Long-range heteronuclear coupling constants in 2,6-disubstituted purine derivatives

Article abstract of DOI:10.1002/mrc.3806

Four- and five-bond heteronuclear J-couplings between the hydrogen H-8 and carbons C-6 and C-2 in a series of 7- and 9-benzyl substituted purine derivaties with variuous substituents in positions 2 and 6 were studied by coupled ¹³ C NMR and H,C

Full text of DOI:10.1002/mrc.3806

Research Article
Received: 2 January 2012
Revised: 7 February 2012
Accepted: 8 February 2012
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI 10.1002/mrc.3806
Long-range heteronuclear coupling constants
in 2,6-disubstituted purine derivatives
Eliška Procházková,a Lucie Čechová,a Petr Jansaa and Martin Dračínskýa*
Four- and five-bond heteronuclear J-couplings between the hydrogen H-8 and carbons C-6 and C-2 in a series of 7- and
9-benzyl substituted purine derivaties with variuous substituents in positions 2 and 6 were studied by coupled 13 C NMR
and H,C-HMBC experiments and by DFT calculations. We have found that for some of the derivatives, the five-bond
coupling H8-C2 is higher than the four-bond H8-C6 coupling, which is also evidenced by a stronger crosspeak in the HMBC.
This finding contradicts the generally accepted opinion that only strong three-bond crosspeaks and one weak four-bond
H8-C6 crosspeak can be observed in the HMBC spectra of purine derivatives. The misinterpretation of HMBC spectra may
lead to an incorrect determination of the purine derivatives’ structure. Copyright © 2012 John Wiley & Sons, Ltd.
Supporting information may be found in the online version of this article.
1
Keywords: NMR; H; 13 C; long-range coupling; purine derivatives; DFT calculations
used for the assignment of the carbon signals. The H8-C6 cou-
Introduction
pling can be identified in the HMBC spectra as a crosspeak with
lower intensity.[22]
Naturally occurring purines are the basic constituents of nucleic
acids; they also interact with enzymes and other proteins as
components of the cofactors and signal molecules.[1] Adenosine
5’-triphosphate (ATP) controls the energy metabolism of cells,
and nicotinamide adenine dinucleotide (NAD+, NADH) and flavin
adenine dinucleotide (FAD, FADH2) are the key cofactors not only
of the cellular citric acid cycle, which is involved in the cellular
oxidation/reduction processes.[2] Another molecule of biological
relevance is acetyl-coenzyme A, which is of central importance
for metabolism.
In this paper, we present a combined experimental and compu-
tational study of the four- and five-bond heteronuclear couplings
H8-C6 and H8-C2. The experimentally studied compounds are
depicted in Fig. 2. We demonstrate that the five-bond coupling is
of a comparable magnitude with the four-bond coupling and a
structure determination based on the HMBC patterns can lead to
incorrect structures and/or the incorrect assignment of signals.
Purine derivatives bearing diverse types of substituents display
a broad spectrum of biological activities,[3] including an inter-
feron-inducing effect or more often an inhibitory effect against
leukotriene A4 hydrolase, sulfotransferase, phosphodiesterase,
kinase, and other enzymes.[4] Modified nucleosides and nucleotides
are very important classes of compounds used in the therapy of
a wide variety of diseases, since they can act as antiviral,[5–8]
antitumor[9,10] or antimicrobial[11] agents.
The distribution of electrons around the purine skeleton affects
not only its chemical properties and reactivity but also the NMR
parameters. The nature of the substituent is reflected in the
NMR chemical shifts and nuclear spin–spin coupling constants,
which makes NMR spectroscopy an excellent tool for investigat-
ing and interpreting the structure, reactivity and intermolecular
interactions in terms of the electron distribution.[12] The 13C and
15 N NMR chemical shifts as well as 1H–X coupling constants
can be used not only to distinguish between different regio-
isomers,[13] but also to reflect equally well the positions of pro-
tons, which enables the study of the tautomeric equilibria.[2,14–20]
The assignment of the carbon signals is usually done using
three-bond H-C heteronuclear J-coupling.[21] A schematic graph-
ical representation of the ‘building blocks’ for the assignment of
purine derivatives signals with C-6 substituents is given in Fig. 1.
