The strength of the 3′-gauche effect dictates the structure of 3′-O-anthraniloyladenosine and its 5′-phosphate, two analogues of the 3′-end of aminoacyl-tRNA

Article abstract of DOI:10.1039/a809350d

Anthranilic acid charged yeast tRNA^Phe or E. coli tRNA^Val are able to form a stable complex with EF-Tu*GTP, hence the 2′-and 3′-O-anthraniloyladenosines and their 5′-phosphate counterparts have been conceived to be the smallest units that are capable of mimicking aminoacyl-tRNA. Since the 3′-O-anthraniloyladenosine also binds more efficiently to the EF-Tu*GTP complex compared to its 2′-isomer, we have herein delineated the stereoelectronic features that dictate the conformation of 3′-O-anthraniloyladenosine and its 5′-phosphate vis-a-vis their 2′-counterparts and we have also addressed how their structures and thermodynamic stabilizations are different from adenosine and 5′-AMP. It has been found that the electron-withdrawing anthraniloyl group exerts gauche effects of variable strengths depending upon whether it is at the 2′-or at 3′-position because of either the presence or absence of O2′-N9 gauche effect, [GE(O2′-C2′-C1′-N9)]. thereby steering the pseudorotation of the constituent sugar moiety either to the North (N)-type (C3′-endo) or South (S)-type (C2′-endo) conformation. The 3′-O-anthraniloyladenosine 5′-phosphate has a relatively more stabilized S-type conformation ΔG° = -4.6 kJ mol^-1) than 3′-O-anthraniloyladenosine itself (ΔC° = -3.9 kJ mol^-1), whereas the ΔG° for 2′-O-anthraniloyl-adenosine and its 5′-monophosphate are respectively -0.9 and -1.8 kJ mol^-1, suggesting that the 3′-gauche effect of the 3′-O-anthraniloyl group is stronger than that of 2′-O-anthraniloyl in the drive of the sugar conformation. Since the EF-Tu can specifically recognize the aminoacylated-tRNA from the non-charged tRNA, we have assessed the free-energy (ΔG°) for this recognition switch to be the least ≈ -2.9 kJ mol^-1 by comparison of ΔG° of the N=S pseudorotational equilibrium for 3′-O-anthraniloyladenosine 5′-phosphate and 5′-AMP. The 3′-O-anthraniloyl-adenosine and its 5′-phosphate are much more flexible than the isomeric 2′-counterparts as is evident from the temperature-dependent coupling constants analysis. The relative rate of the transacylation reaction of 2′(3′)-O-anthraniloyladenosine and its 5′-phosphate is cooperatively dictated by the two-state N=S pseudorotational equilibrium of the sugar, which in turn is controlled by a balance of the stereoelectronic 3′-and 2′-gauche effects as well as by the pseudoaxial preference of the 3′-O-or 2′-O-anthraniloyl group. The reason for the larger stabilization of the 2′-endo conformer for 3′-O-anthraniloyladenosine and its 5′-phosphate lies in the fact that the C3′-O3′ bond takes up an optimal gauche orientation with respect to the C4′-O4′ bond dictating the pseudoaxial orientation of the 3′-anthraniloyl residue, which can be achieved only in the S-type sugar conformation with adenin-9-yl and the 2′-OH groups in the pseudoequatorial geometry, compared to the preferred C3′-endo sugar with a pseudoaxial aglycone and 2′-OH found in the 3′-terminal adenosine moiety in the helical 3′-CCA end of uncharged tRNA.

Full text of DOI:10.1039/a809350d

View Article Online / Journal Homepage / Table of Contents for this issue
The strength of the 3Ј-gauche effect dictates the structure of
3Ј-O-anthraniloyladenosine and its 5Ј-phosphate, two analogues
of the 3Ј-end of aminoacyl-tRNA
Parag Acharya,a Barbara Nawrot,b Mathias Sprinzl,c Christophe Thibaudeaua and
Jyoti Chattopadhyaya*a
a Department of Bioorganic Chemistry, Box 581, Biomedical Centre, University of Uppsala,
S-751 23 Uppsala, Sweden. E-mail: jyoti@bioorgchem.uu.se. Fax: ϩ4618554495
b Laboratory of Bioorganic Chemistry, Polish Academy of Sciences Centre of Molecular and
Macromolecular Studies 90-363 Lodz, Sienkiewicza 112, Poland
c Laboratorium für Biochemie, Universität Bayreuth, 95440 Bayreuth, Germany
Received (in Cambridge) 30th November 1998, Accepted 14th April 1999
Anthranilic acid charged yeast tRNAPhe or E. coli tRNAVal are able to form a stable complex with EF-Tu*GTP,
hence the 2Ј- and 3Ј-O-anthraniloyladenosines and their 5Ј-phosphate counterparts have been conceived to be the
smallest units that are capable of mimicking aminoacyl-tRNA. Since the 3Ј-O-anthraniloyladenosine also binds
more efficiently to the EF-Tu*GTP complex compared to its 2Ј-isomer, we have herein delineated the stereo-
electronic features that dictate the conformation of 3Ј-O-anthraniloyladenosine and its 5Ј-phosphate vis-a-vis
their 2Ј-counterparts and we have also addressed how their structures and thermodynamic stabilizations are
different from adenosine and 5Ј-AMP. It has been found that the electron-withdrawing anthraniloyl group exerts
gauche effects of variable strengths depending upon whether it is at the 2Ј- or at 3Ј-position because of either the
presence or absence of O2Ј–N9 gauche effect, [GE(O2Ј–C2Ј–C1Ј–N9)], thereby steering the pseudorotation of
the constituent sugar moiety either to the North (N)-type (C3Ј-endo) or South (S)-type (C2Ј-endo) conformation.
The 3Ј-O-anthraniloyladenosine 5Ј-phosphate has a relatively more stabilized S-type conformation ∆GЊ = Ϫ4.6 kJ
molϪ1) than 3Ј-O-anthraniloyladenosine itself (∆GЊ = Ϫ3.9 kJ molϪ1), whereas the ∆GЊ for 2Ј-O-anthraniloyl-
adenosine and its 5Ј-monophosphate are respectively Ϫ0.9 and Ϫ1.8 kJ molϪ1, suggesting that the 3Ј-gauche effect of
the 3Ј-O-anthraniloyl group is stronger than that of 2Ј-O-anthraniloyl in the drive of the sugar conformation. Since
the EF-Tu can specifically recognize the aminoacylated-tRNA from the non-charged tRNA, we have assessed the
free-energy (∆GЊ) for this recognition switch to be the least ≈ Ϫ2.9 kJ molϪ1 by comparison of ∆GЊ of the N
S
pseudorotational equilibrium for 3Ј-O-anthraniloyladenosine 5Ј-phosphate and 5Ј-AMP. The 3Ј-O-anthraniloyl-
adenosine and its 5Ј-phosphate are much more flexible than the isomeric 2Ј-counterparts as is evident from the
temperature-dependent coupling constants analysis. The relative rate of the transacylation reaction of 2Ј(3Ј)-O-
anthraniloyladenosine and its 5Ј-phosphate is cooperatively dictated by the two-state N S pseudorotational
equilibrium of the sugar, which in turn is controlled by a balance of the stereoelectronic 3Ј- and 2Ј-gauche effects as
well as by the pseudoaxial preference of the 3Ј-O- or 2Ј-O-anthraniloyl group. The reason for the larger stabilization
of the 2Ј-endo conformer for 3Ј-O-anthraniloyladenosine and its 5Ј-phosphate lies in the fact that the C3Ј–O3Ј bond
takes up an optimal gauche orientation with respect to the C4Ј–O4Ј bond dictating the pseudoaxial orientation of the
3Ј-anthraniloyl residue, which can be achieved only in the S-type sugar conformation with adenin-9-yl and the 2Ј-OH
groups in the pseudoequatorial geometry, compared to the preferred C3Ј-endo sugar with a pseudoaxial aglycone and
2Ј-OH found in the 3Ј-terminal adenosine moiety in the helical 3Ј-CCA end of uncharged tRNA.
