Pentopyranosyl Oligonucleotide Systems: 9th Communication. The β-D-Ribopyranosyl-(4′→2′)-oligonucleotide System ('Pyranosyl-RNA'): Synthesis and Resume of Base-Pairing Properties

Article abstract of DOI:10.1002/hlca.200390349

Pyranosyl-RNA ('p-RNA') is an oligonucleotide system isomeric to natural RNA and composed of the very same building blocks as RNA. Its generational, chemical, and informational properties are deemed to be those of an alternative nucleic acid system that c

Full text of DOI:10.1002/hlca.200390349

4
270
Helvetica Chimica Acta ± Vol. 86 (2003)
Pentopyranosyl Oligonucleotide Systems
9
th Communication1)
The b-d-Ribopyranosyl-(4' ! 2')-oligonucleotide System (−Pyranosyl-RNA×):
Synthesis and Resum e¬ of Base-Pairing Properties
2
a
2b
2c
2d
by Stefan Pitsch ), Sebastian Wendeborn ), Ramanarayanan Krishnamurthy* ), Armin Holzner ), Mark
2
e
2f
2g
2h
2i
2k
Minton ), Martin Bolli ), Christian Miculca ), Norbert Windhab ), Ronald Micura ), Michael Stanek ),
Bernhard Jaun, and Albert Eschenmoser*
Laboratory of Organic Chemistry, Swiss Federal Institute of Technology (ETH), Hˆnggerberg, HCI H309,
CH-8093 Z¸rich
and
The Skaggs Institute for Chemical Biology at The Scripps Research Institute (TSRI),
10550 North Torrey Pines Road, La Jolla, CA-92037, USA
Dulio Arigoni zu seinem 75. Geburtstag gewidmet, in Erinnerung an gute alte Zeiten
Pyranosyl-RNA (−p-RNA× ) is an oligonucleotide system isomeric to natural RNA and composed of the
very same building blocks as RNA. Its generational, chemical, and informational properties are deemed to be
those of an alternative nucleic acid system that could have been a candidate in Nature×s evolutionary choice of
the molecular basis of genetic function. We consider the study of the chemistry of p-RNA as etiologically
relevant in the sense that knowledge of its structural, chemical, and informational properties on the chemical
level offers both a perspective and reference points for the recognition of specific structural assets of the RNA
structure that made it the (supposedly) superior system among possible alternatives and, therefore, the system
that became part of biology as we know it today. The paper describes the chemical synthesis of b-d- (and l)-
ribopyranosyl-(4' ! 2')-oligonucleotide sequences, presents a resume of their structural and chemical properties,
and cautiously discusses what we may and may not have learned from the pyranosyl isomer of RNAwith respect
to the conundrum of RNA×s origin.
1)
For communications No. 1 ± 8 see [1 ± 8]. The label −Chemistry of Pyranosyl-RNA× previously used for the
series of papers on p-RNA [1 ± 8] has been changed (see [9]) into −Pentopyranosyl Oligonucleotide
Systems× as a consequence of the extension of our work on p-RNA to a whole family of diastereoisomeric
pentopyranosyl oligonucleotides. In [9], the present paper had been assigned No. 9 in this series. The
follow-up papers No. 10 [9], 11 [10], 12 [11], and 13 [12] in the series have already appeared. The present
paper also is communication No. 30 in the series −Chemistry of a-Aminonitriles×. The assignment of papers
in this latter series is as follows: No. 29 is [13], [14] counts as No. 28, [15] is No. 27, [16] is No. 26, [17]
counts as No. 25, [8] is No. 24, [7] counts as No. 23, [18] is No. 22, [5] No. 21, [6] No. 20, [4] counts as No. 19,
[3] is No. 18, [19] No. 17, [2] No. 16, [20] No. 15, [21] No. 14, [22] No. 13, [23] No. 12, and [1] is No. 11; the
communication assigned to be No. 10 [24] has not appeared yet; [25] is No. 9, [26] No. 8, [27] No. 7, [28]
No. 6, [29] No. 5, [30] No. 4, [31] No. 3, [32] No. 2, and [33] is No. 1.
Postdoctorates: a) ETH September 1992 ± February 1995; b) ETH November 1992 ± December 1993;
c) ETH July 1992 ± December 1994 and TSRI1996 ±; d) ETH April 1992 ± October 1993; e) ETH
September 1993 ± April 1994; f) ETH and TSRIJanuary 1995 ± June 1997; g) Universit‰t Frankfurt
October 1993 ± M‰rz 1995; h) Universit‰t Frankfurt 1993 ± 1995; i) ETH and TSRIJanuary 1996 ± June
2)
1998; k) TSRIJune 1998 ± June 1999.

Helvetica Chimica Acta ± Vol. 86 (2003)
4271
1. Introduction. ± The b-d-Ribopyranosyl-(4' ! 2')-oligonucleotide system (−pyra-
nosyl-RNA× or −p-RNA×) is a constitutional isomer of RNA, containing its d-ribose
building block in the pyranose (instead of the furanose) form, and the phosphodiester
linkage between the d-ribose C-atoms C(4') and C(2') (instead of C(5') and C(3')).
From a chemical point of view, the type of generational chemistry that allows the
molecular structure of contemporary RNA to be derived from its building blocks (d-
ribose, phosphate and nucleobases) might equally give rise to (and have given rise to)
the pyranosyl isomer of RNA. Therefore, this latter system is deemed to have been a
potentially natural, evolutionary alternative to RNA. Why, then, did Nature choose the
furanose and not the pyranose form of RNA as an informational molecular system?
Any comment to this question has as its absolute pre-requisite a knowledge of p-RNA×s
chemical properties, a quest primarily directed to the property that constitutes the
chemical basis of RNA×s role in Biology, namely, its capability to undergo informa-
tional WatsonÀCrick pairing. Thus, the study of the chemistry of the pyranosyl isomer
of RNA must be part of any experimental approach to a chemical etiology of the
structure type of the natural nucleic acids.
In a series of preliminary communications in this and other journals [1 ± 8], we had
reported on the synthesis, the structure, and the chemical properties of p-RNA. These
studies were later extended to a whole family of pentopyranosyl oligonucleotide
systems [15 ± 17], all of them representing potentially natural nucleic acid alternatives
according to the criteria previously defined [34 ± 40]. While our preliminary
communications on these latter studies have been complemented recently by full
papers in this journal reporting in detail the procedures used in synthesis, purification,
and characterization of the pentopyranosyl oligonucleotides investigated [9 ± 12], the
corresponding paper on p-RNA, describing the synthesis of b-d-ribopyranosyl-(4' !
2
')-oligonucleotides in detail, had been unduly postponed. The present paper closes this
gap.
Work on the pyranosyl-isomer of RNA was initiated in 1992 as a consequence of
observations we had made in studies on the properties of hexopyranosyl-(6' ! 4')
alternatives of RNA. We had found that none of the three investigated members of the
hexopyranosyl-(6' ! 4') family, namely, oligonucleotides of the b-allo-, b-altro- and b-
glucopyranosyl series (Fig. 1), possesses a capacity for WatsonÀCrick pairing that
would be comparable in strength, selectivity, and informational consistency to that of
corresponding RNA oligonucleotides [24][36 ± 38][41][42]. This had led us to
conclude that, for functional reasons, such hexopyranosyl-(6' ! 4') analogues of
3
RNA could not have acted as evolutionary competitors of the RNA structure ). The
inability of these potentially natural (fully hydroxylated) hexopyranosyl systems to
undergo efficient WatsonÀCrick base pairing was in sharp contrast to the remarkable
pairing capability of the −unnatural× 2',3'-dideoxy model system homo-DNA (Fig. 1),
which we had investigated before, and whose base-pairing strength in both the
3)
In a strict sense, this conclusion is valid only for the three investigated members from the eight
diastereoisomers of the hexopyranosyl-(6' ! 4')-oligonucleotide family containing their nucleobases in
equatorial positions. Analogous intrastrand steric hindrance in the pairing conformation is, however, to be
expected for (most of) the remaining members. For pairing properties of hexapyranosyl oligonucleotide
systems with non-nucleosidic (−unnatural×) structure, see [43 ± 45].

4
272
Helvetica Chimica Acta ± Vol. 86 (2003)
Fig. 1. Members of the family of b-d-hexopyranosyl-(6' ! 4')-oligonucleotides dealt with in previous studies
WatsonÀCrick and (reverse) Hoogsteen mode had been found to be greater than that
4
of either DNA or RNA ). This contrast has been rationalized as resulting from
intrastrand steric hindrance in the pairing conformation of the fully hydroxylated
Fig. 2. The two pairing-conformations of the homo-DNA system as deduced by qualitative conformational
analysis [29] and observed by NMR structure determination [26].
4)
The criterion for deciding whether a RNA analogue is to be deemed a −potentially natural× nucleic acid
alternative refers to whether the structure of an analogue can be derived from a (CH O) aldose sugar
n ˆ 4 ± 6) by the same type of (potentially natural) chemistry that allows the structure of RNA to be
derived from ribose [29][40].
2
n
(

Helvetica Chimica Acta ± Vol. 86 (2003)
4273
hexopyranosyl systems. Such hindrance is expected to occur between an equatorial 2'-
OH group of a pyranosyl unit and the neighboring (downstream-positioned)
nucleobase; it is clearly to be inferred from model considerations based on homo-
DNA×s two types of pairing conformations (À g/ À g and t/ À g), which were derived by
5
qualitative conformational analysis [29] ) (Fig. 2), and were found to co-exist in a
homo-DNA duplex by an NMR structure analysis by Otting et al [26]. The results of
experimental studies on 2'-deoxy- and 3'-deoxyallopyranosyl model systems
[
24][46][47] (Table 1) corroborated this interpretation.
Table 1. Relative Base-Pairing Strengths in b-d-Allopyranosyl-Oligonucleotide Derivatives
R3
'
R2'
Relative pairing strength
OH
H
OH
H
very weak [41][38]
very strong [25]
OH
H
H
OH
strong [46][38][37]
weak [46][38][37]
It was this insight into the consequences of the steric bulk of fully hydroxylated
6' ! 4')-hexopyranosyl building blocks that led us to ask whether base pairing might be
(
5)
Homo-DNA×s g/Àg pairing conformation corresponds to the pairing conformation of RNA (and A-type
conformation of DNA) as illustrated by the juxtaposition of the formulae below. Herdewijn×s closely
related hexitol-NA [43] is a constitutional isomer of homo-DNA, it has a non-nucleosidic constitution and
is, therefore, not to be considered a potentially natural nucleic acid alternative. Its importance lies in its
capability to cross-pair with RNA (its nucleobase occupies an axial conformation [43]). Herdewijn×s
observations on the diverging base-pairing capabilities of the two epimeric mono-hydroxylated HNA
derivatives (the axial epimer cross-pairs with RNA [44], the equatorial epimer does not [45]) can equally
be rationalized as being due to intrastrand steric hindrance in the latter [45]. Of special interest to the
conformational analysis of oligonucleotide systems is Wengel×s −LNA× [48], the strongest WatsonÀCrick
oligonucleotide base-pairing system encountered thus far. Its extraordinary pairing strength convincingly
corroborates the relevance of the steric criteria that had been used for deriving the pairing conformation
of homo-DNA (and, implicitly, also of A-type DNA) by qualitative conformational analysis [29][1].

4
274
Helvetica Chimica Acta ± Vol. 86 (2003)
encountered in systems which, while still having their sugar unit in the pyranose form,
would contain only five instead of six C-atoms. In such pentopyranosyl systems, the
phosphodiester groups could not link, however, the pyranose units in the same way as
they link the pentofuranosyl and the hexapyranosyl units in RNA and homo-DNA,
6
respectively. According to a conjecture that, at the time, was believed to be valid ), any
oligonucleotide system expected to possess the capability of informational base pairing
should have in its backbone six covalent bonds per monomer unit, as is the case in the
backbone of the natural nucleic acids (as well as of homo-DNA). Therefore, the
backbone of a pentopyranosyl oligonucleotide system would have to link the pyranosyl
units between the sugar atoms C(4') and C(2') (Fig. 3). Interestingly, such a (4' ! 2')-
positioning of the phosphodiester bridge in a pyranosyl isomer of RNA was highly
reminiscent of an observation we had made earlier in a project relating to a problem in
prebiotic chemistry: base-catalyzed aldomerization of glycolaldehyde phosphate under
appropriate conditions in the presence of (0.5 equiv.) formaldehyde leads to (rac)
ribopyranose-2,4-diphosphate as the most abundant component of a reaction mixture
consisting of all possible diastereoisomeric (rac) tetrose-2,4-diphosphates, pentose-2,4-
7
diphosphates, and hexose-2,4,6-triphosphates [30] ).
Fig. 3. Constitutional, configurational, and conformational formulae representing the b-d-ribopyranosyl-(4' !
2
')-oligonucleotide (p-RNA) system
At the outset of our work on p-RNA stood a qualitative conformational analysis of
the repeating unit of the b-ribopyranosyl-(4' ! 2')-backbone. This analysis was
delineated in pictorial detail in our first communication on p-RNA [1]; it showed
that, within the ensemble of nine least strained (idealized) conformations, one, and
only one, of them is conformationally repetitive and, therefore, a possible pairing
6)
7)
In later studies, it was found that this requirement is of limited validity only; see [10].
For a more detailed discussion of the relationship between these findings and the pentopyranosyl-(4' ! 2')-
type of oligonucleotide structure, see [22].

