Design, synthesis, and antiviral evaluation of some 3′-carboxymethyl- 3′-deoxyadenosine derivatives

Article abstract of DOI:10.1080/15257770701426278

3′-Carboxymethyl-3′-deoxyadenosine derivatives were prepared from 2′-O-TBDMS-3′-[(ethoxycarbonyl)methyl]-3′-deoxyadenosine (1) via simple and efficient procedures. Conversion of 1 to its 5′-azido-5′-deoxy derivative 5 was accomplished via a novel one-pot

Full text of DOI:10.1080/15257770701426278

Nucleosides, Nucleotides, and Nucleic Acids, 26:499–519, 2007
Copyright ꢀC Taylor & Francis Group, LLC
ISSN: 1525-7770 print / 1532-2335 online
DOI: 10.1080/15257770701426278
DESIGN, SYNTHESIS, AND ANTIVIRAL EVALUATION OF SOME
3ꢀ-CARBOXYMETHYL-3ꢀ-DEOXYADENOSINE DERIVATIVES
Matt A. Peterson, Pucheng Ke, Houguang Shi, and Carl Jones
of Chemistry and Biochemistry, Brigham Young University, Provo, Utah, USA
Department
2
Brenda R. McDougall and W. Edward Robinson, Jr. Department of Pathology
2
and Laboratory Medicine, University of California, Irvine, California, USA
3ꢁ-Carboxymethyl-3ꢁ-deoxyadenosine derivatives were prepared from 2ꢁ-O-TBDMS-
2
3ꢁ-[(ethoxycarbonyl)methyl]-3ꢁ-deoxyadenosine (1) via simple and efficient procedures.
Conversion of 1 to its 5ꢁ-azido-5ꢁ-deoxy derivative 5 was accomplished via a novel one-pot
method employing 5ꢁ-activation (TosCl) followed by efficient nucleophilic displacement with
tetramethylguanidinium azide. Compound 5 was converted to 5ꢁ-[(N-methylcarbamoyl)amino]
derivative
8 via one-pot reduction/acylation employing H2/Pd-C followed by treatment
with p-nitrophenyl N-methylcarbamate. N6-phenylcarbamoyl groups were introduced
by treatment with phenylisocyanate, and an efficient new method for lactonization of
2ꢁ-O-TBDMS-3ꢁ-[(ethoxycarbonyl)methyl]-3ꢁ-deoxyadenosines to give corresponding 2ꢁ,3ꢁ-lactones
was also developed. Target compounds were evaluated for anti-HIV and anti-HIV integrase
activities, but were not active at the concentrations tested.
Keywords HIV; HIV integrase; Antivirals; 3ꢁ-C-Branched nucleotides
INTRODUCTION
Branched chain 2ꢁ(3ꢁ)-deoxy- or 2ꢁ,3ꢁ-dideoxynucleosides have long
been of interest due to their possession of unique biological activities
relative to those of unbranched parent nucleosides.[1–7] Antiviral,[8–10]
antitumor,[11–13] and/or antitubercular[14] activities for these compounds
have been reported, and conformational effects induced by sugar branching
have also been investigated.[15] Recently, we became interested by the pos-
sibility that 3ꢁ-carboxymethyl-3ꢁ-deoxyadenosine[4,7] derivatives could serve
Received 29 November 2006; accepted 3 March 2007.
This work was supported in part by a grant from the Public Health Service (1RO1-AI063973 to
WER). Generous support from the BYU Cancer Research Center is also gratefully acknowledged.
Address correspondence to Matt A. Peterson, Department of Chemistry and Biochemistry, Brigham
Young University, Provo, Utah 84602. E-mail: matt peterson@byu.edu
499

500
M. A. Peterson et al.
FIGURE 1
as ideal scaffolds on which to build inhibitors of HIV integrase (Figure 1).
HIV integrase (IN) is one of three enzymes involved in HIV replication and
is essential for covalently joining the viral DNA into the host genome. IN is
a multimeric enzyme (at least tetrameric and probably octameric)[16] and
belongs to a superfamily of polynucleotidyl phosphotransferases in which a
metal dication (Mg2+ or Mn2+) is required for catalysis.[17] The metal dica-
tion plays a critical role in both 3ꢁ-end processing of viral DNA and in strand
transfer (Figure 2). In the 3ꢁ-end processing reaction, IN catalyzes cleavage
of a GT dinucleotide from each 3ꢁ-end of the viral DNA via magnesium-
mediated phosphodiester hydrolysis. In the subsequent strand transfer reac-
tion, each free 3ꢁ-OH on the 3ꢁ-processed viral cDNA undergoes magnesium-
mediated transesterification with host DNA, ultimately giving rise to fully
integrated provirus (Figure 2). We reasoned that appropriately function-
alized 3ꢁ-carboxymethyl-3ꢁ-deoxyadenosine derivatives might bind to the
active-site Mg2+ and active-site amino acid residues, and, thus, potentially
inhibit IN. Binding interactions between the 3ꢁ-terminal deoxyadenosine
of viral DNA and IN had been indicated by ligand-docking calculations[18]
and DNA photo-crosslinking experiments,[19] and these studies supported
the notion that 3ꢁ-carboxymethyl-3ꢁ-deoxyadenosine derivatives 2 might
bind to the active site and provide significant levels of IN inhibition.
Molecular mechanics calculations (MMFF, MacSpartan Pro, Wavefunc-
tion, Inc., Irvine, CA, USA) had also indicated that 3ꢁ-carboxymethyl-3ꢁ-
deoxyadenosine derivative 2a coordinates with Mg(OH)4 to give a complex
with octahedral geometry (Figure 3). Carboxylates are known ligands[20] for
Mg2+, and hydroxide had been used as a solvent surrogate in related model
complexes.[21]
The foregoing considerations suggested that 3ꢁ-carboxymethyl-3ꢁ-
deoxyadenosine framework 2 (metal binding moiety = COO−) might be
a useful construct from which to conduct a structure-based lead discovery
program. Accordingly, an approximately 49,000 member virtual library was
screened in silico for derivatives 2 that exhibited tight binding against IN
crystal structure 1BIU (Protein Data Bank). Crystal structure 1BIU is one

501

502
M. A. Peterson et al.
FIGURE 3
of only two crystal structures for IN catalytic core domains that contain an
intact active-site Mg2+ ion.[22,23] A vast combinatorial library of compounds 2
(metal binding moiety = COO−) varying at R1 and R2 (222 variations at each
position generating a 222 × 222 combinatorial library) was docked against
known active-site residues Asp 64, Thr 66, His 67, Asp 116, Glu 152, Lys 156,
Lys 159, and the active-site Mg2+ ion. The lowest energy binding solution for
each of the 49,000 compounds was determined (FlexX, Sybyl 6.9, Tripos,
Inc., St. Louis, MO, USA), and the highest ranked (lowest FlexX score)
ligand from the entire library was identified (Figure 4). Hydrogen bonding
interactions for the top hit (2b) occurred between His 67 and the 5ꢁ-urea
moiety, Thr 66 and adenine N3, Glu 152 and the C6 urea, and between
the 2ꢁ-OH and the carbonyl oxygen of Asp 64. The 3ꢁ-carboxymethyl group
coordinated with Mg2+ in an equatorial relationship relative to the 2ꢁ-OH, in
harmony with the calculated geometry for the model complex Mg(OH)4·2a
(Figure 3). His 67 underwent a σ-π interaction with the 5ꢁ-urea, and Asn
155 underwent a similar interaction with the adenine heterocycle.
RESULTS AND DISCUSSION
In order to establish the validity of the binding interactions indicated
by the virtual library screening, we prepared compounds 2b–h (Figure 5).
Compounds 2c–h were designed to probe the effects of various groups on
binding activity (structure activity relationship), and 2b had been selected
as the top hit from the virtual library. Compounds 2b–d shared in common
the 5ꢁ-N -methyl and 6-N -phenyl urea moieties postulated to bind with His
67/Lys 159 and Glu 152, respectively. The 5ꢁ-azido group had been identi-
fied as a lower ranking 5ꢁ-substituent from the library docking calculations,
and 2e and 2f were readily derived from intermediates needed to prepare
5ꢁ-N -methylurea derivatives 2b–d. Compounds 2g and 2h were included to

3ꢁ-Carboxymethyl-3ꢁ-Deoxyadenosine Derivatives
503
FIGURE 4
test the possibility that lactones might undergo IN-catalyzed saponification
to generate carboxylates 2b and 2e in the active site of IN.
The synthesis began with compound 1 which was readily prepared[24]
in four steps from 2ꢁ,5ꢁ-bis-O-TBDMS adenosine (Scheme 1). Attempts to
prepare 5ꢁ-chloro-5ꢁ-deoxyadenosine derivative 3 using standard chlorina-
tion conditions (SOCl2/Pyr/CH2Cl2) gave desired product in low yields
(30–40%). Significant amounts of an unisolated polar byproduct (base-
line by TLC) were formed, which we assume derived from the N 3,5ꢁ-
cyclonucleoside salt.[25] Treatment of 1 with TsCl/DMAP[26] in ice-cold
CH2Cl2 gave compound 4 in excellent isolated yields (72–90%). Conver-
sion of compounds 3 (4) to 5ꢁ-azido-5ꢁ-deoxyadenosine derivative 5 using
standard conditions (NaN3/DMF)[27] was complicated by decomposition,
and yields for the desired product ranged from 20–40%. Since it has
FIGURE 5