It has been suggested however3 that the destacking of the
3Ј-terminal adenosine upon aminoacylation is the structural
Introduction
During the process of in vivo protein synthesis an amino acid
attached to the 3Ј-terminal adenosine of transfer RNA (tRNA)
serves as an acceptor for a peptidyl residue. In this process the
aminoacylated-tRNA (aa-tRNA) which is correctly charged by
a cognate aminoacyl tRNA synthetase has to specifically interact
with several enzymes, protein factors such as elongation factor
Tu, and the ribosome. Differences of several orders of magni-
tude exist in the equilibrium dissociation constants between
non-aminoacylated tRNA and aminoacyl-tRNA and these
macromolecules constitute the molecular basis of recognition
of aminoacyladenosine on the 3Ј-end of aminoacyl-tRNA.1
Despite this large difference in the equilibrium dissociation
constants between non-aminoacylated tRNA and aminoacyl-
tRNA in the interaction with elongation factor Tu, no large
structural change was noted in tRNA upon aminoacylation.2
element determining the recognition of aminoacyl-tRNA by
the elongation factor Tu (EF-Tu)*GTP complex. At standard
physiological conditions the migration of the aminoacyl residue
between the 2Ј- and 3Ј-ribose positions (transacylation) in
aa-tRNA takes place.4 This means that EF-Tu must be able to
distinguish between the two isomeric substrates, namely 2Ј- and
3Ј-aminoacyl-tRNA. The 3Јaminoacyl-tRNA is found to bind
to EF-Tu both in the solid5 and solution4b state.
It has been shown that the 3Ј-O-anthraniloyladenosine, com-
pared to its 2Ј-isomer, binds more efficiently to the EF-Tu*GTP
complex from the eubacterium Thermus thermophilus.6 Since
7
8
anthranilic acid charged yeast tRNAPhe or E. coli tRNAVal
are able to form a stable complex with EF-Tu*GTP, the 2Ј- and
3Ј-O-anthraniloyladenosines (Scheme 1, 1a and 2a) and their
5Ј-phosphate (Scheme 1, 1b and 2b) counterparts have been
J. Chem. Soc., Perkin Trans. 2, 1999, 1531–1536
1531

View Article Online
a
Table 1 Thermodynamics (kJ molϪ1
)
of the pseudorotational equilibrium of the isomeric 2Ј- and 3Ј-O-anthraniloyladenosine and their 5Ј-
∆GЊc
phosphates
ϪT∆SЊb
∆∆GЊd,e
Compound
∆HЊb
(298 K)
275 K
298 K
338 K
(298 K)
Adenosine
5Ј-AMP
2Ј-ant-Ado (1a)
3Ј-ant-Ado (2a)
2Ј-ant-AMP (1b)
3Ј-ant-AMP (2b)
Ϫ4.4 (0.2)
Ϫ3.5 (0.6)
Ϫ1.3 (0.5)
Ϫ5.6 (0.7)
Ϫ0.3 (0.5)
Ϫ8.5 (0.8)
2.6 (0.1)
1.8 (0.6)
Ϫ0.5 (0.5)
1.7 (0.7)
Ϫ0.7 (0.5)
3.9 (0.8)
Ϫ1.8 (0.2)
Ϫ1.7 (0.1)
Ϫ1.8 (0.2)
Ϫ3.9 (0.2)
Ϫ0.9 (0.2)
Ϫ4.6 (0.3)
1.8 (0.1)
Ϫ1.7 (0.2)
Ϫ4.0 (0.2)
Ϫ0.9 (0.2)
Ϫ5.0 (0.3)
Ϫ1.4 (0.2)
Ϫ1.7 (0.2)
Ϫ3.7 (0.2)
Ϫ1.0 (0.2)
Ϫ3.9 (0.2)
0.0d
Ϫ2.1d
ϩ0.8e
Ϫ2.9e
a The signs of the thermodynamic parameters are chosen in such a way that the positive and negative values indicate the drive of the N
S
equilibrium towards N-type and S-type conformations, respectively (see ref. 9). b ∆HЊ and ∆SЊ are average values (their standard deviations are
shown in brackets) (see ref. 9 and 13). c ∆GЊ at T = 275 K, 298 K and 338 K have been calculated from the relation: ∆GT = ϪRT ln (xS/xN) where xS
and xN are mole fractions of S- and N-type conformations. These values were comparable to the ∆GЊ calculated using the relation:
∆GT = ∆HЊ Ϫ T∆SЊ. d ∆∆GЊ has been calculated according to the relation: ∆∆GЊ = ∆GЊ (298 K) of 2Ј- or 3Ј-anthraniloyl-Ado Ϫ ∆GЊ (298 K) of
adenosine. e ∆∆GЊ has been calculated according to the relation: ∆∆GЊ = ∆GЊ (298 K) of 2Ј- or 3Ј-anthraniloyl-AMP Ϫ ∆GЊ (298 K) of 5Ј-AMP.
endo) pseudorotational equilibrium10,11 of the sugar moiety
NH2
in nucleos(t)ides is energetically controlled by the interplay of
N
stereoelectronic gauche9a,g,i,j,12 [σC–H→σ*C–O/N orbital mixing]
N
N
and anomeric9 [n(O4Ј)→σ*C1Ј–N orbital mixing] effects as well
NH2
ROH2C
O
ROH2C
N
N
N
as the steric effects of various substituents. It is however the
electronic character, the configuration of the sugar substituents
as well as their environments, in general, that dictate the relative
strengths of these stereoelectronic effects.9
O
O
N
OH
HO
N
O
O
O
NH2
H2N
(B) Difference between the energetics of 3Ј-O-anthraniloyl-
adenosine and its 5Ј-phosphate and their 2Ј-counterparts
The population of S-type conformations in 3Ј-O-anthraniloyl-
adenosine and in its 5Ј-phosphate decreases with increasing
temperature (see Experimental section (B) for details of pseudo-
rotational analyses9,13 of sugar conformations and Experi-
mental section (C) for van’t Hoff plots). There is a net decrease
of 6 and 10% xS for 3Ј-O-anthraniloyladenosine and its 5Ј-
phosphate, respectively, on going from 275 to 338 K.9,13 How-
ever, in the case of the 2Ј-isomer of both adenosine as well as its
5Ј-phosphate, the sugar conformation is locked in a 60:40 ratio
of S- and N-type sugars over the whole temperature range. A
van’t Hoff plot of ln (xS/xN) versus the inverse of temperature
for the 3Ј-O-anthraniloyladenosine as well as its 5Ј-phosphate
gave straight lines.9a,s The slope and the intercept gave ∆HЊ,
Ϫ∆SЊ as well as ∆GЊ (Table 1). Since there was very little
temperature dependency of ln (xS/xN) in 2Ј-O-anthraniloyl-
adenosine and its 5Ј-phosphate, the van’t Hoff plot was much
less accurate and the error in the determination of the thermo-
dynamic quantities was high (Table 1).