Helvetica Chimica Acta ± Vol. 86 (2003)
4275
8
conformation ). Based on this criterion, the b-ribopyranosyl-(4' ! 2')-oligonucleotide
system was expected to be capable of base pairing since, in the homo-DNA series, a
similar expectation that was reached by the same type of conformational reasoning had
been vindicated by experiment [27]. Such experimental support for an expectation
notwithstanding, the remarkable base-pairing properties of p-RNA came eventually
nevertheless as a surprise: p-RNAwas found to be a much stronger and, concerning the
pairing mode, a more selective base-pairing system than the natural (furanosyl) RNA.
We have reported on these properties and some of its consequences in our earlier
communications [1 ± 8] in sufficient detail so that we can concentrate in the present
paper on describing mainly those aspects of p-RNA chemistry that refer to the
synthesis, purification, and characterization of b-d-ribopyranosyl-mononucleotides and
b-d-ribopyranosyl-(4' ! 2')-oligonucleotides and deal with the properties of p-RNA
oligonucleotides in the form of an overview.
2
. Synthesis. ± 2.1. Preparation of the b-Ribopyranosyl-Nucleoside Derivatives
Containing (Protected) Adenine (A), Guanine (G), 2,6-Diaminopurine (D), Isogua-
nine (isoG), Cytosine (C), and (Unprotected) Uracil (U), and Thymine (T) as
Nucleobases. Derivatives of b-d-ribopyranosyl nucleosides containing all five canonical
nucleobases had already been described in the literature by various authors [49],
however, not in the context of a synthesis of corresponding oligonucleotides. Our own
efforts in this field include, in addition to b-d-ribopyranosyl nucleosides containing the
five canonical nucleobases, the corresponding nucleosides of D and I. Scheme 1
summarizes the synthesis of the seven b-d-ribopyranosyl nucleosides 3a ± 3g, all of them
±
with the exception of the U and T derivatives ± in a protected form suited for the
construction of oligonucleotide sequences. In all cases, the nucleosidation methodology
of Vorbr¸ggenÀHilbertÀJohnson [50], previously successfully applied in the homo-
DNA series [28], could be adopted without any difficulties by using an a/b-mixture of
9
d-tetrabenzoyl-ribopyranose [51] as starting material ). As usual (see, e.g., [28]),
nucleosidation in the G series presented more problems and demanded extended
exploratory experimentation. While benzoyl as NH -protection group in the A- and
2
8)
See Scheme 1 in [1]. It seems worth pointing out that an analogously built hexopyranosyl-(4' ! 2')-
oligonucleotide system, e.g., the b-allopyranosyl member depicted below, by the same reasoning, would
not be expected to be a base-pairing system. The (repetitive) pairing conformation would suffer serious
steric hindrance (see arrow) and, therefore, not become populated in competition with less strained (non-
repetitive) conformations.
9)
This starting material, prepared according to a procedure described in [51], contained according NMR
spectroscopy ca. 70% of the b-pyranose and 20% of the a-pyranose form, and ca. 10% of furanose
derivatives. In some of the nucleosidation experiments, crystalline and diastereoisomerically pure (but
capriciously crystallizing) a-(2,3,4-tribenzoyl)-a-ribopyranosyl bromide [52][51] was used as starting
material.

4
276
Helvetica Chimica Acta ± Vol. 86 (2003)
Scheme 1. Preparation of the U, T, C, A, G, isoG, and D Nucleosides in the b-d-Ribopyranosyl Series

Helvetica Chimica Acta ± Vol. 86 (2003)
4277
C-nucleosidation partner did not present any problems, exploratory experiments
towards the synthesis of the guanosine member led to the choice of O-allyl/NH-
isobutyryl protection in the nucleoside and of the 6-chloro-2-(isobutyrylamino)purine
2
[
53][54] (as against the N -isobutyryl-6-oxo derivative) as the nucleosidation partner.
2
Originally, nucleosidation was explored by using N -isobutyryl guanine, but this
procedure suffered from low yields, low N(9)/N(7)-selectivity, and difficult separations
2
exacerbated by poor solubilities. Nucleosidation with 6-chloro-N -isobutyrylpurine
proceeded with high N(9)-regioselectivity with only negligible amounts of (not
isolated) N(7)-isomer formed. The replacement of the Cl-atom at C(6) by the allyloxy
group (2e ! 3e) was most satisfactorily achieved by treatment with a concentrated
(
20% v/v) soln. of anhydrous Et N in dry allyl alcohol containing 1.5 mol-equiv. of
3
DBU between 08 and room temperature for ca. 5 h. These conditions [55][56]
concomitantly remove the three O-Bz groups of 2e by alcoholysis, but leave the N(2)-
isobutyryl protection untouched. The procedure was eventually preferred to that
initially attempted employing anhydrous Cs (CO ) as a base in dry allyl alcohol. The
2
3
isoguanine- and 2,6-diaminopurine-b-ribopyranosyl-nucleosides were derived, as in the
homo-DNA series [25], from the 2,6-dichloropurine nucleoside, in both cases with
benzoyl as the NH -protecting and allyl [55] as the carbonyl-protecting group. The
2
somewhat tricky transformation of the 2,6-dichloropurine nucleoside 2f to the 2,6-
diaminopurine nucleoside 2g via the (non-purified) 2,6-diazido derivative eventually
proceeded satisfactorily under the conditions developed in the homo-DNA series [25].
All nucleosidations (1 ! 2a ± 2f) gave the equatorial b-isomers as the exclusively
isolated reaction products in (mostly) excellent yields. Assignment of their config-
1
uration at the anomeric center is based on the characteristic value of the H-NMR
coupling constants between the anomeric proton HÀC(1') (d ca. 6 to 6.5 ppm) and the
H-atom at C(2'), both occupying axial positions at the pyranose chairs; the observed
range of these values in the nucleoside series 3a ± 3g are consistently found in the range
of 9.3 to 9.7 Hz (in various deuterated solvents).
Nucleosides of the l-ribopyranosyl series had been prepared in connection with our
pairing studies with homochiral p-RNA strands of opposite sense of chirality [3]; they
were made, for the nucleobases C, G, isoG, D, and T, by the same nucleosidation
IV
procedures as described for the d-series. l-Ribose was prepared by Mo -catalyzed a-
epimerization of l-arabinose according to the method described by van Boom and co-
workers [57].
Scheme 2 summarizes the reaction sequence through which the nucleosides 3a ± 3g
containing their three OH groups in unprotected form were converted into their 3'-
OBz-4'-ODMT-protected derivatives 8a ± 8g used as monomer units in the automated
10
synthesis of p-RNA oligonucleotides ). The reaction sequence was originally worked
out in the U and A series [1], later adapted to the nucleosides containing the other
nucleobases [2], and, in the course of our work on the chemistry of p-RNA-
oligonucleotides, optimized and simplified according to the demands of an efficient
production of p-RNA monomers [4].
10)
One of the early (unsuccessful) plans was to achieve a (4-methoxybenzyloxy)-protection of the 3'-OH
function through partial reductive cleavage of a 3',4'-(4-methoxybenzaldehyde) acetal derivative.

4
278
Helvetica Chimica Acta ± Vol. 86 (2003)
Scheme 2. Preparation of the 3'-O-Benzoyl-4'-O-trityl Derivatives of U, T, C, A, G, isoG, and D Nucleosides in
the b-d-Ribopyranosyl Series
The efforts toward a synthesis of b-ribopyranosyl-(4' ! 2')-oligonucleotides had to
focus from the beginning on the problem of finding an adequate protection of the
(axial) 3'-OH function in b-ribopyranosyl-nucleosides. Since this OH group is expected
to be the sterically most hindered among the three OH functions, and since −obvious×
attempts in exploratory experiments to achieve regiodifferentiation through regiose-
lectively removing hydrolytically the Bz groups from the (equatorial) positions 2' and 4'

Helvetica Chimica Acta ± Vol. 86 (2003)
4279
in a 2',3',4'-tribenzoyl-nucleoside derivative were unsuccessful, the protection problem
presented itself for quite a while as the major hurdle to the development of a p-RNA
10
11
synthesis ). Eventually, the hurdle was overcome by exploiting the observation ) that
unexpectedly) 2'-OBz-3'-OH-b-ribopyranosyl-adenine derivatives in basic media can
(
undergo Bz migration from the equatorial 2'-position to the axial 3'-position. In
exploratory experiments aiming at deprotonation of the free 3'-OH group of the easily
6
6
accessible b-N ,N -dibenzoyl-4'-ODMT-3'-OH-2'-OBz-ribopyranosyl-adenine deriva-
tive 7d (see below and Scheme 2) to be followed by O-alkylation (e.g., by 4-
methoxybenzoyl chloride), it was found that deprotonation with NaH in THF, followed
by addition of aq. NH Cl soln., affords the 3'-OBz-2'-OH derivative besides the 3'-OH-
4
2
'-OBz derivative in a ratio of ca. 5 :1 and in moderate yield. This solved our protection
problem in as far we then decided to aim, from now on, at using Bz as the protecting
group of the axial 3'-OH function for the p-RNA-oligonucleotide synthesis. This choice
was close at hand because of our earlier experience in the hexopyranosyl series where
we had found that Bz could be successfully used as protecting group of axial sugar OH
group in the synthesis of b-altropyranosyl-(6' ! 4')-oligonucleotides [41][42].
The preparative conversion of the nucleosides 3a ± 3g to their protected 4'-ODMT-
3
'-OBz-2'-OH derivatives is proceeding via a (2' ! 3')-Bz migration 7 ! 8 as the central
and critical step. Acid-catalyzed acetalization of nucleosides 3a ± 3g with 4-methoxy-
benzaldehyde dimethyl acetal affords ± as mixtures of endo/exo-diastereoisomers in the
ratio of ca. 3 :1 ± the 3',4'-acetals 4a ± 4g as the major products besides small amounts
(
ratio to major product ca. 1:5 ± 1:6) of their 2',3'-regioisomers (slightly more polar on
TLC), which can be separated from the desired product by chromatography on silica
12
gel ). In the exploratory phase of the investigation (working with the U, T, and A
nucleosides), exo- and endo-diastereoisomers of 4 were separated by chromatography
on silica gel for spectroscopic characterization and configurational assignments (e.g.,
d(CH) for exo-acetal of 4 6.38 ppm, for endo-acetal 5.87 ppm, in CDCl [58]). The
3
yield of the acetalization step is highest when the reaction is terminated as soon as the
starting trial has been consumed (TLC). Longer reaction times lead to acetal migration
and mixed diacetal formations [58] (see Exper. Part). In the preparative routine, the
mixtures of diastereoisomeric 4',3'-acetals 4a ± 4g were treated with 2.5 mol-equiv. BzCl
in pyridine with 4-(dimethylamino)pyridine (DMAP) as catalyst at room temperature
in order to benzoylate the 2'-OH group, the acetal groups were cleared by mild
treatment with CF CO H in MeOH to give the 2'-OBz-3',4'-dihydroxy nucleosides 6a ±
3
2
6g and, finally, their (less hindered) 4'-OH groups were regioselectively protected as
DMT derivatives 7a ± 7g. Under the employed benzoylation conditions, the nucleo-
bases of the U, T, A, and isoG nucleosides also undergo N-benzoylation of their imide
NH and amide NH groups, respectively. Since such imide N-Bz and N,N-dibenzoyl
groups are quite labile towards hydrolysis (and prone to Bz transfers), the benzoylation
products in the U, T, and isoG series were routinely not isolated but subjected directly
to the conditions of deacetalization and subsequent tritylation of the equatorial 4'-OH
function. The latter reaction 6 ! 7 proceeds with high regioselectivity in all seven
nucleotide series.
11)
12)
Observation made by S. W.
2',3'-Regioisomers of 4a ± 4g were recycled to triols 3a ± 3g after acid-catalyzed deacetalization.