504
M. A. Peterson et al.
long been recognized that intramolecular alkylation of N3 competes with
intermolecular nucleophilic substitution of 5ꢁ-activated adenosine derivatives
to give cyclonucleosides[28] and derived rearrangement products,[29] we
reasoned that yields for the desired intermolecular substitution reaction
might be enhanced if higher concentrations of the poorly soluble azide
nucleophile could be achieved. This approach represented a fundamental
departure from the more generally employed strategies for suppressing cy-
clonucleoside formation which typically involve reducing the nucleophilic-
ity of N3 via N6 acylation[27,28] or disfavoring the required syn conformation
by employing sterically encumbered N6 protection.[30] While these latter
strategies had proven successful in a number of previous cases, we wished to
avoid any protection/deprotection steps that might unnecessarily lengthen
the synthesis. Accordingly, we undertook an optimization study and were
ultimately delighted to find that treatment of crude 4 with 7 equiv. of tetram-
ethylguanidinium azide (TMGA; [(Me2N)2CNH2]N3) in DMF (65◦C) gave
5 in 83% yield from compound 1. The relatively high solubility of TMGA in
polar aprotic organic solvents has been exploited for high-yield preparations
of amino acids and other target compounds. To the best of our knowledge,
the present report represents the first application of TMGA to problems
involving adenosine cyclonucleoside formation. This new strategy may pro-
vide a generally useful alternative to two-step N6-protection/deprotection
strategies more commonly employed for suppressing N 3,5ꢁ-cyclonucleoside
formation. [Caution: TMGA is known to form explosive byproducts in
halogenated solvents—e.g., diazidomethane from dichloromethane—thus,
halogenated solvents should be avoided.][31]
Compound 5 was saponified to give compound 6 (85%). In-
troduction of the metal-coordinating N -methoxy-N -methylcarboxamide
group was accomplished by treating 6 with carbonyldiimidazole and
N,O-dimethylhydroxylamine to give 7 (84%). Hydrogenation of 5
(H2/Pd–C/EtOH) followed by treatment of the resulting 5ꢁ-amino-5ꢁ-
deoxyadenosine intermediate with p-nitrophenyl N -methylcarbamate gave
compound 8 (92%). Conversion of 5 (7) to 9 (10) was accomplished
by treatment of 5 (7) with phenylisocyanate in CH2Cl2 (95 and 94%,
respectively). One-pot reduction/acylation of 9 (10) using the same con-
ditions employed for 5 → 8 gave targets 2c and 2d (96 and 63%, re-
spectively). Our recently reported high-yield synthesis of p-nitrophenyl N -
methylcarbamate[32] makes this method an attractive alternative to reported
methods for introducing the 5ꢁ-(N -methylcarbamoyl) group which employ
highly toxic methylisocyanate as the acylating reagent.[33]
Conversion of 3ꢁ-carboxymethyl-3ꢁ-deoxy derivatives 1, 2c, 5, 8, and
9 to the corresponding 2ꢁ,3ꢁ-lactones using conditions we had previously
found effective for preparing compound 11 and a related uridine-derived
2ꢁ,3ꢁ-lactone (TBAF/THF),[24] gave products that were very difficult to

3ꢁ-Carboxymethyl-3ꢁ-Deoxyadenosine Derivatives
505
SCHEME 1
purify due to co-elution of tetrabutylammonium salts with the desired
target compounds (Scheme 2). Multiple column chromatographies em-
ploying careful gradient elutions were required to obtain 2ꢁ,3ꢁ-lactone
derivatives that were free from contaminating tetrabutylammonium impu-
rities. An evaluation of alternative desilylating reagents ultimately revealed
that a phase transfer catalyzed biphasic mixture consisting of KF/PhCH2
N(Et)3Cl/CH3CN/H2O gave clean conversion to the 2ꢁ,3ꢁ-lactone products,
which could be purified via a single pass through a silica gel column using an
appropriate eluting solvent. Treatment of compounds 1, 2c, 5, 8, and 9 with
KF/PhCH2N(Et)3Cl/CH3CN/H2O gave 11–13 and 2g (2h) in excellent
yields (79–92%). Saponification of 2ꢁ,3ꢁ-lactones 2g, 2h, 11, and 12 gave
compounds 2a, 2b, 2e, and 2f (81–100%).
With compounds 2b–h in hand, we next turned our attention to an
evaluation of their anti-HIV and IN inhibitory activities (Table 1). Promising
activities were not exhibited by any of the compounds tested. The failure
of these compounds to exhibit IN inhibitory activities may derive from
their possible affinity for binding sites remote from the active-site, or may

506
M. A. Peterson et al.
SCHEME 2
also possibly reflect weaknesses inherent in the incremental construction
algorithm forming the basis of the FlexX ligand docking calculations.[34,35]
Entropic and enthalpic contributions of dissociating water ligands from the
active site Mg2+ are not taken into consideration by the FlexX algorithm,
and ligand conformational energy terms are not included in the FlexX
scoring function.[34] Thus, whereas the docking experiments point to
favorable binding interactions for compound 2b (Figure 4), free energy
contributions from the dissociation of water ligands and/or from the
conformation of the inhibitor may actually disfavor binding as indicated
by the docking. In addition, the underlying success of FlexX docking
experiments strongly depends upon initial placement of base fragments
that serve as anchors for iterative incremental construction and docking
of increasingly more complex ligand fragments.[34] Incorrect selection or
placement of the base fragment can skew ensuing iterative placements of the
more complex fragments generated during the incremental construction,
and heavy weighting of charged ligand/receptor interactions has been
shown to give “false positives” in some instances.[36] It is also true that no

3ꢁ-Carboxymethyl-3ꢁ-Deoxyadenosine Derivatives
507
TABLE 1 Activities of Test Compounds in Biochemical Assays
IC50d (µM)
EPe STf
a
b
Compd
ED50 (µM)
CT50 (µM)
CT5c(µM)
2b
2c
2d
2e
2f
>98
>13
>17
>149
>62
>19
>34
385
37.8
22.6
812
175
21.9
58.5
143
6.2
11.3
162
21
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
2g
2h
9.3
23.2
aInhibitory concentration required to protect MT-2 cells from 50% viral induced cell death.
bCytotoxic concentration required to inhibit cell growth by 50%.
cCytotoxic concentration required to inhibit cell growth by 5%.
dInhibitory concentration required to inhibit IN 3ꢁ-end processing (EP) or strand transfer
(ST)by 50%.
e 3ꢁ-End processing.
f Strand transfer.
full-length IN structure and no structure of an IN-DNA substrate complex
have been reported; thus, our lack of success in identifying potent lead IN
inhibitors may be due to insufficient structural information. Coordination
of a second active-site Mg2+ cation by the third member of the catalytic
triad (Glu 152) has been invoked as being critical for the mechanism of
full-length integrase enzymes.[37] Such coordination was not present in
the two-domain X-ray structure used in this study (1BIU). Furthermore,
Glu 152 resides in what has been shown to be a flexible loop in the IN
primary structure;[38] thus, it is possible that the position of Glu 152 in
crystal structure 1BIU does not accurately reflect its position in solution.
X-ray data for partial IN structures such as 1BIU suggest that for the
full-length enzyme the active site may be relatively solvent accessible and
could thus potentially possess a local dielectric constant >4. Theoretical
calculations of the free energies for dissociating water ligands from Mg2+
aqua complexes support an inverse relationship between the polarity of the
local environment and the favorability of the displacement of a water ligand
by a charged carboxylate (threshold value ε ≤4).[20] If the local dielectric
constant of IN is substantially > 4, the free energies for binding carboxylate
ligands to the active-site Mg2+ of IN may be unfavorable.[20]
CONCLUSION
We have prepared a small library of compounds designed to test the hy-
pothesis that appropriately derivatized 3ꢁ-carboxymethyl-3ꢁ-deoxyadenosine
derivatives might bind to the active-site Mg2+ and active-site amino acid
residues of IN and thus competitively inhibit the enzyme. Although bi-
ological activities were not promising, several important results derive

508
M. A. Peterson et al.
from the synthetic effort: (1) TMGA-promoted nucleophilic substitution
of 5ꢁ-O-p-toluenesulfonyl adenosine derivative 4 gave high yields of 5ꢁ-
azido-5ꢁ-deoxyadenosine derivative 5, thus demonstrating a potentially
general alternative to two-step N6-protection/deprotection strategies cur-
rently used for high yield preparations of 5ꢁ-azido-5ꢁ-deoxyadenosine
derivatives from corresponding 5ꢁ-activated adenosine precursors; (2) the
biphasic reagent/solvent system KF/PhCH2N(Et)3Cl/CH3CN/H2O gives
enhanced yields of adenosine 2ꢁ,3ꢁ-lactone nucleosides with greatly sim-
plified work-up procedures relative to previously reported conditions
(TBAF/THF; painstaking chromatography); and (3) conversion of 5ꢁ-azido-
5ꢁ-deoxyadenosine derivatives 5, 9 and 10 to N -methylurea derivatives 8,
2c, and 2d (respectively) was achieved via an efficient one-pot reduc-
tion/acylation procedure employing p-nitrophenyl N -methylcarbamate as
a safer alternative to methylisocyanate.
EXPERIMENTAL
Moisture sensitive reactions were performed in flame-dried flasks under
an inert atmosphere (N2 or Ar) at ambient temperature unless indicated
otherwise. Solvents (CH2Cl2, pyridine, EtOAc, DMF, Et3N) were dried by
passing through columns of activated alumina under Ar. TLC was per-
formed using Merck Kieselgel 60 F254 sheets, and Flash[39] chromatography
was performed with silica gel (230–400 mesh) and reagent grade solvents.
“Solvent A” for chromatography consisted of the separated organic phase of
EtOAc/i-PrOH/H2O (4:1:2). UV spectra were determined with solutions in
MeOH or H2O. 1H NMR spectra were determined using internal references
at δ 7.27 (CDCl3), and 2.50 (DMSO-d6), and 13C NMR spectra were
measured using internal references at δ 77.3 (CDCl3), and 39.5 (DMSO-d6).
High resolution mass spectra were obtained using fast atom bombardment
(FAB, NaOAc/thioglycerol or thioglycerol matrix) or electrospray (ES) ion-
ization techniques. Commercially available reagents were used as supplied,
and tetramethylguanidinium azide[40] and compound 1[24] were prepared
as previously reported.
2ꢁ-O-(tert-Butyldimethylsilyl)-5ꢁ-chloro-3ꢁ,5ꢁ-dideoxy-3ꢁ-[(ethoxycarbonyl)
methyl]adenosine (3). To a stirred solution of 1 (200 mg, 0.443 mmol)
and pyridine (100 mg, 1.27 mmol) in CH2Cl2 (3.0 mL) at 0◦C was added
thionyl chloride (2 M in CH2Cl2, 1.0 mL, 2.0 mmol). The mixture was
stirred for 30 minutes, then allowed to warm to room temperature and
stirred overnight. Volatiles were removed under reduced pressure and
the residue was partitioned (EtOAc//NaHCO3(aq)). The organic layer
was dried (Na2SO4), filtered, and volatiles were removed under reduced
pressure. Chromatography (5% MeOH/CH2Cl2) gave 3 (62 mg, 30%): UV
(MeOH) λ max 260 nm, λ min 230 nm; 1H NMR (CDCl3, 500 MHz) δ 8.35