It is generally believed that, for derivatives substituted at both
positions 2 and 6, small four-bond interactions H8-C6 can be
Experiment
The syntheses of compounds 1–4 and 6–7 have been described
previously.[23–25] For the selective preparation of 7-substituted
purines, an alternative synthetic approach could be used.[26]
The preparation of compounds 5, 9 and 10 will be described in
a separate paper with the description of their biological activities.
Briefly, compound 5 was obtained after hydrolysis of compound
1; compound 9 was prepared by methanolysis of compound 6,
and compounds 8 and 10 were prepared by microwave-assisted
reactions of compound 6 with 1 M hydrochloric acid (compound
8) or with dibutylamine (compound 10).
The NMR spectra were measured on a Bruker Avance 600
1
(with H at 600.13 MHz and 13 C at a frequency of 150.92 MHz)
and/or Bruker Avance 500 (with 1H at 499.95 MHz and 13 C at
125.71 MHz) using a 5 mm TXI cryoprobe and about 5–10 mg of
sample in 0.6 ml of CDCl3 or DMSO-d6. The chemical shifts are
given in d-scale (with the H and 13 C referenced to TMS or to
1
*
Correspondence to: Martin Dračínský, Institute of Organic Chemistry and
Biochemistry, Academy of Sciences, Flemingovo náměstí 2, 166 10 Prague,
Czech Republic. E-mail: dracinsky@uochb.cas.cz
a Institute of Organic Chemistry and Biochemistry, Academy of Sciences, Flemingovo
náměstí 2, 166 10 Prague, Czech Republic
Magn. Reson. Chem. 2012, 50, 295–298
Copyright © 2012 John Wiley & Sons, Ltd.

E. Procházková et al.
R6
effects reasonably and that implicit solvent models (e.g. PCM) fail
in the calculation of the solvent effects on NMR parameters.[30,31]
Further discrepancy between the calculated and experimental
data can be caused by the inaccuracy of the DFT calculations
and neglecting vibrational averaging.[32]
The NMR spectra signal assignment (C-1’ - C-4’ correspond to
C-ipso, C-ortho, C-meta and C-para of the benzyl substituent) is
given as follows:
7
N
6
5
4
1
N
H
8
9
2
N
H
N
3
R9
9-Benzyl-6-chloro-9H-purin-2-amine (1): 13 C NMR (150.9 MHz,
DMSO): d = 46.3 (CH2); 123.5 (C-5); 127.4 (C-2’), 128.0 (C-4’), 129.0
(C-3’); 136.8 (C-1’); 143.5 (C-8); 149.7 (C-6); 154.3 (C-4); 160.1
Figure 1. Schematic representation of the ‘HMBC building blocks’
(three-bond coupling pathways) for the assignment of purine derivatives
signals with C-6 substituents.
1
(C-2) ppm. H NMR (600.1MHz, DMSO): d = 5.29 (s, 2H, CH2); 6.92
(bs, 2H, NH2); 7.26 – 7.38 (m, 5H, H-2’, H-3’, H-4’), 8.22 (s, H-8) ppm.
9-Benzyl-2,6-dichloro-9H-purine (2): 13 C NMR (125.7 MHz,
CDCl3): d = 48.0 (CH2); 128.0 (C-2’), 129.0 (C-4’), 129.3 (C-3’); 130.6
(C-5); 133.9 (C-1’); 145.5 (C-8); 151.8 (C-6); 153.1 (C-4 and C-2)
R6
7
N
6
R6
5
1
1
ppm. H NMR (499.9 MHz, CDCl3): d = 5.42 (s, 2H, CH2); 7.32 (m,
N
6
7
8
2H, H-2’); 7.36 – 7.41 (m, 3H, H-3’, H-4’), 8.07 (s, H-8) ppm.