1a: R = H
1b: R = OPO3H–
2a: R = H
2b: R = OPO3H–
North (N)-sugar (C3′-endo-C2′-exo)
South (S)-sugar (C3′-exo-C2′-endo)
Pseudoequatorial aglycone
Pseudoaxial aglycone
Phase-angle (P):0°
P
36°
Phase-angle (P):144°
P 180°
Puckering amplitude (ψm) = 38.6°±3°
Puckering amplitude (ψm) = 38.6°±3°
Scheme 1
conceived to be the smallest units that are capable of mimicking
aa-tRNA.6,8 Since 3Ј-O-anthraniloyladenosine binds more
strongly to EF-Tu than the corresponding 2Ј-isomer, it has been
suggested that the structure of the 3Ј-isomer could be a good
starting point to design specific inhibitors6 of the aa-tRNA*-
EF-Tu interaction, and thus the whole bacterial protein bio-
synthesis. It is this latter question that has persuaded us to
examine whether there are any intrinsic differences between the
sugar pseudorotational behaviour of 2Ј-O-anthraniloyl- and 3Ј-
O-anthraniloyladenosines (2Ј-ant-Ado and 3Ј-ant-Ado respect-
ively) and their 5Ј-phosphate counterparts. Related to this, we
have also debated whether the thermodynamics of the sugar
pseudorotational characteristics of 2Ј-O-anthraniloyl and 3Ј-O-
anthraniloyladenosines and their 5Ј-phosphate counterparts
are indeed related to their intrinsic ability to transacylate.
In this work, we have examined the relative thermodynamics
of the pseudorotation of 2Ј- and 3Ј-O-anthraniloyladenosines
and their 5Ј-phosphate counterparts as well as the underlying
stereoelectronic forces that thermodynamically dictate their
conformational preferences in comparison with adenosine and
adenosine 5Ј-phosphate (5Ј-AMP), thereby attempting to
understand the energetic basis of discrimination of uncharged
tRNA vis-a-vis aminoacylated-tRNA by the EF-Tu*GTP
complex.
(C) Comparison of the strengths of the 2Ј- and 3Ј-gauche effects in
2Ј- and 3Ј-O-anthraniloyladenosines and their 5Ј-phosphates
In 2Ј- and 3Ј-O-anthraniloyladenosines and their 5Ј-phosphate
counterparts, the anomeric and inherent steric effects of
adenin-9-yl are present in all four derivatives, and hence can be
disregarded in this relative assessment of the stereoelectronic
forces that drive their respective sugar pseudorotations. There is
however an interplay of three different gauche effects (GE) in
both 2Ј- and 3Ј-isomers: GE[O2Ј–C2Ј–C1Ј–N9], GE[O2Ј–C2Ј–
C1Ј–O4Ј] and GE[O3Ј–C3Ј–C4Ј–O4Ј], which play a key role in
the modulation of the thermodynamics of the two-state
N
S equilibrium. In this dynamic N S conformational
equilibrium, the influence of GE[O2Ј–C2Ј–C3Ј–O3Ј] can also
be assumed to be mutually cancelled, since it is present in both
2Ј- and 3Ј-isomers.
Results and discussion
(A) Interplay of stereoelectronic forces in the drive of the sugar
conformations in nucleos(t)ides
Our earlier NMR studies9 have uniquely shown that the
two-state North (N)-type (C3Ј-endo) or South (S)-type (C2Ј-
We have earlier shown9 that GE[O2Ј–C2Ј–C1Ј–N9] and
GE[O3Ј–C3Ј–C4Ј–O4Ј] drive the sugar conformation towards
the S-type, whereas GE[O2Ј–C2Ј–C1Ј–O4Ј] drives the sugar to
the N-type. Hence the electronic nature governing the strength
of the relative electronegatives of the substituent attached to
1532
J. Chem. Soc., Perkin Trans. 2, 1999, 1531–1536

View Article Online
the 2Ј- or 3Ј-position will either enhance or weaken the strength
of the 2Ј- or 3Ј-GE,9a,g,i,j and this will be the decisive factor in
the outcome of the gauche effect promoted drive of the overall
However, in the case of 3Ј-O-anthraniloyladenosine 5Ј-
phosphate (3Ј-ant-AMP) both GE[O3Ј–C3Ј–C4Ј–O4Ј] and
GE[O2Ј–C2Ј–C1Ј–N9] drive the sugar conformation towards
the S-type, whereas the GE[O2Ј–C2Ј–C1Ј–O4Ј] drives the sugar
conformation to the N-type. Here it is noteworthy that it is the
GE[O3Ј–C3Ј–C4Ј–O4Ј] which is the most dominating driving
force to the preferred S-type conformation (90% S, 10% N in
the N S equilibrium) because of the stronger electro-
negativity of the 3Ј-O-anthaniloyl group, and as a result, we
observe a relatively large stabilization of the S-type conformer
as evident from a ∆GЊ (298 K) value of Ϫ4.6 kJ molϪ1. The drive
of the sugar conformation in 3Ј-O-anthraniloyladenosine (3Ј-
ant-Ado) is also governed by the 3Ј-gauche effect arising from
the 3Ј-O-anthraniloyl group, causing a net preference for the
S-type sugar (∆GЊ = Ϫ3.9 kJ molϪ1, Table 1).
The important characteristic observed here for 3Ј-ant-AMP
or Ado derivatives vis-a-vis 2Ј-O-anthraniloyl counterparts is
that the ∆GЊ is temperature-dependent: The ∆GЊ for 3Ј-ant-
AMP varies from Ϫ5.0 kJ molϪ1 at 275 K to Ϫ3.9 kJ molϪ1 at
338 K, whereas for 3Ј-ant-Ado, ∆GЊ varies from Ϫ4.0 kJ molϪ1
to Ϫ3.7 kJ molϪ1 in the same temperature range, thereby show-
ing an intrinsic structural flexibility of the 3Ј-isomer. It has not
escaped our attention that this inherently flexible character of
3Ј-O-anthraniloyladenosine and its 5Ј-phosphate is perhaps a
useful degree of freedom that can possibly be used by EF-Tu to
steer and tune its conformation to a free-energy minimum for
facilitating its recognition, interaction and binding to EF-Tu.
The second important biological consequence of the energetic
preferences of 3Ј-O-anthraniloyladenosine and its 5Ј-phosphate
compared to adenosine and 5Ј-AMP (compare ∆∆GЊ in Table
1) is that, upon aminoacylation, the bias of the two-state 3Ј-
endo (N) 2Ј-endo (S) pseudorotational equilibrium shifts to
the C2Ј-endo from the preferred C3Ј-endo state as found in the
3Ј-terminal adenosine moiety in the helical 3Ј-CCA end of
tRNA in the free compared to the aminoacylated state. This
shift of the N S equilibrium to the S form is the crucial
structural change of aminoacyl-tRNA compared to free tRNA.
