2'-Deoxy-7-(hydroxymethyl)-7-deazaadenosine: A new analogue to model structural water in the major groove of DNA

Article abstract of DOI:10.1021/ja961540s

A novel deoxynucleoside, 2'-deoxy-7-(hydroxymethyl)-7-deazaadenosine, was synthesized with the intent of using the analogue to mimic the role of 'structural' waters in the major proove of DNA. The target compound was synthesized in four steps starting fro

Full text of DOI:10.1021/ja961540s

J. Am. Chem. Soc. 1996, 118, 10065-10068
10065
2′-Deoxy-7-(hydroxymethyl)-7-deazaadenosine: A New
Analogue to Model Structural Water in the Major Groove of
DNA§
Jason K. Rockhill,†,‡ Scott R. Wilson,‡ and Richard I. Gumport*,†,‡
Contribution from the College of Medicine and Department of Biochemistry and School of
Chemical Sciences, UniVersity of Illinois, 600 South Mathews AVenue, Urbana, Illinois 61801
ReceiVed May 8, 1996X
Abstract: A novel deoxynucleoside, 2′-deoxy-7-(hydroxymethyl)-7-deazaadenosine, was synthesized with the intent
of using the analogue to mimic the role of “structural” waters in the major groove of DNA. The target compound
was synthesized in four steps starting from 2-((ethoxymethylene)amino)-5-bromo-1-(2′-deoxy-3′,5′-di-O-p-toluoyl-
â-D-erythro-pentofuranosyl)pyrrole-3,4-dicarbonitrile. The structure was characterized by X-ray diffraction and proton
NMR spectroscopic analyses. The crystal structure shows an intramolecular hydrogen bond between an amino proton
on N6 and the oxygen of the hydroxymethyl group. When superimposed onto particular adenines in the structure
of the tryptophan (trp) repressor/operator complex, the analogue places the oxygen of the hydroxymethyl group very
near the oxygen contributed by the water in the protein/DNA complex. This analogue may be useful for probing the
role of structural waters in other specific protein/DNA complexes and in DNA bending.
Introduction
Water is thought to play an important role in the specific in-
teractions between proteins and deoxyribonucleic acids and in
DNA bending dynamics. Several structures of proteins com-
plexed with their target DNA sequences, determined both by
X-ray diffraction1a-e and NMR1f spectroscopic analyses, show
“fixed” water molecules acting as bridges between amino acids
and hydrogen bonding sites on the target deoxyribonucleic acid.
A general structural motif for these water-mediated contacts with
adenine is hydrogen bonding to the N7 of the purine ring and
the exocyclic amino group (Figure 1A). In addition, molecular
modeling predicted similar “structural” water molecules bridging
specific hydrogen bond donors and acceptors in d(A)5 tracts in
DNA.2 These waters were postulated to contribute to the
bending of DNA containing such poly d(A)n tracts.
Figure 1. (A) Water-mediated contact with 2′-deoxyadenosine residue
(see text) and (B) hm7c7dA, water-mediating mimic.
incorporate the analogue into oligodeoxyribonucleotides and to
test its effects on the postulated interactions.