4
280
Helvetica Chimica Acta ± Vol. 86 (2003)
Extended optimization studies on the (2' ! 3')-Bz migration 7 ! 8 were carried out
in the adenine (B ˆ N,N-dibenzoyladenine) and uracil (B ˆ N-Bz-uracil) series. They
resulted in the development of a rather complex cocktail of agents (1 mol-equiv.
13
DMAP, 3 equiv. Et N, 2 equiv. 4-nitrophenol, and 20 equiv. i-PrOH ) in 82 equiv.
3
pyridine) that, at 608 during 24 ± 28 h, produced with remarkable reliability (2' ! 3')-
Bz-migration products 8d (B ˆ N-monobenzoyladenine) and 8a (B ˆ uracil) in (e.g.)
6
6
7% yield (isolated). The equilibrium ratio between the components 7d (as N -
monobenzoyl derivative) and 8d under these reaction conditions was estimated to lie
on the side of the 3'-OBz derivatives by a factor of (at least) 1:10. Beyond the NMR-
spectroscopic constitutional assignments given below, it was chemically shown that 7d
(
B ˆ N,N-dibenzoyl-adenine) and the N,N-dibenzoyl analogue of 8d gave the same
tetrabenzoyl derivative when O-benzoylated with BzCl in pyridine in the presence of
DMAP, benzoylation of the free OH group of 7d proceeding thereby much more
slowly than that of the free OH group of the 8d analogue.
6
6
The (7 ! 8)-Bz-migration procedure developed in the N ,N -dibenzoylated adenine
3
and N -benzoylated-uracil series turned out to be comparably efficient in the G, isoG,
and D series. Mixtures of 2'- and 3'-benzoates, when recovered as such in the course of
the chromatographic isolation of pure 3'-benzoates, were recycled in the d-ribopyr-
anosyl-guanine, the l-ribopyranosyl-thymine, and d-ribopyranosyl-cytosine series
(
yields up to 80%). In the C series, routine chromatographic separation of the
migration product from the 2'-OBz-3'-OH derivative on silica gel met first with
difficulties. This gave rise to a separation procedure that consisted in trimethylsilylation
of the mixture×s free OH groups (ratio C(2')ÀOH to C(3')ÀOH ca. 6 :1) and
subsequent regioselective desilylation of the 2'-O-silyloxy component by mild acid
treatment, chromatographic separation of the 3'-OBz-2'-OH derivative 8c from
remaining silylated components (yield of 8c 50%, after recycling up to 60%).
Eventually, it was found that 8c can be crystallized from the solution of a chromatog-
raphically purified mixture of 8c/7c in Et O containing ca. 4% pentane and obtained
2
thereby in pure form in 48% yield (see Exper. Part).
The constitutional assignments for the intermediates 6a ± 6g, 7a ± 7g, and 8a ± 8g rest
1
on the H-NMR data of these intermediates; excerpts from such data are listed in
Table 2. C-Atom C(1') (anomeric center) in all derivatives is recognized by its chemical
shift around 6 ppm and the (large) coupling constant (ca. 9.5 Hz) between its (axial) H-
atom and H ÀC(2'). The spectral identification of the HÀC(3'), HÀC(4'), and
ax
H ÀC(5') and H ÀC(5') centers in 3a ± 3g is based on spin-coupling and NOE-
a
b
correlation data. The position of the O-Bz group in 6a ± 6g, 7a ± 7g, and 8a ± 8g is
corroborated by the chemical-shift pattern of HÀC(3'), HÀC(4'), and HÀC(2') in
these derivatives.
A short comment on the Bz-migration step 7a ± 7g ! 8a ± 8g seems appropriate, not
necessarily with regard to the mechanistic and kinetic aspects of the reaction, but
13)
i-PrOH serves the purpose of accepting the (activated) Bz group of the N-Bz-uracil nucleus and one of the
Bz groups of the N,N-dibenzoyladenine nucleus; this precludes the formation of 2',3'-dibenzoyl derivatives
as by-products (only relevant for these two nucleosides). The presence of 4-nitrophenol has been found to
be beneficial for both rate and product yield; its role is presumably that of a proton donator. In the cases of
B ˆ uracil and thymine, 3 equiv. of EtN(i-Pr)
2 3
was used instead of Et N.

Helvetica Chimica Acta ± Vol. 86 (2003)
4281
Table 2. Chemical Shifts (d/[ppm]) of Non-Exchangeable Sugar H-Atoms of Synthetic Intermediates
Entry
HÀC(1') HÀC(2') HÀC(3') HÀC(4')
H
axÀC(5')
H
eqÀC(5') J(1',2') [Hz]
1
2
3
4
5
6
7
8
9
2a (CDCl
2b (CDCl
3
)
)
6.48
6.48
6.76
6.50
6.44
6.43
6.21
5.69
5.75
)) 5.83
5.92
)) 5.67
5.80
5.99
5.48
6.17
6.18
6.43
6.28
6.75
6.19
6.48
6.20
6.20
6.46
6.26
6.13
6.19
5.89
5.76
5.90
6.03
5.81
5.68
5.71
5.89
5.52
5.50
5.56
6.01
5.93
5.88
6.04
3.95
6.31
6.31
6.34
6.41
6.38
6.39
6.42
4.27
5.54
5.54
5.60
5.70
5.68
5.69
5.67
3.95
4.23
4.25
4.29
4.34
4.33
4.37
9.6
9.7
9.6
3
2c (CDCl
3
)
2d (CDCl
3
)
4.33 ± 4.46 4.33 ± 4.46 9.5
2e (CDCl
3
)
)
)
4.34
4.36
4.35
3.75
4.41
4.45
4.38
3.83
9.3
9.5
9.4
9.7
2f (CDCl
2g (CDCl
3
3
3a (D
3b (CD
3c (DMSO(D
3d (CD OD)
3e (DMSO(D
3f (CD OD)
3g (CD OD)
3h (D O)
6a (CD OD)
6b (CD
6c (CD
2
O)
3
OD)
3.71 ± 3.80 4.17
3.60 ± 3.78 4.02
3.71 ± 3.80 3.71 ± 3.80 3.71 ± 3.80 9.3
3.60 ± 3.78 3.60 ± 3.78 3.60 ± 3.78 9.5
3.90
3.73
3.70 ± 3.90 3.70 ± 3.90 3.70 ± 3.90 9.8
3.75 ± 4.00 3.75 ± 4.00 3.72 ± 4.00 8.8
3.95
4.00
4.03
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
6
3
4.37
4.24
4.35
4.31
4.24
5.26
5.30
5.25
5.52
5.60
5.60
5.57
4.88
4.88
4.84
5.35
5.38
5.25
5.73
3.52
3.52
4.26
4.05
4.24
4.30
4.22
4.41
4.42
4.45
4.54
4.87
4.47
4.50
3.91
3.72
3.80
3.60
3.90
3.67
9.4
9.4
6
3
3
2
3.74 ± 3.76 3.74 ± 3.76 9.4
3.90
3.90
3
3.90
3.90
9.6
9.6
3
OD)
OD)
3
3.87 ± 4.07 3.87 ± 4.07 3.87 ± 4.07 9.5
6d (CDCl
6e (CDCl
6f (CD
6g (CD
7a (CDCl
7b (CDCl
3
)
)
4.08
4.14
4.08
3.96
3.97
3.94
3.85 ± 3.97 3.85 ± 3.97 9.5
3
4.03
3.91
4.09
3.27
3.28
4.24
4.00
4.12
3.91
3.94
9.7
9.4
9.4
9.4
9.5
3
OD)
OD)
3
3
)
)
3
7c (CDCl
7d (CDCl
3
)
3.76 ± 3.79 3.84 ± 3.88 3.25 ± 3.28 3.84 ± 3.88 9.0
3
)
3.74
3.86
3.75
4.44
5.88
5.96
4.05
4.02
4.01
4.00
3.85
3.88
3.38
3.28
3.29
2.49
2.60
2.59
4.02
3.98
3.97
3.61
3.60
3.61
9.6
9.6
9.6
9.2
9.5
9.5
7e (CDCl
3
)
)
7f (CDCl
7g (CDCl
8a (CDCl
3
3
)
)
)
3
8b (CDCl
3
8c (CDCl
8d (CDCl
3
)
3.66 ± 3.76 5.88
3.91 ± 3.95 2.71
3.66 ± 3.76 9.3
3
)
4.39
4.44
4.29
4.45
5.92
6.10
5.92
6.27
4.04
3.95
4.00
4.00
2.79
2.58
2.74
2.47
3.79
3.68
3.72
3.61
9.2
9.2
9.3
9.2
8e (CDCl
3
)
)
)
8f (CDCl
8g (CDCl
3
3
concerning the remarkably one-sided position of the thermodynamic equilibrium.
Base-catalyzed migrations of acyl groups between cis-vicinal OH functionalities in six-
membered ring systems are commonplace, not necessarily so, however, in cases of an
acyl-group migrating from an equatorial to an axial position. Our interpretation, which
involves the qualitative assessment of the contribution of both steric and stereo-
electronic factors, leans toward considering the case as a nice example of the gauche
effect [59] contributing to, if not dominating, the position of a reaction equilibrium:
The violation of the gauche effect by the equatorial 2'-OH function (2'-O and 5'-O
centers antiperiplanar to each other; Fig. 4) is expected to impose a deficit in
conjugative n ! p* stabilization in the 2'-OBz group as compared to a BzO group
whose OH function would not be antiperiplanar to another O center. Ester functions
that are part of a gauche-effect-violating arrangement are expected to be more

4
282
Helvetica Chimica Acta ± Vol. 86 (2003)
Fig. 4. The (2' ! 3')-Bz migration in b-d-ribopyranosyl-nucleosides. Note the difference between the two
isomers with reference to the −gauche effect×.
electrophilic (more prone to transesterification) than ester groups, which ± ceteris
paribus ± obey the gauche effect in such a molecular environment. Thus, an ester
function in a non-violating arrangement is the BzO group in the 3'-position (axial 3'-
OBz antiperiplanar to H ÀC(4)). Therefore, the gauche effect per se could drive the
ax
Bz migration from the 2'- to the 3'-position. In addition, the conjugative s(CÀH) !
d*(CÀO) stabilization expected to be operative in 3'-C(OR)/4'-C(H) part of the
molecule should be more pronounced when the 3'-OH is benzoylated than when it is a
14
free OH group ). The preference of the 3'-OBz over the 2'-OBz position by such
stereoelectronic factor seems clearcut.
Inspecting the (presumably) most stable conformations of the two BzO groups in
7a ± 7g and 8a ± 8g, for their steric constraints, the one imposed on the equatorial 2'-OBz
group by the equatorial nucleobase may be more severe than that on the axial 3'-OBz
group by the equatorial 4'-ODMT-group (with the OÀtrityl bond antiplanar to the
C(3')ÀC(4') bond. Steric factors may, therefore, also contribute to the driving force of
the (2' ! 3')-Bz migration in these molecules.
Scheme 3 depicts the modifications of the nucleosides 8a ± 8g into derivatives
required for the automated solid-support synthesis. These modifications were
performed according to procedures used in the homo-DNA series [28], but with the
following change. The choice of the then used commercially available chloro(cyano-
ethoxy)(diisopropylamino)phosphine as phosphitylating agent had to be re-assessed
because the 3'-OBz protecting group in p-RNA oligonucleotides would have to be
deprotected after the conversion of the phosphotriester linkages into the more stable
phosphodiester linkages so as to avoid strand scission resulting from the cis-
15
configuration of the (4' ! 2')-phosphodiester linkages and free 3'-OH groups ). The
14)
15)
The argument echoes the qualitative rationalization of the gauche effect based on the s ! s* conjugation
concept.
Deprotection under basic conditions would result in the concomitant deprotection of the 3'-OBz group
leading to the free 3'-OH group, which, under the basic conditions, can attack the phosphotriester linkage
leading to stand scission. Such a conclusion was also based on an observation in the case of the hexose-
sugar-derived d-b-allopyranosyl-oligonucleotides, in which there is an all-cis-disposition of 2'-O, 3'-O, and
4
>
'-O groups similar to the p-RNA case: when b-d-allopyranosyl-(AC)
20% decomposition was found to occur even at room temperature; at 508, complete decomposition was
observed within 6 h [41].
6 3
was exposed to ethanolic NH ,