3ꢁ-Carboxymethyl-3ꢁ-Deoxyadenosine Derivatives
509
(s, 1H), 8.18 (s, 1H), 5.97 (s, 1H), 5.59 (br s, 2H), 4.94 (d, J = 4.5 Hz, 1H),
4.37–4.34 (m, 1H), 4.12 (q, J = 7.4 Hz, 2H), 4.01 (dd, J = 3.0, 12.5 Hz,
1H), 3.78 (dd, J = 4.3, 12.8 Hz, 1H), 2.85–2.82 (m, 1H), 2.70 (dd, J = 9.0,
17.0 Hz, 1H), 2.42 (dd, J = 5.8, 16.8 Hz, 1H), 1.26 (t, J = 7.3 Hz, 3 H), 0.90
(s, 9 H), 0.15 (s, 3 H), 0.07 (s, 3 H); 13C NMR (CDCl3, 50 MHz) δ 171.9,
155.8, 153.2, 138.2, 120.4, 91.3, 82.9, 77.5, 61.1, 45.2, 40.7, 30.1, 25.9, 18.1,
14.3, −4.4, −5.4; MS (FAB) m/z 492.1805 (MNa+ [C20H3235ClN5O4SiNa]
= 492.1810).
2ꢁ-O-(tert-Butyldimethylsilyl)-3ꢁ-deoxy-3ꢁ-[(ethoxycarbonyl)methyl]-5ꢁ-
O-(p-toluenesulfonyl)adenosine (4). To a chilled (0◦C) flame-dried flask
containing 1 (378 mg, 0.837 mmol; azeotropically dried via evaporation
of benzene, 5 × 20 mL), p-toluenesulfonylchloride (278 mg, 1.46 mmol),
and DMAP (218 mg, 1.78 mmol) was added ice-cold CH2Cl2 (4.0 mL).
The solution was stirred for 24 hours at 0◦C, then applied directly to a
chromatography column and eluted (80% EtOAc/hexanes → EtOAc).
Appropriate fractions were pooled and volatiles were removed under
reduced pressure (≤20◦C) to give 4 (390 mg, 77%). Compound 4 was
not stable at ambient temperature and underwent decomposition upon
standing either in solution or as a solid amorphous glass. Characterization
was therefore accomplished immediately following isolation, and maximum
purities obtained in this way were approximately 90%. Unambiguous
1
characterization by 13C NMR was complicated by compound instability: H
NMR (CDCl3, 500 MHz) δ 8.30 (s, 1H), 7.95 (s, 1H), 7.77–7.75 (m, 2H),
7.29–7.28 (m, 2H), 5.91 (d, J = 1.0 Hz, 1H), 5.56 (br s, 2H), 4.85 (d, J = 4.0
Hz, 1H), 4.37 (dd, J = 2.0, 8.5 Hz, 1H), 4.27–4.20 (m, 2H), 4.11 (q, J = 7.2
Hz, 2H), 2.82–2.76 (m, 1H), 2.64 (dd, J = 8.8, 16.8 Hz, 1H), 2.42 (s, 3 H),
2.32 (dd, J = 5.5, 17.0 Hz, 1H), 1.19 (t, J = 7.2 Hz, 3H), 0.89 (s, 9 H), 0.14
(s, 3 H), 0.03 (s, 3 H); MS (FAB) m/z 606.2417 (MH+ [C27H40N5O7SSi] =
606.2418).
5ꢁ-Azido-2ꢁ-O-(tert-butyldimethylsilyl)-3ꢁ,5ꢁ-dideoxy-3ꢁ-[(ethoxycarbonyl)
methyl]adenosine (5). To a chilled (0◦C) flame-dried flask containing 1
(360 mg, 0.797 mmol; azeotropically dried via evaporation of benzene, 5
× 20 mL), p-toluenesulfonylchloride (208 mg, 1.10 mmol), and DMAP
(208 mg, 1.70 mmol) was added ice-cold CH2Cl2 (16 mL). The solution
was stirred for 24 hours at 0◦C, after which volatiles were removed under
reduced pressure (≤20◦C). Tetramethylguanidinium azide (880 mg, 5.56
mmol) and DMF (4 mL) were immediately added and the solution was heated
at 65◦C for 7 hours. The mixture was cooled to ambient temperature and
then vigorously stirred while anhydrous Et2O (100 mL) was slowly added.
Precipitated TMGA was removed by filtering through celite. The white solid
mass was triturated, and the filter cake was washed with anhydrous Et2O to
ensure complete transfer of product. Volatiles were removed under reduced
pressure (40◦C) and the residue chromatographed (90% EtOAc/hexanes
→ EtOAc) to give 5 (315 mg, 83%): UV (MeOH) λmax 262 nm, λmin

510
M. A. Peterson et al.
1
233 nm; H NMR (CDCl3, 500 MHz) δ 8.36 (s, 1H), 8.16 (s, 1H), 5.98 (s,
1H), 5.54 (br s, 2H), 4.86 (d, J = 5.0 Hz, 1H), 4.22–4.20 (m, 1H), 4.14 (q, J
= 7.0 Hz, 2H), 3.78 (dd, J = 3.3, 13.8 Hz, 1H), 3.61 (dd, J = 4.8, 13.8 Hz,
1H), 2.85–2.77 (m, 1H), 2.69 (dd, J = 8.3, 16.8 Hz, 1H), 2.37 (dd, J = 5.8,
16.8 Hz, 1H), 1.26 (t, J = 7.3 Hz, 3 H), 0.91 (s, 9 H), 0.17 (s, 3 H), 0.07 (s,
3 H); 13C NMR (CDCl3, 125 MHz) δ 171.6, 155.4, 153.0, 149.4, 138.7, 120.2,
91.1, 82.2, 77.3, 60.9, 52.2, 40.0, 29.9, 25.7, 17.9, 14.1, −4.5, −5.5; MS (FAB)
m/z 499.2214 (MNa+ [C20H32N8O4SiNa] = 499.2214).
5ꢁ-Azido-2ꢁ-O-(tert-butyldimethylsilyl)-3ꢁ-(carboxymethyl)-3ꢁ,5ꢁ-
dideoxyadenosine (6). To a stirred solution of 5 (150 mg, 0.315 mmol) in
THF (2 mL) was added NaOH (200 µL, 5.0 M, 1.0 mmol), and MeOH
(400 µL). The mixture was stirred at ambient temperature until starting
material had been converted to baseline product (6 hours, TLC). Volatiles
were removed under reduced pressure (≤20◦C) and the crude material
was partitioned (CH2Cl2//H2O). Ice was added and the pH was carefully
adjusted to ≈3 via dropwise addition of 1% HCl (aq). The aqueous layer
was washed (CH2Cl2, 5X) until the organic layer was UV transparent
(TLC). The combined organic layers were dried (Na2SO4), filtered, and
evaporated under reduced pressure (≤20◦C) to give 6 (120 mg, 85%): UV
1
(MeOH) λmax 260 nm, λmin 233 nm; H NMR (CDCl3, 500 MHz) δ 8.32
(s, 1H), 8.25 (s, 1H), 7.27 (br s, 2H), 6.02 (s, 1H), 4.76 (d, J = 4.0 Hz, 1H),
4.25 (dd, J = 6.5, 10.5 Hz, 1H), 3.86 (d, J = 13.0 Hz, 1H), 3.63 (dd, J =
3.5, 13.5 Hz, 1H), 2.83–2.80 (m, 1H), 2.71 (dd, J = 8.5, 17.0 Hz, 1H), 2.42
(dd, J = 4.8, 17.3 Hz, 1H), 0.93 (s, 9H), 0.21 (s, 3H), 0.10 (s, 3H); 13C NMR
(CDCl3, 125 MHz) δ 176.1, 155.4, 151.8, 148.9, 138.8, 118.9, 91.1, 82.5,
77.9, 51.9, 39.8, 30.2, 29.7, 25.7, 18.0, −4.5, −5.5; MS (FAB) m/z 471.1902
(MNa+ [C18H28N8O4SiNa] = 471.1901).
5ꢁ-Azido-2ꢁ-O-(tert-butyldimethylsilyl)-3ꢁ,5ꢁ-dideoxy-3ꢁ-[(N-methoxy-N-
methylcarboxamido)methyl]adenosine (7). To a stirred solution of 6 (50 mg,
0.112 mmol) in CH2Cl2 (1.0 mL) at 0◦C was added carbonyl diimidazole
(500 µL of 0.36 M solution in CH2Cl2, 29 mg, 0.18 mol). The ice-bath was
removed and the reaction was allowed to warm to ambient temperature for
1 hour. N ,O-Dimethylhydroxylamine hydrochloride (18 mg, 0.19 mmol),
and Et3N (82 mg, 0.82 mmol), were added and the reaction was followed
by TLC (24 hours). Chromatography (5%MeOH/EtOAc) gave 7 (46 mg,
1
84%): UV (MeOH) λmax 260 nm, λmin 230 nm; H NMR (CDCl3, 500
MHz) δ 8.35 (s, 1H), 8.16 (s, 1H), 5.99 (d, J = 2.0 Hz, 1H), 5.67 (br s, 2H),
4.87–4.86 (m, 1H), 4.25–4.22 (m, 1H), 3.77 (dd, J = 2.8, 13.3 Hz, 1H),
3.70 (s, 3 H), 3.65 (dd, J = 4.5, 13.5 Hz, 1H), 3.16 (s, 3 H), 2.85–2.83 (m,
2H), 2.60–2.52 (m, 1H), 0.90 (s, 9 H), 0.11 (s, 3 H), 0.02 (s, 3 H); 13C NMR
(CDCl3, 125 MHz) δ 172.6, 155.7, 153.2, 149.8, 138.8, 120.3, 91.0, 82.9, 77.8,
61.5, 53.0, 39.9, 32.5, 28.4, 26.0, 18.2, −4.40, −5.10; MS (FAB) m/z 514.2327
(MNa+ [C20H33N9O4SiNa] = 514.2323).