To examine these questions, we synthesized a 7-deazapurine
analogue, 4-amino-7-(2′-deoxy-â-D-ribofuranosyl)-5-(hydroxy-
methyl)pyrrolo[2,3-d]pyrimidine {2′-deoxy-7-(hydroxymethyl)-
7-deazaadenosine (hm7c7dA) (1)}, and characterized its structure
by X-ray diffraction and NMR analyses. Modeling this
analogue into existing structures indicates that it will place an
oxygen atom very close to the position oxygen occupies in the
“fixed” waters of several protein/DNA structures1a-f or of
modeled structures of bent DNA.2 Our ultimate aim is to
Results and Discussion
Synthesis. The synthetic strategy is based on previously
synthesized 7-deazaadenosine compounds,3,4 such as toyoca-
mycin3a (Scheme 1). We synthesized the title compound 1 in
four steps starting with the previously reported â anomer of
glycosylated pyrrole 2.3a Previously, cyclization using NH3/
MeOH provided the pyrrolo[2,3-d]pyrimidine ring system along
with deprotection of the 5′ and 3′ hydroxyls.3a Minor modifica-
tion to this method by reducing the temperature to 4 °C and
the time of reaction to under 2 h yielded the desired cyclized
product 3 without significant deprotection of the 5′ and 3′
hydroxyls. The progress of the reaction was followed by thin
layer chromatography (TLC) (2% MeOH/CH2Cl2 Rf ) 0.30),
and the reaction was stopped before the 5′ and 3′ toluoylic esters
were cleaved. Debromination to 4 was cleanly accomplished
with Pd/C in dioxane with triethylamine and H2 at 50 °C.5 The
nitrile 4 was reductively cleaved to the aldehyde using Raney
† College of Medicine and Department of Biochemistry.
‡ School of Chemical Sciences.
§ A preliminary report of a portion of this work was presented as a poster
at the ASBMB Annual Meeting, New Orleans, LA, June 3-6, 1996.
Rockhill, J. K.; Gumport, R. I. FASEB J. 1996, 10 (6), A1494, no. 2851.
X Abstract published in AdVance ACS Abstracts, September 1, 1996.
(1) (a) Shakked, Z.; Guzikevich-Guerstein, G.; Frolow, F.; Rabinovich,
D.; Joachimiak, A.; Sigler, P. B. Nature 1994, 368, 469-473. (b) Zhang,
H.; Daqing, Z.; Revington, M.; Lee, W.; Jia, X.; Arrowsmith, C.; Jardetzky,
O. J. Mol. Biol. 1994, 238, 592-614. (c) Lin, C. H.; Beale, J. M.; Hurley,
L. H. Biochemistry 1991, 30, 3597-3602. (d) Clark, K. L.; Halay, E. D.;
Lai, E.; Burley, S. K. Nature 1993, 364, 412-420. (e) Feng, J.; Johnson,
R. C.; Dickerson, R. E. Science 1994, 263, 348-355. (f) Qian, Y. Q.;
Otting, G.; Wu¨thrich, K. J. Am. Chem. Soc. 1993, 115, 1189-1190.
(2) (a) Teplukhin, A. V.; Poltev, V. I.; Jernigan, R. L.; Zhurkin, V. B.
Biophys. J. 1995, 68, a357 [abstract]. (b) Teplukhin, A. V.; Zhurkin, V.
B.; Jernigan, R. L.; Poltev, V. I. Mol. Biol. (Engl. transl.) 1996, 30,
75-84.
(3) (a) Bhattacharya, B. K.; Rao, T. S.; Revankar, G. R. J. Chem. Soc.,
Perkin Trans. 1 1995, no. 21, 1543-1550. (b) Ramasamy, K.; Robins, R.
K.; Revankar, G. R. Tetrahedron 1986, 42, 5869-5878.
(4) Swayze, E. E.; Shannon, W. M.; Buckheit, R. W.; Wotring, L. L.;
Drach, J. C.; Townsend, L. B. Nucleosides Nucleotides 1992, 11, (8), 1507-
1527.
S0002-7863(96)01540-5 CCC: $12.00 © 1996 American Chemical Society

10066 J. Am. Chem. Soc., Vol. 118, No. 42, 1996
Rockhill et al.
Scheme 1
Figure 2. SHELXTL (Siemens, 1994) plot showing 35% probability
ellipsoids for non-H atoms and circles of arbitrary size for H atoms.