Helvetica Chimica Acta ± Vol. 86 (2003)
4283
16
reagent allyloxy(chloro)(diisopropylamino)phosphine ) offered itself as an ideal
replacement. The allyl group of O-allyl-phosphotriester-linked oligomers can be
selectively removed under mild neutral conditions to afford the phosphodiester-linked
oligomers [54][55]. Thus, phosphitylations of 8a ± 8g were conducted with allyloxy-
(
chloro)(diisopropylamino)phosphine in CH Cl with EtN(i-Pr) as base to afford,
2 2 2
without any complications, diastereoisomeric mixtures of phosphoramidites 9a ± 9g in
yields greater than 90%. Compounds 8a ± 8g were also the starting materials for putting
the starter units on the CPG solid support to be used in the automated oligomer
synthesis. Based on our previous experience from the homo-DNA series [25][28],
nucleosides 8a ± 8g were coupled with bis(4-nitrophenyl) heptane-1,7-dioate [60] in the
presence of DMAP in pyridine to afford the activated 2'O-(4-nitrophenyl)heptane-1,7-
dioates 10a ± 10g in yields ranging from average (66% for 10a) to high (90% for 10d).
The activated 2'-O-ester derivatives 10a ± 10g were immobilized on the CPG solid
17
support ) by treatment with long-chain alkylamine ± controlled pore glass (LCAA-
CPG) in the presence of EtN(i-Pr) in DMF/pyridine, and the derivatized solid
2
supports were capped with excess Ac O in pyridine to afford 11a ± 11g. The loadings of
2
the 11a ± 11g CPG solid supports were estimated by trityl-cation assay (treating ca. 2 ±
3
4
mg of 11a ± 11g with 70% aq. HClO /MeOH 1:1) and measuring the absorption at
4
À1 À1
À1
98 nm (e of the DMT cation 70000 cm
l
mol ).
2
.2. Synthesis of b-d-Ribopyranosyl-(4' ! 2')-oligonucleotides (p-RNAs). Scheme 4
delineates the overall pathway followed in the automated solid-support synthesis of
Scheme 3. Preparation of the 2'-O-Phosphoramidate and 2'-O-CPG Derivatives of U, T, C, A, G, isoG, and D
Nucleosides in the b-d-Ribopyranosyl Series
16)
The concept of using the allyl group as a phosphodiester protecting group in oligonucleotide synthesis was
first demonstrated by Noyori and co-workers [55]. They used the (allyloxy)[bis(diisopropylamino)]pho-
sphine as the phosphitylating reagent.
17)
Via the formation of an amide bond resulting from the attack of the NH
activated esters 10a ± 10g.
ÀCPG onto the 4-nitrophenyl-
2

4
284
Helvetica Chimica Acta ± Vol. 86 (2003)
Scheme 4. Overview of the Synthesis, Deprotection, Purification, and Isolation of b-d-Ribopyranosyl
Oligonucleotides
p-RNA oligonucleotides, and Table 3 lists all the sequences synthesized over the course
of this project. The syntheses were carried out on 1 ± 10-mmol scale, initially on an
automated DNA synthesizer Gene Synthesizer Plus (Pharmacia), with the phosphor-

Helvetica Chimica Acta ± Vol. 86 (2003)
4285
Table 3. HPLC and MS Data of p-RNA Oligonucleotides
Entry p-RNA
b-d-pr-Oligonucleotide
Deprotectionb)
Analytical HPLC
Reverse phasec) Ion Exchanged)
gradient, t [min] gradient, t [min]
MALDI-TOF-MSe)
À
À
Methods
[M À H] [M À H]
a
system )
R
R
(obs.)
(calc.)
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
A
A
A
A
A
A
A
A
A
A
A
A
A
A
2
3
4
4
6
7
8
A, F
A, F
B, D, F
0 ! 40% B in 30, 15.0
0 ! 40% B in 30, 17.3
±
923.3
1335.3
1317.8
1913.8
±
924.7
-2'-phosphate
-2',3'-cyclophosphate
0 ! 100% B in 30, 11.0
0 ! 100% B in 30, 11.7
0 ! 100% B in 30, 11.1
1342.9
1324.9
1913.3
2241.5
2570.7
3888.5
5133.2
5132.3
2535.7
2478.6
2632.7
±
2185.4
1341.8
1323.8
1967.3
3499.2
1968.3
3933.5
3915.4
B, F
A, F
A, F
C, G
B, F
B, D, F
A, F
A,E
B, D, F
A, E
B, F
f
0 ! 40% B in 30, 20.6 10 ! 65% B in 30, 25.4 ) 2240.1
0 ! 40% B in 30, 21.0
2574.6
3890.0
5127.9
5101.5
2532.9
2480.4
2632.5
±
2187.3
1343.6
1326.3
1967.2
3502.8
1967.0
3934.8
3917.4
12
0 ! 100% B in 30, 19.3
0 ! 100% B in 30, 24.0
0 ! 100% B in 30, 25.1
6
4
4
4
3
2
T
T
T
2
3
4
A
ATAT
2
T
6
1
1
1
1
1
1
1
1
1
1
2
2
2
2
AT
2
A
0 ! 40% B in 30, 20.3 0 ! 100% B in 30, 17.0
0 ! 40% B in 30, 18.2 0 ! 100% B in 30, 15.4
0 ! 100% B in 30, 18.0
U
4
TA
3
T-2'-phosphate
UAU
GA
3
A
0 ! 40% B in 30, 18.5 0 ! 100% B in 30, 16.4
0 ! 100% B in 30, 16.3
AC
ACG
ACG
AG
AGCGT
AGTG
ATA
ATA
phosphate
(AT)
ATATG
ATCG-2'-phosphate
ATCG-2',3'-cyclophosphate
AT-(l)C-G-2'-phosphate
AT-(l)C-G-2',3'-cyclo-
phosphate
ATC-(l)G-2'-phosphate
ATC-(l)G-2',3'-cyclo-
phosphate
AT-(l)C-(l)G-2'-phosphate
AT-(l)C-(l)G-2',3'-cyclo-
phosphate
A-(l)T-CG-2'-phosphate
A-(l)T-CG-2',3'-cyclo-
phosphate
A-(l)T-C-(l)G-2'-phosphate
A-(l)T-C-(l)G-2',3'-cyclo-
phosphate
A-(l)T-(l)C-G-2'-phosphate
A-(l)T-(l)C-G-2',3'-cyclo-
phosphate
3
2
2
-2'-phosphate
B, D, F
0 ! 100% B in 30, 17.7
2
-2',3'-cyclophosphate
0 ! 100% B in 30, 21.0
4
T
B, F
B, F
C, G
B, D, F
0 ! 100% B in 30, 23.4
4
CGC
0 ! 100% B in 30, 24.9
3
0 ! 100% B in 30, 25.0
5
TATCG-2'-phosphate
TATCG-2',3'-cyclo-
0 ! 100% B in 30, 21.3
5
0 ! 100% B in 30, 23.0
23
24
25
26
27
28
4
A, F
B, F
B, D, F
0 ! 40% B in 30, 19.9 0 ! 100% B in 30, 17.5
0 ! 100% B in 30, 23.0
2534.1
3220.1
1315.9
1298.3
1315.6
1297.4
2535.7
3223.1
1316.8
1298.8
1316.8
1298.8
2
T
2
AT
0 ! 100% B in 30, 16.7
0 ! 100% B in 30, 19.2
B, D, F
B, D, F
B, D, F
B, D, F
B, D, F
B, D, F
0 ! 100% B in 30, 16.6
0 ! 100% B in 30, 19.1
2
3
9
0
0 ! 100% B in 30, 16.5
0 ! 100% B in 30, 19.1
1318.3
1297.4
1316.8
1298.8
3
3
1
2
0 ! 100% B in 30, 16.4
0 ! 100% B in 30, 19.0
1318.3
1297.4
1316.8
1298.8
3
3
3
4
0 ! 100% B in 30, 16.7
0 ! 100% B in 30, 19.1
1320.9
1300.2
1316.8
1298.8
3
3
5
6
0 ! 100% B in 30, 16.3
0 ! 100% B in 30, 19.2
1317.4
1297.4
1316.8
1298.8
3
3
7
8
0 ! 100% B in 30, 16.3
0 ! 100% B in 30, 19.2
1315.6
1297.4
1316.8
1298.8
3
4
9
0
A-(l)T-(l)C-(l)G-2'-phosphate B, D, F
A-(l)T-(l)C-(l)G-2',3'-cyclo-
phosphate
0 ! 100% B in 30, 16.6
0 ! 100% B in 30, 19.2
1319.1
1300.2
1316.8
1298.8
4
4
1
2
ATCGATCG-2'-phosphate
ATCGATCG-2',3'-cyclo-
phosphate
B, D, F
0 ! 100% B in 30, 21.2
0 ! 100% B in 30, 24.3
2613.6
2596.6
2616.6
2598.6
4
4
3
4
ATCGATCGATCG-2'-
phosphate
ATCGATCGATCG-2',3'-
cyclophosphate
B, D, F
0 ! 100% B in 30, 23.7
0 ! 100% B in 30, 26.5
3919.6
3898.0
3916.4
3898.4
4
4
5
6
ATGC-2'-phosphate
ATGC-2',3'-cyclophosphate
B, D, F
0 ! 100% B in 30, 16.0
0 ! 100% B in 30, 17.9
1318.2
1300.2
1316.8
1298.9

4
286
Helvetica Chimica Acta ± Vol. 86 (2003)
Table 3 (cont.)
Entry p-RNA
Deprotec-
tion )
Methods
Analytical HPLC
MALDI-TOF-MSe)
b
Reverse phasec)
gradient, t [min]
Ion Exchanged)
gradient, t [min]
[M À H]
À
[M À H]
À
b-d-pr-Oligonucleotide
a
system )
R
R
(obs.)
(calc.)
4
4
7
8
A-(l)T-(l)G-(l)C-2'-phosphate B, D, F
A-(l)T-(l)G-(l)C-2',3'-cyclo-
phosphate
0 ! 100% B in 30, 16.2
0 ! 100% B in 30, 18.1
1316.3
1300.2
1316.8
1298.9
49
50
51
52
53
54
55
ATGCAT
ATGCATG
B, F
B, F
A, F
B, D, F
0 ! 100% B in 30, 17.6
0 ! 100% B in 30, 21.8
1887.5
2234.5
5147.6
1316.0
1298.0
2605.8
2593.2
1887.2
2232.5
5133.3
1315.8
1297.8
2606.6
2588.6
AT
AT
AT
AT
AT
2
2
2
2
2
AT
2
AT
3
A
4
0 ! 40% B in 30, 17.8
A-2'-phosphate
A-2',3'-cyclophosphate
0 ! 100% B in 30, 15.8
0 ! 100% B in 30, 17.8
0 ! 100% B in 30, 20.4
0 ! 100% B in 30, 22.4
AT
AT
2
AT-2'-phosphate
AT-2',3'-cyclo-
B, D, F
2
phosphate
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
AT
AT
AT
AT
(AU)
AU AUA
CACAC
2
3
3
3
CAGCG
-2'-phosphate
-2'-phosphate
C, G
B, F
B, D, F
0 ! 100% B in 30, 22.7
0 ! 100% B in 30, 18.9
0 ! 100% B in 30, 16.8
0 ! 100% B in 30, 19.6
2537.0
2623.3
1305.5
1288.5
2480.4
±
4049.0
1871.4
628.1
611.0
918.1
4037.0
±
1239.7
1220.1
1238.2
1770.6
2378.5
2456.3
2441.5
±
2537.6
2624.7
1306.8
1288.8
2478.6
±
4048.5
1870.2
627.4
609.4
915.6
4039.5
±
1237.2
1219.2
1237.2
1768.14
2379.5
2457.3
2439.3
±
1277.8
1259.8
2458.6
1888.2
4026.5
1847.2
667.4
2538.0
1913.2
3205.0
1317.8
1299.8
1888.2
3474.2
2537.6
1614.0
1596.0
1317.8
1299.8
1317.8
1299.8
1317.8
1299.8
A
4
-2',3'-cyclophosphate
4
A, E
A, E
C, G
B, F
0 ! 40% B in 30, 18.2 0 ! 100% B in 30, 15.4
0 ! 40% B in 30, 18.6 0 ! 100% B in 30, 16.5
0 ! 100% B in 30, 23.5
3
2
3
G
2
CTC
2
CATATA
0 ! 100% B in 30, 15.7
C
C
2
-2'-phosphate
-2',3'-cyclophosphate
B, D, F
0 ! 100% B in 30, 6.5
2
0 ! 100% B in 30, 6.5
3',4'-cyclophosphate-C
3
B, F
C, G
A, F
B, F
0 ! 40% B in 30, 10.4 0 ! 100% B in 30, 9.5
0 ! 100% B in 30, 23.4
C
3
C
4
C
4
C
4
ACTCG
2
CTC
2
f
0 ! 40% B in 30, 11.5 0 ! 100% B in 30, 8.8 )
-2'-phosphate
-2',3'-cyclophosphate
0 ! 40% B in 30, 12.4 0 ! 100% B in 30, 11.7
0 ! 40% B in 30, 12.1 0 ! 100% B in 30, 11.7
0 ! 40% B in 30, 7.1 0 ! 100% B in 30, 10.9
4'-phosphate-C
4
B, F
B, F
B, F
B, F
f
C
6
C
8
C
8
C
8
0 ! 20% B in 20, 13.4 5 ! 70% B in 30, 15.7 )
0 ! 30% B in 30, 14.4
-2'-phosphate
-2',3'-cyclophosphate
0 ! 100% B in 30, 15.5
0 ! 40% B in 30, 13.9 0 ! 100% B in 30, 16.6
0 ! 40% B in 30, 13.9 0 ! 100% B in 30, 15.3
0 ! 100% B in 30, 14.9
4'-phosphate-C
8
B, F
B, D, F
C
3
C
3
C
3
C
3
C
3
C
2
G-2'-phosphate
G-2',3'-cyclophosphate
GC
G
TC
1278.2
1261.4
2460.9
0 ! 100% B in 30, 17.0
3
G
B, F
B, F
B, F
B, F
B, D, F
B, F
C, G
B, F
0 ! 100% B in 30, 19.2
f
3
0 ! 40% B in 30, 20.6 10 ! 100% B in 30, 15.1 ) 1889.6
3
G
2
CTC
2
0 ! 100% B in 30, 22.3
0 ! 100% B in 30, 16.2
0 ! 100% B in 30, 11.8
4027
1847.5
667.5
2538.0
1915.0
3204.0
1320.0
1301.9
TATA
CG-2'-phosphate
CGA CG
CGAGCG
CGCA GCG
CGCG-2'-phosphate
CGCG-2',3'-cyclophosphate
(CG)
CGCGT
CGCTGA
CGCT
CGCT
(l)C-G
(l)C-G
CG
CG
CG
CG
2
T
2
0 ! 30% B in 30, 18.5
0 ! 100% B in 30, 21.6
0 ! 100% B in 30, 21.8
0 ! 100% B in 30, 17.2
0 ! 100% B in 30, 20.1
4
B, D, F
f
3
B, F
B, F
C, G
B, D, F
0 ! 40% B in 30, 17.5 10 ! 100% B in 30, 14.8 ) 1889.6
4
CGC
0 ! 100% B in 30, 23.8
0 ! 100% B in 30, 22.2
0 ! 100% B in 30, 18.1
0 ! 100% B in 30, 20.6
0 ! 100% B in 30, 16.6
0 ! 100% B in 30, 18.6
0 ! 100% B in 30, 16.9
0 ! 100% B in 30, 18.8
0 ! 100% B in 30, 16.9
0 ! 100% B in 30, 18.7
3483.4
2537.0
1614.7
1597.5
1316.4
1299.3
1318.0
1299.2
1316.4
1298.4
2
T
2
-2'-phosphate
2
-2',3'-cyclophosphate
2
-2'-phosphate
B, D, F
B, D, F
B, D, F
2
-2',3'-cyclophosphate
2
2
2
2
C-2'-phosphate
C-2',3'-cyclophosphate
-(l)C-2'-phosphate
-(l)C-2',3'-cyclo-
phosphate