3ꢁ-Carboxymethyl-3ꢁ-Deoxyadenosine Derivatives
511
2ꢁ-O-(tert-Butyldimethylsilyl)-3ꢁ,5ꢁ-dideoxy-3ꢁ-[(ethoxycarbonyl)methyl]-
5ꢁ-[(N-methylcarbamoyl)amino]adenosine (8). A solution of 5 (613 mg, 1.29
mmol) and 10% Pd–C (220 mg) in EtOAc (11 mL) was vigorously stirred
overnight under an atmosphere of H2 (balloon pressures). p-Nitrophenyl
N -methylcarbamate (440 mg, 2.24 mmol) and anhydrous Na2CO3 (440 mg,
4.15 mmol) were added, and the resulting mixture was stirred for 5 hours
under N2. Solids were filtered (celite), the filter cake washed with EtOAc,
and volatiles were evaporated under reduced pressure. Chromatography
(10% MeOH/CH2Cl2) gave 8 (600 mg, 92%): UV (MeOH) λmax 260 nm,
λmin 229 nm; 1H NMR (CDCl3, 500 MHz) δ 8.37 (s, 1H), 7.88 (s, 1H), 6.02
(br s, 1H), 5.78 (d, J = 4.0 Hz, 1H), 5.57 (br s, 2H), 4.95–4.93 (m, 1H),
4.51–4.38 (m, 1H), 4.24–4.22 (m, 1H), 4.15 (q, J = 7.2 Hz, 2H), 3.71–3.66
(m, 1H), 3.49 (dd, J = 4.0, 15.0 Hz, 1H), 2.84–2.80 (m, 1H), 2.80 (d, J =
5.0 Hz, 3 H), 2.69 (dd, J = 6.8, 17.3 Hz, 1H), 2.49 (dd, J = 6.8, 17.3 Hz,
1H), 1.28 (t, J = 7.0 Hz, 3 H), 0.84 (s, 9 H), −0.07 (s, 3 H), −0.14 (s, 3
H); 13 C NMR (CDCl3, 125 MHz) δ 172.1, 159.5, 155.8, 152.8, 149.2, 139.4,
120.4, 91.4, 83.7, 76.2, 60.7, 41.9, 39.7, 30.4, 27.1, 25.6, 17.8, 14.1, −4.80,
−5.40; MS (ES) m/z 508.2699 (MH+ [C22H38N7O5Si] = 508.2698).
5ꢁ-Azido-2ꢁ-O-(tert-butyldimethylsilyl)-3ꢁ,5ꢁ-dideoxy-3ꢁ-[(ethoxycarbonyl)
methyl]-N6-(N-phenylcarbamoyl)adenosine (9). To a stirred solution of 5
(633 mg, 1.33 mmol) in CH2Cl2 (16 mL) was added phenylisocyanate (190
mg, 1.60 mmol). The mixture was stirred at ambient temperature until
TLC indicated complete conversion of 5 to desired product (5 days). The
mixture was added directly to a chromatography column and eluted (10
→ 40% EtOAc/hexanes) to give 9 (755 mg, 95%): UV (MeOH) λmax 279
1
nm, λmin 243 nm; H NMR (CDCl3, 500 MHz) δ 11.74 (s, 1H), 8.62 (s,
1H), 8.39 (s, 1H), 8.11 (s, 1H), 7.65 (d, J = 8.5 Hz, 2H), 7.39–7.36 (m,
2H), 7.14–7.12 (m, 1H), 6.04 (s, 1H), 4.86 (d, J = 5.0 Hz, 1H), 4.24–4.22
(m, 1H), 4.14 (q, J = 7.2 Hz, 2H), 3.81 (dd, J = 2.8, 13.3 Hz, 1H), 3.63
(dd, J = 4.3, 13.3 Hz, 1H), 2.81–2.79 (m, 1H), 2.69 (dd, J = 8.5, 17.0 Hz,
1H), 2.39 (dd, J = 5.3, 17.3 Hz, 1H), 1.26 (t, J = 7.3 Hz, 3 H), 0.93 (s, 9
H), 0.19 (s, 3 H), 0.07 (s, 3 H); 13C NMR (CDCl3, 125 MHz) δ 171.5, 151.4,
150.8, 150.0, 149.9, 141.5, 138.1, 129.0, 123.8, 120.2, 91.3, 82.5, 77.5, 60.9,
52.2, 40.1, 29.7, 25.7, 18.0, 14.1, −4.5, −5.5; MS (FAB) m/z 596.2772 (MH+
[C27H38N9O5Si] = 596.2765).
5ꢁ-Azido-2ꢁ-O-(tert-butyldimethylsilyl)-3ꢁ,5ꢁ-dideoxy-3ꢁ-[(N-methoxy-N-
methylcarboxamido)methyl]-N6-(N-phenylcarbamoyl)adenosine (10). To a
solution of 7 (46 mg, 0.094 mmol) in CH2Cl2 (1.0 mL) was added
phenylisocyanate (12 mg, 0.10 mmol). The mixture was stirred at ambient
temperature until TLC indicated complete conversion of 7 to desired
product (7 days). The mixture was added directly to a chromatography
column and eluted (80% EtOAc/hexanes → EtOAc) to give 10 (54 mg,
1
94%): UV (MeOH) λmax 279 nm, λmin 242 nm; H NMR (CDCl3, 500
MHz) δ 11.77 (s, 1H), 8.63 (s, 1H), 8.40 (s, 1H), 8.13 (s, 1H), 7.66 (d,