Each tRNA specific for a particular amino acid needs a cognate
synthetase for its recognition, whereas once tRNA is amino-
acylated, all tRNA molecules are transported by the same
protein (EF-Tu) to the ribosome. Thus, the free-energy of the
structural changes of the 3Ј-terminal adenosine moiety which
occur upon aminoacylation of tRNA must serve as a recogni-
tion element for EF-Tu. The ∆GЊ for this switch (at 298 K)
has been assessed to be ≈ Ϫ2.9 kJ molϪ1, obtained by subtract-
ing the ∆GЊ at 298 K of the N S equilibrium of 3Ј-O-
anthraniloyladenosine 5Ј-phosphate from that of 5Ј-AMP.
N
S equilibrium.
Due to its stronger electron-withdrawing nature, the
anthraniloyl group exerts gauche effects of variable strengths
depending upon whether it is located at the 2Ј- or 3Ј-position
because this will influence the presence or absence of GE[O2Ј–
C2Ј–C1Ј–N9]. The strength of the 2Ј- or 3Ј-gauche effects of the
anthraniloyl group in 1a,b and 2a,b, and thereby the effect of the
2Ј- or 3Ј-aminoacylation in general, can be assessed in com-
parison with the hydroxy group by comparing the effect of the
former with the latter in adenosine or 5Ј-AMP. Owing to the
stereoelectronic nature of a pure gauche interaction,12 it should
be independent of the overall entropy of the system, which is
difficult to delineate in practice. As accurate estimate of the
relative magnitude of the pure stereoelectronic 3Ј- and 2Ј-
gauche effects in 1a,b and 2a,b in comparison with adenosine or
5Ј-AMP can be simply derived9a,i,j by subtracting the respective
∆HЊ values to give ∆∆HЊ (eqns. (1a) and (2a)), whereas the
overall effect of substitution of the anthraniloyl group in 1a,b
and 2a,b compared with 2Ј/3Ј-OH in adenosine or 5Ј-AMP can
only be assessed by the relative change of stabilization of
S- over N-type pseudorotamers by subtracting their respective
∆GЊ values to give ∆∆GЊ (eqns. (1b) and (2b)). Thus the relative
strengths of the 2Ј- or 3Ј-gauche effects of the anthraniloyl
group (∆∆GЊ) in 1a,b and 2a,b can be quantified.
∆∆HЊ = [∆HЊ (298 K) of 2Ј- or 3Ј-ant-Ado] Ϫ
[∆HЊ (298 K) of adenosine] (1a)
∆∆GЊ (at 298 K) = [∆GЊ (298 K) of 2Ј- or 3Ј-ant-Ado] Ϫ
[∆GЊ (298 K) of adenosine] (1b)
∆∆HЊ = [∆HЊ (298 K) of 2Ј- or 3Ј-ant-AMP] Ϫ
[∆HЊ (298 K) of 5Ј-AMP] (2a)
∆∆GЊ (at 298 K) = [∆GЊ (298 K) of 2Ј- or 3Ј-ant-AMP] Ϫ
[∆GЊ (298 K) of 5Ј-AMP] (2b)
Thus, ∆∆HЊ (∆∆GЊ) values for 2Ј-O-anthraniloyladenosine
5Ј-phosphate and 2Ј-O-anthraniloyladenosine are respectively
3.2 (ϩ0.8) and 3.1 (0.0) kJ molϪ1, whereas ∆∆GЊ for their 3Ј-
counterparts are respectively Ϫ5.0 (Ϫ2.9) and Ϫ1.2 (Ϫ2.1) kJ
molϪ1, showing that the 3Ј-O-anthraniloyl group exerts a
stronger gauche effect than the 2Ј-O-anthaniloyl group. The
reason for the weaker gauche effect of the 2Ј-O-anthraniloyl
group is that its actual strength is reduced by the competing
GE[O2Ј–C2Ј–C1Ј–N9] and GE[O2Ј–C2Ј–C1Ј–O4Ј], which are
absent for the 3Ј-O-anthraniloyl group.
(D) The temperature-dependency of the molar ratio of the 2Ј- and
3Ј-anthraniloyl isomers
In the case of 2Ј-O-anthraniloyladenosine 5Ј-phosphate (2Ј-
ant-AMP), a relatively stronger GE[O2Ј–C2Ј–C1Ј–N9] driving
the sugar to the S-type conformation is counteracted by the
GE[O2Ј–C2Ј–C1Ј–O4Ј], which drives the sugar to the N-type
conformation. Their magnitudes are relatively similar as is evi-
dent from the 60:40 xS/xN sugar population, which translates
into ∆GЊ = Ϫ0.9 kJ molϪ1 (at 298 K, including the contribution
of the anomeric effect) for the drive of the N S equilibrium to
the S-type. Similarly in the case of 2Ј-O-anthraniloyladenosine
(2Ј-ant-Ado), the relatively stronger GE[O2Ј–C2Ј–C1Ј–N9]
driving the sugar to the S-type conformation is counteracted
by the GE[O2Ј–C2Ј–C1Ј–O4Ј], which drives the sugar to the
N-type conformation. Their magnitudes are relatively com-
parable in 2Ј-ant-Ado, and we find a ∆GЊ (298 K) = Ϫ1.8
kJ molϪ1 (Table 1) favouring slightly an S-type conformation.
Remarkably, the ∆GЊ of the N S equilibrium of both 2Ј-
ant-AMP and 2Ј-ant-Ado are temperature independent over
the 275–338 K range, thereby showing poor flexibility in its
structural make-up, which is also evidenced from counteracting
values for ∆HЊ and ϪT∆SЊ (Table 1).
It has been observed that the molar ratio of 2Ј-O-anthraniloyl-
adenosine (or its 5Ј-phosphate) gradually increases over that of
the corresponding 3Ј-isomer as the temperature increases from
275 K to 338 K (see Experimental section (D)). The plot of the
logarithm of the ratio of 3Ј- over 2Ј-isomer as a function of the
inverse of the temperature gives a straight line (see Experi-
mental section (E)), thereby giving the thermodynamics of the
isomerization process: ∆HЊ = Ϫ1.6 kJ molϪ1, ϪT∆SЊ at 298
K = 0.4 kJ molϪ1, and ∆GЊ = Ϫ1.2 kJ molϪ1 for 2Ј- and 3Ј-ant-
AMP, and ∆HЊ = Ϫ1.0 kJ molϪ1, ϪT∆SЊ at 298 K = Ϫ0.7 kJ
molϪ1, and ∆GЊ = Ϫ1.7 kJ molϪ1 for 2Ј- and 3Ј-ant-Ado.