Helvetica Chimica Acta ± Vol. 86 (2003)
4287
Table 3 (cont.)
Entry p-RNA
Deprotectionb)
Analytical HPLC
Reverse phasec) Ion Exchanged)
gradient, t [min] gradient, t [min]
MALDI-TOF-MSe)
b-d-pr-Oligonucleotide systema) Methods
[M À H] [M À H]
À
À
R
R
(obs.)
(calc.)
1
1
00
01
CG-(l)G-C-2'-phosphate
CG-(l)G-C-2',3'-cyclo-
phosphate
C-(l)G-GC-2'-phosphate
C-(l)G-GC-2',3'-cyclo-
phosphate
B, D, F
B, D, F
B, D, F
0 ! 100% B in 30, 17.0 1316.4
0 ! 100% B in 30, 18.7 1298.4
1317.8
1299.8
1
1
02
03
0 ! 100% B in 30, 17.0 1315.5
0 ! 100% B in 30, 18.9 1297.5
1317.8
1299.8
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
CG
3
CG
3
CG
3
CG
6
-2'-phosphate
-2',3'-cyclophosphate
0 ! 100% B in 30, 19.2 1357.8
0 ! 100% B in 30, 23.5 1340.6
0 ! 100% B in 30, 25.9 2619.3
0 ! 100% B in 30, 25.4 2621.1
0 ! 100% B in 30, 20.3 2806
0 ! 100% B in 30, 17.4 1861.7
1357.8
1339.8
2618.6
2618.6
2505.8
1861.2
2004.0
2004.0
2691.0
2691.0
1331.8
1313.8
1331.8
1313.8
2646.6
2628.6
CG
C
3
B, F
B, F
B, F
B, F
B, F
B, F
B, F
B, F
B, D, F
3 2
CTC G CT
CT ATA
2
D
6
0 ! 40% B in 20, 14.8 0 ! 70% B in 30, 24.2 2003.6
(l)D
6
8
0 ! 40% B in 20, 15.1
0 ! 40% B in 20, 15.6
0 ! 30% B in 20, 17.3
2003.5
2690.0
2869.6
D
8
(l)D
DTCG-2'-phosphate
0 ! 100% B in 30, 17.1 1331.1
0 ! 100% B in 30, 20.0 1314.0
0 ! 100% B in 30, 17.0 1330.8
0 ! 100% B in 30, 19.7 1315.5
0 ! 100% B in 30, 22.1 2645.3
0 ! 100% B in 30, 26.1 2630.9
DTCG-2',3'-cyclophosphate
(l)D-TCG-2'-phosphate
(l)D-TCG-2',3'-cyclophosphate
DTCGDTCG-2'-phosphate
DTCGDTCG-2',3'-cyclo-
phosphate
B, D, F
B, D, F
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
DTGC-2'-phosphate
DTGC-2',3'-cyclophosphate
B, D, F
0 ! 100% B in 30, 16.7 1330.2
0 ! 100% B in 30, 18.6 1314.9
0 ! 100% B in 30, 17.8 1569.0
0 ! 100% B in 30, 18.5 1649.0
0 ! 100% B in 30, 20.8 1629.0
0 ! 100% B in 30, 20.9 1912.0
0 ! 100% B in 30, 21.6 1993.0
0 ! 100% B in 30, 24.7 1976.0
0 ! 100% B in 30, 29.4 3934.0
0 ! 100% B in 30, 28.6 3589.0
0 ! 100% B in 30, 21.2 1913.0
0 ! 100% B in 30, 21.3 2234.0
0 ! 100% B in 30, 11.5 668.7
0 ! 100% B in 30, 13.2 649.8
0 ! 100% B in 30, 13.1 973.3
0 ! 100% B in 30, 14.3 955.4
0 ! 100% B in 30, 13.0 1276.1
0 ! 100% B in 30, 15.3 1261.5
0 ! 100% B in 30, 14.4 1277.0
0 ! 100% B in 30, 15.4 1263.7
0 ! 100% B in 30, 14.2 1275.2
0 ! 100% B in 30, 15.2 1259.3
0 ! 100% B in 30, 14.1 1277.9
0 ! 100% B in 30, 15.8 1260.2
0 ! 100% B in 30, 14.2 1277.6
0 ! 100% B in 30, 15.9 1257.5
0 ! 100% B in 30, 15.0 1278.0
0 ! 100% B in 30, 16.0 1259.3
0 ! 100% B in 30, 19.6 2453.6
0 ! 100% B in 30, 19.3 2457.6
0 ! 100% B in 30, 20.7 2519.0
0 ! 100% B in 30, 19.9 2538.3
0 ! 100% B in 30, 20.3 2540.9
0 ! 100% B in 30, 21.8 3718.8
0 ! 100% B in 30, 18.1 1318.7
1331.8
1313.8
1568.0
1648.0
1629.9
1913.2
1993.2
1975.2
3934.5
3589.3
1913.2
2232.5
667.4
GAGC
GAGC
GAGC
GAGC
GAGC
GAGC
GAGC
GAGC
GAGC
2
2
2
2
2
2
2
2
2
C, G
C, G
-2'-phosphate
-2',3'-cyclophosphate
G
G-2'-phosphate
G-2',3'-cyclophosphate
C, G
C, G
G
G
3
TGTG
TGTG
C, G
C, G
C, G
B, F
2
-pl(G)
GATGCAT
GC-2'-phosphate
GC-2',3'-cyclophosphate
B, D, F
649.4
972.6
954.6
GC
2
GC
2
GC
3
GC
3
-2'-phosphate
-2',3'-cyclophosphate
-2'-phosphate
B, D, F
B, D, F
B, D, F
B, D, F
B, D, F
B, D, F
B, D, F
1277.8
1259.2
1277.8
1259.8
1277.8
1259.8
1277.8
1259.8
1277.8
1259.8
1277.8
1259.8
2458.6
2458.6
2520.5
2538.5
2538.5
3719.3
1317.8
-2',3'-cyclophosphate
(l)G-C
(l)G-C
3
-2'-phosphate
-2',3'-cyclophosphate
3
G-(l)C-C
G-(l)C-C
2
-2'-phosphate
2
-2',3'-cyclophosphate
GC-(l)C-C-2'-phosphate
GC-(l)C-C-2',3'-cyclophosphate
GC
GC
2
-(l)C-2'-phosphate
-(l)C-2',3'-cyclophosphate
2
(l)G-(l)C
(l)G-(l)C
3
-2'-phosphate
-2',3'-cyclophosphate
3
GC
6
GC
3
GC
3
GC
3
GC
3
GC
3
GC
2
G
B, F
B, F
GC
GC
GC
GC
GC
3
3
3
3
3
-2',3'-cyclophosphate
-2'-phosphate
-3'-phosphate
B, D, F
B, D, F
B, F
GC
3
G-2'-phosphate
B, D, F

4
288
Helvetica Chimica Acta ± Vol. 86 (2003)
Table 3 (cont.)
Entry p-RNA
Deprotectionb)
Analytical HPLC
Reverse phasec) Ion Exchanged)
gradient, t [min] gradient, t [min]
MALDI-TOF-MSe)
À
À
b-d-pr-Oligonucleotide
Methods
[M À H] [M À H]
a
system )
R
R
(obs.)
(calc.)
1
1
1
55
56
57
GC G-2',3'-cyclophosphate
2
0 ! 100% B in 30, 19.9
0 ! 100% B in 30, 17.3
0 ! 100% B in 30, 20.1
1299.2
1317.3
1300.2
1299.8
1317.8
1299.8
(l)G-C
(l)G-C
2
2
G-2'-phosphate
G-2',3'-cyclo-
B, D, F
B, D, F
B, D, F
B, D, F
B, D, F
phosphate
1
1
58
59
G-(l)C-CG-2'-phosphate
G-(l)C-CG-2',3'-cyclo-
phosphate
GC-(l)C-G-2'-phosphate
GC-(l)C-G-2',3'-cyclo-
phosphate
0 ! 100% B in 30, 17.0
0 ! 100% B in 30, 19.7
1317.3
1297.5
1317.8
1299.8
1
1
60
61
0 ! 100% B in 30, 16.9
0 ! 100% B in 30, 19.7
1315.5
1301.1
1317.8
1299.8
1
1
62
63
GC
GC
2
-(l)G-2'-phosphate
-(l)G-2',3'-cyclo-
0 ! 100% B in 30, 17.2
0 ! 100% B in 30, 19.8
1316.4
1297.5
1317.8
1299.8
2
phosphate
1
1
64
65
GC
GC
2
G
G
2
C
C
2
G-2'-phosphate
G-2',3'-cyclo-
0 ! 100% B in 30, 23.0
0 ! 100% B in 30, 25.8
2617.0
2599.4
2618.6
2600.6
2
2
2
phosphate
(GC)
f
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
3
B, F
B, F
B, D, F
0 ! 40% B in 30, 19.8 10 ! 100% B in 30, 15.6 ) 1888.3
1888.2
1277.8
1357.8
1339.8
1317.8
1299.8
2529.6
2849.8
3170.0
3499.2
3474.2
3490.2
4771.8
1653.0
1636.0
1332.8
1314.8
1965.3
707.4
GCG
GCG
GCG
2
2
2
0 ! 100% B in 30, 18.8
0 ! 100% B in 30, 20.9
0 ! 100% B in 30, 23.5
0 ! 100% B in 30, 16.7
0 ! 100% B in 30, 18.8
0 ! 100% B in 30, 23.3
0 ! 100% B in 30, 23.9
0 ! 100% B in 30, 24.4
0 ! 100% B in 30, 25.3
0 ! 100% B in 30, 24.5
0 ! 100% B in 30, 25.1
0 ! 100% B in 30, 26.8
0 ! 100% B in 30, 20.4
0 ! 100% B in 30, 23.6
0 ! 100% B in 30, 22.2
0 ! 100% B in 30, 18.7
0 ! 100% B in 30, 17.2
0 ! 100% B in 30, 15.7
0 ! 100% B in 30, 19.5
0 ! 100% B in 30, 22.0
0 ! 100% B in 30, 20.7
0 ! 100% B in 30, 21.7
0 ! 100% B in 30, 23.6
0 ! 100% B in 30, 23.2
0 ! 100% B in 30, 19.3
0 ! 100% B in 30, 20.7
1276.9
1358.0
1340.5
1320.9
1302.8
2529.7
2851.9
3173.3
3500.0
3473.3
3489.1
4771.8
1655.8
1638.3
1333.6
1312.8
1963.2
708.1
-2'-phosphate
-2',3'-cyclophosphate
GCGC-2'-phosphate
GCGC-2',3'-cyclophosphate
B, D, F
GCGT
GCGT
GCGT
GCGT
GCGT
GCGT
GCGT
GCGT
GCGT
2
3
4
4
4
5
8
2
2
CGC
CGC
CGC
CGCA
B, F
B, F
B, F
B, F
B, F
B, F
B, F
B, D, F
CGC
CGC
2
CGCT
-2'-phosphate
-2',3'-cyclophosphate
GCTG-2'-phosphate
GCTG-2',3'-cyclophosphate
B, D, F
GD
4
C
B, F
B, D, F
G
2
G
2
G
2
G
2
G
2
G
2
G
2
G
3
G
3
G
3
G
3
G
3
G
3
G
3
G
4
G
4
-2'-phosphate
-2',3'-cyclophosphate
±
±
A
2
C
2
G
C, G
B, F
C, G
2243.0
1916
2242.4
1916.3
1993.2
1975.2
2258.4
1357.8
1339.8
1888.2
2617.4
±
3959.5
2008.3
1397.2
1219.2
1397.2
1397.2
±
AGC
AGC
AGC
AGC
2
2
2
2
-2'-phosphate
-2',3'-cyclophosphate
G
1992.0
1974.0
2258.0
1357.4
1339.9
C, G
B, D, F
C-2'-phosphate
C-2',3'-cyclophosphate
f
C
3
B, F
B, F
B, F
B, F
A, F
B, F
B, F
B, F
B, F
B, F
B, F
B, F
B, F
B, F
0 ! 40% B in 30, 22.2 10 ! 100% B in 30, 15.9 ) 1889.6
CG
CG
CG
3
3
3
C
C
CG
0 ! 40% B in 30, 15.3 0 ! 100% B in 30, 24.8
0 ! 100% B in 30, 25.8
2617.5
±
3
C
0 ! 100% B in 30, 28.5
3960.1
2010.0
1398.7
1220.9
1397.0
1397.6
±
f
I
3
0 ! 40% B in 30, 15.7 0 ! 60% B in 30, 21.2 )
-2'-phosphate
-2',3'-cyclophosphate
0 ! 100% B in 30, 20.4
0 ! 40% B in 30, 14.5 0 ! 100% B in 30, 24.3
0 ! 100% B in 30, 20.4
(l)G
4
-2'-phosphate
4'-phosphate-G
4
0 ! 100% B in 30, 20.7
f
G
G
5
0 ! 40% B in 30, 17.2 10 ! 100% B in 30, 10.2 )
f
6
0 ! 40% B in 30, 15.1 10 ! 100% B in 30, 13.7 ) 2008.2
2008.3
2008.3
2697.4
2697.4
(l)G
6
8
0 ! 40% B in 30, 15.4 0 ! 100% B in 30, 24.5
0 ! 40% B in 30, 16.2 0 ! 100% B in 30, 27.2
0 ! 40% B in 30, 16.1 0 ! 100% B in 30, 27.2
2008.7
2698.4
2699.6
G
8
(l)G