512
M. A. Peterson et al.
J = 8.0 Hz, 2H), 7.40–7.37 (m, 2H), 7.15–7.11 (m, 1H), 6.05 (s, 1H), 4.88
(m, 1H), 4.28–4.26 (m, 1H), 3.82 (d, J = 10.5 Hz, 1H), 3.71–3.66 (m, 1H),
3.70 (s, 3 H), 3.17 (s, 3 H), 2.86–2.53 (m, 2H), 2.56–2.53 (m, 1H), 0.90 (s, 9
H), 0.15 (s, 3 H), 0.08 (s, 3 H); 13C NMR (CDCl3, 125 MHz) δ 172.1, 151.2,
150.8, 149.9, 141.1, 138.0, 129.0, 123.8, 120.8, 120.3, 91.0, 82.8, 77.7, 61.2,
52.5, 39.6, 32.2, 29.7, 27.9, 25.7, 17.9, −4.6, −5.4; MS (ES) m/z 633.2695
(MNa+ [C27H38N10O5SiNa] = 633.2694).
3ꢁ-(Carboxymethyl)-3ꢁ-deoxyadenosine-2ꢁ,3ꢁ-lactone (11). To a stirred so-
lution of 1 (50 mg, 0.11 mmol) in CH3CN (1.0 mL) were added PhCH2
N(Et)3Cl (5 mg, 0.022 mmol), KF (15 mg, 0.26 mmol), and H2O (40
µL). The mixture was vigorously stirred at ambient temperature until TLC
indicated that 1 had been consumed (42 hours). Silica gel was added and
volatiles were evaporated under reduced pressure (≤20◦C). The dried silica
gel was poured onto the top of a chromatography column packed with
CH2Cl2 and eluted (5 → 10% MeOH/CH2Cl2). Evaporation of pooled
fractions gave 11 (26 mg, 80%). 1H and 13C NMR and UV data agreed with
reported values.[24]
5ꢁ-Azido-3ꢁ-(carboxymethyl)-3ꢁ,5ꢁ-dideoxyadenosine-2ꢁ,3ꢁ-lactone (12). To
a stirred solution of 5 (50 mg, 0.105 mmol) in CH3CN (1.0 mL) were added
PhCH2N(Et)3Cl (5 mg, 0.022 mmol), KF (15 mg, 0.26 mmol), and H2O
(80 µL). The mixture was vigorously stirred at ambient temperature until
TLC indicated that 5 had been consumed (72 hours). Silica gel was added
and volatiles were evaporated under reduced pressure (≤20◦C). The dried
silica gel was poured onto the top of a chromatography column packed with
CH2Cl2 and eluted (2.5 → 10% MeOH/CH2Cl2). Evaporation of pooled
fractions gave 12 (27 mg, 81%): UV (MeOH) λmax 259 nm, λmin 236 nm;
1H (CDCl3, 500 MHz,) δ 8.35 (s, 1H), 7.91 (s, 1H), 6.17 (s, 1H), 5.61 (dd, J
= 1.0, 6.5 Hz, 1H), 5.54 (br s, 2H), 4.14–4.10 (m, 1H), 3.82–3.79 (m, 1H),
3.61 (dd, J = 4.8, 12.8 Hz, 1H), 3.55 (dd, J = 5.5, 13.0 Hz, 1H), 2.96 (dd, J
= 8.8, 18.3 Hz, 1H), 2.55 (dd, J = 1.0, 18.0 Hz, 1H); 13 C NMR (DMSO-d6,
125 MHz) δ 175.6, 156.2, 152.9, 148.8, 139.9, 119.1, 88.0, 86.6, 84.1, 51.8,
40.8, 31.6; MS (ES) m/z 317.1110 (MH+ [C12H13N8O3] = 317.1111).
3ꢁ-(Carboxymethyl)-3ꢁ,5ꢁ-dideoxy-5ꢁ-[(N-methylcarbamoyl)amino]adeno-
sine-2ꢁ,3ꢁ-lactone (13). To a stirred solution of 8 (26 mg, 0.051 mmol) in
CH3CN (1.0 mL) were added PhCH2N(Et)3Cl (30 mg, 0.13 mmol), KF
(15 mg, 0.26 mmol), and H2O (80 µL). The mixture was vigorously stirred
at ambient temperature until TLC indicated that 8 had been consumed
(9 hours). The reaction mixture was added directly to a column and
chromatographed (solvent A) to give 13 (14 mg, 79%): UV (MeOH) λmax
1
260 nm, λmin 239 nm; H (DMSO-d6, 500 MHz) δ 8.32 (s, 1H), 8.17 (s,
1H), 7.34 (br s, 2H), 6.24 (d, J = 1.5 Hz, 1H), 6.07 (t, J = 5.8 Hz, 1H),
5.78 (q, J = 4.7 Hz, 1H), 5.51 (dd, J = 2.3, 7.3 Hz, 1H), 3.97–3.93 (m, 1H),
3.28–3.24 (m, 1H), 2.94 (dd, J = 8.5, 18.0 Hz, 1H), 2.53 (d, J = 5.0 Hz,
13
3 H), 2.51–2.46 (m, 2H); C NMR (DMSO-d6, 125 MHz) δ 175.7, 158.6,

3ꢁ-Carboxymethyl-3ꢁ-Deoxyadenosine Derivatives
513
156.1, 152.8, 148.8, 139.6, 119.1, 87.8, 86.8, 84.5, 41.8, 40.9, 31.8, 26.3; MS
(ES) m/z 348.1416 (MH+ [C14H18N7O4] = 348.1420).
3ꢁ-(Carboxymethyl)-3ꢁ,5ꢁ-dideoxyadenosine (2 a). To a solution of 11 (21
mg, 0.072 mmol) in THF:MeOH [0.6 mL, (5:1)] was added NaOH (80 µL
of 1.0 M, 0.080 mmol). The mixture was stirred at 65◦C until TLC showed
conversion of 11 to baseline product. Volatiles were removed under reduced
pressure to give 2a (24 mg, quant). The crude residue was dissolved in
H2O (100 µL). Silica gel and solvent A were added, and volatiles were
evaporated under reduced pressure (≤20◦C). The dried silica gel was added
to a column and chromatographed (solvent A) to give 2a (18 mg, 81%): UV
(MeOH) λmax 261 nm, λmin 229 nm; 1H NMR (DMSO-d6, 500 MHz) δ 8.57
(br s, 1H), 8.42 (s, 1H), 8.12 (s, 1H), 7.22 (br s, 2H), 5.84 (d, J = 2.5 Hz,
1H), 5.52 (br s, 1H), 4.32 (d, J = 4.5 Hz, 1H), 4.01–3.98 (m, 1H), 3.69 (d,
J = 12.0 Hz, 1H), 3.62–3.59 (m, 1H), 3.50 (d, J = 12.0 Hz, 1H), 2.24 (dd, J
= 7.5, 14.5 Hz, 1H), 2.17 (dd, J = 5.3, 14.8 Hz, 1H), 1.77–1.75 (m, 1H); 13C
NMR (DMSO-d6, 125 MHz) δ 173.4, 156.0, 152.4, 148.6, 138.6, 119.1, 90.4,
84.3, 75.4, 60.7, 37.5, 29.6; MS (ES) m/z 310.1144 (MH+ [C12H16N5O5] =
310.1151).
3ꢁ-(Carboxymethyl)-3ꢁ,5ꢁ-dideoxy-5ꢁ-[(N-methylcarbamoyl)amino]-N6-
(N-phenylcarbamoyl)adenosine sodium salt (2b). To a solution of 2h (54
mg, 0.12 mmol) in DMSO (0.5 mL) was added NaOH (0.20 mL of 1.0 M,
0.20 mmol). The mixture was stirred at ambient temperature until TLC
showed conversion of 2h to baseline product. Volatiles were removed under
reduced pressure to give 2b (64 mg, quant). This material was >98% pure
1
as determined by reverse phase HPLC and H NMR: UV (MeOH) λmax
1
279 nm, λmin 243 nm; UV (H2O) λmax 278 nm (ε 22,600); H NMR
(D2O:DMSO-d6 (1:9), 500 MHz) δ 8.25 (s, 1H), 8.11 (s, 1H), 7.55 (d, J
= 8.0 Hz, 2H), 7.22 (t, J = 7.8 Hz, 2H), 6.90–6.87 (m, 1H), 5.83 (d, J =
3.0 Hz, 1H), 4.50 (dd, J = 2.0, 6.0 Hz, 1H), 3.90–3.87 (m, overlaps with
solvent), 3.41 (dd, J = 3.0, 14.5 Hz, 1H), 3.16 (dd, J = 6.5, 14.0 Hz, 1H),
2.52 (s, 3 H), 2.42–2.38 (m, 1H), 2.31 (dd, J = 8.0, 14.8 Hz, 1H), 2.15
13
(dd, J = 5.3, 14.8 Hz, 1H); C NMR (D2 O:DMSO-d6 (1:9), 125 MHz) δ
177.5, 161.6, 159.8, 159.1, 152.3, 148.5, 141.5, 138.2, 129.3, 124.7, 121.8,
119.2, 90.3, 83.8, 76.8, 42.8, 42.0, 35.0, 26.9; MS (ES) m/z 507.1711 (MH+
[C21H24N8O6Na] = 507.1717).
2ꢁ-O-(tert-Butyldimethylsilyl)-3ꢁ,5ꢁ-dideoxy-3ꢁ-[(ethoxycarbonyl)methyl]-
5ꢁ-[(N-methylcarbamoyl)amino]-N6-(N-phenylcarbamoyl)adenosine (2c). A
solution of 9 (100 mg, 0.168 mmol) and 10% Pd–C (50 mg) in EtOAc
(2 mL) was vigorously stirred for 15 hours under an atmosphere of H2
(balloon pressures). p-Nitrophenyl N -methylcarbamate (45 mg, 0.23 mmol)
and anhydrous Na2CO3 (45 mg, 0.42 mmol) were added, and the resulting
mixture was stirred for 4 hours under N2. Solids were removed via filtration
(celite/EtOAc), and volatiles were evaporated under reduced pressure.
The crude residue was chromatographed (5 → 10% MeOH/CH2Cl2) to