(E) The ratio of mole fractions of 2Ј- over 3Ј-isomers of the
anthraniloyl group and relative populations of S- and
N-conformers are correlated
We then examined whether the relative rate of 2Ј/3Ј-isomeriz-
ation of the anthraniloyl group is in some way related to the
change of the population of the sugar conformation. Hence, we
plotted the logarithm of the ratio of mole fractions of 2Ј/3Ј-
J. Chem. Soc., Perkin Trans. 2, 1999, 1531–1536
1533

View Article Online
isomers of both 3Ј-ant-Ado and 3Ј-ant-AMP versus the
logarithm of the ratio of the population of S- and N-type sugar
pseudorotamers, which gave straight lines (see Experimental
section (F)). These correlation plots show that as well as the
increase of the population of the N-type conformer with
increasing temperature, the ratio of the 2Ј-isomer increases (see
Experimental section (F)). Similarly, as the temperature
decreases, the population of the S-type conformer increases
with an increase of the 3Ј-isomer population, because the bulky
and polar 2Ј-anthraniloyl group takes up a pseudoaxial pos-
ition in the N-type conformation whereas the 3Ј-anthraniloyl
group becomes pseudoaxial in the S-type conformation
(Scheme 1). Thus, it is shown for the first time that the prefer-
ence of a sugar conformation and the preponderance of a
specific aminoacylated isomer are mutually interdependent. It
therefore follows that the 2Ј/3Ј-aminoacyl isomer equilibrium
can be shifted more to the 2Ј-isomer by shifting the pseudo-
rotational equilibrium to the N-type conformation by retuning
the anomeric effect of the adenin-9-yl moiety in the aa-tRNA
by protonation or metallation or complexation with a electro-
philic ligand under given physiological conditions. This may in
turn act as a switch for promoting the dissociation of enzymes
and protein factors that recognize and bind to the 3Ј-aminoacyl
isomer and the associated C2Ј-endo conformation.
and the σ* orbital of the C4Ј–O4Ј (σ→σ* orbital mixing),
resulting in an optimal gauche orientation of C3Ј–O3Ј with
respect to C4Ј–O4Ј, dictates the pseudoaxial orientation of the
3Ј-anthraniloyl residue in the S-type sugar conformation,
thereby placing both the adenin-9-yl moiety and the free 2Ј-OH
group in the pseudoequatorial positions. This simple conform-
ational switch, orchestrated by the dynamic change of the
2Ј 3Ј transacylation equilibrium, has at least three important
biological consequences: (i) the aminoacylation influences the
N
S equilibrium of the 3Ј-terminal adenosine, (ii) the S-type
conformation of the 3Ј-end of aminoacyl-tRNA fits to the
binding pocket of EF-Tu,7 (iii) ligands with a predominant
N-type sugar conformation, such as puromycin, cannot serve as
analogs of aminoacyl-tRNA for interaction with EF-Tu.8
Experimental
(A) Temperature-dependent 1H NMR spectra at 600 MHz
1H NMR spectra were recorded for both compounds [2Ј(3Ј)-O-
anthraniloyladenosine and 2Ј(3Ј)-O-anthraniloyladenosine 5Ј-
phosphate] at 5 mM concentration at 600 MHz (Bruker DRX
600) in 0.5 ml of the following buffer solutions: 100 mM NaCl,
10 mM NaH2PO4, 10 mM Na2HPO4, 20 mM MgCl2, 10 µM
EDTA, pD 7.2 in 99.9% D2O with 0.05 ml of CD3OD as co-
solvent (δCH CN = 2.00 ppm as internal reference) between 275
3
K and 338 K at 5 K intervals. The pD value corresponds to
the reading on a pH meter equipped with a calomel electrode
calibrated with two standard buffers (pH 4 and 7) and is not
corrected for the deuterium isotope effect. All spectra have been
recorded using 64 K data points and 100 scans. The accurate
3JHH ( 0.1 Hz) values (Table 2) were obtained through simu-
lation using the DAISY program package (supplied by Bruker
spectrospin, Germany) and have been used for the pseudo-
rotational analyses [vide infra, section (B)].
Conclusions
(1) It is evident from the relative temperature dependency of
the N S population and from the relative preference for the
S-type conformation in the 3Ј- and 2Ј-isomers of anthraniloyl-
AMP or adenosine derivatives that the 3Ј-isomers have a
relatively more stabilized S-type conformation (∆GЊ = Ϫ4.6 kJ
molϪ1 and ∆GЊ = Ϫ3.9 kJ molϪ1 for 3Ј-isomers of AMP and
adenosine respectively), driving the 3Ј-O-anthraniloyl group
into the pseudoaxial orientation because of the intrinsic energy-
minimizing gauche orientation of C3Ј–O3Ј and C4Ј–O4Ј bonds.
The 3Ј-aminoacyl isomers are also endowed with a larger stabil-
ization of the 2Ј-endo conformer, and also are more inherently
flexible, which allows them to play key roles in more effective
enzymatic recognition by EF-Tu compared to the 2Ј-isomer.
(2) It has been found that the relative migration of the
anthraniloyl group from the 3Ј- to the 2Ј-position is nearly the
same for both the anthraniloyladenosines and their 5Ј-
phosphate derivatives. However, the percentage change of the
S-type conformation (Table 1) over the temperature range 275–
338 K and ∆GЊ are higher for 3Ј-ant-AMP (Ϫ4.6 kJ molϪ1)
than 3Ј-ant-Ado (Ϫ3.9 kJ molϪ1). This suggests that the struc-
tural modification of 3Ј-ant-AMP will most probably produce a
better inhibitor of the aa-tRNA*EF-Tu interaction which is
necessary for bacterial protein biosynthesis since the magnitude
of the stabilization of the C2Ј-endo conformation most prob-
ably serves as a recognition element by elongation factor Tu.
(3) Upon aminoacylation, the bias of the two-state 3Ј-endo
(N) 2Ј-endo (S) pseudorotational equilibrium shifts to the
C2Ј-endo state in the 3Ј-terminal adenosine moiety in the helical
3Ј-CCA end of tRNA. This shift of the N S equilibrium to
the S form is the crucial structural change that takes place in the
aminoacyl-tRNA compared to the uncharged tRNA. Each
tRNA specific for a particular amino acid needs a cognate
synthetase for its recognition, whereas once tRNA is amino-
acylated, all tRNA molecules are transported by the same
protein (EF-Tu) to ribosome. Thus, the free-energy of the struc-
tural changes of the 3Ј-terminal adenosine moiety which occur
upon aminoacylation of tRNA must serve as a recognition
element for EF-Tu. The energy penalty for this switch (∆GЊ at
298 K) has been assessed to be ≈Ϫ2.9 kJ molϪ1 by comparison
of ∆GЊ of the N S pseudorotational equilibrium of 3Ј-O-
anthraniloyladenosine 5Ј-phosphate and 5Ј-AMP.
(B) Conformational preferences of sugar moieties in 2Ј- and 3Ј-
isomers of anthraniloyladenosine and its 5Ј-phosphate derivatives
by pseudorotational analyses with PSEUROT program
3
3
3
The experimentally measured vicinal JHH (3J1Ј2Ј, J2Ј3Ј, J3Ј4Ј
)
values have been used (Table 2) as input for the PSEUROT
(version 5.4) program13 to calculate the temperature-dependent
bias of the two-state N S equilibrium of the sugar moieties in
the 2Ј- and 3Ј-isomers of anthraniloyladenosine and its 5Ј-
phosphate derivatives. The PSEUROT program13 calculates the
iterative best fit between the experimentally measured 3JHH and
the five pseudorotational parameters needed to describe the
two-state equilibrium9–11 [i.e. the phase angles of the N (PN)
and S (PS) pseudorotamers, their respective puckering ampli-
tudes Ψm(N) and Ψm(S), as well as the mole fractions of one of
them].