Helvetica Chimica Acta ± Vol. 86 (2003)
4289
Table 3 (cont.)
Entry p-RNA
Deprotectionb)
Analytical HPLC
Reverse phasec) Ion Exchanged)
gradient, t [min] gradient, t [min]
MALDI-TOF-MSe)
À
À
b-d-pr-Oligonucleotide
Methods
[M À H] [M À H]
a
system )
R
R
(obs.)
(calc.)
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
223
224
225
226
227
228
229
G
G
G
G
G
G
10
A, F
C, G
B, F
B, F
B, F
C, G
B, F
B, D, F
0 ! 40% B in 30, 6.3
3390.7
3390.1
2304.5
1606.1
1926.3
1984.3
1959.3
2232.5
1292.8
1274.8
1332.8
1314.8
1901.2
2008.3
2008.3
2698.7
2698.7
2009.3
2008.3
2553.7
5132.3
2567.6
2549.6
3
2
2
2
2
TGTG
TA
TATA
TG
TGTG
0 ! 100% B in 30, 27.7 2305.0
0 ! 100% B in 30, 18.8 1605.3
0 ! 100% B in 30, 20.8 1925.4
0 ! 100% B in 30, 26.0 1986
0 ! 100% B in 30, 26.2 1958.0
0 ! 100% B in 30, 21.9 2234.0
0 ! 100% B in 30, 16.5 1295.7
0 ! 100% B in 30, 18.0 1276.9
0 ! 100% B in 30, 18.8 1332.7
0 ! 100% B in 30, 21.9 1313.7
0 ! 100% B in 30, 19.8 1901.4
2008.8
2
3
GTACGTA
GTC
GTC
GTCG-2'-phosphate
GTCG-2',3'-cyclophosphate
2
-2'-phosphate
-2',3'-cyclophosphate
2
B, D, F
GT
2
ATA
B, F
A, F
B, F
B, F
B, F
A, F
A, F
C, G
B, D, F
B, D, F
I
6
0 ! 40% B in 30, 14.8
0 ! 100% B in 30, 10.3
0 ! 18% B in 20, 17.7
0 ! 25% B in 20, 8.6
(l)I
6
2008.0
2699.3
2700.1
f
I
8
(l)I
(IG)
8
3
0 ! 40% B in 30, 15.8 0 ! 60% B in 30, 18.3 ) 2009.0
I
3
G
3
0 ! 40% B in 30, 14.7
2010.4
0 ! 100% B in 30, 18.9 2555.0
0 ! 100% B in 30, 23.6 5140.3
0 ! 100% B in 30, 20.8 2564.7
0 ! 100% B in 30, 23.4 2555.5
TA
TA
TAC
TAC
phosphate
TAC CG
4
2
TA
TA
T
T
2
2
3 4
TATA T
2
2
CG-2'-phosphate
CG-2',3'-cyclo-
2
2
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
2
T
2
2
TA
2
B, F
0 ! 100% B in 30, 24.6 4160.7
0 ! 100% B in 30, 16.2 1316.3
0 ! 100% B in 30, 18.9 1300.2
0 ! 100% B in 30, 16.1 1316.3
0 ! 100% B in 30, 18.9 1299.3
0 ! 100% B in 30, 16.4 1318.2
0 ! 100% B in 30, 18.9 1301.9
0 ! 100% B in 30, 18.2 1887.3
0 ! 100% B in 30, 21.8 2233.8
0 ! 100% B in 30, 15.6 1317.2
0 ! 100% B in 30, 17.3 1300.2
0 ! 100% B in 30, 15.7 1317.2
0 ! 100% B in 30, 17.4 1299.3
0 ! 100% B in 30, 25.0 2622.6
0 ! 100% B in 30, 16.4 1953.0
0 ! 100% B in 30, 17.9 1934.3
0 ! 100% B in 30, 21.0 3801.7
0 ! 100% B in 30, 23.7 3887.6
0 ! 100% B in 30, 21.5 3758.1
0 ! 100% B in 30, 16.9 1925.4
0 ! 100% B in 30, 18.1 1910.8
0 ! 100% B in 30, 21.1 4088.4
0 ! 100% B in 30, 26.2 3916.6
0 ! 100% B in 30, 20.6 2006.0
0 ! 100% B in 30, 23.4 1988.2
0 ! 100% B in 30, 21.4 2327.1
0 ! 100% B in 30, 24.3 2311.1
0 ! 100% B in 30, 17.0 1637.1
0 ! 100% B in 30, 19.3 1621.0
0 ! 100% B in 30, 17.7 1942.0
0 ! 100% B in 30, 19.5 1924.5
0 ! 100% B in 30, 19.9 1981.5
0 ! 100% B in 30, 23.0 1962.7
0 ! 100% B in 30, 24.7 3863.6
4156.6
1316.8
1298.9
1316.8
1298.8
1316.2
1298.8
1887.2
2232.5
1316.8
1298.9
1316.8
1298.9
2617.7
1950.2
1932.2
3803.4
3883.5
3756.4
1926.2
1909.2
4085.6
3916.5
2006.2
1988.2
2327.4
2309.4
1636.0
1618.0
1941.2
1923.2
1981.2
1963.2
3865.5
TACG-2'-phosphate
B, D, F
TACG-2',3'-cyclophosphate
(l)T-ACG-2'-phosphate
(l)T-ACG-2',3'-cyclophosphate
TAC-(l)G-2'-phosphate
TAC-(l)G- 2',3'-cyclophosphate
TACGTA
TACGTAG
TAGC-2'-phosphate
TAGC-2',3'-cyclophosphate
(l)T-AGC-2'-phosphate
(l)T-AGC-2',3'-cyclophosphate
B, D, F
B, D, F
B, F
B, F
B, D, F
B, D, F
TAG
4
TA
B, F
B, D, F
TATA
TATA
TATA
TATA
2
2
2
2
C-2'-phosphate
C-2',3'-cyclophosphate
C
2
ATATA
B, F
B, F
B, F
B, D, F
G
2
ATATA
TATAC
4
TATA
TATACC-2'-phosphate
TATACC-2',3'-cyclophosphate
TATAC
4
TAT
2
B, F
B, F
B, D, F
TATAG
TATAG
TATAG
TATAG
TATAG
4
2
2
2
2
TATA
-2'-phosphate
-2',3'-cyclophosphate
T-2'-phosphate
T-2',3'-cyclophosphate
B, D, F
B, D, F
B, D, F
B, D, F
B, F
TATAT-2'-phosphate
TATAT-2',3'-cyclophosphate
TATATC-2'-phosphate
TATATC-2',3'-cyclophosphate
TATATG-2'-phosphate
TATATG-2',3'-cyclophosphate
2 2
TATATG T ATA