514
M. A. Peterson et al.
give 2c (101 mg, 96%): UV (MeOH) λmax 279 nm (ε 22,700), λmin 242
1
nm; H NMR (CDCl3, 500 MHz) δ 12.31 (s, 1H), 10.13 (br s, 1H), 8.86 (s,
1H), 8.64 (s, 1H), 7.57 (d, J = 7.5 Hz, 2H), 7.42–7.39 (m, 2H), 7.21–7.18
(m, 1H), 5.94 (s, 1H), 5.78 (t, J = 6.3 Hz, 1H), 5.06–5.03 (m, 2H), 4.20
(d, J = 10.5 Hz, 1H), 4.11–4.07 (m, 2H), 3.85–3.83 (m, 1H), 3.49 (d, J =
13.0 Hz, 1H), 2.79 (dd, J = 4.5, 17.0 Hz, 1H), 2.62 (d, J = 5.0 Hz, 3 H),
2.62–2.50 (m, 1H), 2.49–2.48 (m, 1H), 1.24 (t, J = 7.0 Hz, 3 H), 0.94 (s, 9
H), 0.27 (s, 3 H), 0.11 (s, 3 H); 13 C NMR (CDCl3, 125 MHz) δ 172.0, 159.4,
153.3, 149.9, 149.8, 142.8, 137.3, 129.1, 124.6, 121.2, 92.0, 84.7, 77.2, 60.3,
39.7, 38.5, 28.8, 26.7, 25.7, 17.9, 14.0, −4.3, −5.8; MS (FAB) m/z 649.2899
(MNa+ [C29H42N8O6SiNa] = 649.2894).
2ꢁ-O-(tert-Butyldimethylsilyl)-3ꢁ,5ꢁ-dideoxy-3ꢁ-[(N-methoxy-N-methylcarb-
oxamido) methyl]-5ꢁ-[(N-methylcarbamoyl)amino]-N6-(N-phenylcarbamoyl)
adenosine (2d). A solution of 10 (50 mg, 0.082 mmol) and 10% Pd–C
(50 mg) in EtOAc (1 mL) was vigorously stirred for 18 hours under an
atmosphere of H2 (balloon pressures). p-Nitrophenyl N -methylcarbamate
(25 mg, 0.13 mmol) and anhydrous Na2CO3 (50 mg, 0.47 mmol) were
added, and the resulting mixture was stirred for 4 hours under N2. Solids
were removed via filtration (celite/EtOAc), volatiles were evaporated
under reduced pressure, and the residue was chromatographed (10%
MeOH/EtOAc) to give 2d (33 mg, 63%): UV (MeOH) λmax 279 nm (ε
1
22,200), λmin 245 nm; H NMR (CDCl3, 500 MHz) δ 12.32 (s, 1H), 10.14
(br s, 1H), 8.90 (s, 1H), 8.61 (s, 1H), 7.58 (d, J = 7.5 Hz, 2H), 7.40 (t, J =
7.5 Hz, 2H), 7.19–7.16 (m, 1H), 5.96 (s, 1H), 5.85 (br s, 1H), 5.07 (d, J =
4.0 Hz, 1H), 5.02 (d, J = 3.5 Hz, 1H), 4.25 (d, J = 10.5 Hz, 1H), 3.78–3.75
(m, 1H), 3.73 (s, 3 H), 3.58 (d, J = 11.5 Hz, 1H), 3.13 (s, 3 H), 2.78 (d, J
= 5.0 Hz, 2H), 2.61 (d, J = 4.5 Hz, 3 H), 2.50–2.46 (m, 1H), 0.94 (s, 9 H),
13
0.28 (s, 3 H), 0.10 (s, 3 H);
C NMR (CDCl3, 125 MHz) δ 172.7, 159.3,
153.2, 150.04, 150.01, 149.9, 142.8, 137.5, 129.1, 124.5, 121.2, 92.1, 84.8,
77.6, 61.1, 40.3, 38.4, 32.1, 29.7, 26.8, 25.8, 18.0, −4.4, −5.5; MS (ES) m/z
642.3182 (MH+ [C29H44N9O6Si] = 642.3184).
5ꢁ-Azido-3ꢁ-(carboxymethyl)-3ꢁ,5ꢁ-dideoxy-N6-(N-phenylcarbamoyl)
adenosine sodium salt (2e). To a solution of 2g (29 mg, 0.067 mmol) in
DMSO (0.5 mL) was added NaOH (0.10 mL of 1.0 M, 0.10 mmol). The
mixture was stirred at ambient temperature until TLC showed conversion
of 2g to baseline product. Volatiles were removed under reduced pressure
to give 2g (32 mg, quant). This material was >98% pure as determined
1
by reverse phase HPLC and H NMR: UV (MeOH) λmax 279 nm, λmin
1
241 nm; UV (H2O) λmax 278 nm (ε 22, 100); H NMR (D2 O:DMSO-d6
(1:9), 500 MHz) δ 8.24 (s, 1H), 8.13 (s, 1H), 7.49 (dd, J = 1.8, 7.3 Hz,
2H), 7.26–7.16 (m, 2H), 6.87–6.83 (m, 1H), 5.86 (d, J = 2.5 Hz, 1H), 4.54
(dd, J = 2.0, 6.0 Hz, 1H), 4.03–4.00 (m, overlaps with solvent), 3.54 (dd, J
= 2.3, 13.8 Hz, 1H), 3.42 (dd, J = 5.8, 13.8 Hz, 1H), 2.45–2.44 (m, 1H),

3ꢁ-Carboxymethyl-3ꢁ-Deoxyadenosine Derivatives
515
2.26 (dd, J = 7.8, 15.3 Hz, 1H), 2.08 (dd, J = 5.5, 15.0 Hz, 1H); 13 C NMR
(D2 O:DMSO-d6 (1:9), 125 MHz) δ 177.3, 160.7, 158.3, 152.4, 148.9, 141.2,
138.7, 138.6, 129.4, 124.2, 122.2, 119.5, 90.3, 83.3, 76.5, 53.1, 41.8, 34.4; MS
(ES) m/z 476.1404 (MH+ [C19H19N9O5Na] = 476.1407).
5ꢁ-Azido-3ꢁ-(carboxymethyl)-3ꢁ,5ꢁ-dideoxyadenosine (2f). To a solution
of 12 (22 mg, 0.070 mmol) in THF:MeOH [0.6 mL, (5:1)] was added
NaOH (80 µL of 1.0 M, 0.080 mmol). The mixture was stirred at 65◦C
until TLC showed conversion of 12 to baseline product. The mixture was
added directly to a chromatography column and chromatographed (5 →
10%MeOH/CH2Cl2; 25 → 50% solvent A/EtOAc). Pooled fractions were
evaporated under reduced pressure (≤20◦C) to give 2f (20 mg, 85%): UV
(MeOH) λmax 260 nm, λmin 228 nm; UV (H2O) λmax 260 nm (ε 14,500);
1H NMR (DMSO-d6, 500 MHz) δ 8.27 (s, 1H), 8.17 (s, 1H), 7.30 (br s, 2H),
5.96 (d, J = 2.0 Hz, 1H), 4.64 (dd, J = 2.0, 5.5 Hz, 1H), 4.10–4.07 (m, 1H),
3.70–3.66 (m, 2H), 3.33 (br s, 1H), 2.77–2.71 (m, 1H), 2.57 (dd, J = 8.8,
13
17.3 Hz, 1H), 2.43 (dd, J = 5.3, 17.3 Hz, 1H); C NMR (DMSO-d6, 125
MHz) δ 173.3, 156.1, 152.6, 149.0, 138.7, 119.1, 90.4, 82.2, 74.8, 52.2, 39.8,
29.6; MS (ES) m/z 335.1230 (MH+ [C12H15N8O4] = 335.1216).
5ꢁ-Azido-3ꢁ-(carboxymethyl)-3ꢁ,5ꢁ-dideoxy-N6-(N-phenylcarbamoyl)
adenosine-2ꢁ,3ꢁ-lactone (2 g). To a stirred solution of 9 (73 mg, 0.123 mmol)
in CH3CN (2.0 mL) were added PhCH2 N(Et)3Cl (5 mg, 0.022 mmol), KF
(15 mg, 0.26 mmol), and H2O (40 µL). The mixture was vigorously stirred
at ambient temperature until TLC indicated that 9 had been consumed (4
days). Silica gel was added and volatiles were evaporated under reduced
pressure (≤20◦C). The dried silica gel was poured onto the top of a
column packed with 75% EtOAc/hexanes and product was eluted (75%
EtOAc/hexanes → EtOAc). Evaporation of pooled fractions gave 2g (46
mg, 86%): UV (MeOH) λmax 279, λmin 240; UV (H2O) λmax 278 nm
1
(ε 20,500); H (DMSO-d6, 500 MHz,) δ 11.70 (s, 1H), 10.21 (s, 1H), 8.72
(s, 1H), 8.66 (s, 1H), 7.63 (d, J = 7.5 Hz, 2H), 7.38–7.35 (m, 2H), 7.08
(t, J = 7.5 Hz, 1H), 6.43 (d, J = 2.0 Hz, 1H), 5.65 (dd, J = 1.8, 6.8 Hz,
1H), 4.28–4.24 (m, 1H), 3.73 (dd, J = 3.0, 13.5 Hz, 1H), 3.55–3.49 (m,
13
2H), 2.98 (dd, J = 8.5, 18.0 Hz, 1H), 2.69 (dd, J = 1.5, 18.0 Hz, 1H);
C
NMR (DMSO-d6, 125 MHz) δ 175.3, 151.0, 150.7, 150.0, 142.6, 138.4, 128.9,
123.2, 120.5, 119.4, 88.2, 86.4, 84.3, 51.7, 40.6, 31.5; MS (ES) m/z 436.1483
(MH+ [C19H18N9O4] = 436.1482).
3ꢁ-(Carboxymethyl)-3ꢁ,5ꢁ-dideoxy-5ꢁ-[(N-methylcarbamoyl)amino]-N6-
(N-phenylcarbamoyl)adenosine-2ꢁ,3ꢁ-lactone (2h). To a stirred solution of
2c (82 mg, 0.131 mmol) in CH3CN (3.0 mL) were added PhCH2 N(Et)3Cl
(50 mg, 0.22 mmol), KF (22 mg, 0.38 mmol), and H2O (80 µL). The
mixture was vigorously stirred at ambient temperature until TLC indicated
that 2c had been consumed (60 hours). Silica gel was added and volatiles
were evaporated under reduced pressure (≤20◦C). The dried silica gel was