9-Benzyl-6-chloro-9H-purine (3)13 C NMR (125.7 MHz, DMSO):
d = 47.3 (CH2); 127.9 (C-2’), 128.3 (C-4’), 129.0 (C-3’); 131.0 (C-5);
136.2 (C-1’); 147.7 (C-8); 149.4 (C-6); 151.9 (C-2); 152.0 (C-4) ppm.
1H NMR (499.9 MHz, DMSO): d = 5.42 (s, 2H, CH2); 7.32 (m, 2H,
H-2’); 7.36 – 7.41 (m, 3H, H-3’, H-4’), 8.07 (s, H-8) ppm.
5
N
1N
4
9
2
R2
N
N
8
3
4
R2
N
2
N9
3
R2
NH2
2 Cl
R6
Cl
Cl
Cl
NHCH3
OH
R2
Cl
R6
9-Benzyl-N6-methyl-9H-purin-2,6-diamine (4): 13 C NMR
(125.7 MHz, DMSO): d = 27.2 (CH3); 45.6 (CH2); 113.6 (C-5); 127.2
(C-2’), 127.6 (C-4’), 128.8 (C-3’); 137.3 (C-8); 137.9 (C-1’); 151.0
6
1
Cl
NH2
OH
Cl
7
Cl
8
3 H
4
5
OH
OCH3
NBu2
1
9
NH2
NH2
(C-4); 155.6 (C-6); 160.6 (C-2) ppm. H NMR (499.9 MHz, DMSO):
10
Cl
d = 5.42 (s, 2H, CH2); 7.32 (m, 2H, H-2’); 7.36 – 7.41 (m, 3H, H-3’,
H-4’), 8.07 (s, H-8) ppm.
9-Benzylguanine (5): 13 C NMR (150.9 MHz, DMSO): d = 47.8
(CH2); 110.4 (C-5); 128.2 (C-2’), 128.8 (C-4’), 129.5 (C-3’); 135.8
(C-1’); 137.9 (C-8); 150.7 (C-4); 155.0 (C-6); 155.7 (C-2) ppm. 1H
NMR (600.1 MHz, DMSO): d = 5.30 (s, 2H, CH2); 7.05 (bs, 2H, NH2);
7.28 – 7.40 (m, 5H, H-2’, H-3’, H-4’), 8.87 (s, H-8) ppm.
Figure 2. The studied compounds and their numbering.
DMSO using d(DMSO) 2.50 and 39.7 ppm, respectively). The typi-
cal experimental conditions for the 1H NMR spectra were 32
scans, a spectral width of 6 kHz, and an acquisition time of 5 s,
yielding 60 K data points. The FIDs were zero-filled to 128 K data
points. The coupled 13 C NMR spectra were acquired using a
gated decoupling pulse sequence (decoupling during d1). The
13 C spectra were also acquired with selective decoupling of
H-8. The coupled 13 C and selectively decoupled 13 C experiments
(30 mg of a sample dissolved in 0.6 ml of a solvent) were acquired
with a spectral width of 70 ppm, offset of 135 ppm, d1 = 3 s, ac-
quisition time of 3.7 s yielding 64 K data points, which were
zero-filled to 128 K data points. The 2D-homonuclear (H,H-COSY)
and 2D-heteronuclear (H,C-HSQC and H,C-HMBC) experiments
were performed for the structural assignments of the 1H and
13 C signals (using standard 2D-NMR pulse sequences of Bruker
software).