4
290
Helvetica Chimica Acta ± Vol. 86 (2003)
Table 3 (cont.)
Entry p-RNA
Deprotectionb)
Analytical HPLC
Reverse phasec) Ion Exchanged)
MALDI-TOF-MSe)
À
À
b-d-pr-Oligonucleotide
Methods
[M À H]
[M À H]
a
system )
gradient, t
R
[min]
gradient, t
R
[min]
(obs.)
(calc.)
264
265
266
267
268
269
270
271
272
273
274
275
276
278
279
280
281
282
283
284
285
286
TAT
TAT
2
g)
g)
g)
1567.3
1564.0
2526.6
2544.0
1534.0
1574.0
2535.7
5133.3
5437.3
1331.8
1313.8
1331.8
1313.8
2498.4
3780.4
±
4
A
2
C, G
C, G
B, F
B, F
A, F
A, F
A, F
B, D, F
0 ! 100% B in 30, 21.8 2527.0
0 ! 100% B in 30, 18.9 2544.7
0 ! 100% B in 30, 17.4 1539.4
0 ! 100% B in 30, 19.9 1572.4
T
2
T
2
T
2
T
4
T
4
4
A TA
CGC
GCG
A
A
4
0 ! 40% B in 30, 20.7 0 ! 100% B in 30, 18.4 2533.2
3
TA
2
TA
2
T
0 ! 40% B in 30, 17.5
0 ! 40% B in 30, 17.9
5133.8
5438.1
d(T)-pr(T
TDCG-2'-phosphate
TDCG-2',3'-cyclophosphate
TDGC-2'-phosphate
TDGC-2',3'-cyclophosphate
4 3 3 2
A TA TA T)
0 ! 100% B in 30, 17.1 1331.6
0 ! 100% B in 30, 19.9 1315.0
0 ! 100% B in 30, 17.1 1331.7
0 ! 100% B in 30, 19.3 1312.7
B, D, F
T
T
8
A, F
C, G
A, E
A, E
A, E
A, E
A, E
A, E
A, E
A, E
0 ! 40% B in 30, 19.6 0 ! 100% B in 30, 14.0 2497.6
0 ! 100% B in 30, 26.5 3782.0
12
UA
3
UAU
2
0 ! 40% B in 30, 18.6 0 ! 100% B in 30, 16.2
±
(UA)
4
0 ! 40% B in 30, 18.4 0 ! 100% B in 30, 16.0 2481.4
2478.6
549.4
±
U
U
U
U
U
U
2
2
3
4
7
8
0 ! 40% B in 30, 9.6
0 ! 100% B in 30, 2.9
549.0
±
855.7
2480.0
2080.8
AUA
3
U
0 ! 40% B in 30, 18.9 0 ! 100% B in 30, 16.4
0 ! 40% B in 30, 11.9 0 ! 100% B in 30, 6.2
0 ! 40% B in 30, 19.0
856.6
A
4
2478.6
2081.2
2386.4
0 ! 40% B in 30, 18.9
0 ! 40% B in 30, 15.7 0 ! 100% B in 30, 13.4 2386.1
a) All sequences refer to d-p-RNA (d-pr) unless indicated otherwise; ꢀl) ˆ l-prˆ l-p-RNA; pl ˆ d-pyranosyl-
lyxoNA; d(T) ˆ Thymidine. All cyclophosphate containing sequences were prepared from their respective 2'-
b
or 3'- or 4'-phosphate precursor sequences by treatment with DEC ¥ HCl. ) Method A: Allyl deprotection for
0
−
(
Trityl-off× sequences: 360 ml (3.7 mmol) of BuNH
225 mmol) of Ph P, 140 ml (3.7 ml) of HCO H in 4.5 ml of THF, 558, 1.5 h; 4.5 ml of 0.1m aq. NaCS
B: Allyl deprotection for −Trityl-on× sequences: 20 mg (148 mmol) of Et NH CO [54] in 0.75 ml of CH
Cl , room temperature, 2 h (12 ± 24 h
. Method C: Allyl deprotection for −Trityl-off× sequences:
2
, 20 mg (22 mmol) of (dibenzylideneacetone)
3
Pd , 60 mg
2
3
2
2
NEt
. Method
Cl
2
2
2
3
2
0
1
1.6 mg (10 mmol) of (Ph
for isoG sequences); 3.5 ml of 0.1m aq. NaCS
60 ml (4.5 mmol) of Et NH , 177 mg (171 mmol) of (dibenzylideneacetone)
2.7 mmol) of Ph P, 360 ml (9.6 mmol) of HCO H in 6.5 ml of THF, 508, 3 ± 4 h; 4.5 ml of 0.1m aq. NaCS
Method D: Detachment from CPG/acyl deprotection: sat. aq. NH /EtOH, 3 : 1, room temperature, 3 ± 4 h.
Method E: Detachment from CPG/acyl deprotection: 4.5 ml of 0.2m MeONH ¥ HCl in aq. conc. NH /EtOH,
room temperature, 20 h. Method F: Detachment from CPG/acyl deprotection: 1.5 ml of 15 ± 25% aq. NH NH
O, 48, 20 ± 30 h (room temperature, 48 h for D-containing sequences). Method G: Detachment from CPG/
acyl deprotection: 5 ml of 40% (w/w) aq. MeNH /conc. aq. NH
1 : 1, room temperature, 5 ± 6 h. c) Aquapore
3
P)
4
Pd ,1.3 mg (5 mmol) of Ph
3
P in 1.0 ml of CH
2
2
2
NEt
2
0
4
(
2
2
3
Pd ÀCHCl
3
adduct [61], 432 mg
2
NEt .
3
2
2
3
2
3
2
2
¥
H
2
2
3
RP-300 C-8 Brownlee, 220 Â 4.6 mm, 7 mm, flow 1 ml/min. Mobile phase: eluant A: 0.1m Et
3 2
N, 0.1m AcOH, H O,
d
pH 7.0; eluant B: 0.1m Et
3
N, 0.1m AcOH, H
2
O/MeCN 1 : 4. ) MONO-Q HR 5/5 Pharmacia, 10 Â 0.5 cm, flow
1
ml/min. Mobile phase: eluant A: 10 mm Na HPO , H O, pH 11.5; eluant B: 10 mm Na HPO , 1m NaCl, H O,
2
4
2
2
4
2
e
pH 11.5. ) Matrix-assisted laser-desorption ionization time-of-flight mass spectroscopy; matrix: 3-hydro-
picolinic acid or a-cyanohydroxycinnamic acid or 2,4,6-trihydroxyacetophenone and ammonium citrate buffer.
f
)
Nucleogen-DEAE 60-7 Macherey-Nagel, 125 Â 4 mm, flow 1 ml/min eluant A: 10 mm Na
2 4 2
HPO , H O/MeCN
g
1
: 4, pH 11.5; eluant B: 10 mm Na
2
HPO
4
, 1m NaCl, H
2
O/MeCN 1 : 4, pH 11.5. ) From Hoechst, Z¸rich.
amidites 9a ± 9g and the suitably derivatized CPG solid supports 11a ± 11g, utilizing the
protocols that were developed earlier for the homo-DNA-oligonucleotide synthesis
and adapting them to the specific demands of the p-RNA synthesis [28]. The following
were the conditions used: a) detritylation with 6% Cl CHCO H in ClCH CH Cl for a
2
2
2
2
period of 5 ± 7 min, b) coupling for 45 ± 60 min with 10% (w/v) soln. of phosphor-
amidites in MeCN (160 ± 400 ml per coupling) in the presence of a mixture of 0.15m
5-(4-nitrophenyl)-1H-tetrazole and 0.35m 1H-tetrazole in MeCN (180 ± 600 ml per
coupling), c) capping for 1.5 min with 0.45m DMAP in MeCN plus a mixture of

Helvetica Chimica Acta ± Vol. 86 (2003)
4291
collidine/Ac O/MeCN 3 :2 :5; and d) oxidation for 1 min with 0.01m I in MeCN/
2
2
collidine/H O 100 :9 :45. This procedure led to average coupling yields of > 99% (by
2
trityl assay). Later in the study, the oligonucleotide sequences were also synthesized
with Expedite 8909 Gene Assembler (PE Biosystems), with the following changes to the
synthesis protocol: a) the use of 0.1m soln. of phosphoramidites 9a ± 9g in MeCN, b) use
of a 0.35m soln. of 2-(ethylsulfanyl)-1H-tetrazole in the coupling step, c) coupling time
of 60 min, and d) the use of a 6% Cl CHCO H soln. in ClCH CH Cl for detritylation
2
2
2
2
over a 5-min period. Average coupling efficiencies were >98% (by tritylation assay).
In both instances, the majority of the oligonucleotides were synthesized in the −Trityl-
on× mode, and, in the early phase of the investigation sequences, in the −Trityl-off×
18
mode ).
Prior to the detachment from the CPG solid support and deprotection of acyl
groups, the p-RNA-oligonucleotide strands attached to the solid support were first
0
treated with Pd catalyst under the Noyori conditions [54][55] to remove the allyl
groups, thus converting the labile phosphotriester linkage to the more stable
phosphodiester linkage. In the early trials of the study, allyl deprotection was
performed with the tris(dibenzylideneacetone)dipalladium(0) in a two-phase medium
by using BuNH and HCO H at 508 [55], a procedure which was cumbersome from an
2
2
19
operational standpoint ). A more convenient approach was the modification
0
developed by Noyori and co-workers [54] where (Ph P) Pd and Et NH ¥ HCO were
3
3
2
3
used resulting in a more homogeneous reaction mixture leading to a complete reaction
within 3 ± 4 h at room temperature. In the case of G- and isoG-containing sequences,
this treatment led to the simultaneous de-allylation of the allyl protecting groups on the
0
purines as well. The Pd catalyst was removed by reacting with 0.1m aq. NaSC(S)NEt
2
leaving the acyl-protected p-RNA-oligonucleotides on the solid support.
While detachment from the solid support could be achieved without complications
under standard deprotection conditions (conc. aq. NH in H O/EtOH 3 :1, room
3
2
temperature, 40 h), the acyl deprotection of p-RNA-oligonucleotides is more difficult
20
because of the steric hindrance towards debenzylation in the axial 3'-OBz groups ).
This can lead to partial acyl deprotection with substantial strand scission. The latter is a
consequence of the all-cis-disposition of the 2'-O, 3'-O, and 4'-O groups that facilitates
the formation of 2',3'- or 3',4'-cyclic phosphates (see below). However, treatment with
1
5 ± 25% aq. NH NH ¥ H O at 48 over a period of 20 ± 48 h resulted in the desired
2 2 2
concomitant solid-support detachment and acyl-group deprotection with minimal
strand scission. Attempts to remove NH NH by evaporation on a rotavap resulted also
2
2
in strand scission. However, it was observed that sequences still containing the end Tr
18)
The terms −Trityl-off× and −Trityl-on× mode refer to whether the final 4'-O-DMT group is removed before
or after taking the oligomer off the solid support. The sequences that were prepared in the −Trityl-on×
mode were conveniently handled in purification, when the base-deprotection step involved treatment with
2 2 2
NH NH ¥ H O (see Exper. Part, Sect. 4.3).
This method was incompatible with oligonucleotides containing the end Tr group. Also the use of BuNH
has its drawbacks due to its high nucleophilicity [54].
19)
20)
2
In one case (pr-C
.8m soln. of NH
6
), deprotection of the 3'-OBz groups under two different methods were tested: a) with
3
4
Cl in NH
3
/EtOH 3:1, at room temperature for 28 h, which resulted in substantial strand
¥ HCl in NH /EtOH 3:1, at
scission without complete 3'-OBz deprotection; b) with 0.24m soln. of MeONH
2
3
room temperature for 28 h, which gave a smaller percentage of strand scission, but also greater percentage
of incomplete deprotection of 3'-OBz as compared to Method a.

4
292
Helvetica Chimica Acta ± Vol. 86 (2003)
group could be best separated from NH NH by passing through a Sep Pak C18 (Waters)
2
2
column. This became the reason for synthesizing and isolating p-RNA-oligonucleotide
sequences with an end Tr group (−Trityl-on× mode). For D-containing sequences, the
hydrazine deprotection had to be carried out at room temperature for at least 48 h to obtain
21
fully deprotected oligomers ). In the case of U-containing sequences, not surprisingly,
the NH NH treatment fails due to the sensitivity of the pyrimidine ring towards the
2
2
reactive nucleophile; this problem was overcome partially by the use of MeONH in
2
22
ethanolic NH under carefully controlled conditions ). In contrast, the pyrimidine ring of
3
thymine is sufficiently stable toward the hydrazinolytic conditions of deprotection. Table 3
lists the specific deprotection methods applied to the various oligonucleotides synthesized.
Sequences without an end Tr group (−Trityl-off× mode) were directly taken to the
next step of HPLC purification. Oligomers synthesized with the end Tr group (−Trityl-
on× mode) were detritylated with 80% aq. HCO H at room temperature. Purification of
2
the crude oligonucleotides was carried out by reverse-phase HPLC (Spherisorb-S10X
RP-C18, 220 Â 12 mm, with 0.1m aq. Et NH ¥ AcOH as buffer) or ion-exchange HPLC
3
(
Nucleogen DEAE 60-7 or Mono Q HR 5/5, with 10 mm Na HPO as buffer) according
2 4
to their preparative history. Table 3 lists all synthesized p-RNA oligonucleotides with
their specific purification protocol. In the initial stages of the project, the crude
oligonucleotides containing the end Tr group were first purified by preparative reverse-
phase HPLC, followed by detritylation (80% aq. HCO H, room temperature). Later, it
2
was found that the initial reverse-phase purification could be bypassed and
detritylation, followed by purification on ion-exchange HPLC, was efficient enough
to provide oligonucleotides with purity of > 95%. For oligonucleotides containing a 2'-
or 3'-phosphate or 2',3'- or 3',4'-cyclophosphate group (see below), ion-exchange HPLC
was the purification method of choice.
0
The efficiency of the (Ph P) Pd /Et NH ¥ HCO /hydrazinolysis deprotection
3
3
2
3
procedure employed as described above was fairly adequate for sequences of short
lengths (e.g., octamer or less), but, in some cases, the deprotection procedure gave
unsatisfactory results, as typified by the HPLC traces in Fig. 5. This variation in yields
depending on the sequences worsened when extended to longer sequences (e.g.,
dodecamers), especially for mixed sequences (see, e.g., Fig. 6).
Optimized Procedure for p-RNA-Oligonucleotide Deprotection. While the above
mentioned synthesis methods were satisfactory, there was a need to improve the
deprotection protocol, which would lead to increased efficiency of the purification
process (by reducing the difficulties due to the occurrence of interfering peaks in the
chromatogram). Towards this end, it was discovered that a 1:1 combination of 40%
(
w/w) aq. MeNH in conc. aq. NH at room temperature for 5 ± 6 h could efficiently
2
3
detach the oligonucleotides from the solid support and concomitantly remove the acyl
protecting groups. Since NH and MeNH were easily removed by evaporation without
3
2
any deleterious side effects, a side benefit was the riddance of the necessity of having to
21)
22)
The Bz group at the N -position of D is removed very slowly (compared to the N -Bz group).
The deprotection was monitored by HPLC and stopped at a stage judged to have the maximum amount of
the fully deprotected sequence before degradation began to occur. The partially deprotected sequences
were collected, combined, and subjected again to the deprotection reaction conditions. Such a tedious
operation prompted us to replace U with T in our future studies.
2
6

Helvetica Chimica Acta ± Vol. 86 (2003)
4293
Fig. 5. Ion-exchange HPLC traces (Mono-Q) of selected p-RNA-oligonucleotide sequences before and after
purification. The traces of the crude products document the wide variability in the efficiency of the deprotection
0
protocol employed (deprotection of allyloxy groups by Pd complex [55], followed by subsequent
(concomitant) detachment from CPG and acyl group deprotection by hydrazinolysis). See also Fig. 6,a. For
HPLC conditions, see Exper. Part, Sect. 6.