516
M. A. Peterson et al.
poured onto the top of a column packed with 5% MeOH/CH2Cl2 and
eluted (5 → 10% MeOH/CH2Cl2). Evaporation of pooled fractions gave
2h (56 mg, 92%): UV (MeOH) λmax 279 nm (ε 23,200), λmin 240 nm;
1H NMR (DMSO-d6, 500 MHz) δ 11.74 (s, 1H), 10.18 (br s, 1H), 8.71 (s,
1H), 8.66 (s, 1H), 7.63 (d, J = 8.0 Hz, 2H), 7.38–7.35 (m, 2H), 7.09 (t, J =
7.5 Hz, 1H), 6.37 (d, J = 2.0 Hz, 1H), 6.05 (t, J = 6.0 Hz, 1H), 5.77 (dd,
J = 4.5, 8.5 Hz, 1H), 5.57 (dd, J = 1.8, 7.3 Hz, 1H), 4.03–3.99 (m, 1H),
3.41–3.36 (m, 2H), 2.98 (dd, J = 8.5, 18.0 Hz, 1H), 2.55 (d, J = 5.0 Hz,
13
3 H); C NMR (DMSO-d6, 125 MHz) δ 176.3, 159.3, 151.8, 151.6, 150.8,
143.3, 139.2, 129.7, 123.9, 121.4, 120.1, 88.8, 87.5, 85.7, 42.4, 41.5, 40.7,
32.5, 27.1; MS (ES) m/z 467.1795 (MH+ [C21H23N8O5] = 467.1791).
CELLS AND VIRUS
MT2 and H9 cells, both CD4+ lymphoblastoid cell lines, were cultured
in RPMI-1640 containing HEPES and supplemented with 11.5% heat-
inactivated fetal bovine serum and 2 mM L-glutamine. HIV-containing su-
pernatant fluids were clarified of cells by low-speed centrifugation followed
by filtration through 0.45 µm cellulose acetate filters.
50% Cytotoxic and 50% Effective Concentrations
The 50% cytotoxic/cytostatic concentration (CT50) of each compound
was determined as described, previously.[41–43] Briefly, in triplicate wells
of a 96 well plate each compound was serially diluted. MT2 cells were
added to each well and the cells plus compound were incubated for
72 hours at 37◦C. Cells were resuspended, transferred to a poly-L-lysine
coated plate, and stained with Finter’s neutral red dye, a vital dye. Cells
were incubated for 1 hour at 37◦C to stain and to adhere. Cells were
washed with phosphate buffered saline and lysed in acid alcohol. A540
was determined and the percent viable cells were calculated relative to 8
replicates without compound (100% viable) and 8 blank wells (0% viable).
The CT50 was calculated using CalcuSyn for Windows software. The 5%
cytotoxic/cytostatic dose (CT5), a nontoxic concentration where 95% of
the cells were viable, was also calculated for each of the compounds.
The 50% effective concentration (EC50) was determined in essentially
the same manner as the CT50 and as described, previously.[41–43] However,
after a one hour incubation of cells and compounds, HIVLAI was added
to each well. The cells were harvested at 72 hours and the percentage of
protection from HIV-induced cytopathic effects was calculated using Finter’s
neutral red dye. The percentage of viable cells relative to 8 cell control
replicates (cells but no virus or compound, 100% viable) and 8 virus control

3ꢁ-Carboxymethyl-3ꢁ-Deoxyadenosine Derivatives
517
replicates (cells and virus but no compound, 0% viable) was determined.
The EC50 was calculated using CalcuSyn for Windows software.
Inhibition of HIV Integrase
We previously have documented the practical application of screening
compounds at 10 µM for inhibition of integrase in vitro.[44] Compounds
which inhibit HIV replication by 50% or more at 10 µM are further sub-
jected to dilution from 10 µM to 30 nM in 1/2 log10 dilutions to calculate the
50% inhibitory concentration (IC50). None of the compounds were active
at 10 µM in the 3ꢁ-endprocessing or strand transfer reactions; therefore,
IC50 analyses were not performed. Recombinant integrase from HIVNL4–3
was expressed in and purified from Escherichia coli. Recombinant integrase
was incubated with compound and 32P-labeled oligonucleotide substrate
homologous to the HIV long terminal repeat DNA for 1 hour at 37◦C.
Reactions were stopped by the addition of EDTA to a final concentration of
18 mM. The reaction products were separated from substrate by denaturing
polyacrylamide gel electrophoresis. The percent conversion of substrate
to products was quantified using a Storm phosphorimager (Molecular
Dynamics, Sunnyvale, CA, USA). The conditions for these reactions have
been well-described, previously.[44–48] Controls in all experiments include
25 µM L-chicoric acid (100% inhibition) and 25 µM L-tartaric acid (0%
inhibition).
REFERENCES
1. Li, N.-S.; Piccirilli, J.A. Efficient synthesis of 2ꢁ-C-β-methylguanosine. J. Org. Chem. 2006, 71,
4018–4020.
2. Ye, J.-D.; Liao, X.; Piccirilli, J.A. Synthesis of 2ꢁ-C-difluoromethylribonucleosides and their enzy-
matic incorporation into oligonucleotides. J. Org. Chem. 2005, 70, 7902–7910.
3. Li, N.-S.; Tang, X.-Q.; Piccirilli, J.A. 2ꢁ-C-Branched ribonucleosides. 2. Synthesis of 2ꢁ-C-β-
trifluoromethyl pyrimidine ribonucleosides. Org. Lett. 2001, 3, 1025–1028.
4. Robins, M.J.; Doboszewski, B.; Timoshchuk, V.A.; Peterson, M.A. Glucose-derived
3ꢁ-(carboxymethyl)-3ꢁ-deoxyribonucleosides and 2ꢁ,3ꢁ-lactones as synthetic precursors for
amide-linked oligonucleotide analogues. J. Org. Chem. 2000, 65, 2939–2945.
5. Harry-Oꢁkuru, R.E.; Smith, J.M.; Wolfe, M.S. A short, flexible route toward 2ꢁ-C-branched ribonu-
cleosides. J. Org. Chem. 1997, 62, 1754–1759.
6. Jung, P.M.J.; Burger, A.; Biellmann, J.-F. Diastereofacial selective addition of ethynylcerium reagent
and Barton-McCombie reaction as the key steps for the synthesis of C-3ꢁ-ethynylribonucleosides
and of C-3ꢁ-ethynyl-2ꢁ-deoxyribonucleosides. J. Org. Chem. 1997, 62, 8309–8314.
7. Robins, M.J.; Sarker, S.; Xie, M.; Zhang, W.; Peterson, M.A. Synthesis of 2ꢁ, 3ꢁ-fused (3.3.0)
γ -butyrolactone-nucleosides and coupling with amino nucleosides to give amide-linked nucleotide-
dimer analogues. Tetrahedron Lett. 1996, 37, 3921–3924.
8. Clark, J.L.; Hollecker, L.; Mason, J.C.; Stuyver, L.J.; Tharnish, P.M.; Lostia, S.; McBrayer, T.R.;
Schinazi, R.F.; Watanabe, K.A.; Otto, M.J.; Furman, P.A.; Stec, W.J.; Patterson, S.E.; Pankiewicz, K.W.
Design, synthesis, and antiviral activity of 2ꢁ-deoxy-2ꢁ-fluoro-2ꢁ-C-methylcytidine, a potent inhibitor
of hepatitis C virus replication. J. Med. Chem. 2005, 48, 5504–5508.
9. Eldrup, A.B.; Prhavc, M.; Brooks, J.; Bhat, B.; Prakash, T.P.; Song, Q.; Bera, S.; Bhat, N.; Dande, P.;
Cook, P.D.; Bennett, C.F.; Carroll, S.S.; Ball, R.G.; Bosserman, M.; Burlein, C.; Colwell, L. F.; Fay, J.F.;