The geometry optimizations and NMR parameters calculations
were conducted using the Gaussian 09 software package.[27] All
of the structures were optimized at the DFT level of theory using
the B3LYP functional.[28,29] For geometry optimization, the stan-
dard 6-31 + G(d,p) basis set was used, and for the shielding and
coupling constants, a slightly larger 6-311++G(d,p) basis set was
utilized. The vibrational frequencies and free energies were calcu-
lated for all of the optimized structures, and the stationary-point
character (a minimum) was thus confirmed. The optimizations
and shielding- and coupling-constant calculations were done in
vacuum. We have previously explored various approaches for
modeling the solvent effects on the calculation of NMR param-
eters. We have shown that a large number of molecular dynamics
snapshot geometries have to be averaged to model the solvent
7-Benzyl-2,6-dichloro-7H-purine (6): 13 C NMR (125.7 MHz,
CDCl3): d = 50.9 (CH2); 121.7 (C-5); 127.0 (C-2’), 129.0 (C-4’), 129.4
(C-3’); 134.1 (C-1’); 143.9 (C-6); 150.4 (C-8); 153.2 (C-2); 163.5
1
(C-4) ppm. H NMR (499.9 MHz, CDCl3): d = 5.69 (s, 2H, CH2); 7.18
(m, 2H, H-2’); 7.36 – 7.42 (m, 3H, H-3’, H-4’), 8.27 (s, H-8) ppm.
7-Benzyl-6-chloro-7H-purin-2-amine (7): 13 C NMR (125.7 MHz,
DMSO): d = 49.4 (CH2); 115.0 (C-5); 126.6 (C-2’), 128.0 (C-4’), 129.0
(C-3’); 137.4 (C-1’); 142.6 (C-6); 150.2 (C-8); 160.3 (C-2); 164.6
(C-4) ppm. 1H NMR (499.9 MHz, DMSO): d = 5.56 (s, 2H, CH2);
6.64 (bs, 2H, NH2); 7.13 (m, 2H, H-2’); 7.27 -7.36 (m, 3H, H-3’,
H-4’), 8.55 (s, H-8) ppm.
7-Benzylxanthine (8): 13 C NMR (125.7 MHz, DMSO): d = 49.0
(CH2); 106.2 (C-5); 127.7 (C-2’), 128.1 (C-4’), 128.8 (C-3’); 137.3
(C-1’); 142.8 (C-8); 149.7 (C-4); 151.4 (C-2); 155.7 (C-6) ppm. 1H
NMR (499.9 MHz, DMSO): d = 5.41 (s, 2H, CH2); 7.27 – 7.36 (m, 5H,
H-2’, H-3’, H-4’), 8.13 (s, H-8); 10.87 (bs, 1H) and 11.58 (bs, 1H, H-1
and H-3) ppm.
7-Benzyl-2-chloro-6-methoxy-7H-purin (9): 13 C NMR (125.7MHz,
DMSO): d = 50.2 (CH2); 55.1 (CH3); 111.8 (C-5); 127.5 (C-2’), 128.2
(C-4’), 128.9 (C-3’); 136.8 (C-1’); 148.5 (C-8); 151.3 (C-2); 157.5 (C-6);
162.8 (C-4) ppm. 1H NMR (499.9MHz, DMSO): d = 4.05 (s, 3H, CH3);
5.54 (s, 2H, CH2); 7.27 – 7.31 (m, 5H, H-2’, H-4’), 7.35 (m, 2H, H-3’);
8.73 (s, H-8) ppm.
7-Benzyl-N,N-dibutyl-2-chloro-7H-purin-6-amine (10): 13 C NMR
(125.7 MHz, DMSO): d = 13.8 (C-4”); 19.5 (C-3”); 29.1 (C-2”); 49.5
(C-1”); 50.9 (CH2); 113.8 (C-5); 126.3 (C-2’), 128.0 (C-4’), 128.9 (C-3’);
137.0 (C-1’); 149.9 (C-8); 151.9 (C-2); 155.6 (C-6); 163.6 (C-4) ppm.
wileyonlinelibrary.com/journal/mrc
Copyright © 2012 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2012, 50, 295–298

Heteronuclear coupling constants in purine derivatives
1H NMR (499.9MHz, DMSO): d = 0.75 (t, 6H, J4”,3” = 7.3, H-4”); 1.01
(m, 4H, H-3”); 1.35 (m, 4H, H-2”); 3.39 (m, 4H, H-1”); 5.56 (s, 2H,
CH2); 6.97 (m, 2H, H-2’); 7.24 – 7.32 (m, 3H, H-3’, H-4’), 8.62 (s, H-
8) ppm.