The PSEUROT program is based on the pseudorotation con-
cept14 and on the new generalised Karplus equation,15a,b which
3
links JHH coupling constants to the corresponding proton–
proton torsion angles on the basis of the “λ” substituent
parameters scale.15c,d The following λ values have been used for
the substituents attached to the H–C–C–H fragments: λ (C1Ј) =
λ (C2Ј) = λ (C3Ј) = λ (C4Ј) = 0.62, λ (C5Ј) = 0.68, λ (O4Ј) = 1.27,
λ (OH) = 1.26, λ (glycosyl-N) = 0.58, λ [OC(᎐O)R] = 1.15.15c,d
᎐
The proton–proton torsion angles in the pentofuranose moiety
can be translated into the corresponding endocyclic angles (νi,
i = 0–4) by the PSEUROT program using a relationship of the
type ΦHH = B ϩ Aνi. A and B parameters for all compounds are
as follows: A = 1.03 and B = 121.4 for Φ1Ј2Ј, A = 1.06 and
B = 2.40 for Φ2Ј3Ј, A = 1.09 and B = Ϫ124.0 for Φ3Ј4Ј. The
PSEUROT analyses have been performed in two ways: (i)
Ψm(N) and Ψm(S) were first assumed to be identical and con-
strained to the same value in a certain range (vide infra) during
the PSEUROT calculation, and (ii) when either the N or the S
pseudorotamers is significantly preferred over the other (i.e. by
(4) In summary, the culmination of the optimal antiperi-
planar interaction between the σ orbital of the C3Ј–H bond
1534
J. Chem. Soc., Perkin Trans. 2, 1999, 1531–1536

View Article Online
3
Table 2 JHH couplings (in Hz) for all compounds over the temperature range 275–338 K
Compound
Coupling 275 K 278 K 283 K 288 K 293 K 298 K 303 K 308 K 313 K 318 K 323 K 328 K 333 K 338 K
a
a
a
2Ј-ant-ADO (1a)
3J1Ј2Ј
3J2Ј3Ј
3J3Ј4Ј
3J1Ј2Ј
3J2Ј3Ј
3J3Ј4Ј
3J1Ј2Ј
3J2Ј3Ј
3J3Ј4Ј
3J1Ј2Ј
3J2Ј3Ј
3J3Ј4Ј
6.1
5.5
3.8
7.2
5.4
2.3
5.4
5.4
4.3
7.6
5.2
1.7
6.1
5.5
3.8
7.2
5.5
2.4
5.4
5.4
4.3
7.5
5.2
1.8
6.1
5.5
3.8
7.2
5.5
2.4
5.4
5.4
4.3
7.5
5.2
1.9
6.1
5.6
3.8
7.1
5.5
2.4
5.4
5.4
4.3
7.5
5.2
1.9
6.1
5.6
3.8
7.1
5.5
6.1
5.6
3.8
7.1
5.5
2.6
5.5
5.5
4.3
7.4
5.3
2.1
6.1
5.6
3.8
7.0
5.5
2.6
5.4
5.5
4.4
7.3
5.3
2.1
6.0
5.6
3.9
7.0
5.5
2.7
5.4
5.6
4.4
7.3
5.3
2.2
6.0
5.6
3.9
6.9
5.5
2.7
5.4
5.6
4.4
7.3
5.3
2.2
6.0
5.6
3.9
6.9
5.6
2.8
5.4
5.6
4.4
7.2
5.4
2.3
6.0
5.6
3.9
6.9
5.6
2.8
5.5
5.6
4.4
7.1
5.4
2.5
6.0
5.7
3.9
6.8
5.6
2.9
5.4
5.6
4.5
7.1
5.4
2.5
5.9
5.7
4.0
6.8
5.6
2.9
5.5
5.6
4.5
7.0
5.4
2.6
3Ј-ant-ADO (2a)
2Ј-ant-AMP (1b)
3Ј-ant-AMP (2b)
7.1
5.5
2.5
5.4
5.5
4.3
7.4
5.3
2.0
2.5
a
a
a
7.5
5.3
1.9
a H2Ј resonance at this temperature was under the residual HOD signal, and these coupling constants could therefore not be determined.
≥70%), P and Ψm of the minor conformer have been frozen to
different sets of values (vide infra), while the geometry of the
major conformer was optimized freely during the PSEUROT
fitting process.
during these PSEUROT analyses, are Ϫ18Њ < PN < 39Њ and
163Њ < PS < 131Њ.
3Ј-ant-AMP. The population of S-type conformers (%S) at
different temperatures is as follows (error is shown in paren-
theses): 90 (1.2) at 275 K, 88 (1.2) at 288 K, 87 (1.2) at 298 K, 85
(1.2) at 308 K, 85 (1.2) at 318 K, 81 (1.2) at 328 K, 80 (1.2) at
338 K. Ψm(N) and Ψm(S) have been first constrained to the
same value from 30Њ to 45Њ in 1Њ steps. PN and PS are optimized
during these PSEUROT analyses. We have also subsequently
constrained PN to Ϫ15Њ < PN < 30Њ in 15Њ steps with Ψm(N)
fixed at 35Њ, 37Њ and 40Њ which resulted in 145Њ < PS < 158Њ with
33Њ < Ψm(S) < 38Њ (∆Jmax < 0.3 Hz and rms < 0.2 Hz). The
resulting mole fractions of N-type and S-type conformers were
used to draw the same number of van’t Hoff plots.
The ensemble of PSEUROT analyses performed for each
compound has been designed in order to examine the hyper-
space of geometries that are accessible to the N-type and S-type
pseudorotamers and in such way that the deviations between
the experimental and back-calculated 3JHH (they are reflected in
the values of ∆Jmax, vide infra) as well as the rms of the analyses
remain minimal (vide infra). We have also investigated the influ-
ence of the error ( 0.1 Hz) of the experimentally measured
3JHH by performing various sets of PSEUROT analyses with
3
randomization of JHH coupling values by allowing them to
vary 0.1 Hz from experimental values over the entire tempera-
ture range for all compounds. For each compound, the different
sets of temperature-dependent mole fractions of the N-type or
the S-type pseudorotamers were calculated during various
PSEUROT optimization, which were used to make the van’t
Hoff plots (typically 10000 plots). The different slopes and
intercepts of these van’t Hoff plots were subsequently averaged
to get average contributions to the N S equilibrium as well
as their associated standard deviations.
(C) Van’t Hoff plots of mole fractions of sugar conformations vs.
1/T for 1a and 1b
The plot of the logarithm of ratio of mole fractions of S-type
conformation (xS) and N-type conformation (xN) [ln(xS)/(xN)]
as a function of the inverse of the temperature (1000/T) over
the temperature range of 275–338 K for 3Ј-O-anthraniloyl-
adenosine (1a) and 3Ј-O-anthraniloyladenosine-5Ј-monophos-
phate (1b) give straight lines with slope = 0.68 (σ = 0.01),
intercept = Ϫ0.70 (σ = 0.05) and Pearson’s correlation coef-
ficient (R) = 0.99 for the former, and slope = 1.02 (σ = 0.03),
intercept = Ϫ1.57 (σ = 0.12) and Pearson’s correlation coef-
ficient (R) = 0.98 for the latter.