4
294
Helvetica Chimica Acta ± Vol. 86 (2003)
synthesize sequences with end Tr group (−Trityl-on×) and thereby eliminating the need
for an additional separation step (removal of NH NH with the Sep Pak C18 cartridge).
2
2
The next improvement resulted from applying the procedure of Noyori and co-workers
23
[
61] for more efficient allyl deprotection ). Thus, the CPG solid support containing the
oligonucleotide strand was treated with a homogenous soln. of the catalyst prepared
from tris(dibenzylideneacetone)dipalladium(0)ÀCHCl adduct, Ph P, Et NH , and
3
3
2
2
24
HCO H in THF at 508 for 3 ± 4 h ). After removal of the Pd catalyst by washing with
2
0
.1m aq. NaSC(S)NEt N soln., CPG containing the oligonucleotide sequence was
2
subjected to 40% aq. MeNH /30% aq. NH 1:1, at room temperature for 5 ± 6 h, which
2
3
resulted in concomitant detachment and acyl-group deprotection. Concentration in
vacuo and ion-exchange purification of the crude oligonucleotides (over a Mono Q HR
5/5 column) afforded oligonucleotides with >95% purity with improved yields in the
range of 2 ± 10 fold (Table 4). Fig. 6 demonstrates the effectiveness of this optimized
deprotection method over the previously used hydrazine-based procedure for a 13-mer
mixed sequence.
Table 4. Comparison of the Efficiency of the Two Deprotection Methods for p-RNA Oligonucleotides
Entry
p-RNA sequence
Deprotection
Method )
O. D. 260 nm (yield %)
a
1
2
3
4
5
6
pr-GAGCC-2'-phosphate
pr-GAGCC-2'-phosphate
pr-CACACCCGGCTCC
pr-CACACCCGGCTCC
pr-GGAGCCG
A
B
A
B
A
B
7.5 (2.1)
26.4 (8.0)
7.1 (1.5)
88.0 (15)
21.1 (10)
88.7 (18)
pr-GGAACCG
a) Method A: 1. Pd(PPh
O 4 : 1, 258, 0.25 h. Method B: 1. Pd
08, 3.5 h; 2. MeNH /aq. NH 1 : 1, 258, 5.5 h.
)
, PPh
, Et
NH
¥ HCO
in CH
Cl
, 258, 4.5 h; 2. NH
NH
/H
O, 48, 42 h; 3. HCO
NH, HCO H in THF,
3
4
3
2
2
3
2
2
2
2
2
2
H/
H
5
2
2
(dibenzylideneacetone)
3
ÀCHCl adduct, PPh
3
3
, Et
2
2
2
3
Synthesis of Phosphate and Cyclophosphate Derivatives of p-RNAOligonucleotides.
Studies on the capability of nonenzymatic template-directed ligation and replication
necessitated the synthesis of the suitable 2'- or 4'-phosphate and 2',3'- or 3',4'-
cyclophosphate derivatives of p-RNA oligonucleotides. The 4'-phosphate oligonucleo-
tide derivatives were synthesized by the following procedure: after the final
detritylation step on the synthesizer, the CPG-bound oligonucleotide was coupled
with 0.1m (allyloxy)[bis(diisopropylamine)]phosphine [55] in MeCN in the presence
of the activator, affording the 4'-O-phosphitylated oligonucleotide derivative. Sub-
sequent capping and oxidation followed by the usual deprotection of the allyl groups
under the Noyori conditions [54] resulted in the 4'-phosphorylated p-RNA oligonu-
23)
Initiated by M. S. It was thought that there is incomplete allyl deprotection under the previously used
Noyori conditions, which led to a greater percentage of strand scission in the subsequent acyl-deprotection
step under basic conditions.
24)
The progress of allyl deprotection could be monitored by HPLC after base-catalyzed release/deprotection
2 3
step (Et NH/conc. aq. NH treatment).

Helvetica Chimica Acta ± Vol. 86 (2003)
4295
Fig. 6. Ion-exchange HPLC traces (Mono Q) in the synthesis of the sequence pr-(CACACCCGGCTCC)
documenting the efficiency of the optimized allyl and base deprotection procedure B over a previously used
deprotection protocol A. a) HPLC Trace resulting from the old deprotection protocol (1. Pd(PPh
Et NH ¥ HCO in CH Cl , 258, 4.5 h; 2. NH NH /H O, 48, 42 h; 3. HCO H/H O 4 :1, 258, 0.25 h). b) HPLC
Trace resulting from the optimized deprotection protocol (1. Pd (dibenzylideneacetone) adduct, PPh
ÀCHCl
Et NH, HCO H in THF, 508, 3.5 h; 2. MeNH /aq. NH , 258, 5.5 h). For HPLC conditions, see Exper. Part. Sect. 6.
3 4 3
) , PPh ,
2
2
3
2
2
2
2
2
2
2
2
3
3
3
,
2
2
2
3
cleotides (see Exper. Part., Sect. 5.1). Oligonucleotide-2'-phosphate (and 3'-phosphate)
derivatives were synthesized by automated solid-support synthesis by using a
25
universal solid support derivatized with a sulfone linker [62] ). The 2'-end (or the
25)
Initially CPG-derivatized with 2'-benzoyl-1'-benzyl-5'-(4,4'-dimethoxytrityl)-3'-succinyl-b-d-ribofurano-
side was used; but this was later abandoned for the more convenient approach outlined above. We thank
Dr. H. Moser, Dr. R. H‰ner, and Dr. F. Natt (Novartis AG, Basel) for a generous gift of the sulfone linker
[62].

4
296
Helvetica Chimica Acta ± Vol. 86 (2003)
26
3
'-end) ) of the starting nucleoside was coupled onto this solid support. Applying the
Noyori protocol [54][55] removed the allyl groups; the base-deprotection procedure,
while removing the acyl protecting groups, causes the simultaneous detachment of the
oligonucleotide strand from the solid support (via the elimination of sulfone linker),
yielding the p-RNA-oligonucleotide-2'- or 3'-phosphate to be subjected to HPLC-
purification (see Exper. Part., Sect. 5.2).
p-RNA Oligonucleotides containing 3',4'- and 2',3'-cyclophosphate end groups were
prepared by treating the corresponding 4'- or 2'-phosphate sequences (0.1 ± 0.5 mm aq.
soln.), with a large excess of N-[3-(dimethylamino)propyl]-N'-ethylcarbodiimide ¥ hy-
drochloride (DEC ¥ HCl; 0.5 ± 1.0m aq. soln.) at room temperature for 3 ± 4 h
(
Scheme 5). The reactions were monitored by ion-exchange HPLC and stopped when
ca. 90% complete. The excess DEC ¥ HCl was removed by desalting, and the cyclic
phosphate group containing oligomers were isolated by ion-exchange HPLC. The cyclic
phosphates are stable under the conditions of isolation and purification (see Exper.
Part., Sect. 5.3).
Scheme 5. Preparation of 2',3'-Cyclophosphates of b-d-Ribopyranosyl Oligonucleotides from the Correspond-
ing 2'-Phosphate Derivatized Oligonucleotides
3
. Properties of p-RNA Oligonucleotides. ± As already mentioned in the
Introduction, most results of our experimental studies on the structural and chemical
properties of p-RNA oligonucleotides have been described in much detail in our
preliminary communications [1 ± 8], so that there is no need to fully reproduce and re-
discuss these results here. Therefore, this chapter is restricted to an overview of the
observations made in these many studies. However, some of the experiments on
26)
The preparation of the p-RNA-mononucleoside 3'-phosphoramidites is described in [10].

Helvetica Chimica Acta ± Vol. 86 (2003)
4297
p-RNA properties that have remained unpublished (or are only partially published)
thus far are incorporated in this chapter in more detail.
3
.1. Structure. A comprehensive NMR structure determination by Jaun and
Schlˆnvogt and co-workers [4] of the duplex formed by the self-complementary p-
RNA-octamer sequence pr(CGAATTCG), accompanied by a MD-based molecular
modeling by Wolf and co-workers [4], has shown that this duplex has a left-handed
(weakly) helical ladder structure held together in antiparallel strand orientation by
canonical WatsonÀCrick base pairs with a large inclination of the (averaged) backbone
axis relative to the (averaged) base-pair axes [1][8] and, as a consequence thereof, with
a mode of base stacking that is purely interstrand (in sharp contrast to the pure
intrastrand stacking made of neighboring nucleobases in B-type DNA) (Fig. 7). The
structure analysis confirmed in essence the conjectures on the p-RNA duplex structure
suggested by a qualitative conformational analysis of a p-RNA single strand based on
idealized conformations by using steric and stereoelectronic selection criteria [1]. Fig. 8
recalls the two most pronounced deviations between the idealized and the exper-
imental structure [4]; they occur in the torsion angles b and e, and refer to the two
bonds that connect the phosphodiester group to the ribopyranose positions C(4') and
C(2'). These deviations are interpreted to be the consequence of −Newman strain× [64],
which is to be expected for any phosphodiester group bound to a secondary C-atom in
an idealized (strictly staggered) conformation. Specifically in this case, the strain is due
to the steric clash between the equatorial H-atom at C(5') and the phosphodiester O-
atom bound to C(2') (partially relieved by the deviation of the torsion angle b from its
idealized value of 1808), and to an analogous steric clash between the equatorial H-
atom at C(3') and one of the (negatively charged) phosphodiester O-atom of the C(2')-
bound phosphodiester group (relieved by the deviation of torsion angle e from its
idealized value À608). The intra-phosphodiester torsion angles a (ca. 718 ±778) and z
(
À1748 ± À 1778) [4] deviate only slightly from their idealized values; they reflect two
well-behaved À g/t phosphodiester conformations. The occurrence of g/t conformations
in p-RNA duplexes documents that there can hardly be a significant difference in
intrinsic stability of g/g and g/t phosphodiester conformations, as some earlier
(theoretical) studies had been proposing [65], and as our original version of predicting
the base-pairing potential of homo-DNA oligonucleotide systems by qualitative
27
conformational analysis had assumed to exist [29] ). Simulated annealing calculations
of the structure of the p-RNA duplex derived from pr(CGAATTCG) based on the
NMR data of [4] by Schwalbe and co-workers [66] corroborated the conclusions drawn
by Jaun and Schlˆnvogt [4]. Complementary information on, and a further illustration
of, such a quasi-linear type of oligonucleotide structure comes from a NMR structure
analysis carried out by Schwalbe et al. [67] of a duplex derived from the partially self-
complementary AATAT sequence in the series of Quinkert×s d-peptide analogues of p-
RNA [68].
3
.2. Base Pairing. p-RNA-Oligonucleotide sequences containing the canonical
nucleobases undergo base-pairing with strictly antiparallel strand orientation in the
27)
However, the postulate that t/t phosphodiester conformations are −forbidden conformations× formally
deduced from the same type of stereoelectronic reasoning [29] seems to hold; no such conformation seems
to have been observed in phosphodiester or oligonucleotide structures so far.

Fig. 7. NMR Structure of p-RNA duplex derived from the self-complementary sequence pr-(CGAATTCG) [4]. a) Time-averaged and refined structure of octamer duplex
derived from self-complementary p-RNA sequence 4'-CGAATTCG-2', based on a 1000-ps MD calculation (Amber) by R. M. Wolf [39] and an NMR structure analysis by
Jaun et al. [39]. Ball-and-stick representation as well as excerpts on upper right and lower left illustrate the interstrand base stacking characteristic to pentopyranosyl-
(
2' ! 4')-oligonucleotide duplexes (purine-pyrimidine and purine-purine stacking: upper right, pyrimidine-pyrimidine −non-stacking×: lower left). Courtesy of Dr. R. M.
Wolf, Novartis Pharma AG, Basel. b) Excerpt of duplex structure referred to in a), illustrating interstrand base stacking in side view (a former cover picture of Helv. Chim. Acta).

Helvetica Chimica Acta ± Vol. 86 (2003)
4299
Fig. 7 (cont.)
canonical WatsonÀCrick pairing mode [1][2][4]. Their base-pairing strength is
uniformly greater than that of corresponding RNA and DNA oligonucleotide
sequences (Tables 5 and 6). This behavior is not specific to the ribopyranosyl
oligonucleotide system, the entire family of pentopyranosyl-(4' ! 2')-oligonucleotide
systems shows WatsonÀCrick base pairing that is stronger than that of RNA and DNA