The uncertainty in the signal assignment based on the HMBC
patterns may lead to incorrect structure determination. For some
7-substituted purine derivatives, the spatial proximity obtained
from the NOE experiments can be used to determine the struc-
ture unequivocally. This approach is demonstrated on compound
10. In the ROESY spectrum of compound 10, we observed cross-
peaks between the benzyl part of the molecule and the dibutyl-
amino part, which is possible only for the 7-benzyl-6-dibutylamino
derivative. In some cases, the assignment of the carbon signal
could be obtained from the HMBC spectra observing the cross-
peaks of the NH2, HNCH3 or OCH3 groups. For some derivatives,
the spectral assignment could be done only with the help of
calculated chemical shifts.
We performed geometry optimization and shielding- and
coupling-constant calculations for a series of purine derivatives
with various substituents in positions 2 and 6 and a methyl
substituent in position 7 or 9. We believe that the change of
the benzyl substituent to a methyl group can cause only minor
differences in the calculated coupling constants, while the
calculations are much less demanding. We confirmed this for
compound 6. The differences in the calculated H8-C2 and
H8-C6 coupling constants between compound 6 (7-benzylderi-
vative) and 2,6-dichloro-7-methyl-7H-purine were lower than
0.01 Hz.
Results and Discussion
In a typical HMBC spectrum of 2,6-disubstituted purine deriva-
tives (refer to Fig. 3), we can observe strong three-bond cross-
peaks between the hydrogen H-8 and carbons C-4 and C-5. The
strong crosspeaks are accompanied by one or two weak cross-
peaks corresponding to H8-C6 and/or H8-C2. The stronger cross-
peak from these two low-intensity crosspeaks is believed to be
the H8-C6 correlation.[22] However, we have observed that for
many purine derivatives, the long-range crosspeaks are of a
comparable magnitude, and in some cases, the magnitude of
the H8-C2 crosspeak is larger than that of the H8-C6 crosspeak.
The intensity of a HMBC crosspeak is dependent on the match
between heteronuclear coupling constant and the experimental
delay of the anti-phase evolution (1/2 J). For heteronuclear cou-
plings lower than 1 Hz, the evolution delay should be larger than
500 ms, which leads to a strong signal suppression by relaxation.
We performed optimization of the evolution delay and we have
observed that for longer delays, the gain of intensity of the weak
crosspeaks was overridden by the relaxation losses. Therefore, all
the HMBC spectra reported in this paper were measured with
evolution delay of 50 ms. Using this setup, the intensity of the
long-range HMBC crosspeaks is mainly governed by the value
of the J-coupling. The HMBC spectra of all of the studied com-
pounds are shown in Supporting information.
The long-range heteronuclear coupling can be observed in
hydrogen-coupled 13 C spectra. Because of the small values of the
four- and five-bond coupling constants, line narrowing and longer
acquisition times are necessary, which makes this technique possi-
ble only for concentrated samples or 13 C labeled compounds. We
measured the coupled 13 C spectra of compounds 2, 5 and 6. A part
of the coupled spectrum of compound 5 depicting the signals of
C-2 and C-6 is shown in Figure S11 in the SI. For comparison, the
spectrum obtained with a selective decoupling of H-8 is also shown
in the figure. The experimental values of the long-range coupling
constants obtained from the coupled 13 C spectra are discussed
below. Given the experimental difficulties in obtaining the coupled
spectra, we estimated qualitatively the couplings and their relative
intensities from the HMBC spectra.