2Ј-ant-Ado. The population of S-type conformers (%S) at
different temperatures are as follows (error is shown in paren-
theses): 68 (1.7) at 275 K, 68 (1.7) at 288 K, 68 (1.7) at 298 K, 68
(1.7) at 308 K, 66 (1.7) at 318 K, 66 (1.7) at 328 K, 65 (1.7) at
338 K. Ψm(N) and Ψm(S) have been first constrained to the
same value from 30Њ to 45Њ in 1Њ steps. PN and PS, optimized
during these PSEUROT analyses, are Ϫ39Њ < PN < 40Њ and
125Њ < PS < 154Њ.
(D) The temperature-dependency of the molar ratio of 2Ј- and 3Ј-
anthraniloyl isomers
The molar ratio of 2Ј- to 3Ј-isomers of the anthraniloyladeno-
sine and its 5Ј-phosphate derivatives at various temperatures
have been achieved by integrating the H1Ј of the 3Ј-isomer
while integration of H1Ј proton resonance of the 2Ј-isomer has
been referenced as 1.0. They are as follows: 2Ј-(3Ј)-ant-AMP
(2b)/(1b) 1.71 at 275 K, 1.65 at 288 K, 1.61 at 298 K, 1.56 at 308
K, 1.58 at 318 K, 1.52 at 328 K, 1.51 at 338 K. 2Ј-(3Ј)-ant-Ado
(2a)/(1a) 1.94 at 275 K, 1.98 at 288 K, 1.94 at 298 K, 1.95 at 308
K, 1.91 at 318 K, 1.85 at 328 K, 1.81 at 338 K.
3Ј-ant-Ado. The population of S-type conformers (%S) at dif-
ferent temperatures is as follows (error is shown in parentheses):
85 (1.3) at 275 K, 84 (1.3) at 288 K, 83 (1.3) at 298 K, 82 (1.3) at
308 K, 81 (1.3) at 318 K, 80 (1.3) at 328 K, 79 (1.3) at 338 K.
Ψm(N) and Ψm(S) have been first constrained to the same value
from 30Њ to 45Њ in 1Њ steps. PN and PS, optimized during these
PSEUROT analyses are Ϫ20Њ < PN < 40Њ and 152Њ < PS < 164Њ.
We have also subsequently constrained PN to Ϫ15Њ < PN < 30Њ
in 15Њ steps with Ψm(N) fixed at 34Њ, 36Њ and 38Њ which resulted
in 152Њ < PS < 163Њ with 35Њ < Ψm(S) < 38Њ (∆Jmax < 0.3 Hz
and rms < 0.2 Hz). The resulted mole fractions of the N-type
and S-type conformers were used to make the van’t Hoff plots.
(E) Van’t Hoff plots of mole fractions of 2Ј/3Ј-O-anthraniloyl
isomers vs. 1/T for 1a and 1b
The plot of the logarithm of the ratio of mole fractions of the
3Ј-isomer (3Ј) and the 2Ј-isomer (2Ј) [i.e. ln(3Ј)/(2Ј)] as a func-
tion of the inverse of the temperature (1000/T) over the tem-
perature range of 275–338 K for 3Ј-O-anthraniloyladenosine
(1a) and for 3Ј-O-anthraniloyladenosine 5Ј-monophosphate
(1b) give straight lines with a slope = 0.12 (σ = 0.01), inter-
cept = 0.27 (σ = 0.04) and Pearson’s correlation coefficient
(R) = 0.88 for the former, and with slope = 0.19 (σ = 0.01),
2Ј-ant-AMP. The population of S-type conformers (%S) at
different temperatures is as follows (error is shown in paren-
theses): 59 (1.5) at 275 K, 59 (1.5) at 288 K, 59 (1.5) at 298 K, 57
(1.5) at 308 K, 59 (1.5) at 318 K, 60 (1.5) at 328 K, 59 (1.5) at
338 K, Ψm(N) and Ψm(S) have been first constrained to the
same value from 30Њ to 45Њ in 1Њ steps. PN and PS, optimized
J. Chem. Soc., Perkin Trans. 2, 1999, 1531–1536
1535

View Article Online
Chem. Commun., 1994, 537; (i) C. Thibaudeau, J. Plavec, N. Garg,
A. Papchikhin and J. Chattopadhyaya, J. Am. Chem. Soc., 1994, 116,
4038; (j) J. Plavec, C. Thibaudeau and J. Chattopadhyaya, J. Am.
Chem. Soc., 1994, 116, 6558; (k) C. Thibaudeau, J. Plavec and
J. Chattopadhyaya, J. Am. Chem. Soc., 1994, 116, 8033; (m) J. Plavec,
PhD Thesis, Department of Bioorganic Chemistry, Uppsala
University, Sweden, 1995; (n) J. Plavec, C. Thibaudeau and
J. Chattopadhyaya, Tetrahedron, 1995, 51, 11775; (p) C. Thibaudeau,
J. Plavec and J. Chattopadhyaya, J. Org. Chem., 1996, 61, 266;
(q) J. Chattopadhyaya, Nucleic Acids Symp. Ser., 1996, 35, 111; (r)
J. Plavec, C. Thibaudeau and J. Chattopadhyaya, Pure Appl. Chem.,
1996, 68, 2137; (s) I. Luyten, C. Thibaudeau and J. Chattopadhyaya,
Tetrahedron, 1997, 53, 6433; (t) C. Thibaudeau, A. Földesi and
J. Chattopadhyaya, Tetrahedron, 1997, 53, 14043; (u) C. Thibaudeau,
A. Földesi and J. Chattopadhyaya, Tetrahedron, 1998, 54, 1857; (v)
I. Luyten, C. Thibaudeau and J. Chattopadhyaya, J. Org. Chem.,
1997, 62, 8800; (w) C. Thibaudeau and J. Chattopadhyaya,
Nucleosides Nucleotides, 1998, 17 1589; (x) I. Luyten, J. Matulic-
Adamic, L. Beigelman and J. Chattopadhyaya, Nucleosides
Nucleotides, 1998, 17, 1605; (y) C. Thibaudeau and J. Chatto-
padhyaya, Nucleosides Nucleotides, 1997, 16, 523; (z1) C.
Thibaudeau, J. Plavec and J. Chattopadhyaya, J. Org. Chem., 1998,
63, 4967; (z2) C. Thibaudeau, A. Kumar, S. Bekiroglu, A. Matsuda,
V. E. Marquez and J. Chattopadhyaya, J. Org. Chem., 1998, 63,
5447.
intercept = Ϫ0.16 (σ = 0.02) and Pearson’s correlation coef-
ficient (R) = 0.97 for the latter.