Dashed lines show H-bonding scheme. Equivalent positions O(1′A)
Ni and sodium hypophosphate in a buffered solution (pyridine/
acetic acid/water 2:1:1).6 The aldehyde was reduced, and the
toluoyl protecting groups were removed using an excess (4
equiv) of sodium borohydride.7 Nucleoside 1 was isolated in
the aqueous layer after extraction of unreacted starting material
and byproducts with ethyl acetate. After the solution was
concentrated under reduced pressure, compound 1 and the
associated inorganic salts were slurried with methanol (two times
for 30 min) and filtered to remove the inorganic salts. The
compound was further purified by column chromatography and
recrystallization. The low yield (7.0%) for the last step may
be attributed to the multiple manipulations necessary to obtain
pure product. The yield of the nucleoside, per se, might be
increased if the reduction and deprotection were performed
separately.
1
3
at /2 + x, /2 - y, -z; N(3A) at x, 1 + y, z; O(12A) at 1 - x, -1/2 +
y, 1/2 - z; and N(1A) at -1/2 + x, 3/2 - y, -z complete the H-bonding
scheme. The O(11)-N(6) distance is 3.153(8) Å. The N(6)-H(62)-
O(11) angle is 157(7)°.
Table 1. Distances (Å) and Angles (deg) for Hydrogen Bond
Interactions Involving the Amino Group and the Hydroxylmethyl
Group
atoms
distance (Å)b
atoms
angles (deg)b
H62-O11
H61-O1′ a
H121-O11
H11-N3a
2.18 (0.10)
2.24 (0.09)
2.14 (0.07)
1.95 (0.01)
N6-H61-O11
N6-H62-O1′ a
O12-H121-O11
O11-H11-N3 a
157 (7)
155 (7)
161 (7)
172 (7)
a Hydrogen bonds are with another monomer in the unit cell. b Errors
are given in parentheses.
Structural Determination. The structure of hm7c7dA was
determined by X-ray diffraction analysis (Figure 2). The
structure revealed a possible intramolecular hydrogen bond
between the amino and hydroxymethyl groups. The distance
between the exocyclic amino group hydrogen (H62) and the
oxygen (O11) of the hydroxymethyl group is 2.18((0.10) Å
with an angle of 157((7)°. This distance is close enough to
allow hydrogen bonding between the hydroxymethyl group and
the exocyclic amino group. An extensive hydrogen-bonding
network exists within the unit cell. The hydroxyl group (O11,
H11) is involved in other hydrogen bond interactions: the N3
nitrogen of another nucleoside monomer and a water molecule.8
(Table 1 lists distances and angles for all hydrogen bonds
involving the amino (H61, H62) and hydroxymethyl (O11, H11)
groups.) This extensive hydrogen-bonding network may con-
tribute to the observed conformation.
Preliminary studies of the interaction between the hydroxy-
methyl group and the amino group in solution were performed
by proton NMR spectroscopy. The water-exchangeable pro-
ton spectrum of hm7c7cA revealed neither a downfield shift
nor a splitting of the amino protons (H61 and H62), suggest-
ing that interaction between the hydroxymethyl group and the
amino protons is transient on the NMR time scale.9 However,
the residence half-life of the exocyclic amino protons from
hm7c7dA was about half that of the corresponding amino protons
in 2′-deoxyadenosine. Assuming a first-order process, this small
2-fold rate enhancement may result from the hydroxyl group
acting as a base to facilitate proton exchange with solvent.
The results of the structural determination by X-ray crystal-
lography and NMR provide support for the existence of an
intramolecular hydrogen bond. The proton exchange data from
the solution studies suggest that the hydrogen bond between
H62 and O11 exists in the absence of the other hydrogen bond
interactions involving O11 that are observed in the crystal
structure. If the hydrogen bond between the amino proton and
the oxygen (O11) had been fixed, the amino protons signals
would have split. Since this was not the case, the hydroxy-
methyl group is likely not restricted to one conformation in
solution. The observed water-mediated contacts with adenine
residues in the X-ray crystal structures of protein/DNA
complexes,1a-e place the waters above, in, and below the plane
of the purine ring. The mobility of the hydroxymethyl may
allow it to adopt an energetically favorable fit when substituted
into particular protein/water/adenine complexes.