Both the calculated J(H8-C2) and J(H8-C6) were always posi-
tive. Interestingly, the calculated values of J(H2-C8) were always
negative, and the absolute values were close to 0.2 Hz, which is
in agreement with the lack of crosspeaks between H-2 and
C-8 in the HMBC spectra of the purine derivatives. For all deriva-
tives with the same substituents in both positions 2 and 6 except
for the dihydroxyderivatives, the calculated chemical shift of C-6
was lower than that of C-2 by 1–6 ppm for 9-methyl- and by
11–19 ppm for the 7-methylderivatives.
In the HMBC spectrum of compound 1 (Fig. 3), we observed
weak crosspeaks to both C-2 and C-6 with the C-2 crosspeak
being slightly more intensive. This observation is in agreement
with the calculated coupling constants (Table 1, Entry 8), where
the H8-C2 coupling was 0.61 Hz compared to 0.32 Hz of the
H8-C6 coupling. The signals of the carbon atoms C2 and C4 are
unfortunately overlapped in the spectrum of compound 2, and
therefore, the H8-C2 crosspeak (if present) is overlapped by the
strong H8-C4 crosspeak in the HMBC spectrum. We measured
the coupled 13 C NMR spectrum and ascertained that the experi-
mental long-range couplings 0.4 and 0.3 Hz were very close to
the calculated values (0.52 and 0.19 Hz, Table 1, Entry 4). In agree-
ment with the small value of the H8-C6 coupling constant, no
H8-C6 crosspeak was observed in the HMBC spectrum. Similarly,
in the HMBC spectrum of compound 3, we did not observe any
H8-C6 crosspeak, and the H8-C2 crosspeak (if present) was over-
lapped by the strong H8-C4 signal. This finding is consistent with
the rather small calculated couplings (Table 1, Entry 3). In the
HMBC spectra of compound 4, the H8-C2 crosspeak was weaker
than the H8-C6 crosspeak, which is again in agreement with the
calculated coupling constants. We measured the coupled 13 C
NMR spectrum of compound 5, and the experimental coupling
constants (1.2 Hz for the H8-C6 and 0.4 for the H8-C2 coupling)
were in excellent agreement with the calculated values 0.92
and 0.34 Hz, respectively (Table 1, Entry 10). In addition, the
H8-C2 crosspeak was missing in the HMBC spectrum.
We obtained the experimental H8-C2 and H8-C6 coupling con-
stants of compound 6 from the coupled 13 C spectrum, and again,
the calculated values (0.86 and 0.73 Hz, Table 1, Entry 4) agreed
Figure 3. The H-8 region of the HMBC spectrum of compound 1.
Magn. Reson. Chem. 2012, 50, 295–298
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/mrc

E. Procházková et al.
Table 1. The calculated heteronuclear coupling constants J(H8-C2)
and J(H8-C6) in substituted 9- and 7-methylpurines
and signal assignment could be also the comparison of calcu-
lated and experimental chemical shifts.[20]
9-isomer
7-isomer
Acknowledgement
Entry
R2
R6
J(H8-C2) J(H8-C6) J(H8-C2) J(H8-C6)
We are grateful to the Grant Agency of Academy of Sciences of the
Czech Republic (Project KJB400550903) for supporting this work.
1
H
H
0.32
0.48
0.35
0.52
0.35
0.57
0.57
0.61
0.49
0.34
0.36
0.16
0.49
0.54
0.56
0.36
0.34
0.21
0.19
0.55
0.46
0.62
0.32
0.58
0.92
0.75
0.85
0.44
0.64
0.66
0.61
0.79
0.65
0.86
0.60
0.68
0.68
0.75
0.76
0.41
0.27
0.36
0.80
0.79
0.67
0.93
0.89
0.89
0.73
0.81
0.96
0.84
0.79
0.80
1.2
2
Cl
H
References
3
H
Cl
Cl
4
Cl
[1] H. Rosemeyer. Chem. Biodivers. 2004, 1, 361–401.