(F) The correlation plot of sugar conformation with the ratio of
2Ј/3Ј-isomers
The plot of the logarithm of the ratio of the mole fractions of
the 3Ј-isomers (3Ј) and the 2Ј-isomer (2Ј) [ln(3Ј)/(2Ј)] as a func-
tion of the logarithm of the ratio of mole fractions of the S-
type conformation (xS) and N-type conformation (xN) [ln(xS)/
(xN)] over the temperature range of 278–333 K for 3Ј-O-
anthraniloyladenosine (1a) and 3Ј-O-anthraniloyladenosine 5Ј-
monophosphate (1b) give straight lines with slope = 0.19
(σ = 0.01), intercept = Ϫ0.16 (σ = 0.02) and Pearson’s corre-
lation coefficient (R) = 0.82 for the former, and slope = 0.18
(σ = 0.01), intercept = 0.15 (σ = 0.02) and Pearson’s correlation
coefficient (R) = 0.96 for the latter.
Acknowledgements
We thank the Swedish Board for Technical Development,
Swedish Natural Science Research (NFR) Council, Swedish
Board for Technical Development (NUTEK), the Deutsche
Forschungsgemeinschaft (Sp 243/5-3) and Fonds der
Chemischen Industrie for generous financial support. Thanks
are due to the Wallenbergstiftelsen, Forskningsra˚dsnämnden
and University of Uppsala for funds for the purchase of 500
MHz and 600 MHz Bruker DRX NMR spectrometers.
10 (a) H. P. M. de Leeuw, C. A. G. Haasnoot and C. Altona, Isr. J.
Chem., 1980, 20, 108; (b) C. Altona and M. Sundaralingam, J. Am.
Chem. Soc., 1972, 94, 8205; (c) C. Altona and M. Sundaralingham,
J. Am. Chem. Soc., 1973, 95, 2333.
11 (a) J. Feigon, A. H. J. Wang, G. A. van der Marel, J. H. van Boom
and A. Rich, Nucleic Acids Res., 1984, 12, 1243; (b) S. Tran-Dinh,
J. Taboury, J.-M. Neumann, T. Huynh-Dinh, B. Genissel, B.
Langlois d’Estaintot and J. Igolen, Biochemistry, 1984, 23, 1362;
(c) P. W. Davis, K. Hall, P. Cruz, I. Tinoco and T. Neilson, Nucleic
Acids Res., 1986, 14, 1279; (d) P. W. Davis, R. W. Adamiak and
I. Tinoco, Biopolymers, 1990, 29, 109; (e) P. Agback, A. Sandstrom,
S.-I. Yamakage, C. Sund, C. Glemarec and J. Chattopadhyaya,
J. Biochem. Biophys. Methods, 1993, 27, 229; ( f ) P. Agback, C.
Glemarec, L. Yin, A. Sandstrom, J. Plavec, C. Sund, S.-I.
Yamakage, G. Wiswanadham, B. Rousse, N. Puri and J. Chatto-
padhyaya, Tetrahedron Lett., 1993, 34, 3929.
12 (a) N. S. Zefirov, Zh. Org. Khim., 1970, 6, 1761; (b) N. S. Zefirov,
L. G. Gurvich, A. S. Shashkov, M. Z. Krimer and E. A. Vorob’eva,
Tetrahedron, 1976, 32, 1211; (c) R. Hoffmann, Acc. Chem. Res.,
1971, 4, 1; (c) N. D. Epiotis, S. Sarkanen, D. Bjorkquist,
L. Bjorkquist and R. Yates, J. Am. Chem. Soc., 1974, 96, 4075;
(d) K. B. Wiberg, Acc. Chem. Res., 1996, 29, 229; (e) W. K. Olson
and J. L. Sussman, J. Am. Chem. Soc., 1982, 104, 270; ( f ) W. K.
Olson, J. Am. Chem. Soc., 1982, 104, 278.
13 (a) F. A. A. M. De Leeuw and C. Altona, J. Comput. Chem., 1983, 4,
428 and PSEUROT, QCPE program No 463; (b) C. A. G. Haasnoot,
F. A. A. M. de Leeuw and C. Altona, Tetrahedron, 1980, 36, 2783.
14 J. E. Kilpatrick, K. S. Pitzer and R. Spitzer, J. Am. Chem. Soc., 1947,
69, 2483.
15 (a) E. Diez, J. S. Fabian, J. Guilleme, C. Altona and L. A. Donders,
Mol. Phys., 1989, 68, 49; (b) L. A. Donders, F. A. A. M. de Leeuw
and C. Altona, Magn. Reson. Chem., 1989, 27, 556; (c) C. Altona,
J. H. Ippel, A. J. A. W. Hoekzema, C. Erkelens, G. Groesbeek and
L. A. Donders, Magn. Reson. Chem., 1989, 27, 564; (d) J. van Wijk,
unpublished results.
References
1 F. Janiak, V. A. Dell, J. K. Abrahamson, B. S. Watson, D. L. Miller
and A. E. Johnson, Biochemistry, 1990, 29, 4268.
2 R. O. Potts, N. C. Ford, Jr. and M. J. Fournier, Biochemistry, 1981,
20, 1653.
3 B. Nawrot, W. Milius, A. Ejchart, S. Limmer and M. Sprinzl,
Nucleic Acids Res., 1997, 25, 948.
4 (a) M. Taiji, S. Yokoyama and T. Miyazawa, Nucleic Acids Res.,
1982, 11, 161; (b) M. Taiji, S. Yokoyama and T. Miyazawa,
Biochemistry, 1985, 98, 1447.
5 P. Nissen, M. Kjeldgaard, S. Thirup, G. Polekhina, L. Reshetnikova,
B. F. C. Clark and J. Nyborg, Science, 1995, 270, 1464.
6 S. Limmer, M. Vogtherr, B. Nawrot, R. Hillenbrand and M. Spinzl,
Angew. Chem., Int. Ed. Engl., 1997, 36, 2485.
7 L. Servillo, C. Balestrieri, L. Quagliuolo, L. Iorio and A. Giovane,
Eur. J. Biochem., 1993, 213, 583.
8 B. Nawrot and M. Sprinzl, Nucleosides Nucleotides, 1998, 17, 815.
9 (a) J. Plavec, W. Tong and J. Chattopadhyaya, J. Am. Chem. Soc.,
1993, 115, 9734; (b) J. Plavec, N. Garg and J. Chattopadhyaya,
J. Chem. Soc., Chem. Commun., 1993, 1011; (c) J. Plavec, L. H.
KooleandJ.Chattopadhyaya,J.Biochem.Biophys.Methods,1992,25,
253; (d) L. H. Koole, H. M. Buck, A. Nyilas and J. Chattopadhyaya,
Can. J. Chem., 1987, 65, 2089; (e) L. H. Koole, H. M. Buck,
H. Bazin and J. Chattopadhyaya, Tetrahedron, 1987, 43, 2289; ( f )
L. H. Koole, J. Plavec, H. Liu, B. R. Vincent, M. R. Dyson, P. L.
Coe, R. T. Walker, G. W. Hardy, S. G. Rahim and J.
Chattopadhyaya, J. Am. Chem. Soc., 1992, 114, 9934; (g) J. Plavec,
C. Thibaudeau, G. Viswanadham, C. Sund and J. Chattopadhyaya,
J. Chem. Soc., Chem. Commun., 1994, 781; (h) C. Thibaudeau,
J. Plavec, K. A. Watanabe and J. Chattopadhyaya, J. Chem. Soc.,
Paper 8/09350D
1536
J. Chem. Soc., Perkin Trans. 2, 1999, 1531–1536