We used the molecular modeling program QUANTA10 to
compare the structure of hm7c7dA to the corresponding adenine
residues in the trp repressor/operator complex (PDB no. 1TRO)
that have water-mediated contacts between the protein and the
DNA.11 The analogue was superimposed onto the relevant
adenines (A5 and A-7) by aligning corresponding purine ring
atoms (N1, C2, N3, C4, C5, C6, C8, and N9). The resulting
superimposed structures are portrayed in Figure 3. We note
(5) Sharma, M.; Bloch, A.; Bobek, M. Nucleosides Nucleotides 1993,
12, (6), 643-648.
(6) Flanagan, D. M.; Joullie´, M. M. Synth. Commun. 1990, 20, 459-
467.
(7) Brown, M. S.; Rapoport, H. J. Org. Chem. 1963, 28, 3261-3263.
(8) A complete listing of all hydrogen bonds within the unit cell is
included in the supporting information.
(10) The calculations were done using QUANTA 4.1, which is available
from Molecular Simulation, Inc., San Diego, CA.
(9) The water exchange spectra were obtained in 90% H2O/D2O at 2
°C. The experimental details and a spectrum of the region of interest are
included in supporting information.
(11) Otwinowski, Z.; Schevitz, R. W.; Zhang, R.-G.; Lawson, C. L.;
Joachimiak, A. J.; Marmorstein, R.; Luisi, B. F.; Sigler, P. B. Nature 1988,
335, 321.

2′-Deoxy-7-(hydroxymethyl)-7-deazaadenosine
J. Am. Chem. Soc., Vol. 118, No. 42, 1996 10067
Conclusion
We synthesized 2′-deoxy-7-(hydroxymethyl)-7-deazaadenos-
ine. We are currently incorporating the analogue into DNA,
and we plan to examine its effects on the trp repressor/operator
interactions and on the bending of DNA. In summary, the
structure of the analogue suggests that it may serve as a useful
probe for “structural” water in adenine-containing biomolecules
and their complexes.
Experimental Section
General Considerations. Most of the chemicals were purchased
from Aldrich Chemical Co. and used without further purification. The
2′-deoxyadenosine for the proton exchange NMR was purchased from
Sigma. 1H NMR spectra were recorded on either a GE-300 (300 MHz)
or Varian Unity-500 (500 MHz) spectrometer. Chemical shifts are
reported relative to TMS in CDCl3 or the residual DMSO peaks in
DMSO-d6. Reported melting points were taken on a Thomas Hoover
melting apparatus and are uncorrected. Analytical thin layer chroma-
tography was performed on Analtech silica gel plates with F-254
indicator. Column chromatography was carried out using Merck silica
gel (50-200 µm) under 3-5 psi of air. Mass spectroscopy was
performed by the University of Illinois Mass Spectrometry Laboratory,
and elemental analyses were performed by the University of Illinois
Microanalytical Laboratory.
4-Amino-6-bromo-7-(2′-deoxy-3′,5′-di-O-p-toluoyl-â-D-erythro-
pentofuranosyl)pyrrolo[2,3-d]pyrimidine-5-carbonitrile (3). A satu-
rated solution of ammonia in MeOH was cooled to -78 °C. To this
was added 1.57 g of 2-((ethoxymethylene)amino)-5-bromo-1-(2′-deoxy-
3′,5′-di-O-p-toluoyl-â-D-erythro-pentofuranosyl)pyrrole-3,4-dicarboni-
trile (2).3a The solution was stirred and warmed to 4 °C. The reaction
was allowed to proceed until completion as determined by TLC (Rf )
0.30 in 2% MeOH/CH2Cl2). The solution was then cooled to -78 °C,
and the precipitation of a white solid was observed. The solid 3 was
removed by filtration and used without further purification (1.04 g,
70.0%): mp 188-189 °C; 1H NMR (300 MHz, CDCl3) δ 2.39 (3H, s,
Ph-CH3), 2.43 (3H, s, Ph-CH3), 2.60 (1H, m, H2′), 3.90 (1H, m, H2′),
4.60 (1H, m, H4′), 4.65, 4.80 (2H, m, H5), 5.53 (2H, bs, NH2), 6.01
(1H, m, H3′), 6.58 (1H, t, J ) 6.6 Hz, H1′), 7.18-7.28, 7.90-7.97
(8H, m, 2 Ph), 8.25 (1H, s, H8); MS (FAB+), m/e 592, 512 (-Br).