[2] T. Bartl, Z. Zacharová, P. Sečkářová, E. Kolehmainen, R. Marek. Eur. J.
Org. Chem. 2009, 1377–1383.
[3] I. Collins, J. J. Caldwell, 10.11 - Bicyclic 5–6 Systems: Purines, in Com-
prehensive Heterocyclic Chemistry III, vol. 10 (Eds: A. R. Katritzky, C. A.
Ramsden, E. F. V. Scriven, J. K. Taylor), Elsevier, Oxford, 2008 pp.
525–597.
[4] M. Legraverend, D. S. Grierson. Bioorg. Med. Chem. 2006, 14, 3987–4006.
[5] E. De Clercq, A. Holý, I. Rosenberg, T. Sakuma, J. Balzarini, P. C. Maudgal.
Nature 1986, 323, 464–467.
5
H
NH2
6
NH2
NH2
NH2
Cl
H
7
NH2
8
Cl
9
NH-CH3
OHa
OHa
OHa
10
11
12
13
14
15
NH2
OHa
H
1.1
1.1
[6] A. Holý. Curr. Pharm. Des. 2003, 9, 2567–2592.
Cl
OCH3
N(CH3)2
NH-CH3
0.73
0.74
0.86
[7] E. De Clercq, A. Holý. Nat. Rev. Drug Discovery 2005, 4, 928–940.
[8] C. Simons, Q. P. Wu, T. T. Htar. Curr. Top. Med. Chem. 2005, 5, 1191–1203.
[9] M. Kidwai, R. Venkataramanan, R. Mohan, P. Sapra. Curr. Med. Chem.
2002, 9, 1209–1228.
[10] N. R. Kode, S. Phadtare. Molecules 2011, 16, 5840–5860.
[11] J. N. Kim, K. F. Blount, I. Puskarz, J. Lim, K. H. Link, R. R. Breaker. ACS
Chem. Biol. 2009, 4, 915–927.
Cl
NH2
aKeto forms of the guanine, hypoxanthine and xanthine bases were
used in the calculations.
[12] S. Standara, K. Maliňáková, R. Marek, J. Marek, M. Hocek, J. Vaara,
M. Straka. Phys. Chem. Chem. Phys. 2010, 12, 5126–5139.
[13] R. Marek, J. Brus, J. Toušek, L. Kovacs, D. Hocková. Magn. Res. Chem.
2002, 40, 353–360.
[14] O. Tsikouris, T. Bartl, J. Tousek, N. Lougiakis, T. Tite, P. Marakos,
N. Pouli, E. Mikros, R. Marek. Magn. Reson. Chem. 2008, 46, 643–649.
[15] M. T. Chenon, R. J. Pugmire, D. M. Grant, R. P. Panzica, L. B. Townsend.
J. Am. Chem. Soc. 1975, 97, 4627–4636.
[16] J. Kongsted, K. Aidas, K. V. Mikkelsen. Phys. Chem. Chem. Phys. 2010,
12, 761–768.
[17] N. C. Gonnella, H. Nakanishi, J. B. Holtwick, D. S. Horowitz, K. Kanamori,
N. J. Leonard, J. D. Roberts. J. Am. Chem. Soc. 1983, 105, 2050–2055.
[18] M. Dračínský, P. Jansa, J. Chocholoušová, J. Vacek, S. Kovačková,
A. Holý. Eur. J. Org. Chem. 2010.
[19] M. Dračínský, P. Jansa, K. Ahonen, M. Buděšínský. Eur. J. Org. Chem.
2011, 1544–1551.
[20] E. Procházková, M. Šála, R. Nencka, M. Dračínský. Magn. Res. Chem.
2012 in press. DOI: 10.1002/mrc.2864.
