, L.; Dieder-

ichsen, U.; Zimmermann, K.; Holzner, A.; Leumann, C.;

Eschenmoser, A. Helv. Chim. Acta 1998, 81, 375.

7. Beier, M.; Reck, F.; Wagner, T.; Krishnamurthy, R.;

Eschenmoser, A. Science 1999, 283, 699.

25. (a) Tinoco, I., Jr.; Sauer, K.; Wang, J. C. In Physical

Chemistry: Principles and Applications in Biological Sciences;

Young, D., Cavanaugh, D., Eds.; Prentice Hall: Upper Saddle

River, NJ, 1995; p 559. (b) Marky, L. A.; Breslauer, K. J.

8. Jungmann, O.; Wippo, H.; Stanek, M.; Huynh, H. K.;

Krishnamurthy, R.; Eschenmoser, A. Org. Lett. 1999, 10,

1527.

2428

H. Wippo et al. / Bioorg. Med. Chem. 9 (2001) 2411–2428

Biopolymers 1987, 26, 1601. (c) Cantor, C. R.; Schimmel, P. R.

Biophysical Chemistry; Freeman: San Francisco, CA, 1980;

Part III, p 1135.

personal communication)] have shown to have this conforma-

tion by NMR spectroscopy. For the lyxopyranosyl-(4⁰!3⁰)-

system, the assignment rests on the reasoning that the inverted

chair would force the two (charged) vicinal phosphodiester

groups into sterically and electrostatically unfavorable contact

in the di-equatorial (syn-clinal) conformation.

30. Micura, R.; Kudick, R.; Pitsch, S.; Eschenmoser, A.

Angew. Chem. Int. Ed. 1999, 38, 680.

31. Schlonvogt, I.; Pitsch, S.; Lesueur, C.; Eschenmoser, A.;

Jaun, B. Helv. Chim. Acta 1996, 79, 2316.

32. In the idealized conformation used in Fig. 4, the nucleosi-

dic torsion angle (which is an important determinant for the

backbone/base-pair-axes inclination³⁰is taken to be À120^ꢀ.

The inclinations deduced from these vertical projections of

conformational formulas are rough approximations only.

More signiﬁcant assignments are obtained using a quantiﬁable

deﬁnition of the backbone/base-pair-axes inclination sug-

gested in the context of this problem by M. Egli and P. Lubini

(to be published).³³

33. Private communication. M. Egli, E. Vanderbilt Uni-

versity.

34. Schoning, K.-U.; Scholz, P.; Guntha, S.; Wu, X.;

Krishnamurthy, R.; Eschenmoser, A. Science 2000, 290,

1347.

26. The corresponding p-RNA duplex in the all-(4⁰!2⁰) series

melts at 82 ^ꢀC under similar conditions: Krishnamurthy, R.;

Pitsch, S.; Minton, M.; Miculka, C.; Windhab, N.; Eschen-

moser, A. Angew. Chem., Int. Ed. Engl. 1996, 35, 1537 In an

exploratory experiment, attempts to prepare pl(4⁰!3⁰)(G₈)

failed due to low coupling yields on the oligonucleotide syn-

thesizer and diﬃculties encountered during the attempted

HPLC-puriﬁcation of the deprotected oligonucleotide mix-

ture.

27. For further ﬁgures (T_mcurves, CD and mixing curves)

documenting the pairing behavior of the (4⁰!3⁰)-lyxopyr-

anosyl system, see ref 13.

28. One of the limiting factors was diﬃculties encountered in

the preparation of pure G-containing pl-(4⁰!3⁰)-oligomers.

29. This reasoning assumes that both the (4⁰!2⁰)- and the

(4⁰!3⁰)-lyxopyranosyl systems have the same type of chair

conformation with the equatorial positioning of the nucleo-

base. For the former, the presence of the conformation is

strongly indicated by the system’s cross-pairing with all other

members of pentopyranosyl-(4⁰!2⁰)-family, two of which

[ribo-³¹and arabino-pyranosyl (B. Jaun and O. Ebert, ETH,

Article Doi

Pentopyranosyl oligonucleotide systems. Part 11: Systems with shortened backbones: (D)-β-ribopyranosyl-(4′→3′)- and (L)-α-lyxopyranosyl-(4′→3′)-oligonucleotides

DOI: 10.1016/S0968-0896(01)00220-6

Source and publish data:

Authors:

Article abstract of DOI:10.1016/S0968-0896(01)00220-6

Full text of DOI:10.1016/S0968-0896(01)00220-6

Products guided by the article

R&D Labs maybe for 389089-98-9

Relevant to this article

Hot Product