Anal. Calcd for C28H24N5O5Br‚1/2H2O: C, 56.10; H, 4.20; N, 11.69;
Br, 13.33. Found: C, 55.98; H, 4.16; N, 11.42; Br, 13.60.
A larger-scale synthesis of compound 3 was completed using a crude
preparation of the glycosylated pyrrole 2.3a The saturated ammonia/
MeOH (100 mL) was cooled to 0 °C. The cooled ammonia/MeOH
was transferred to a 250 mL round bottom flask that contained
approximately 30 g of a crude, dried mixture containing compound 2
(cooled to 0 °C). The stirred reaction was allowed to warm to 4 °C,
and the progress of the reaction was followed by TLC. Upon
completion of the reaction, the mixture was concentrated under reduced
pressure to a black tar. The tar was repeatedly washed with MeOH
until the odor of ammonia could not be detected. The tar was dissolved
in hot EtOAc and allowed to cool. After the solution was cooled to 4
°C, a precipitate formed and was collected by filtration to yield 6.0 g
of a yellow solid. The filtrate was concentrated under reduced pressure
and absorbed onto approximately 2.0 g of silica gel. Flash chroma-
tography (CH2Cl2/MeOH, 1-5% MeOH) followed by recrystallization
in acetone/ethanol (70:30) yielded 3.05 g of yellow crystals. Both crops
of crystals had the same physical and spectral properties observed for
the small-scale synthesis. The overall yield for the three steps, based
on the starting material 2-amino-5-bromo-3,4-dicyanopyrrole, was 32%
(9.05 g).
Figure 3. Modeling hm7c7dA into the binding pocket for water-
mediated contacts to adenine residues in the trp repressor/operator
complex. Protein surface calculated by Grasp (v.1.2).13 (A) Analogue
superimposed on adenine (A5) of the operator that is recognized through
a water-mediated contact with the backbone amide nitrogen of Ala 80.
(B) Analogue superimposed on adenine (A-7) of the operator that is
recognized through a water-mediated contact with Thr83. The oxygens
of the waters of interest are yellow, the DNA is green, and the analogue
is aqua blue. The other strand of DNA and most of the protein were
removed for clarification.
that neither the oxygen of the waters in the protein/DNA
complex nor the oxygen of the hydroxymethyl group are
coplanar with the purine ring.12 The distances between the
oxygen of the water and the oxygen of the hydroxymethyl group
is 0.712 Å for the water mediating a contact between A5 and
the backbone nitrogen of Ala 80 and 0.835 Å for the contact
mediating an interaction between A-7 and the hydroxyl group
of Thr 83. Although the overlap is not exact, these distances
are less than the resolution of the solved structure, suggesting
that O11 of the analogue may be in a position suitable for
mediating the hydrogen bond contacts between the protein
and the DNA. A possible concern is the introduction of the
slightly bulky methylene. In both relevant instances (Ala80 and
Thr83) the nearest neighbor to carbon atom C10 is over 3 Å
away suggesting a lack of steric clashes. The modeling used
data from the X-ray diffraction analysis without energy mini-
mization.
4-Amino-7-(2′-deoxy-3′,5′-di-O-p-toluoyl-â-D-erythro-pentofura-
nosyl)pyrrolo[2,3-d]pyrimidine-5-carbonitrile (4). A 500 mL Parr
bottle was charged with 6.0 g (10.2 mmol) of compound 3, 16 mL of
triethylamine, 2.0 g of Pd/C (10%), 50 mL of dioxane, and 50 psi of
H2 gas. The solution was shaken for 4 h at 50 °C. The solution was
filtered through Celite and washed two times with 50 mL of hot EtOH.