[21] T. Dieckmann, J. Feigon. Curr. Opin. Struc. Biol. 1994, 4, 745–749.
[22] R. Marek, V. Sklenář. Annu. Rep. NMR Spectrosc. 2004, 54, 201–242.
[23] A. Holý, J. Günter, H. Dvořáková, M. Masojídková, G. Andrei, R. Snoeck,
J. Balzarini, E. De Clercq. J. Med. Chem. 1999, 42, 2064–2086.
[24] C. Dalby, C. Bleasdale, W. Clegg, M. R. J. Elsegood, B. T. Golding, R. J. Griffin.
Angew. Chem. Int. Ed. 1993, 32, 1696–1697.
excellently with the experimental ones (0.8 and 0.7 Hz). Cross-
peaks of equal intensity were observed in the HMBC spectrum
of this compound. Similarly, in the HMBC spectra of compounds
7, 9 and 10, both crosspeaks H8-C2 and H8-C6 were observed,
and their intensities were almost identical, which is in line with
the calculated coupling constant values 0.7–0.8 Hz for all the cou-
plings. In contrast to that the calculated value of the H8-C2
coupling constant (0.27 Hz) in compound 8 is much smaller than
the H8-C6 coupling constant (1.1 Hz), which is, indeed, mani-
fested in the HMBC spectrum by a missing H8-C2 crosspeak.
Based on a careful analysis of the calculated long-range cou-
pling constants, we can conclude that the value of H8-C2 cou-
pling depends primarily on the nature of the substituent in the
position 2. The following order of the coupling values depending
on the C-2 substituent was observed: H < Cl < NH2. Exceptions
from this rule are the keto forms of guanine, hypoxanthine and
xanthine, which have always very low H8-C2 coupling. Similarly,
the H8-C6 coupling value is dominated by the nature of the C-6
substituent, with the keto forms of guanine, hypoxanthine and xan-
thine having the highest values of the coupling. 7-methylisomers
always have both H8-C2 and H8-C6 coupling constants values
higher than the 9-methylisomers.
[25] L. Čechová, P. Jansa, M. Šála, M. Dračínský, A. Holý, Z. Janeba. Tetra-
hedron 2011, 67, 866–871.
[26] V. Kotek, N. Chudikova, T. Tobrman, D. Dvorak. Org. Lett. 2010, 12,
5724–5727.
[27] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R.
Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H.
Nakatsuji, X. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G.
Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J.
Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T.
Vreven, J. J. A. Montgomery, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J.
Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J.
Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi,
M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C.
Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin,
R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G.
Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D.
Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox,
Gaussian 09, Revision A.02, Gaussian, Inc., Wallingford CT, 2009.
[28] A. D. Becke. J. Chem. Phys. 1993, 98, 5648–5652.
Conclusions
We have demonstrated both experimentally and by DFT calcula-
tions that the four- and five-bond heteronuclear J-couplings of
the hydrogen H-8 with carbon atoms C-6 and C-2 may be of a
comparable size. Depending on the substituents attached to
the purine skeleton, the values of the coupling constants can
change significantly. For proper structure determination and
signal assignment, care must be taken with this issue and one
cannot rely on small crosspeaks in the HMBC spectra, which can
be caused by both H8-C2 and H8-C6 interactions. The DFT calcu-
lated coupling constants were shown to agree very well with the
experimental data. An alternative for the structure determination
[29] C. T. Lee, W. T. Yang, R. G. Parr. Phys. Rev. B 1988, 37, 785–789.
[30] M. Dračínský, P. Bouř. J. Chem. Theory Comput. 2010, 6, 288–299.
[31] M. Dračínský, J. Kaminský, P. Bouř. J. Phys. Chem. B 2009, 113, 14698–14707.
[32] M. Dračínský, J. Kaminský, P. Bouř. J. Chem. Phys. 2009, 130, 094106.
wileyonlinelibrary.com/journal/mrc
Copyright © 2012 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2012, 50, 295–298