518
M. A. Peterson et al.
Flores, O.A.; Getty, K.; LaFemina, R.L.; Leone, J.; MacCoss, M.; McMasters, D.R.; Tomassini, J.E.;
Von Langen, D.; Wolanski, B.; Olsen, D.B. Structure-acitivity relationship of heterobase-modified
2ꢁ-C-methyl ribonucleosides as inhibitors of hepatitis C virus RNA replication. J. Med. Chem. 2004,
47, 5284–5297.
10. Lee-Ruff, E.; Ostrowski, M.; Ladha, A.; Stynes, D.V.; Vernik, I.; Jiang, J.-L.; Wan, W.-Q.; Ding, S.-F.;
Joshi, S. Synthesis and HIV inhibition of 2ꢁ,3ꢁ-dideoxy-3ꢁ-C-hydroxymethyl nucleosides. J. Med. Chem.
1996, 39, 5276–5280.
11. Franchetti, P.; Cappellacci, L.; Pasqualini, M.; Petrelli, R.; Vita, P.; Jayaram, H.N.; Horvath, Z.;
Szekeres, T.; Grifantini, M. Antitumor activity of C-methyl-β-D-ribofuranosyladenine nucleoside
ribonucleotide reductase inhibitors. J. Med. Chem. 2005, 48, 4983–4989.
12. Ichikawa, S.; Shuto, S.; Minakawa, N.; Matsuda, A. Nucleosides and nucleotides. 163. Synthesis
of 3ꢁ-β-branched uridine derivatives via intramolecular Reformatsky-type reaction promoted by
samarium diiodide. J. Org. Chem. 1997, 62, 1368–1375.
13. Hattori, H.; Tanaka, M.; Fukushima, M.; Sasaki, T.; Matsuda, A. Nucleosides and Nucleotides. 158. 1-
(3-C-ethynyl-β-D-ribo-pentofuranosyl)cytosine, 1-(3-C-ethynyl-β-D-ribo-pentofuranosyl)uracil, and
their nucleobase analogues as new potential multifunctional antitumor nucleosides with a broad
spectrum of activity. J. Med. Chem. 1996, 39, 5005–5011.
14. Vanheusden, V.; Munier-Lehmann, H.; Froeyen, M.; Dugue, L.; Heyerick, A.; De Keukeleire, D.;
Pochet, S.; Busson, R.; Herdewijn, P.; Calenbergh, S. V. 3ꢁ-C-Branched-chain-substituted nucleosides
and nucleotides as potent inhibitors of mycobacterium tuberculosis thymidine monophosphate
kinase. J. Med. Chem. 2003, 46, 3811–3821.
15. Meldgaard, M.; Hansen, F.G.; Wengel, J. 3ꢁ-C-Branched LNA-type nucleosides locked in an
N-type furanose ring conformation: Synthesis, incorporation into oligodeoxynucleotides, and
hybridization studies. J. Org. Chem. 2004, 69, 6310–6322.
16. Cherepanov, P.; Maertens, G.; Proost, P.; Devreese, B.; Van Beeumen, J.; Engelborghs, Y.; De Clercq,
E.; Debyser, Z. HIV-1 integrase forms stable tetramers and associates with LEDGF/p75 protein in
human cells. J. Biol. Chem. 2003, 278, 372–381.
17. Dyda, F.; Hickman, A.B.; Jenkins, T.M.; Engelman, A.; Craigie, R.; Davies, D.R. Crystal structure
of the catalytic domain of HIV-1 integrase: Similarity to other polynucleotidyl transferases. Science
1994, 266, 1981–1986.
18. Adesokan, A.A.; Roberts, V.A.; Lee, K.W.; Lins, R.D.; Briggs, J.M. Predictions of HIV-1 inte-
grase/viral DNA interactions in the catalytic domain by fast molecular docking. J. Med. Chem. 2004,
47, 821–828.
19. Jenkins, T.M.; Esposito, D.; Engelman, A.; Craigie, R. Critical contacts between HIV-1 integrase
and viral DNA identified by structure-based analysis and photo-crosslinking. EMBO J . 1997, 16,
6849–6859.
20. Dudev, T.; Cowan, J.A.; Lim, C. Competitive binding in magnesium coordination chemistry: Water
versus ligands of biological interest. J. Am. Chem. Soc. 1999, 121, 7665–7673.
21. Rawat, D.S.; Zaleski, J.M. Mg2+-Induced thermal enediyne cyclization at ambient temperature. J.
Am. Chem. Soc. 2001, 123, 9675–9676.
22. Goldgur, Y.; Dyda, F.; Hickman, A.B.; Jenkins, T.M.; Craigie, R.; Davies, D.R. Three new structures
of the core domain of HIV-1 integrase: An active site that binds magnesium. Proc. Natl. Acad. Sci.
USA 1998, 95, 9150–9154.
23. Maignan, S.; Guilloteau, J.-P.; Zhou-Liu, Q.; Clement-Mella, C.; Mikol, V. Crystal structures of the
catalytic domain of HIV-1 integrase free and complexed with its metal cofactor: High level of
similarity of the active site with other viral integrases. J. Mol. Biol. 1998, 282, 359–368.
24. Peterson, M.A.; Nilsson, B. L.; Sarker, S.; Doboszewski, B.; Zhang, W.; Robins, M.J. Amide-linked
ribonucleoside dimers derived from 5ꢁ-amino-5ꢁ-deoxy- and 3ꢁ-(carboxymethyl)-3ꢁ-deoxynucleoside
precursors. J. Org. Chem. 1999, 64, 8183–8192.
25. Clark, V.M.; Todd, A.R.; Zussman, J. Nucleotides. Part VIII. Cyclonucleoside salts. A novel
rearrangement of some toluene-p-sulphonylnucleosides. J. Chem. Soc. 1951, 2952–2958.
26. Davisson, V.J.; Davis, D.R.; Dixit, V.M.; Poulter, C.D. Synthesis of nucleotide 5ꢁ-diphosphates from
5ꢁ-O-tosyl nucleosides. J. Org. Chem. 1987, 52, 1794–1801.
27. Liu, F.; Austin, D.J. A general synthesis of 5ꢁ-azido-5ꢁ-deoxy-2ꢁ,3ꢁ-O-isopropylidene nucleosides. J.
Org. Chem. 2001, 66, 8643–8645.

3ꢁ-Carboxymethyl-3ꢁ-Deoxyadenosine Derivatives
519
28. Liu, F.; Austin, D.J. Synthesis of 5ꢁ-functionalized adenosine: Suppression of cyclonucleoside
formation. Tetrahedron Lett. 2001, 42, 3153–3154.
29. Anzai, K. Ring cleavage and modification at C-2 of a 3,5ꢁ-cyclonucleoside derivative of adenosine.
Agric. Biol. Chem. 1976, 40, 373–376.
30. Nyilas, A.; Glemarec, C.; Chattopadhyaya, J. Synthesis of [3ꢁ-(O) → (5ꢁ(C)]-oxyacetamido linked
nucleosides. Tetrahedron 1990, 46, 2149–2164.
31. Li, C.; Shih, T.-L.; Jeong, J.U.; Arasappan, A.; Fuchs, P.L. The use of tetramethylguanidinium
azide in non-halogenated solvents avoids potential explosion hazards. Tetrahedron Lett. 1994, 35,
2645–2646.
32. Peterson, M.A.; Shi, H.; Ke, P. A simple and efficient biphasic method for the preparation of 4-
nitrophenyl N-methyl- and N-alkylcarbamates. Tetrahedron Lett. 2006, 47, 3405–3407.
33. Montgomery, J.A.; Thomas, H.J. Nitrosoureidonucleosides. J. Med. Chem. 1979, 22, 1109–1113.
34. Rarey, M.; Kramer, B.; Lengauer, T. Multiple automatic base selection: Protein–ligand docking
based on incremental construction without manual intervention. J. Comp. Mol. Des. 1997, 11,
369–384.
35. Kramer, B.; Rarey, M.; Lengauer, T. Evaluation of the flexX incremental construction algorithm for
protein-ligand docking. Protein. Struct. Funct. Gen. 1999, 37, 228–241.
36. Stahl, M.; Rarey, M. Detailed analysis of scoring functions for virtual screening. J. Med. Chem. 2001,
44, 1035–1042.
37. Bujacz, G.; Alexandratos, J.; Wlodawer, A.; Merkel, G.; Andrake, M.; Katz, R.A.; Skalka, A.M. Binding
of different divalent cations to the active site of avian sarcoma virus integrase and their effects on
enzymatic activity. J. Biol. Chem. 1997, 272, 18161–18168.
38. Bujacz, G.; Alexandratos, J.; Zhou-Liu, Q.; Clement-Mella, C.; Wlodawer, A. The catalytic domain
of human immunodeficiency virus integrase: ordered active site in the F185 H mutant. FEBS Letters
1996, 398, 175–178.
39. Still, W.C.; Kahn, M.; Mitra, A. Rapid chromatographic technique for preparative separations with
moderate resolution. J. Org. Chem. 1978, 43, 2923–2925.
40. Papa, A.J. Synthesis and azidolysis of 2-chlorotetramethylguanidine. Synthetic utility of hexa- and
tetramethylguanidinium azide. J. Org. Chem. 1966, 31, 1426–1430.
41. Montefiori, D.C.; Robinson, W.E., Jr.; Schuffman, S. S.; Mitchell, W.M. Evaluation of antiviral
drugs and neutralizing antibodies against human immunodeficiency virus by a rapid and sensitive
microtiter infection assay. J. Clin. Microbiol. 1988, 26, 231–235.
42. Robinson, W.E., Jr.; Cordeiro, M.; Abdel-Malek, S.; Jia, Q.; Chow, S.A.; Reinecke, M.G.; Mitchell,
W.M. Dicaffeoylquinic acid inhibitors of human immunodeficiency virus integrase: Inhibition of
the core catalytic domain of human immunodeficiency virus integrase. Mol. Pharmacol. 1996, 50,
846–855.
43. Robinson, W.E., Jr.; Reinecke, M.G.; Abdel-Malek, S.; Jia, Q.; Chow; S. A. Inhibitors of HIV-1
replication that inhibit HIV integrase. Proc. Natl. Acad. Sci., USA 1996, 93, 6326–6331.
44. King, P.J.; Ma, G.; Miao, W.; Jia, Q.; McDougall, B.R.; Reinecke, M.G.; Cornell, C.; Kuan, J.; Kim,
T.R.; Robinson, W.E., Jr. Structure-activity relationships: Analogues of the dicaffeoylquinic and
dicaffeoyltartaric acids as potent inhibitors of human immunodeficiency virus type 1 integrase and
replication. J. Med. Chem. 1999, 42, 497–509.
45. King, P.J.; Lee, D.J.; Reinke, R.A.; Victoria, J.G.; Beale, K.; Robinson, W.E., Jr. Human immunode-
ficiency virus type 1 integrase containing a glycine to serine mutation at position 140 is attenuated
for catalysis and resistant to integrase inhibitors. Virology 2003, 306, 147–161.
46. Lee, D.J.; Robinson, W.E., Jr. Human immunodeficiency virus type 1 (HIV-1) integrase: Resistance
to diketoacid integrase inhibitors impairs HIV-1 replication, integration, and confers cross-
resistance to L-chicoric acid. J. Virol. 2004, 78, 5835–5847.
47. Lee, D.J.; Robinson, W.E., Jr. Preliminary mapping of a putative inhibitor-binding pocket for human
immunodeficiency virus type 1 integrase inhibitors. Antimicrob. Agents Chemother. 2006, 50, 134–142.
48. Reinke, R.A.; King, P.J.; Victoria, J.G.; McDougall, B.R.; Ma, G.; Mao, Y.; Reinecke, M.G.; Robinson,
W.E., Jr. Dicaffeoyltartaric acid analogues inhibit human immunodeficiency virus type 1 (HIV-1)
integrase and HIV-1 replication at non-toxic concentrations. J. Med. Chem. 2002, 45, 3669–3683.