The filtrate was concentrated under reduced pressure. The product was
purified by flash chromatography (CHCl3/EtOAc 4:1) to yield a white
solid (4) (4.14 g, 79.9%): mp 152-154 °C; 1H NMR (300 MHz,
(12) The water associated with adenine residue A5 is shifted out of the
plane of the purine ring by 1.38 Å toward G6 such that the water appears
to be shared between A5 and G6. The water associated with adenine residue
A-7 is shifted 0.671 Å toward the 3′ terminus. For hm7c7dA, the oxygen
is 1.029 Å away from the plane of the purine ring. The plane was defined
by the atoms N1, C2, N3, C4, C5, C8, and N9.
(13) Nicholls, A.; Sharp, K. A.; Honig, B. Proteins 1991, 11, 281-296.

10068 J. Am. Chem. Soc., Vol. 118, No. 42, 1996
Rockhill et al.
CDCl3) δ 2.43 (3H, s, Ph-CH3), 2.44 (3H, s, Ph-CH3), 2.71-2.83 (2H,
m, H2′), 4.62-4.79 (3H, m, H4′ and H5′), 5.63 (2H, bs, NH2), 5.72
(1H, m, H3′), 6.70 (1H, t, J ) 7.7 Hz, H1′), 7.26-7.30 (4H, m, Ph),
7.68 (1H, s, H8), 7.90-7.98 (4H, m, Ph), 8.35 (1H, s, H2); MS (FAB+),
m/e 512. Anal. Calcd for C28H25N5O5‚1/2H2O: C, 64.40; H, 4.73; N,
13.35. Found: C, 64.60; H, 5.03; N, 13.46.
(2H, d, J ) 5.0 Hz (CH2OH), H101, H102), 5.06 (1H, t, J ) 5.49 Hz,
(CH2OH), H5′O), 5.23 (1H, d, J ) 4.0 Hz, (CH2OH), H3′O), 5.72 (1H,
t, J ) 5.0 Hz, (CH2OH), H11), 6.46 (1H, dd Jab ) 6.0 Hz, Jab′ ) 2.4
Hz, H1′), 6.92 (2H, bs, NH2), 7.27 (1H, s, H2), 8.04 (1H, s, H8)
(hydroxyl and amino proton shifts were verified by the disappearance
from the spectra of these signals when D2O is added to the DMSO
solution); HRMS (FAB+) found 281.1250, calcd for C12H16N4O4,
281.1245. Anal. Calcd for C12H16N4O4: C, 51.4; H, 5.76; N, 20.0.
Found: C, 51.20; H, 5.62; N, 19.61.
4-Amino-7-(2′-deoxy-â-D-ribofuranosyl)-5-(hydroxymethyl)pyr-
role[2,3-d]pyrimidine (1). A pyridine/acetic acid/water (2:1:1) buffer
(200 mL) was cooled to 0 °C to which 4.0 g (9.8 mmol) of nitrile 4
was added along with 14.5 g of sodium hypophosphate monohydrate.
The flask was purged with argon, and 2.35 g of Raney Ni (60% in
mineral oil) was added. The reaction was heated to 55-65 °C for 6 h.
The solution was cooled to room temperature and filtered through Celite.
The Celite was washed two times with hot ethanol (100 mL). The
filtrate was concentrated under reduced pressure. The resulting oil was
taken up in 150 mL EtOAc and to this was added 200 mL of a 0.5 M
solution of citric acid. The organic layer was washed one time with
200 mL of saturated sodium bicarbonate and two times with 200 mL
of saturated sodium chloride. The organic layer was concentrated,
leaving a tan solid. Thin layer chromatography (Rf ) 0.45, 5% MeOH/
CH2Cl2) analysis revealed an incomplete reaction. Proton NMR
(CDCl3) spectroscopy on the crude mixture gave a singlet peak at 9.40
ppm corresponding to the aldehyde proton. The solid was dissolved
in absolute isopropyl alcohol (20 mL) and evaporated to dryness two
times. The solid was dissolved in 100 mL of absolute isopropyl alcohol,
and the flask was flushed with argon. Under an inert atmosphere,
sodium borohydride (371.6 mg, 9.78 mmol, 4 equiv) was added and
the flask was warmed to 50 °C. The solution was stirred for 30 min
and then cooled to room temperature, and 10 drops of 1.0 M HCl were
added slowly. The solution was evaporated to dryness and dissolved
in 100 mL of EtOAc and 30 mL of 10% sodium bicarbonate solution.
The organic layer was separated, and the aqueous layer was re-extracted
with 20 mL of EtOAc. The aqueous layer was concentrated under
reduce pressure resulting in a white solid. (The organic layer contains
a mixture of partially deprotected products. None of the fully
deprotected nucleoside could be observed.) This solid was crushed
and then stirred with 50 mL of absolute methanol for 30 min. The
inorganic salts were removed by filtration, and the procedure was
repeated. To the combined filtrates was added approximately 1.5 g of
silica gel, and the solvent was removed in Vacuo. The product was
filtered through silica gel (20% MeOH/CH2Cl2), the combined filtrates
were concentrated, and the procedure was repeated. Final purification
was achieved by recrystallization from methanol to yield 218.9 mg
(7.05%) of 1: mp 211-212 °C, TLC Rf ) 0.55 (20% MeOH/CH2-
Cl2); λmax (H2O)/nm 274 (pH ) 7), 278 (pH ) 1), 270 (pH ) 12); 1H
NMR (500 MHz/DMSO-d6) δ 2.09 (1H, m, H2′), 2.44 (1H, m, H2′),
3.50 (2H, m, H5′1/H5′2), 3.79 (1H, m, H4′), 4.31 (1H, m, H3′), 4.57
X-ray Structure Determination. Compound 1 was crystallized
from methanol to give 2′-deoxy-7-(hydroxymethyl)-7-deazaadenosine
dihydrate: MW ) 316.32, C12H20N4O6, crystal dimensions of 0.10 ×
0.14 × 0.36 mm, orthorhombic, space group P212121, a ) 7.0764(11)
Å, b ) 8.0669(13) Å, c ) 24.681(4) Å, V ) 1408.9(4) Å3, Z ) 4,
ρcalcd ) 1.491 mg/m3, θ ) 3.30-22.64°, Mo KR radiation (λ ) 0.71073
Å), and T ) 198 K. Data were collected using a Siemens SMART/
CCD, and the structure was solved Via direct methods. Full-matrix
least squares refinement on F2 using SHELXTL-V5.0 converged with
a final R1 ) 0.074 and wR2 ) 0.175 for 1413 observed (I > 2σ(I))
reflections. Positions and displacement parameters for water and amino
H atoms were independently refined, torsion angles for idealized
hydroxyl H atoms were refined, and the remaining H atoms were
included as fixed idealized contributors. Isotropic U values for idealized
H atoms were assigned as 1.2 times the Ueq of the adjacent atom.
Acknowledgment. This paper is dedicated to Nelson J.
Leonard on the occasion of his 80th birthday. This work was
supported, in part, by NIH grant GM25621 to R.I.G. We thank
Dr. Balikrishen Bhat for assistance with the synthesis, Dr.
Howard Robinson for obtaining the water-exchange NMR
spectra, and Dr. Fred Lakner for critical reading of the
manuscript.
Note Added in Proof: The compound 5-(hydroxymethyl)-
tubercidin, the ribonucleoside version of the title compound,
was previously synthesized starting from the natural product
sangivamycin (5-carboxamidetubercidin) using completely dif-
ferent chemistry. Uematsu, T.; Suhadolnik, R. J. J. Med. Chem.
1973, 16 (12), 1405-1407.
Supporting Information Available: X-ray diffraction data
and proton NMR data in water (8 pages). See any current
masthead page for ordering and Internet access information.
JA961540S